^ The 12th Informal Conference on Photochemistry v<* June 28 - July 1 U.S. Department of Commerce National Bureau of Standards Gaithersburg, Maryland Extended Abstracts Digitized by the Internet Archive in 2012 with funding from LYRASIS Members and Sloan Foundation http://www.archive.org/details/12thinformalconf00info MONDAY, JUNE 28, 1976 RED AUDITORIUM A.M. 9:00 Introductory Remarks - M. J. Kurylo (Physical Chemistry Division, National Bureau of Standards, Washington, DC 20234) 9:05 Welcoming Address - J. D. Hoffman (Director, Institute for Material Research, National Bureau of Standards, Washington, DC 20234) SESSION A: Actinometric and Radiometric Measurement CHAIRMAN: T. W. Martin (Department of Chemistry, Vanderbilt University, Nashville, TN 37235) 9:20 Al. Invited Lecture: A Modern Approach to Accurate Radiometry - E. F. Zalewski (Physical Chemistry Division, National Bureau of Standards, Washington, DC 20234) 9:50 A2. Temperature Dependence of the Fluorescence Quantum Yields of Rhodamine Derivatives - R. E. Schwerzel , N. E. Klosterman, and C. M. Verber (Battelle Columbus Laboratories, Columbus, OH 43201) 10:10 A3. Invited Lecture: National Bureau of Standards Ultraviolet Radiometric Standards - W. R. Ott (Optical Physics Division, a. National Bureau of Standards, Washington, DC 20234) tj |* 10:40 A4. New Methods of Measuring Light Intensities - D. G. Taylor, t J . N . Demas , W. D. Bowman, E. W. Harris, and R. P. McBride (Chemistry Department, University of Virginia, Charlottesville 2 VA 22901) 11:00 A5. Infrared Photography at 5 and 10 urn - G. F. Frazier, T. D. Wilkerson , and J. M. Lindsay (Versar Inc. , Springfield, VA 22151) 11:20 COFFEE BREAK SESSION B: Chemiluminescent Processes CHAIRMAN: J. P. Simons (Chemistry Department, The University, Birmingham B15 2TT England) 11:40 Bl. Kinetics of the Chemiluminescence Accompanying Metal Atom Oxidation Reactions - W. Felder and A. Fontijn (Aerochem Research Laboratories, Princeton, NJ 08540) P.M. 12:00 B2. A Study of the Chemiluminescence of the Pb + Reactions - M. J. Kurylo, W. Bratm, S. Abramowitz , and M. Krauss (Physical Chemistry Division, National Bureau of Standards Washington, DC 20234) 12:20 B3. Chemiluminescence in Reactions of Ozone - S . Toby , F. S'. Toby, and B. Kaduk (School of Chemistry, Rutgers University, New Brunswick, NJ 08903) 12:40 B4. Dioxetane Chemistry in the Gas Phase: UV - Visible Chemiluminescence from the Reactions of ( A ) with Olefins - D . J . Bogan (Chemical Dynamics Branch, Naval Research Laboratory, Washington, DC 20375) 1 : 00 LUNCH MONDAY, JUNE 28, 1976 GREEN AUDITORIUM SESSION C: Photochemical Isotope Separation CHAIRMAN: R. D. Deslattes (Optical Physics Division, National Bureau of Standards, Washington, DC 20234) A.M. 9:20 CI. Plenary Lecture: Molecular Dissociation by High Intensity Infrared Laser Radiation - K. H. Welge (Universitaet Bielefeld, Bielefeld, West Germany) 10:10 C2. Invited Lecture: Photochemistry of Formaldehydes Past and Present - E. K. C. Lee , R. S. Lewis, and R. G. Miller (Department of Chemistry, University of California, Irvine, CA 92717) 10:40 C3. Decomposition of Formic Acid by Infrared Laser Radiation - R. Corkum, C. Willis , M. H. Back, and R. A. Back (Chemistry Division, National Research Council of Canada, Ottawa Canada) 11:00 C4. The Laser Augmented Decomposition of D B Adducts - K. R. Chien and S . H ; Bauer (Department of Chemistry, Cornell University, Ithaca, NY 14853) 11:20 COFFEE BREAK li A.M. 11:40 C5. Infrared Laser Isotope Separation - J . L . Lyman and S. D. Rockwood (Los Alamos Scientific Laboratory, P. 0. Box 1663, Los Alamos, NM 87545) P.M. 12:00 C6. CO TE Laser Induced Photochemical Enrichment of Carbon Isotopes - J. J. Ritter and S. M. Freund (National Bureau of Standards, Washington, DC 20234) 12:20 C7. Measurement of Mean Lives in Atomic Uranium - J. Z. Klose (Optical Physics Division, National Bureau of Standards, Washington, DC 20234) 12:40 C8. Principles of Photochemical Separation of Isomeric Nuclides - G. C. Baldwin , H. M. Clark, D. Hakala, and R. R. Reeves (Department of Chemistry, Rensselaer Polytechnic Institute, Troy, NY 12181) 1 : 00 LUNCH MONDAY, JUNE 28, 1976 RED AUDITORIUM SESSION D: Primary Photochemical Processes (Large Molecules and Energy Transfer) CHAIRMAN: 0. P. Strausz (Department of Chemistry, University of Alberta, Edmonton, Canada T6G 2E1) P.M. 1:40 Dl. Invited Lecture: The Early Years of Photochemistry in the Vacuum Ultraviolet - W. Groth (Institut fur Physikalische Chemie, der Universitat, Bonn, West Germany) 2:10 D2. Fluorescence and Decomposition of Tertiary Amines in a Glow Discharge - F. Lahmani and R. Srinivasan (International Business Machine Research Center, Yorktown Heights, NY 10598) 2:30 D3. Photochemistry and Flash Photolysis of 5 Nitroquinoline - A. C. Testa and A. Cu (Department of Chemistry, St. John's University, Jamaica, NY 11439) 2:50 D4. Photochemistry and Photophysics of Nitrogen Heterocycles - M. S. Henry and M. Z. Hoffman (Department of Chemistry, Boston University, Boston, MA 02215) in P.M. 3:10 D5. The Photophysics of Several Condensed Ring Heteroaromatic Compounds - F. S. Wettack , R. Klapthor, A. Shedd, M. Koeppe, K. Janda, P. Dwyer, and K. Stratton (Department of Chemistry, Hope College, Holland, MI 49423) 3:30 D6. Molecular Weight Dependence of Triplet - Triplet Processes in Poly (2-Vinylnaphthalene) - N. F. Pasch and S. E. Webber (Department of Chemistry, University of Texas, Austin, TX 78712) 3:50 COFFEE BREAK 4:10 D7. Singlet and Triplet Precursors of ( A ) - B. Stevens and J. A. Ors (Department of Chemistry , University of South Florida, Tampa, FL 33620) 4:30 D8. The Heavy Atom Effect on the Photochemical Aycloaddition Processes of Acenaphthylene - B. F. Plummer and L. J. Scott (Trinity University, Department of Chemistry, San Antonio, TX 78284) MONDAY, JUNE 28, 1976 GREEN AUDITORIUM SESSION E: Photochemistry with Lasers CHAIRMAN: R. P. Wayne (Physical Chemistry Laboratory, Oxford University, Oxford 0X1 3QZ England) P.M. "V 1 1:50 El. Laser Induced Fluorescence Emission Spectroscopy of H ? C0(A, A ) K. Y. Tang and E. K. C. Lee (Department of Chemistry, University of California, Irvine, CA 92717) 2:10 E2. Laser Excited NO Fluorescence Lifetime Studies in the 600 nm Region - V. M. Donnelly and F. Kaufman (Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260) 3 + 2:30 E3. Laser Fluorescence Studies, Including the B H(0 ) States of BrF, IF, and IC1 - M. A. A. Clyne , A. H. Curran, and I. S. McDermid (Department of Chemistry, Queen Mary College, London E14NS England) iv P.M. 2:50 E4. Laser Induced Fluorescence of CN Radicals Produced by Photodissociatlon of RCN - R. J. Cody and M. J. Sabety- Dzvonik (NASA/Goddard Space Flight Center, Greenbelt, MD 20771) 3:10 E5. Photofragment Spectroscopy of A-^-X Transition in CN Using Two Pulsed Lasers - A. Baronavski and J. R. McDonald (Chemistry Division, Naval Research Laboratory, Washington, DC 20375) 3:30 E6. Spectroscopic and Photochemical Studies of Gaseous Ions Using Tunable Dye Lasers - J. R. Eyler (Department of Chemistry, University of Florida, Gainesville, FL 32611) 3:50 COFFEE BREAK 4:10 E7. Photodissociation Spectra of Positive Ions with Time-of- Flight Analysis - T. F. Thomas and J. F. Paulson (Air Force Geophysics Laboratory, Hanscom AFB, Bedford, MA 01731) 4:30 E8. Vibrational States of Molecules In the Visible Region by Thermo- optical Spectroscopy and the Local Mode Model - A. C. Albrecht (Department of Chemistry, Cornell University, Ithaca, NY 14853) TUESDAY, JUNE 29, 1976 RED AUDITORIUM SESSION F: Environmental Photochemistry (Upper Atmosphere) CHAIRMAN: R. F. Hampson (Physical Chemistry Division, National Bureau of Standards, Washington, DC 20234) A.M. 9:00 Fl. Invited Lecture: The In-Situ Measurement of Atoms and Radicals in the Upper Atmosphere - J. G. Anderson (University of Michigan, Ann Arbor, MI 48109) 9:30 F2. The 0( S) Air glow - New Laboratory Results - T. Slanger and G. Black (Stanford Research Institute, Menlo Park, CA 94025) 9:50 F3. A Quantum Yield Determination for 0("T)) Production from Ozone via Laser Flash Photolysis - D. L. Philen , R. T. Watson, and D. D. Davis (Department of Chemistry, University of Maryland, College Park, MD 21201) A.M. 10:10 F4. Absence of N_0 Photolysis in the Troposphere - D. H. Stedman , R. J. Cicerone, W. L. Chameides, and R. B. Harvey (University of Michigan, Ann Arbor, MI 48109) 10:30 F5. Temperature Dependence of 0( u) Reactions of Atmospheric Importance - J. A. Davidson and H. I. Schiff (York University, Toronto, Canada) G. E. Streit, A. L. Schmeltekopf , and C. L. Howard (NOAA, Boulder, CO 80302) 10:50 COFFEE BREAK 11:10 F6. Invited Paper: Modeling Stratospheric Photochemistry and Kinetics - J. McAfee and P. J. Crutzen (NCAR, Boulder, CO 80303) 11:40 F7. The Rate Constant for + NO + M from 217-500K in Five Heat Bath Gases - J. V. Michael , W. A. Payne, and D. A. Whytock (NASA/Goddard Space Flight Center, Greenbelt , MD 20771) P.M. 12:00 F8. Determination by the Phase Shift Technique of the Temperature Dependence of the Reactions of 0( P) with HC1, HBr, and HI - D. L. Singleton and R. J. Cvetanovic (Division of Chemistry, National Research Council of Canada, Ottawa Canada K1A 0R9) 12 : 20 LUNCH TUESDAY, JUNE 29, 1976 GREEN AUDITORIUM SESSION G: Inorganic Photochemistry CHAIRMAN: M. Z. Hoffman (Department of Chemistry, Boston University, Boston, MA 02215) A.M. 9:00 Gl. Invited Lecture: High Energy Pulsed Laser Photolysis of Some Chromium (III) and Cobalt (III) Complexes - A. W. Adams on , A. R. Gutierrez, R. E. Wright, and R. T. Walters (Department of Chemistry, University of Southern California, Los Angeles, CA 90007) vi A.M. 9:30 G2. Studies Using a Combination of Flash Photolysis and Pulsed Magnetic Induction: Application to the NO Radical in Aqueous Acid Solution at 25°C - T. W. Martin (Vanderbilt University, Nashville, TN 37235) and M. V. Stevens (Methodist Hospital, Memphis, TN 38104) 9:50 G3. Luminescence of Some Metal Trisdithioacetylacetonate Complexes - M. K. De Armond and J. Merrill (Department of Chemistry, North Carolina State University, Raleigh, NC 27607) 10:10 G4. Luminescent Transition Metal Complex Photosensitizers - J. N. Demas , J. W. Addington, R. P. McBride, E. W. Harris, and S. Peterson (Chemistry Department, University of Virginia, Charlottesville, VA 22901) 10:30 G5. Picosecond Studies of Transition Metal Complexes - P. E. Hoggard (Department of Chemistry, Polytechnic Institute of New York, Brooklyn, NY 11201), A. D. Kirk (Department of Chemistry, University of Victory, B.C., Canada), G. B. Porter (Department 6£ Chemistry, University of British Columbia, Vancouver, Canada), M. G. Rockley (Department of Chemistry, Oklahoma State University, Stillwater, OK 74074) and M. W. Windsor (Department of Chemistry, Washington State University, Pullman, WA 99163) 10:50 COFFEE BREAK 11:10 G6. Invited Lecture: Electron Transfer Properties of Excited States of Transition Metal Complexes - V. Balzani , F. Bolletta, M. Maestri, A. Juris, and N. Serpone (Universita di Bologna, Bologna, Italy) 11:40 G7. A Comparison of the Excited-State Electron-Transfer Reactions of Ru(bipy) and 0S(bipy)„ - P. Fisher, E. Finkenberg, and H. D. Gafney (Department of Chemistry, City University of New York, Flushing, NY 11367) P.M. 12 : 20 LUNCH TUESDAY, JUNE 29, 1976 RED AUDITORIUM SESSION H: Primary Photochemical Processes (Small Molecules) CHAIRMAN: G. W. Mains (Department of Chemistry, Oklahoma State University, Stillwater, OK 74074) vix P.M. 1:20 HI. Invited Lecture: Photodissociation of Simple Polyatomic Molecules - J. P. Simons (Chemistry Department, The University, Birmingham, B15 2TT England) 1:50 H2. Energy Distribution in the Photodissociation of Methylketene at 215 nm - M. E. Umstead , R. G. Shortridge , and M. C. Lin (Chemistry Division, Naval Research Laboratory, Washington, DC 20375) 2:10 H3. The Production and Reactions of 1,1,2 ,2-Tetrachloroethane - M. H. J. Wijnen (Chemistry Department, Hunter College, New York, NY 10021) 2:30 H4. Measurement of Branching Ratios for the + CS -* OCS + S Reaction - R. E. Graham and D. Gutman (Department of Chemistry, Illinois Institute of Technology, Chicago, IL 60616) 2:50 H5. Photodissociation of Molecular Beams of Metallic Iodides - M. Kawasaki , H. Litvak, S.-J. Lee, and R. Bersohn (Department of Chemistry, Columbia University, New York NY 10027) 3:10 H6. The Photochemistry of 2-Furaldehyde in the Gas Phase - A. Gandini, P. A. Hackett , J. M. Parsons, and R. A. Back (Division of Chemistry, National Research Council of Canada, Ottawa Canada) 3:30 COFFEE BREAK 4:00 H7. Photoinitiated Decomposition of Monosilane - E . R . Aus t in and F. W. Lampe (Department of Chemistry, Pennsylvania State University, University Park, PA 16802) 4:20 H8. Excitation of HNO by ( A ) - T. Ishiwata, H. Akimoto, and I . Tanaka (Department of Chemistry, Tokyo Institute of Technology, Tokyo, Japan) 4:40 H9. Collision and Photoinduced Dissociation of NH and HO - J. Masanet, J. Fournier, and C. Vermeil (Equipe de Recherche, C.N.R.S., Paris, France 75231) TUESDAY, JUNE 29, 1976 GREEN AUDITORIUM SESSION I: Photochemical Conversion of Solar Energy CHAIRMAN: A. W. Adamson (Department of Chemistry, University of Southern California, Los Angeles, CA 90007) Vlll P.M. 1:20 II. Invited Lecture: Photogalvanic Cells - M. D. Archer and M. I. C. Ferreira (The Royal Institution, London W1X 4BS England) and W. J. Albery and W. R. Bowen (Oxford University, Oxford 0X1 3QZ England) 1:50 12. The Importance of Intermediate Partitioning in Energy Storing Photoreactions - G. Jones II , W. R. Bergmark, and M. Santhanam (Department of Chemistry, Boston University, Boston, MA 02215) 2:10 13. Optimization of the Iron - Thionine Photogalvanic Cell; Photochemical Aspects - P. D. Wildes , N. N. Lichtin, and M. Z. Hoffman (Department of Chemistry, Boston University, Boston, MA 02215) 2:30 14. A. Biomimetic Approach to Solar Energy Conversion - T. R. Janson and J. J. Katz (Chemistry Division, Argonne National Laboratory, Argonne, IL 60439) 2:50 15. Invited Lecture: Light Energy Conversion via Photoredox Processes in Micellar System - M. GrStzel (Hahn-Meitner- Institut fur Kernforschung, Berlin, Germany) 3:20 16. Photoelectrolysis of Water by Solar Energy - D. I. Tcehernev (Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, MA 02173) 3:40 COFFEE BREAK SESSION J: Photophysical Processes CHAIRMAN: C. S. Parmenter (Department of Chemistry, Indiana University, Bloomington, IN 47401) 4:00 Jl. Recent Work in Exciplex Photophysics - D. V. O'Connor and W. R. Ware (Department of Chemistry, University of Western Ontario, London, Canada) 4:20 J2. Charge Transfer and Hydrogen Atom Transfer Reactions of Excited Aromatic Hydrocarbon - Amine Systems - N. Mataga , T. Okada, and T. Mori (Department of Chemistry, Osaka University, Toyonaka, Osaka 560, Japan) 4:40 J3. Protolysis Equilibria of Lumichrome (6,7-Dimethyl-Alloxazine) In the Lowest Excited Singlet State - B. Holmstrbm , S. Bergstrom, and H. B. Larson (Chalmers Institute of Technology and the University of Gothenburg, Gothenburg, Sweden) 5:00 J4. Photochemistry of Pheophytins and Porphyrins - D. C. Brune, J. Fajer , and S. P. Van (Division of Molecular Sciences, Brookhaven National Laboratory, Upton, NY 11973) ix A.M. WEDNESDAY, JUNE 30, 1976 RED AUDITORIUM SESSION K: Environmental Photochemistry (Photochemical Smog) CHAIRMAN: H. Niki (Ford Motor Company, Dearborn, MI 48121) 9:00 Kl. Invited Lecture: A Kinetic Study of CH and (CH ) CO Radical Reactions by Kinetic Flash Spectroscopy - M. R. Whitbeck, J. W. Bottenheim, S. Z. Levine, and J. G. Calvert (Department of Chemistry, Ohio State University, Columbus, OH 43210) 9:30 K2. The Reaction of OH Radicals with C H, and C H, - R. Overend and G. Paraskevopoulos {Division of Chemistry, National Research Council of Canada, Ottawa, Canada K1A OR9) 9:50 K3. On the Triplet State (s) of Sulfur Dioxide - J. W. Bottenheim , F. Su, D. L. Thorsell, J. G. Calvert, and E. Damon (Department of Chemistry, Ohio State University, Columbus, OH 43210) 3 10:10 K4. The SO ( B ) Photosensitized Isomerization of Cis and Trans-I,2-Dichloroethylene - F . B . Wamp ler (Los Alamos Scientific Laboratory, Los Alamos, NM 87545) and J. W. Bottenheim (Department of Chemistry, Ohio State University, Columbus, OH 43210) 10:30 K5. Gas Phase Photochemical Synthesis of Peroxyacyl Nitrates via Chlorine - Aldehyde Reaction - B. W. Gay Jr. , R. C. Noonan, J. J. Bufalini and P. L. Hanst (EPA, Research Triangle Park, NC 27711) 10:50 COFFEE BREAK 11:10 K6. Invited Lecture: Chemical Control of Photochemical Smog - J. Heicklen (Department of Chemistry, Pennsylvania State University, University Park, PA 16802) 11:40 K7. Photochemically Induced Free Radical Reactions in Nitrogen Dioxide - Acetaldehyde Mixtures - E. R. Allen (Atmospheric Sciences Research Center, State University of New York, Albany, NY 12222) P.M. 12:00 K8. Photooxidation of Toluene-NO -0 -N System in Gas Phase - H. Akimoto, M. Hoshino, G. Inoue, M. Okuda, and N. Washida (National Institute for Environmental Studies, P.O. Yatabe, Ibaraki 300-21 Japan) P.M. 12:20 K9. Effects of Solar Energy Distribution on Photochemical Smog Formation - G. Z. Whitten , J. P. Killus, and H. Hogo (Systems Applications Inc., San Rafael, CA 94903) and M. C. Dodge (EPA, Research Triangel Park, NC 27711) 12:40 LUNCH WEDNESDAY, JUNE 30, 1976 GREEN AUDITORIUM SESSION L: Photochemical and Photophysical Processes (Condensed Phases) CHAIRMAN: B. Stevens (Department of Chemistry, University of South Florida, Tampa, FL 33620) A.M. 9:00 LI. Invited Lecture: Photophysics of Bound and Dissociative States of Small Molecules in Condensed Phases - L. E. Brus and V. E. Bondybey (Bell Laboratories, Murray Hill, NJ 07974) 9:30 L2. Spectroscopy and Photochemistry of Matrix Isolated Metal Hexaf luorides - W. F. Coleman and R. T. Paine (Department of Chemistry, University of New Mexico, Albuquerque, NM 87131) and R. B. Lewis, R. S. McDowell, L. H. Jones, and L. B. Asprey (Los Alamos Scientific Laboratory, Los Alamos, NM 87544) 9:50 L3. Mercury Photosensitized Production of Trapped Radicals in Organic Glasses at <_ 77K - N. Bremer, B. J. Brown, G. H. Morine, and J. E. Willard (Department of Chemistry, University of Wisconsin, Madison, WI 53706) 10:10 L4. Magnetic Field Effects on Triplet - Triplet Annihilation in Crystals - S. H. Tedder and S. E. Webber (Department of Chemistry, University of Texas, Austin, TX 78712) 10:30 L5. The Contribution of the Physical and Chemical Defects to the Photochemistry of Crystalline Durene at Very Low Temperatures - A. Despres, V. Lejeune, and E. Migirdicyan (C.N.R.S., Universite Paris - Sud, 91405 Orsay, France) 11:20 L6. Photochemistry of Electrons Trapped in Organic Glasses - G. C. Dismukes, S. L. Hager, D. P. Lin, G. H. Morine and J. E. Willard (Department of Chemistry, University of Wisconsin, Madison, WI 53706) XI A.M. 11:40 L7. Spin Labels and the Mechanism of the S -► T Nonradiative Process in Dyraldehyde; Possible Manifestation of Pseudo Jahn-Teller Forces on Nonradiative Processes - A. Campion and M. A. El-Sayed (Department of Chemistry, University of California, Los Angeles, CA 90024) P.M. 12:00 L8. Photodesorption from Metals: Measured Desorption Rates in Comparison with a MO Treatment - N. Trappen (Universitat des Saarlandes, Saarbrucken, West Germany) 12:20 L9. Matrix Isolation Spectroscopic Studies of Free Radicals and Molecular Ions Produced by Collisions of Molecules with Excited Argon Atoms - M. E. Jacox (Physical Chemistry Division, National Bureau of Standards, Washington, DC 20234) 12:40 LUNCH WEDNESDAY, JUNE 30, 1976 RED AUDITORIUM SESSION M: Environmental Photochemistry (Upper Atmosphere) CHAIRMAN: A. H. Laufer (Physical Chemistry Division, National Bureau of Standards, Washington, DC 20234) P.M. 1:40 Ml. Invited Lecture: Atmospheric Photochemistry of Chlorof luoro- carbons - M. J. Molina (Department of Chemistry, University of California, Irvine, CA 92717) 2:10 M2. Ultraviolet Photoabsorption Cross-Sections of CF CK and CFC1 as a Function of Temperature - A. M. Bass and A. E. Ledford Jr. (Physical Chemistry Division, National Bureau of Standards, Washington, DC 20234) 2:30 M3. Photodissociation of CCL - R. E. Rebbert and P. Ausloos (Physical Chemistry Division, National Bureau of Standards, Washington, DC 20234) 2:50 M4. The Photolysis of CIO - S. Jaffe (California State College, Los Angeles, CA 90032 J and W. B. DeMore (Jet Propulsion Laboratory, Pasadena, CA 91103) xn P.M. 3:10 M5. Stratospheric Reactions of Chlorofluoromethanes - R. J. Donovan and H. M. Gillespie (University of Edinburgh, Edinburgh Scotland) and J. Wolf rum and K. Kaufman (Max-Planck- Institut fur Stromungsforschung, 34 Gottingen, West Germany) 3:30 COFFEE BREAK SESSION N: Environmental Photochemistry (Photochemical Smog) CHAIRMAN: J. R. McNesby (Office of Air and Water Measurement, National Bureau of Standards, Washington, DC 20234) 3:50 Nl. Invited Lecture: Some Fundamental and Applied Aspects of the Atmospheric Reactivity of Selected Organic Molecules - J. N. Pitts , R. Atkinson, K. R. Darnall, A. C. Lloyd, R. Perry, and A. M. Winer (Statewide Air Pollution Research Center, University of California, Riverside, CA 92502) 4:20 N2. Invited Lecture: IR Fourier-Transform Spectroscopic Studies of Atmospheric Reactions - H. Niki , P. Maker, C. Savage, and L. Breitenbach (Ford Motor Company, Dearborn, MI 48121) 4:50 N3. Invited Lecture: Photochemically Generated Ozone from Isolated Strong Point Sources - D . D . Davis and W. Keifer (Chemistry Department, University of Maryland, College Park, MD 21201) 5:20 N4. Evidence for Alkaxy Radical Isomerization in C.-C, Alkanes in NO -Air Systems - W. P. L. Carter , K. R. Darnall, A. C.^Lloyd, A. M. Winer, and J. N. Pitts (Statewide Air Pollution Research Center, University of California, Riverside, CA 92502) WEDNESDAY, JUNE 30, 1976 GREEN AUDITORIUM SESSION 0: Photophysical Processes CHAIRMAN: L. E. Brus (Bell Laboratories, Murray Hill, NJ 07974) 1:40 01. Invited Lecture: Collisional Destruction of Rovibronic Levels in S Glyoxal: Electronic, Vibrational and Rotational State Changes - L. G. Anderson, A. E. W. Knight, and C. S. Parmenter (Department of Chemistry, Indiana University, Bloomington, IN 47401) Xlll P.M. 2:10 02. The Vibronic Dependence of Glyoxal Photodissociation - G. H. Atkinson and C. G. Venkatesh (Department of Chemistry, Syracuse University, Syracuse, NY 13210) 2:30 03. Behavior of Benzene in Low Vibrational Levels - S. A. Lee, J. M. White, and W. A. Noyes Jr. (Department of Chemistry, University of Texas, Austin, TX 78712) 2:50 04. Dual Lifetime Fluorescence from Pyrimidine - K. G. Spears and M. El-Manguch (Department of Chemistry, Northwestern University, Evanston, IL 60201) 3:10 05. Time Resolved Emission Spectra of Low Pressure Aromatic Molecules - M. D. Swords and D. Phillips (Department of Chemistry, The University, Southampton S09 5NH England) 3:30 COFFEE BREAK 3:50 06. Studies in the Mechanism of Radiationless Conversion of Electronic Energy - T. A. Gregory and S . Lip sky (Department of Chemistry, University of Minnesota, Minneapolis, MN 55455) A: 10 07. Photosensitization of the 2-Butenes by Benzaldehyde in the Gas Phase - A. J. Yardwood (Mc Master University, Hamilton, Ontario, Canada) and G. R. De Mare and M. Termonia (Universite Libre de Bruxelles, B-1050 Brussels, Belgeium) THURSDAY, JULY 1, 1976 RED AUDITORIUM SESSION P: Elementary Reaction Processes (Thermal Systems) CHAIRMAN: D. Volman (Department of Chemistry, University of California, Davis, CA 95616) A.M. 9:00 PI. Plenary Lecture - Near Infrared Detection of Peroxyl Radicals in Mercury Photosensitized Reactions - H. E. Hunziker (International Business Machines Research Laboratory, San Jose, CA 95193) 9:50 P2. Reactions of Hydrogen Atoms with Fluorinated Ketones - D. W. Grattan and K. 0. Kutschke (Division of Chemistry, National Research Council of Canada, Ottawa, Canada K1A 0R6) 10:10 P3. The Reaction of NH with Olefins Studied by Flash Photolysis ■ R. Lesclaux and Pham Van Khe (Universite 1 de Bordeaux, 33405 Talence, France) xiv A.M. 3 - 10:30 P4. Detection and Reactions of NH (X E ) Radicals in the Vacuum UV Flash Photolysis of NH Using Resonance Fluorescence - I. Hansen, K. Hoinghaus , C. Zetzsch, and F. Stuhl (Ruhu-Universitat , 4630 Bochum, West Germany) 10:50 COFFEE BREAK 11:10 P5. Invited Lecture: Interpretation of the Arrhenius Plots for Reactions of Oxygen Atoms with Olefins - R. J. Cvetanovic (Division of Chemistry, National Research Council of Canada, Ottawa Canada K1A 0R6) 3 11:40 P6. Absolute Rate of the Reaction of 0( P) with Hydrogen Sulfide - D. A. Whytock and R. B. Timmons (Chemistry Department, Catholic University of American, Washington, DC 20017) and J. H. Lee, J. V. Michael, W. A. Payne, and L. J. Stief (NASA/Goddard Space Flight Center, Greenbelt, MD 20771) P.M. 12:00 P7. Kinetics of the Reaction OH + NO (+M) -> HNO (+M) Over a Wide Range of Temperature and Pressure - C. Anastasi and I. W. M. Smith (University Chemical Laboratories, Cambridge CB2 1EP England) and R. Zellner (Institute fur Physikalische Chemie der Universitat Gottingen, 34 Gottingen, West Germany) 12:20 P8. Vacuum UV Flash Photolysis of Phosphine: Rate of the Reaction H + PH„ and Implications for the Photochemistry of the Atmosphere of Jupiter - J. H. Lee, J. V. Michael, W. A. Payne, D. A. Whytock, and L. J. Stief (NASA/Goddard Space Flight Center, Greenbelt, MD 20771) 12 : 40 LUNCH THURSDAY, JULY 1, 1976 GREEN AUDITORIUM SESSION Q: Energy Transfer CHAIRMAN: R. K. Hancock (ERDA, Washington, DC 20545) A.M. 2 9:10 Ql. Energy-Dependent Cross Section for Quenching of Li( P) and Na( P) - J. R. Barker, S-M Lin, and R. E. Weston (Chemistry Department, Brookhaven National Laboratory, Upton, NY 11973) xv A.M. 2 9:30 Q2. Electronic-to-Vibrational Energy Transfer Reactions: Na(3 P) + CO (X I , V = 0) - D. S. Y. Hsu and M. C. Lin (Chemistry Division, Naval Research Laboratory, Washington, DC 20375) 9:50 Q3. Energy Transfer in the Collision of Metastable Excited Ar( P ? ) Atoms with Ground State H( S) Atoms - P. B. Monkhouse , K. D. Bayes, and M. A. A. Clyne (Department of Chemistry, Queen May College, London El 4NS England) 10:10 Q4. Quenching Rate Constants for Ar( 3 P ), Kr( 3 P ) and Xe( 3 P ) by Halogen-containing Molecules, and Branching Ratios for XeF and KrF Formation - J. E. Velazco , J. H. Kolts, and D. W. Setser (Department of Chemistry, Kansas State University, Manhattan, KA 66506) 10:30 Q5. V-V Energy Transfer in H - Additive Gas Mixtures Using A Stimulated Raman Excitation Technique - R. G. Miller and J. K. Hancock (Chemistry Division, Naval Research Laboratory, Washington, DC 20375) 10:50 COFFEE BREAK THURSDAY, JULY 1, 1976 RED AUDITORIUM SESSION R: Elementary Reaction Processes (Excited State Dynamics and Energy Transfer) CHAIRMAN: M. C. Lin (Chemistry Division, Naval Research Laboratory, Washington, DC 20375) P.M. 1:40 Rl. Experiments Concerning the Laser Enhanced Reaction Between and NO - K-K Hui and T. A. Cool (School of Applied and Engineering Physics, Cornell University, Ithaca, NY 14853) 2:00 R2. Infrared Laser Enhanced Reactions: Chemistry of N0(u = 1) with - J. C. Stephenson and S. M. Freund (National Bureau of Standards, Washington, DC 20234) 2:20 R3. Kinetic Energy, - and Internal State Dependence of the NO + -► NO + Reaction - A.E. Redpath and M. Menzinger (Department of Chemistry, University of Toronto, Toronto Canada) xvi P.M. 2:40 R4. The Effect of Infrared Laser Excitation on Reaction Dynamics: + C H and + OCS - R. G. Manning , W. Braun, and M. J. Kurylo ^Physical Chemistry Division, National Bureau of Standards, Washington, DC 20234) 2 t 3:00 R5. Reaction of Flash Photolytically Produced CN(X E ,u) Radicals with 0( P) Atoms - K. J. Schmatjko and J. Wolf rum (Max-Planck-Institut fur Stromungsforschung, D-3400 Gottingen, West Germany) 3:20 COFFEE BREAK 3:40 R6. Vibrational Photochemistry: The Relaxation of HCl(u = 1) and DCl(u = 1) by Bromine Atoms - R. D. H. Brown, I. W. M. Smith , and S. W. J. Van der Merwe (University Chemical Laboratories, Cambridge CB2 1EP England) 4:00 R7. Vibrational Relaxation of HF(u = 1,2,3) In the Presence of H_, N_ , and CO - J. F. Bott (The Aerospace Corporation, Los Angeles, CA 90009) ? 4:20 R8. Quenching of N0(B II ) 1 = Produced by the Reaction of N( D) with NO - G. r BLack , R. L. Sharp less, and T. G. Slanger (Stanford Research Institute, Menlo Park, CA 94025) xvii A Modern Approach to Accurate Radiometry Edward Zalewski Optical Physics Division National Bureau of Standards Washington, D. C. 20234 The many recent advances in electro-optic technology have produced several new and very powerful tools for accurate photon flux measurements. The two advancements that have the greatest immediate impact on photochemistry are the electro-optic amplitude stabilization of high-powered cw lasers and the electrically cali- brated pyroelectric null radiometer. The first has provided a high intensity radiant power source that is monochromatic, highly colli- mated and very stable - long-term drifts of less than 0.05% have been obtained. Such amplitude stabilized lasers facilitate precise control of the number of photons entering a chemical reaction. The second advancement, the electrically calibrated pyroelectric null radiometer, has brought a new order of accuracy, sensitivity and convenience to the measurement of radiant power, thereby facilitating the accurate measurement of photon flux. The tremendous improvement in amplitude stabilization of cw lasers has been accomplished by using a stable solid state photo- detector to monitor the laser beam and, in turn, control the transmittance of an electro-optic modulator. A block diagram of a laser amplitude stabilization system is shown in Fig. 1. The box labeled "beam uniformity optics" refers to apertures, a spatial Al-1 FEEDBACK CONTROL ELECTRO-OPTIC MODULATOR SILICON MONITOR DETECTOR BEAMSPLITTER Figure 1. Block diagram of a cw laser stabilization system. The silicon monitor detector supplies a sense signal to a feedback loop. This controls the transmittance of the electro-optic modulator to reduce the fluctuations in the amplitude of the cw laser beam. GOLD BLACK SURFACE PYROELECTRIC - ->_Aj-lAj-l_ DETECTOR OUTPUT ^-J~L n ELECTRICAL INPUT Figure 2. Schematic diagram of the principle of operation of the electrically calibrated pyroelectric null radiometer. The pyro- electric material is alternately heated by radiant and electrical input pulses. The areas under the two types of pulses in the resulting wave train are equal when the heating from the two modes is equivalent. Al-2 filter and a beam expanding telescope used to obtain a uniform photon density in the cross section of the beam. Such a beam shaping system does result in a considerable loss of radiant power but is sometimes necessary to avoid saturation effects in certain experiments. The fundamental principle of the electrically calibrated pyro- electric null radiometer is illustrated in Fig. 2. The idea of electrically calibrating a thermal detector is not new. The new features are the pyroelectric detector material and the ac null-balance approach to the measurements. The pyroelectric, like a thermopile or a bolometer, is a thermal detector. However, it is an ac device rather than dc, responding to temperature changes instead of the steady-state value. This property permits the use of synchronous amplification, increasing the signal-to-noise of the measurements, and consequently improving the sensitivity over that of previous types of electrically calibrated detector radiometers. It also allows the electrical heating pulses to be applied to the detector alternately with the radiant heating pulses. Alternate pulses in the output signal are electronically inverted and amplified by a wide-band lock-in amplifier. The amplitude of the electrical heating pulses are adjusted to match the effect of the radiant heating pulses in order to obtain a zero value for the dc component of the amplifier output. That is, when the null condition is reached the areas under the electrical and radiant heating pulses are equal. The question of whether the electrical and radiant heating modes in the detector are really equivalent can be answered without recourse Al-3 to blackbody measurements. From a study of the physics of these detectors it is found that several types of inequivalences occur between the two different heating modes. From the physics one also obtains the methods whereby the magnitudes of the inequivalences can be evaluated in auxiliary experiments. With a knowledge of the magnitudes of these effects it is then possible to apply the appropriate correction factors to assure the accuracy of the measurement. With presently existing electrically calibrated pyroelectric null radiometers the best accuracies obtainable are between 0.5 and 1%. This is over a range of power from lOyW to 100 mW and a wavelength range from 200 to 10,000 nm. Al-4 Temperature Dependence of the Fluorescence Quantum Yields of Rhodamine Derivatives by Robert E. Schwerzel , Nancy E. Klosterman, and Carl M. Verber Battelle, Columbus Laboratories The temperature dependence of the fluorescence quantum yields of several rhodamine derivatives has been examined in detail. The results of this study, and their implications for the selection of visible fluorescence standards, will be discussed. A2-1 NBS Ultraviolet Radiometric Standards W. R. Ott National Bureau of Standards Washington, D.C. 2023 1 + Introduction In the wavelength region between 5-^00 nm, NBS offers a variety of standard sources and detectors which may be used to calibrate the radiant power of unknown sources or the response of radiation detectors and spectral radiometers. The energy range corresponding to these wavelengths spans from about 3 eV to 300 eV. This energetic radiation is of great interest not only in basic photochemistry research, e.g., in photo absorption and fluorescence studies, but also in industrial applications of photochemistry and photobiology. For example, extreme doses of ultraviolet radiation can cause skin cancer; controlled doses can heal. Continuous exposure to ultraviolet radiation results in the fading of paints and dyes; controlled exposures in specific processes are used to produce protective coatings on materials. Some documented examples of the various applications of ultraviolet radiation as both a natural and artificial element in our home and work environment are shown in Table 1. Radiometric Quantities In general, there are two ways to determine the radiant power of an unknown light source: (a) through the use of standard sources and (b) through the use of standard detectors. Standard sources are most useful when it is desired to know the emission characteristics of an unknown source. Standard detectors are most useful, on the other A3-1 Vhotocherois Cry Thin fllci product ion Tooth decay prevention Textile dyes Faint curing Industrial finishes Cast hardening Instant oscillographs Material degradiation UV photosensitive paper Counterfeit money detection Bacteriological Germicidal lamps in hospitals, schools, and offices Germicidal laaps in industry Environmenta l Studies Atmospheric sciences Ecology and die ozone l^yer Oil spill identification Water purification Water pollution Smog Cauga Medical and Therapeutic Jaundice treatments Calcium deficiency Skin disease treatments Wrinkling Drug detection Medical research Fusion Research National controlled thermo- nuclear fusion efforts Space Science Communications Astrophysics Sky lab Space shuttle Reentry and rocket exhaust VW and X-Ra y Laser s Integrated circuits Plasma probes Isotope separation Molecular synthesis High resolution holography Tumor therapy Laser fusion Photo biology NBS benchmark experiments Plasma Chemis try ■ Thin film deposition Plasma arc steel furnace Recycling of alloys in steel production Table 1. List of applications requiring ultraviolet radiation measurements. hand, when it is desired to know the radiant power at the location of a detector. For example, if the quantity of interest is spectral —2 —1 —1 radiance [watts cm nm sr ] , that is, if one is concerned with the power radiated by a specified emitting surface (cm ) in a certain wavelength band (nm) at a given solid angle (sr), then the light source to be investigated as well as the standard source may be set up in such a manner that the radiation from both passes through the same optical-spectrometric arrangement, thereby eliminating all geometric and other specific factors of the instrumentation. In short, the calibration is effected by a simple substitution of sources in the experimental arrangement. On the other hand, if the quantity of —2 —1 interest is spectral irradiance [watts cm nm ] , that is, if one is concerned with the radiant power incident on a specified surface area A3-2 2 (cm ) in a certain wavelength band (nm), then ideally a standard detector and filter arrangement is placed at the location of the irradiated surface to measure the power. Alternatively, if such a standard radiometer is not available, a standard source of spectral irradiance may be used in conjunction with a suitable diffusing element (to account for variations in the geometries of the standard and the unknown sources ) to determine the response function of the user's dif fuser-radiometer system. At NBS, standard sources of both ultraviolet spectral radiance and spectral irradiance are available. In the case of NBS standard detectors for the ultraviolet, the calibrated quantity is the absolute quantum efficiency [photoelectrons per incident photon] as a function of wavelength. Standard Sources The three primary source standards being used at NBS are the gold point blackbody cavity for which the spectral radiance is given by Planck's law, the wall-stabilized hydrogen arc for which the spectral radiance is given by accurately known quantum mechanical absorption coefficients for atomic hydrogen, and the electron storage ring facility for which the spectral radiance is given by the theory of electron synchrotron radiation. The selection of one of these standards for a calibration application is influenced mostly by the specified wavelength region. The primary standards are most often used to calibrate secondary or transfer standards of spectral radiance which can then be issued to customers. The secondary radiance standards are also used to generate A3-3 spectral irradiance standards. The following ultraviolet standard sources, listed in order of decreasing wavelength, are available from NBS: the tungsten filament quartz-halogen lamp (above 250 nm); the tungsten strip lamp (above 225 nm); the low pressure mercury vapor lamp (253.7 nm); the deuterium arc lamp ( 165-350 nm); the argon "mini- arc" (115-^00 nm); and the synchrotron radiation source, SURF- I I ( 5- U00 nm) . The relative strengths and limitations of these radiometric standards with respect to accuracy, reliability, convenience, and intensity and wavelength range will be discussed in the presented paper. Standard Detectors The primary standard for detector calibrations at NBS is a double-ionization chamber, a gas-filled detector in which each photon absorbed produces one electron-ion pair which is collected by a simple arrangement of parallel plates used to set up the collecting field. The transfer to the actual photodiodes , which are available to cus- tomers, is accomplished through the use of a uniformly grey (percent of radiation absorbed is independent of wavelength) thermopile, whose efficiency has been calibrated with the ionization chamber at short wavelengths and checked with a spectral irradiance standard source at 253. T nm. Windowed photodiodes calibrated in this manner have been used to evaluate the response of other detectors, such as photomulti- pliers, and the response of radiometers, such as "hazard meters", which are designed to have a spectral response equivalent to a specified erythemal curve. Calibrated photodiodes are available throughout the wavelength region 20-320 nm. A3-4 New Methods of Measuring Light Intensities D. G. Taylor, J. N. Demas , W. D. Bowman, E. W. Harris and R. P. McBride, Chemistry Department, University of Virginia, Charlottesville, Virginia 22901, USA. An inexpensive, easily constructed large area bolometer has been developed. The prototype has a l"xl" sensitive surface, but smaller sizes could be easily built to give greater sensitivity and faster response times. The l"xl" version has a maximum deviation of sensitivity of 37 Q from the center to the edge and is considerably more uniform over the central 3/4" diameter. Sensitivity is ~ 5 u-W ( D + 1 2 00H (2) 1 2 + (CH 3 ) 2 C=C(CH 3 ) 2 — > CH2=C(CH 3 )C(CH 3 )2 (3) where D is Ru(bipy^ 3 2 and 1 2 is singlet oxygen. The net effect of absorption of radiation by Ru(bipy) 3 2 is con- sumption of O2 . The actinometer system is closed and O2 consumption is monitored on a gas buret. The effective quantum yield for oxygen uptake, 0' , is given by = 0.855 j * -\ 0X K sv [0 2 ] [TME] (4) 1 + K SV [02]J([TME] + p where K gv is the Stern-Volmer quenching constant for O2 quenching of *Ru(bipy) 3 2 and equals 1400 M* 1 , p = 0.0027 M, and [O2 ] and [TME] are the average 2 and TME concentrations during the photolysis. The actinometer is intrinsically quantum flat and usable for A ^. ~ 520 nm which makes it ideal for use with the ionized Ar laser lines. An apparently previously unreported artifact with the 5 ferrioxalate actinometer has been discovered. With aged phenanthroline developing reagent and following the recom- mended procedure, intensity errors of -40 to +207o can A4-3 result if the phenanthroline is added before the buffer. If the buffer is added before the phenanthroline, the errors are smaller but can still be significant (~ 57 ). If fresh (^ 3 weeks) phenanthroline solutions are used, however, development is instantaneous (< 5 min) and independent of the order of addition of reagents. A preliminary calibration of the 0.15 F ferrioxalate at the 458 nm ionized Ar laser line indicates that the true yield may be ~ 10% lower (0.86^ than the accepted one (~ 0.96). A cooperative effort now in progress with the NBS (Gaithersburg) radiometry group will supply a definitive answer to this question by the time of the conference. 1) J. N. Demas, E. W. Harris, C. M. Flynn, Jr., and D. Diemente, J. Amer. Chem. Soc, 97_, 3838 (1975). 2) J. N. Demas, E. W. Harris, and R. P. McBride, in preparation . 3) J. N. Demas, E. W. Harris and R. P. McBride, "Lasers in Physical Chemistry and Biophysics", Elsevier, 1975, p. 477. 4) J. N. Demas, R. P. McBride and E. W. Harris, submitted. 5) W. D. Bowman and J. N. Demas, submitted. A4-4 6) C. G. Hatchard and C. A. Parker, Proc. Roy. Soc. (London), 255A , 518 (1956). Table I Absolute Spectral Sensitivity of Quantum Counters' A nm 590 570 550 530 510 490 470 450 430 410 390 Rhodamine B 5 g/1 Ethylene Glycol Rhodamine B 8 g/1 MeOH 1.001 .999 .996 1.000 .996 1.000 .990 .999 1.001 .998 .007 .015 .016 .006 .003 .002 .005 .012 .007 .004 .999 .973 1.001 1.005 1.010 999 002 995 000 999 997 005 008 001 996 982 982 996 001 999 000 991 Methylene Blue 1 g/1 MeOH .983 .992 .996 .997 .994 .992 .987 .996 1.002 1.005 1.016 1.023 1.024 1.029 1.042 1.053 1.053 .984 .921 1.006 .991 Nile Blue A 2 g/1 MeOH 1.010 1.012 1.019 1.028 1.021 1.014 1.003 1.001 .988 .970 .963 .957 .948 .970 1.002 1.008 1.016 1.015 1.019 1.011 1.002 A4-5 .987 .980 .980 .988 .989 .982 .982 1.001 .977 .980 .965 1.000 a 370 Relative to bolometer, measured via rhodamine B (5 g/1) MeOH calibrated counter. Table II a b Relative Sensitivity of Quantum Counters ' Rhodamine B (2 g/l) MeOH 1.71 Rhodamine B (5 g/l) MeOH 1.00 Rhodamine B (8 g/l) MeOH 0.49 Rhodamine B (5 g/l) Ethylene Glycol 0.92 Rhodamine B (8 g/l) Ethylene Glycol 0.61 Rhodamine 6G (5 g/l) MeOH 1.67 Methylene Blue (1 g/l) MeOH 0.064 Nile Blue A (2 g/l) MeOH 0.15 Mean sensitivity 360-590 nm compared to standard (Rhodamine B (5 g/l) MeOH) counter. All counters viewed from rear. a A4-6 Infrared Photography at 5 and lOym Gene F. Frazier, T. D. Wilkerson and J. M. Lindsay Versar Inc. , Springfield, Virginia 22151 A new technique for infrared photography is described involving the sensitization and desensitization of silver halide films by IR irradiation. Since a visible exposure follows the IR exposure, the time order is opposite to the well known Herschel effect. Also we record images at much longer wavelengths, 5 and lOym, even though the films are conventional commercial types such as Polaroid 55 P/N. This film is particularly useful for quantitative densitometry, some of which is reported here. For the image power 2 density range 0.1-1 W/cm , IR exposure times are about 10-1 second, respectively. CO and CO lasers (cw) have been the principal IR light sources, and the visible flash is variable (.001-1 sec) and is chosen to fog the film to a modest level of density. Exposure times of 1/10-1/100 second are possible at higher power density, opening up the possibility of inexpensive IR cinematography of the transient behavior of high power laser mode patterns. Exposure time sequences are indicated in Figure 1. The long IR exposure (a) yields desensitization of a negative and blackened print. The short IR exposure (b) can give the opposite or the same effects, depending for example on the color of the visible flash V. A delay of order 5 seconds or more (c) yields no IR effect on the latent image subsequently formed by V. A5-1 tOLU CO 10 m sec" 1 flow velocities required for Sn transport, demand [ Sn ] ~ 10 ml" . We have already successfully titrated [Sn] in the 10 -10 ml" range as part of the Sn quenching measurements (vide infra). The overall rate coefficient of the Sn/N 2 reaction is obtained from 286.3 nm absorption measurements of the rate of change of the relative [Sn] as a function of reaction time and [N 2 0] using standard techniques which are reviewed in detail in Refs. 2 and 3. Quenching rate coefficients are obtained from the observed variations in photon yield as a function of quencher concentrations via Stern-Volmer type analyses. Quenching by reagents is a problem inherent to excited species produced by chemi-excitation and required the development of methods suitable for the corresponding rate coefficient measurements, which will be briefly described. Results The reaction Sn + N 2 -SnO(all states) + N 2 (1) has been found to be strongly temperature dependent. A cursory Bl-3 determination yielded kj = (1 .0 +> 4) X 10" 11 exp [ (-3000 ± 1000/T ] ml molecule" sec" 1 over the 500-930 K range. In this range only SnO (a-X) and(A-X) band system emission could be definitely established. The total photon yield, $ , defined as the fraction of Sn oxidation events that leads to emission of a photon, increases by a factor of only s 1.25 between 500-930 K, i.e. from « . 54 to 0.69. This increase is essen- tially all due to an increase in the A-X system intensity ($ (A-X) in- creases from 0.04 to 0.16 over this range), while $ (a-x) remains constant at s;0.5. The rate coefficient for photon emission- -defined by dhv/dt = k hv [ Sn] [ N z O ]-- is a strong function of T , since k hy (T) = kj(T)*(T). The large $ (a-X) makes SnO (a 3 Z ) particularly attractive as a candidate for the upper state in a potential electronic transition chemical laser. Quenching experiments have concentrated on SnO(a 3 Z! ). The fol- lowing upper limit rate coefficients (based on the estimated 7 radiative lifetime of ss 1 X 10" 3 sec) were measured: k < 1 x 10" 12 ; i N 2° -13 i N 2 15 , Ar 16 • , ■, i -1 k Q 2 < 1 X 10 13 ; k Q 2 < i x 10 " 15 ; k Q < 2 x 10 m ml molecule 1 sec" units . Discuss ion The magnitudes of the observed rate coefficients differ strongly from those observed in several other simple metal atom oxidation re- action systems, e.g. Ba/N 2 0. 8 " 12 Probable reasons for these differ- ences will be discussed. It will be shown that these differences make Bl-4 Sn/N 2 a very attractive electronic transition chemical laser candidate system. Acknowledgment Work sponsored by the Defense Advanced Research Projects Agency (ARPA) and monitored by the Air Force Weapons Laboratory, Air Force Systems Command, United States Air Force, Kirtland Air Force Base, New Mexico 87117. References 1. Fontijn, A., Kurzius, S.C., Houghton, J.J., and Emerson, J. A. Rev. Sci. Instr. 43, 726 (1971). 2. Fontijn A. , Felder W. , and Houghton, J.J. , Fifteenth Sympo- sium (international) on Combustion, (The Combustion Institute, Pittsburgh, 1975) p. 775. 3. Fontijn, A., AIAA Paper 76-131, January 1976. 4. Felder, W. and Fontijn, A., Chem. Phys . Lett. _34> 398 (1975). 5. Fontijn, A., Meyer, C.B., and Schiff, H.I., J. Chem. Phys. 40, 64 (1964). 6. DeZafra, R.L. and Marshall, A., Phys. Rev. A170 , 28 (1968). 7. Meyer, B., Smith, J.J., andSpitzer, K. , J. Chem. Phys. J53, 3616 (1970). 8. Jonah, C.D., Zare , R.N., and Ottinger, Ch. , J. Chem. Phys. 56_, 263 (1972). 9. Jones, C.R. and Broida, H.P. , J. Chem. Phys. 6_0, 4369 (1974). 10. Palmer, H.B., Krugh, W.D. and Hsu, C.J., Fifteenth Sympo- sium (international) on Combustion , (The Combustion Institute, Pittsburgh, 1975) p. 951. Bl-5 11. Felder, W., Gould, R.K., and Fontijn, A., Second Summer Colloquium on Electronic Transition Lasers, September 1975, MIT Press (to be published). 12. Edelstein, S.A. , Eckstrorrii D. J . , Perry, B.E., and Benson, S. W. , ibid. OXIDANT O WINDOW A-ALUMINA TUBE B- SLIDING 0-RING SEAL C-TO POWER SUPPLY D- WINDOW SWEEPER GAS E- OSOANT INLET Md THERMO- COUPLE WELL F-ZIRCAR INSULATION G- SOURCE HOLDER H-Pt-40%flh RESISTANCE WIRE J- ALUMNA HEAT SHIELD K- BRASS VACUUM HOUSING L- COOLING COIL SOURCE GAS O- D ETECTOR C REACTION TUBE DETECTOR A Pt/Rh RESISTANCE WIRE DETECTOR B ZIRCAR THERMAL INSULATION HOLLOW CATHODE LAMP" CHOPPER LENS ALUMINA SIGHT TUBE c Figure 1. HTFFR for Kinetic Studies in the Intermediate (ca 400 - 1400 K) Temperature Range . Figure 2. Optical Arrangement and Cross-Section of HTFFR. Detectors A and B are spectro- meters used for metal atom ab- sorption and chemiluminescence spectral measurements , respec- tively. Detector C is a PMT fil- ter combination for chemilumin- escence intensity and titration measurements. Reaction tube 2 . 5 cm i.d., reaction tube observa>- tion ports 1 cm diam. Bl-6 A Study Of The Chemiluminescence Of The Pb + 0„ Reactions M. J. Kurylo, W. Braun, S. Abramowitz , M. Krauss National Bureau of Standards Institute for Materials Research Washington, DC 20234 Introduction Recent studies on laser enhanced reactions have shown that vibrational energy in a reactant triatomic molecule can appear as excitation in a product triatomic molecular species. Because of the spectroscopic complexities of triatomic molecules we are investigating a chemiluminescent reaction producing a diatomic product. The metal atom-oxidant systems represent a class of such reactions which have been investigated because of their potential as chemical lasers. We report herein some observations on the reaction Pb + 3 + PbO + r (1) While the information obtained from studying the infrared laser enhanced reaction component is minimized by the overall reaction complexity, it does nevertheless provide some additional insight into this reaction system. This information coupled with new high pressure spectroscopic results complement the detailed low pressure investiga- tion by Oldenborg, Dickson, and Zare (ODZ) . Experimental The furnace, reaction cell, fast flow pumping system, CO- laser, and spectrometer are shown schematically in figure 1. Lead vapor, produced from a resistively heated crucible containing lead metal enters the glass reaction chamber in an Ar diluent stream. The 4% B2-1 0_ in 0„ effluent from a commercial 0. generator is mixed with Ar and flows past the cell windows (to eliminate window deposits) into the cell. There it diffusively mixes with the lead-argon flow. The temperature of the crucible (as measured by a thermocouple probe) ranges from 900-1000 K while the temperature in the flame reaction zone varies from 500-600 K. The total pressure in the cell varies from 1 to 5 torr with the flow through the furnace being anywhere from 20 to 50% of the total flow. The 0.5 cm diameter beam from a CO laser tuned to the 9.6 ym P(30) transition is square wave chopped and traverses the flame exciting v_, the asymmetric stretching mode of (1043 cm ) . The chemiluminescence from the Pb + 0_ reaction is monitored through a spectrometer-photo- multiplier assembly. The photomultiplier output is fed through a series of pulse amplifiers and voltage discriminator into a dual counter, one channel of which records the "laser-on" signal and the other the "laser-off". A printout from these counters is synchronized with the spectrometer wavelength-scan-drive thereby facilitating the recording of the modulation spectrum vs wavelength. The normal spectra ("laser- off") emission is automatically obtained from the "laser-off" counter. Results In contrast to the spatially sharp diffusion flame observed for the Ba + (or NO) reaction, the Pb - 0„ flame is quite diffuse (apparently reaction limited) . It is brightest in the high temperature zone at the furnace nozzle, diminishes with decreasing temperature (increasing distance from the nozzle), and persists for some time two or three feet into the pumping system. These observations suggest a B2-2 reaction rate for Pb + which is considerably slower than gas kinetic. The spectrum which we observe at several torr total pressure partly resembles the low pressure spectrum obtained by ODZ and reproduced by us in a similar quasi-beam apparatus. The short wave- length end of the more complete spectrum (fig. 2) can be identified from the low (submicron) pressure spectrum in that it is less diffuse. The long wavelength portion between 480 nm and 595 nm is almost entirely the new a state recently characterized by ODZ. At still longer wavelengths, our spectrum differs from that of ODZ in that we observe a series of strong lines which do not agree with a •> X, A ■*■ X, or B + X, but rather appear to originate from a new state observed via six weak lines by ODZ and identified as b by them. This b state is seen more intensely in our high pressure spectrum. This series of lines in our spectrum can be fit to the expression: v, , ,„ = 16315 + 441. 0v' - [717.7v" - 3.53v" ] ( v , v ) where the lower state constants have been taken from Rosen's compendium. Part of the modulation spectrum is reproduced in figure 3. This spectrum, which has been smoothed, represents only that component of the chemilumine scent emission which varies with the laser excitation of 0- (i.e. only that component which either increases or decreases due to a change in reaction rate with reactant vibrational excitation) . Thus, if the emission originated only from a single electronic state, the modulation spectrum would appear as a quasi -continuum with possibly some small structure due to sharp changes in population of one vibrational line relative to an adjacent one. The large amount of structure in figure 3 indicates the presence of emission from at least B2-3 two electronic states. The peaks (or valleys) in the spectrum do not agree well with either the a -> X or b -*■ X states, but rather with the A •+■ X transitions. The results will be discussed and possible mechanisms proposed taking into account both our results and other investigations of this system. B2-4 PHOTOMUITIPIIER MONOCHROMATOR LENS IASER CHOPPER FIGURE 1. Schematic of the Chemiluminescent Flow Reaction Apparatus FIGURE 2 . Chemlluminescence Spectrum for the Pb(v) + 0,, Reaction FIGURE 3 . Smoothed Modulation Spectrum (the component of the emission which varies with laser excitation of 0,) of the Pb(v) + 0- Reaction B2-5 Chemiluminescence In Reactions of Ozone S. Toby , F. S. Toby and B. Kaduk School of Chemistry, Rutgers University, New Brunswick, N.J. 08903 Chemiluminescence (CL) can provide highly sensitive analytical meth- ods for the estimation of reactants, it serves as a unique probe for the identification of intermediates, and it yields data which provide information on energy transfer and partitioning. Gas phase ozone sys- tems have been particularly rich sources of CL, and much work has been done on O3 + alkenes as an adjunct to kinetic studies. Pitts and co- workers (1) using ozonized oxygen obtained CL from a variety of alkenes and tentatively identified the emissions as from excited formaldehyde and a-diketones. A more extensive investigation (2) confirmed emissions from CH 2 ( 1 A 2 ) and (CH0) 2 ( 3 A U ) and reported that in the absence of 2 Meinel bands (0H(X 2 tt)) were also seen. The presence of 2 had no effect on the CH 2 0( 1 A 2 ) emission in the alkenes studied (2) nor in the case of O3 + propadiene (3). However 2 quenches the methylglyoxal phosphor- escence from O3 + tetramethylethylene and the quenching has been shown to be 10 times as rapid by 2 as by N 2 (4) . An intriguing parallel is the fact that added 2 reduces the reaction rate of O3 + alkenes by a- bout an order of magnitude under conditions normally employed (5) . A similar effect has been noted in the O3 + C 2 F 1+ system: added 2 quenches the CL (6) and also slows the reaction (7) . Reported here are six new CL spectra, measured with apparatus pre- viously described (3,6). The spectra were measured at room tempera- ture using a Jarrell-Ash 0.25-m monochromator with a spectral slit width of 3 nm and a cooled EMI 9683QKB photo multiplier. The 3 flowed B3-1 in a helium carrier gas at a total pressure of approximately 50 torr. The total intensities uncorrected for spectral response and relative to C 2 H 4 = 1 are as follows: trans- C ? F ? H ? 30, C 2 F 3 C1 350, C 2 F 3 H 45, thiophene 50, CS 2 540, HCCCN 22. Trans-C 2 F 2 H 2 , C 2 F 3 C1 and C 2 F 3 H The CL spectra from the reactions of these fluoroalkenes with O3 are shown below 3 +CzFz$t% ^A^A^k 1 1 1 \ t 1 < 1 < . 1 . 1 » 1 i 1 1 , , 1 ! 20Q 1$0 %0 oo 10O The emission appears to be entirely due to excited SO2. The kinetics of this reaction have been studied by Olszyna and Heicklen (9) who did not report CL but postulated the formation of excited SO2 from the re- action CS + 3 -»■ CO + S0 2 *. B3-3 CYANOACETYLENE This unusual spectrum from 3 + HC3N extends from 340 to 780 nm and we have not been able to identify the emitter(s). 3 -hHC 3 l Wm The relationship of CL with the kinetics and energetics of some of these reactions is discussed. References (1) J. N. Pitts, Jr., W. A. Kummer, R. P. Steer and B. J. Finlayson, Adv. in Chem. (ACS Symposium) No. 113, 246 (1972). (2) B. J. Finlayson, J. N. Pitts, Jr., and R. Atkinson, J. Am. Chem. Soc, 96, 5356 (1974). (3) S. Toby, J. Luminescence, 8, 94 (1973). (4) U. Schurath, H. Gus ten and R. D. Penzhorn, J. Photochem., 5_, 33 (1976). (5) F. S. Toby, S. Toby and H. E. O'Neal, Int. J. Chem. Kins., 8, 25 (1976). (6) R. S. Sheinson, F. S. Toby and S. Toby, J. Am. Chem. Soc, 9_7, 6593 (1975). (7) F. S. Toby and S. Toby, submitted for publication. (8) K. H. Becker, M. Inocencio and U. Schurath, Int. J. Chem. Kins., Symp. 1, 205 (19 75). (9) K. J. Olszyna and J. Heicklen, J. Phys . Chem., _74, 4188 (1970). B3-4 DIOXETANE CHEMISTRY IN THE GAS PHASE; UV - VISIBLE CHEMILUMINESCENCE FROM THE REACTIONS OF 0„ ( A ) WITH OLEFINS 2 g Denis J. Bogan# Chemical Dynamics Branch, Code 6180 Naval Research Laboratory Washington, D. C. 20375 INTRODUCTION AND BACKGROUND Dioxetanes (4 membered cyclic peroxides) are the most notable of a small number of strained cyclic molecules which are known to undergo thermal unimolecular reactions yielding electronically excited pro- ducts. O O O O I I || Cnn*) + || R C C R" > C C I I (E a ~25) /\ /\ R' R" ' R R' R' " R" The quantum yields for production of one excited carbonyl product range from 0.01 to perhaps 0.5 with a preferential production of triplet (nil*) states. Such quantum yields are orders of magnitude greater than the statistically expected yields based upon equi- partitioning of energy to all accessible product states. Recent pervasive interests in chemiluminescence, chemical lasers and elec- tronically non adiabatic reactions have combined to make dioxetane chemistry the subject of considerable attention by both experiment- alists ' and theoreticians. ' Previous to our work, studies have involved synthesis of the dioxetane of interest followed by thermal decomposition. The latter has almost always been done in solution where it is complicated by solvent interactions and hence inability to obtain high resolution spectra of the primary excited products. More common than direct chemiluminescence are studies involving energy transfer to dyes which flouresce efficiently (e.g., 9, 10 dibromo anthracene), and product analysis. Much important information has been gained, for example the singlet-triplet titration experiments of Turro and coworkers; however, the condensed phase techniques, by their very nature, pre- clude knowledge of the disposition of energy in the primary excited products. Direct 1, 2 cyclo addition of O ( A) to olefins in solution has been used to prepare dioxetanes from mono olefins lacking allylic B4-1 A 1 hydrogen. ' This technique is known as low temperature photosensi- tized oxygenation and involves production of ( A) via energy transfer from an organic dye, D. D + hv > 1 D* > 3 D * 3 D* + 2 ( 3 £) >D + 2 ( 1 Z) or 2 ( 1 A) 2 ( 1 S) + Quencher — >0 2 ( 1 A) Only the dioxetanes of electron rich olefins have been made by this technique. We have found that the reactions of olefins with ( A) in a low pressure discharge flow system result in the observation of elec- tronically excited products which are expected from the corresponding dioxetanes. The experimental apparatus is shown as Figure 1. A microwave discharge through pure or 3% in He doped with a trace of Hg was used to produce ( A) free of atomic oxygen. The olefin was added downstream (following 2 light traps) and chemiluminescence was observed axially in a cell of 30 cm path length with a 1/4 meter grating spectrometer and cooled S-5 photomultiplier in a single photon counting configuration. There is no evidence of complications from atomic oxygen or ozone. Typical pressure and residence time were 4 torr and 0.1 sec. respectively. The results of our experiments with ethyl, methyl and n butyl vinyl ethers, ' and ethylene have been reported. These and other experiments now underway have centered on two major areas. First, and discussed in Section II, is the dynamics of the reaction, particularly orbital symmetry considerations govern- ing the correlation to excited products and the vibronic motions which serve to promote the reaction. The second main thrust, discussed in Section III, is the production and study of emission spectra which have not been observed previously. II. DYNAMICS, CAPSULE SUMMARY The initial motivation for doing gas phase studies was to obtain high resolution spectra under conditions which would minimize energy transfer and other quenching processes. We soon learned, however, that there is an important fundamental difference between gas phase addition of O ( A) to olefins and liquid phase addition effected by photosensitization. Thermochemistry and chemiluminescence activation energies (Table I) showed that the initially formed adduct of olefin Plus (0 ( A) contains 45 kcal, or more, of excess vibrational energy. ' Experiments were done measuring the rate of collisional stabilization versus the gas kinetic collision frequency; no stabilization was found and we concluded that the mean lifetime of the vibrationally excited dioxetane was less than 10 sec. ' Thus the B4-2 overall reaction is a chemical activation process (la) , followed by rapid fragmentation (lb) . 0=0 ( X A) o — o o R^ X R" (\^ 10-20) /" C \p,„ ^ R C C R" ... ; > C C R' N R (la) R' R" ' R R' R'" R" -12 The chemically activated dioxetane must have a lifetime >10 sec. since dioxetanes have been isolated and characterized from low temperature solution photosensitization. Many reactions give formaldehyde (A A -*■ X A ) emission. A hot band representing excitation in V' (the out of plane bend) occurs with intensity which correlates positively with available energy and nega- tively with the number of vibrational modes of the expected dioxetane. The correlation crudely approximates an RRK relationship. The princi- pal geometry difference between H CO in dioxetane and H CO* ( A ) is in V , which suggests that chemical Franck-Condon factors are operative and that the hot band intensity is inversely related to the lifetime of the vibrationally excited dioxetane. In contrast to the lifetimes, electronic energy partitioning of chemically activated dioxetanes is non statistical. Methyl, ethyl and n-butyl vinyl ethers give only H CO* emission and exhibit egual quantum yields for H CO* emission at all temperatures of study. Many other unsymmetrical olefins give only one of the two possible product emissions. The only apparent exception is C F H which is discussed in the following section. The energy partitioning and hot band information (which could not have been obtained from condensed phase studies) are being used to formulate an orbital symmetry - vibronic interaction theory of the dynamics of these reactions. III. PREVIOUSLY UNOBSERVED EMISSIONS The study of A - X (nIT*-ground) electronic transitions of carbonyl compounds has been a subject of major and continuing interest in both photochemistry and molecular spectroscopy. Neither F CO nor HFCO (A-X) emissions have been characterized. The A (nil*) state of HFCO predissociates very readily and HFCO is the basis of a photodissocic tion chemical laser. The A (nil*) state of F CO is of interest to atmospheric modellers because F CO is a likely secondary product of f luorochlorocarbon photodissociation in the stratosphere and it can, in turn, photodissociate to give F atoms which can then destroy ozone. B4-3 .a- Figure 2 shows spectra which were obtained from the reactions, (2) 2 ( 1 A) + C 2 F 4 > F 2 CO* + F 2 CO (3) 2 ( 1 A) + C 2 F 3 H >HFCO* + F 2 CO >F 2 CO* + HFCO We are working on detailed assignments of these spectra; at present, some assignments have been postulated for F CO. The.barrier to inversion in the A state of F CO has been es- timated as 4000 cm , hence we expect the behavior of an undoubled well in V' since the lower vibrational levels of 2 deep wells will not communicate effectively. On thisbasis the reported origin in the absorption spectrum at 42084 cm is the energy of the transition 4 , and the 0-0 band origin is estimated at 41310 (this will be exact only if V' = v" = 774) . We have tentatively assigned a few transitions and these indicate activity in V" (carbonyl stretch) , v" (symmetric stretch) , V" (out of plane bend) and v" (in plane rock) . The strong 4 O groups of bands appearing in the region 270-320 nm each consist of 4 bands forming series designated A, B.» C, D. The successive members of each series are separated by 660 cm (reading error ca ± 25) . The separations of the series are related as follows; B series - 150 cm to the red of A C series - 325 cm to the red of A D series - 522 cm to the red of A Reasonably convincing indirect evidence that the emission is indeed F CO (A-*X) is; a) our emission begins at 235 nm as compared to the known origin of absorption at 237.6, b) the observed emission intensity is kinetically first order in both 0_ ( A) and C F . The spectrum becomes diffuse and steadily diminishes in intensity from 350 to 520 nm. , much like the long A bands of H„CO (A-X) . The spectrum from the reaction of C F H shows clear evidence that the postulated F CO* emission is present as well as a new very compli- cated emission. The emission is first order in C F H, however argument b (above) cannot be applied without ambiguity. A spectrum was obtained using C F and C F H both at the concentrations used for their respective pure spectra and this is a simple sum of the pure component spectra. This establishes that the postulated HFCO* emission from reaction 3 is not excited by energy transfer from F CO*. If detailed assignments support the postulated assignments, this chemiluminescence technique will be established as a means of characterizing previously unobserved carbonyl (A-X) emissions. The region of the excited state phase space which can be reached by light excitation of a polyatomic molecule is severely restricted by the B4-4 Franck-Condon principle. The volume of phase space accessible by chemical reaction is much less restricted and it is not surprising that regions can be reached from which radiative transitions can be observed. This is particularly significant for excited states which predissociate near the electronic origin. REFERENCES # National Research Council, NRC-NRL Associate, 1974-1976. 1. N. J. Turro et al, Accounts Chem. Res., ]_, 97 (1974). 2. C. Mumford, Chemistry in Britain, .LI, 402 (1975). 3. M. J. S. Dewar and W. Thiel, J. Am. Chem. Soc . , 97_, 3978 (1975). 4. a) Y-N. Chiu, J. Chem. Phys., in press, 64_, number 7, (1976). b) Y-N. Chiu, D. J. Bogan and R. S. Sheinson, to be submitted. 5. N. J. Turro and P. Lechtken, J. Am. Chem. Soc. 94_, 2886 (1972) . 6. W. Adam, Chemiker Zeitung, 99_, 142 (1975). 7. P. D. Bartlett and A. P. Schaap, J. Am. Chem. Soc, 9_2, 3223 (1970) . 8. D. J. Bogan, R. S. Sheinson, R. G. Gann and F. W. Williams, J. Am. Chem. Soc, 97, 2560 (1975). 9. D. J. Bogan, R. S. Sheinson and F. W. Williams, paper F-9, 8th International Conference on Photochem. , Edmonton, Canada, Aug. 1975; abstract including spectra to be published in J. Photochem. 10. D. J. Bogan, R. S. Sheinson and F. W. Williams, J. Am. Chem. Soc 98, 1034 (1976) . 11. M. J. Berry, Chem. Phys. Lett., 29_, 329 (1974). 12. D. E. Klimek and M. J. Berry, Chem. Phys. Lett., 20_, 141 (1973). 13. a) F. S. Rowland and M. J. Molina, Revs. Geophys and Space Phys., 13_, 1 (1975). b) C. C. Chou, H. V. Ruiz, G. Crescentini, F. S. Rowland, paper PHSC 5, First Chemical Congress of North America (and 171st National ACS), Mexico City, Mexico, December 1975. 14. D. A. Condirston and D. C. Moule, Theor. Chim. Acta, 29_, 133 (1973) . 15. G. L. Workman and A. B. F. Duncan, J. Chem. Phys., 52_, 3204 (1970) . 16. a) S. W. Benson, "Thermochemical Kinetics", Wiley, New York, 1968. b) S. W. Benson and H. E. O'Neal, "Kinetic Data on Gas Phase Uni- molecular Reactions", NSRDS-NBS 21, U. S. Dept. of Commerce, 1970. B4-5 TABLE I. Data Summary for Reactions of C> (*A ) with Olefins, 2 g Olefin H C=CHOMe H C=CHOEt H C=CHOnBu E (kcal/mole) Emission -AH — (kcal/mole) 3. i\ 12±1 H„CO 2 107 H C=CH 2 2 EtOHC=CHOEt H COO 2 H C=CHF H 2 C=CF 2 HFC=CF„ F C=CF„ 2 2 H C=CHC1 2 H C=CC1 2 2 CI C=CC1 2 2 21±1 15±1 H CO 2 90.! b none— H 2 CO 129 H 2 CO 136 H 2 CO 122 HFCO, F CO 175- F 2 CO 171 H 2 CO 129 H CO 2 103 e none— 124 H C=CHCH0 2 none MeHC=CMe. Me C=CMe none— 104 c none- 109 a.) AH is calculated for ground state carbonyl products. Thermocnemical data are from reference 16, reference 8 for vinyl ethers. b.) HCOOEt* is believed to rearrange quantitatively to C H plus HCOOH via a Norrish Type II process, see ref . 16b. c.) The formation of allylic hydroperoxides has E a ~3-6 kcal and is favored over dioxetane formation. d.) Our estimate. e.) We infer that CL CO* is lost completely by predissociat- ion and/or other non radiative processes. B4-6 NLET V Y y R MICROWAVE C ) CAVITY ^ \y PUMP FIGURE 1. Discharge Flow System Figure Captions Figure 1; R = heated flow reactor with removeable quartz windows, M = 1/4 meter monochromator with detector described in text. Figure 2; Chemiluminescence spectra from reactions of ( A ) with flourinated olefins. Bottom panel is C F H only, middle panel is C F only, top panel is C F H plus C F . Conditions were; P (total) = 4.1 torr (approx. 0.2 of olefin and 3.9 of oxygen), T = 800 deg K, reactor residence time = 0.1 sec, spectral slit = 0.4 nm. , scan rate = 12.5 nm./min., time const- ant = 1 sec, full scale intensity = 10K counts/sec Figure 2 appears on the following page. B4-7 250 I I I I I 300 350 nm 400 Figure 2 Chemiluminescence spectra from the reactions of 0o( A ) with fluorinated olefins. £■ a B4-8 Molecular Dissociation by High-Intensity Infrared Laser Radiation K. H. Welge Fakultaet fuer Physik, Universitaet Bielefeld, Bielefeld, West Germany The dissociation of molecules by high-intensity laser radiation in the infrared has recently gained substantial interest, particularly since it has been found that the process is highly state and isotope selective. However these processes have not yet been investigated in any detail and are not yet well understood. This paper will be concerned with a review of this subject with emphasis on our own studies in this area. We have been concerned with three kinds of experiments using line selected pulsed CO -laser radiation at energies °o 100 Joule/pulse and intensities % 10 G 2 Watt /cm : 1) Dissociation and photolysis of molecules at pressures down to collisionless conditions using laser fluorescence spec- troscopy for in-situ diagnostics of fragments. Experiments have been concerned with NH and its isotopic compounds; 2) Crossed molecule / CO -laser beam experiments on the kine- tics and dynamics of the high-intensity field dissociation process by photodissociation fragment spectroscopy; 3) Experiments on the isotope selectivity by measurement of the isotope separation factor as function of parameters such as number of pulses per sample, gas pressure, laser energy etc.. Also, the quantum yield of separation has been investigated. Cl-1 Photochemistry of Formaldehydes: Past and Present Edward K.C. Lee, Roger S. Lewis and Richard G. Miller Department of Chemistry University of California, Irvine, California 92717 In the past, photode composition of formaldehyde has been studied extensively because of general interest in understanding the primary photochemical prbcesses for the simple carbonyl compounds. Recent interest in its study certainly stems from the need to evaluate the role that formaldehyde plays in the photochemical air pollution, particularly due to the formation of H- atoms and consequent reactions 1 2 in the photooxidation processes. ' More recent and current interest is generated from the suggested practicality of photochemical laser 3 4 isotope separation of deuterium and other isotopes. ' Since quantum yields of the fluorescence, the radical products (H and HCO) and the molecular elimination products were not known at various single vibronic levels (SVL) earlier, we have undertaken a systematic study in which the mechanism and rates of these primary processes in H 2 CO, HDCO, and D 2 CO can be established by various photo- physical and photochemical techniques. We have recently reported the observation of a large variation of the radiative lifetimes (t ) and the non-radiative lifetimes ( r NR ) of selected SVL's of the first excited state of H 2 CO (S b A %). 5 New results on the fluorescence quantum yields (3> ) of H 2 CO, HDCO, and E^CO vs. excitation wavelength (A ) have been obtained by fluores- cence excitation spectroscopy of these isotopic formaldehydes at low pressures, and they are shown in Figure 1. The values of for the first excited vibrational level (4 1 ) having one quantum excitation of the out-of-plane wagging mode (v 4 ' = 1) are -0.035 for H 2 CO, * This work has been supported by the Office of Naval Research and the National Science Foundation. C2-1 ~0.05 for HDCO, and 1.0 for D 2 CO, indicating the absence of photo- chemical activity at the 4 1 level of D 2 CO; the fluorescence decay times (tO at low limiting pressures are 0.082, 0.141 and 4.5 ^sec, respectively. In general, the rates of non-radiative transitions increase exponentially with the increasing excitation energy. The fact that the values of <1> and t f are slightly higher in HDCO than in H2CO predicts similar rates of deactivation for this isotopic pair, whereas very different photochemical rates are expected for D 2 CO. The rate of non-radiative transition in D 2 CO (4 1 ) is at least 500 times slower than that in H 2 CO (4 1 ), showing an extremely large deuterium isotope effect. Furthermore, operational complications in photochemical isotope separation due to colli sional electronic quenching and vibra- tional energy transfer in H 2 CO and HDCO should be negligible below 1 torr pressure. It has been well recognized that the radical formation process 3 4 could seriously lower the isotope separation efficiency. ' In order to evaluate the quantum yield of hydrogen atom formation from various SVL's, we have tried to devise a sensitive, quantitative and convenient spectroscopic method for detecting photochemical H-atoms by observing the near infrared chemiluminescence emission from the * 7 HNO formed by recombination of H + NO(+M). We have found that the chemiluminescence photoexcitation spectrum observed in the mixture of H 2 CO and NO shows the electronic absorption features of H 2 CO A A 2 < — X x Ai transition, since the photoexcited H 2 CO gives off H-atoms by radical decomposition process. The preliminary o results are illustrated in Figure 2. The chemiluminescence emission from HNO /DNO was confirmed by the analysis of the 760 nm emission band from the mixture of HDCO and NO excited with o a cadmium laser (325.0 nm). The threshold for the H-atom C2-2 formation from H2CO appears to be near 330 nm. The quantitative implications of the above observation to the atmospheric chemistry and the laser isotope separation will be presented at the conference. References 1. (a) B.A. DeGraff and J.G. Calvert, J. Amer. Chem. Soc, 89, 2247 (1967); (b) R.D. McQuigg and J.G. Calvert, J. Amer. Chem. Soc, 91, 1590 (1969). 2. J.G. Calvert, J. A. Kerr, K.-L. Demeriian, and R.D. McQuigg, Science, 175, 751 (1972). 3. (a)E.S. YeungandC.B. Moore, J. Chem. Phys., 58, 3988 (1973); (b)C.B. Moore, Accounts Chem. Res., 6, 323 (1973). 4. J.B. Marling, Chem. Phys. Lett., 35, 84 (1975). 5. R.G. Miller and E.K.C. Lee, Chem. Phys. Lett., 33, 104 (1975). 6. R.G. Miller and E.K.C. Lee, Chem. Phys. Lett., 00, 0000 (1975). 7. (a)M.A.A. ClyneandB.A. Thrush, Disc. Faraday Soc, 33, 139 (1962); (b) T. Ishiwata, H. Akimoto and I. Tanaka, Chem. Phys. Lett., 21, 322 (1973). 8. R.S. Lewis, K.Y. Tang and E.K.C. Lee, unpublished results. C2-3 350 340 330 320 310 300 0.5 0.2 0. * F 2 ,-2 10 2 -3 5h D 2 C0 \ >. HDCO q »\a \ A \ ■ \ \ •\b a o t \ \ \ i l \ I -o .0 o.i 10 -2 10 -3 360 350 340 330 320 310 300 X ex (nm) Figure 1: Fluorescence quantum yield vs. excitation wavelength: 0.045 torr D2CO (filled); 0.20 torr HDCO (half -filled); 0.098 torr H 2 CO (open). SVL assignment: squares = 2 n 4 1 progression; circles = 2 n 4 3 progression; triangle 2 n 5* progression where n = 0, 1, 2, . . . C2-4 O 3 o o u 0) c o •r-i CO CQ 0) CO CO d C o I— 1 3 u Tjisuajui C2~5 Decomposition of Formic Acid Vapor by Infrared Laser Radiation R. Corkum*, C. Willis , M. H. Back* and R. A. Back PIE Group, Chemistry Division, National Research Council of Canada and *Department of Chemistry, University of Ottawa, Ottawa, Canada The infrared induced decomposition of formic acid vapor has been studied using the P. 8 and P p 8 lines from a line tunable, pulsed HF laser. The study was undertaken both to investigate the parameters involved in the induced decomposition and to assess the potential of the system for photochemical isotope enrichment. The beam from the laser was focused into a photolysis cell fitted with infrasil quartz windows. After irradiation the cell contents were condensed at -196°C and the non-condensible products measured by gas burette and analysed either by gas chromatography or mass spectrometry. The only non-condensible product observed was CO and in a few experiments it was demonstrated that an equivalent amount of H„0 was formed. Less than 3% of the decomposition led to C0 ? [l] HCOOH + I.R. •*- H 2 + CO > 91% [2] HCOOH + I.R. -»■ Hp + C0 2 < 3% and Hp. Earlier studies have shown (l) that both channels are important in the pyrolysis of formic acid vapor. The laser beam was focused into the cell with an f = 7-5 cm lens. If this was removed no decomposition was detected. Using a lon£ focal length lens, f = 150 cm, to focus the beam into the cell gave some decomposition but with a yield of only about 1% of that with the C3-1 f = 7. 5 cm lens. These observations are indicative that the decom- position has a relatively sharp intensity threshold. This was confirmed using neutral-density filters in conjunction with the f = 7.5 cm lens. The effect of temperature, pressure, added inert gas and cell path length were investigated. Increasing temperature in the range 25-l80°C led to a slight increase in yield but this could be attributed to a shift in the dimer-monomer equilibrium, with only the monomer absorbing the HF radiation. Increasing the pressure in the range 10 - Uo Torr led to an almost proportional increase in relative quantum yield. Addition of N p or CCU did not affect the amount of IR absorbed but dramatically reduced the observed yield of CO. This effect is interpreted as collisional deactivation of vibrationally excited formic acid. Photolysis was performed in cells 10 cm and 1 cm long. The amount of CO formed in the short cell was only slightly smaller than that formed in the long cell, which is consistent with decomposition occurring only close to the focus of the lens. A lower limit of (J)™ $- 6x10 '" could be set for the quantum yield of the decomposition based on the light absorbed in the short cell. This is a surprisingly large value particularly when it is noted that the activation energy of 29 ± 1 kcal/mole found in the pyrolysis is equivalent to about 3 photons of the incident radiation, and that the quantum yield is still increasing with increasing pressure. Unfortunately, the formic acid system offers little promise for isotope separation. Since HpO and not Hp is the hydrogen-bearing product, any separation in the photolysis would be rapidly lost by C3-2 re-equilibration "with the formic acid substrate. Various mixtures of 12 13 H COOH and H COOH have been irradiated but in each experiment the isotopic ratio in the CO produced faithfully reflected the ratio of 12 the starting material, despite the fact that H COOH absorbs the P. 8 13 line nearly twice as effectively as H COOH. The laser radiation is absorbed by the OH stretching mode of monomeric formic acid and two limiting mechanisms for the subsequent decomposition can be considered. In the first, individual molecules are excited to levels above their dissociation energy by stepwise absorption of quanta in the 0-H stretching vibration. Collisional or power broadening of the higher levels must be invoked to circumvent energy mismatch due to anharmonicity, although since only 3 or k quanta are required to exceed the dissociation energy, this problem is not serious. Unimolecular dissociation of vibrationally excited formic acid molecules would follow the usual internal reorganization of vibrational energy into the mode leading to dissociation to CO + HpO in collisional deactivation. In the second limiting mechanism collisional relaxation of the initially excited OH stretching vibration is fast enough that stepwise excitation to higher levels is not required. Collisional relaxation into other modes results in a small volume of hot vapor which decomposes through a normal thermal decomposition mechanism in competition with cooling by loss to the surrounding cold vapor. Either mechanism could probably account for the relatively high quantum efficiency of the decomposition and both would show a strong dependence on intensity as observed. The lack of isotopic C3-3 selectivity and the pressure dependence suggests that the second mechanism is operative, but uncertainties in the details of the thermal decomposition make it impossible to make a clear choice between the two. The mechanisms suggested are limiting cases, the first favored at low pressure and the second at high pressure. The present experiments may correspond to a mixture of the two with an intermediate degree of relaxation of the OH stretching excitation into other vibrational modes . Reference (l) P. G. Blake and C. Hinshelwood, Proc. Roy. Soc . (London) A255_, kkk (I960). C3-4 The Laser Augmented Decomposition of D B Adducts K. R. Chien and S. H. Bauer Department of Chemistry, Cornell University, Ithaca, New York 14853 We have completed the investigation of the laser augmented decomposition of H BPF and extended these studies to D BPF . The low power, low pressure absorption coefficients (a ) of D BPF for many lines in the C0 9 J.U o o di laser are somewhat larger than the corresponding magnitudes for H BPF , o o and the entire band is shifted to higher frequencies. The band shapes and relative positions follow the trends predicted from a normal mode analysis. The P(J) dependence of the augmentation factor is most interesting. For the tri-deutero compound, the highest photolytic yield is produced by the P(24) line; here a is approximately 1/5 as large as that at the absorption band peak, which occurs at P(4); for the tri-hydrido compound, the most effective line is P(32), where the absorption coefficient (« 1ft ) is approximately 1/4 of that at the band peak, which occurs at P(22). The H/D separations are readily achieved by line selection. With P(24) the augmented decomposition of D Q BPF 11 preferentially selects the B isotope by « 3 5% in a 5 minute run. Also, D BPF is decomposed preferentially, relative to HD BPF ; the H/D separa- tion is substantial, and increases with irradiation time . The (10/11) speci- ficity did not show up in H BPF , in agreement with the normal mode analysis. The augmented rate [H BPF at P(32)] is linearly dependent on the pressure from 2-15 mtorr. At 25 mtorr, and more so at 50 mtorr, the degree of augmentation is increasingly greater. This may be due to the decrease in the mean free path within the reaction cell and consequent v-T heating of the reagent, such that a thermal component is superposed on the photolytic rate. The rate is clearly not linearly dependent on the power . Plots of log(% con- version) vs log( laser intensity) gave slopes »3.1, indicating that at least three photons are required per molecule to augment decomposition. The time dependence of the decomposition, at low % conversion, proved to be unexpect- ed! ed. Whereas the % decomposition is linear with time at the high power levels (120 watts), or for lower powers at higher temperatures (60 w at 36 C), at low powers (50 watts) the % decomposition is approximately quadratic with time but becomes linear at longer time periods. This is strongly suggestive of a multi- photon process . For runs below 15 mtorr the rate constants at several laser powers were calculated on the basis of a first order rate law since these rates are inde- pendent of pressure; the reaction time was kept constant at 12 minutes. "Activation energies" (E ) were deduced from pairs of rate constants at two wall temperatures: 25 C and 38 C. The magnitudes of E varied from 2.5 to 5 . 9 kcal/mole , with a mean value of 3 . 5 kcal/mole . At this stage we assumed that the increment in average gas temperature (which must be somewhat higher than the wall temperature) is equal to the increment in wall temperature . To determine the photon conversion efficiency we measured changes in laser output levels due to intra-cavity power absorption by the sample, at laser power levels used for the moderate photolysis runs (approximately 60 watts). The mean absorption coefficients (a ) were calculated following the analysis of Kaufman and Oppenheim. They found that laser output decreased essentially linearly with (1-e ) until lasing was quenched. Write J /J = l-f3(l-e ); -Li Xj a D = fti where p = 3 . 33 for our laser and a = (a)T> at these o-i) + j l /j low pressures. The (a) coefficients for selected lines were thus obtained. The gross photochemical efficiency was then estimated. The laser power ex- tracted by the sample in the cavity per passage through the cell is 20 J (1-e ), where J is the power read with the monitoring meter, JLi J_i through the output mirror (95% reflecting). The number of photons absorbed is 7 40 J • (a )pD- 10 AtAl> . The ratio of photons absorbed /molecule decomposed 4 o is ~ 4 x 10 for 60 w of laser power and p = 3 mtorr (298 K). Then ~ 30% is decomposed in 12 minutes. This low efficiency and the observed low value for E (mean: 3.5 ±1.1 kcal/mole), in contrast to the overall thermal value of C4-2 29. 3 kcal/mole, may be accounted for by considering: (a) the geometry of the reaction cell, (b) the previously estimated thermal activation energy, ■Hi E (lb) «5 kcal/mole, and (c) the slope of the log(% dec) vs log(power) line, a A semi -quantitative account follows. The two elementary steps which account for the thermal reaction are: k D BPF , =± DB + PF AH° w 24. 5 kcal/mole (la ; -la) 3 3 t o o X k -i D„B + D„BPF„ -r— »- B„D„ + PF„ AH° » -13 . 4 kcal/mole (lb) 3 3 3 K, 2 6 3 b It is plausible to assume that the vibrationally excited adducts do dissociate to a greater extent than via the thermal reaction. Under high flux irradiation, the D Q BPF (v ; V ) leave the illuminated core and collide with unexcited 3 3 3 9 adducts and dissociation fragments in the larger encompassing volume. How- ever, since BD 's recombine to B.D„ and D BPF„ with essentially kinetic 3 2 6 3 3 cross -sections, its concentration level in the encompassing volume is but an order of magnitude higher than under thermal conditions. The upper limit temperature in the encompassing volume is ~ 45 C because of the absence of detectable decomposition of admixed H BCO (in a 12 min. run). The aug- mented rate we observe is due to D BPF which enter the annular region around the illuminated core and there encounter BD 's. A conventional steady-state condition on [BD ] leads to: [BD 3 ! S s " (iT L ) yf +( k 1 /kl)a-t) < 2 > : -l ;y?+(k b /k -l )(1 " ?> decomposed at the speci at early times, when [BD ] ~ [PF ] to unity, when O o -J where 4 is the fraction of D BPF decomposed at the specified time, and y _ 3 3 ranges from 2 [PF 3 ] »[BD 3 1. Under laser irradiation both k and k, increase. Thus, for those adducts that accumulate 3y , about 3 kcal/mole reside in B-P extension. Since E, =4.8 kcal/mole and E = 24. 5 kcal/mole, we postulate that: C4-3 (3l/ 3 ) n A 1A 15 f 21,500"! , 3' „ . „ rt i5 r k = 2.4 x 10 exp I - RT J (3) , (3l/ 3 ) _ 1A 12 r 18001 *b =3xl ° exp L"-iFJ (4) * -11 Application of Eq. (2) gives [BD ] «1.2xl0 " moles /cc, when £ -* 0. 1, o 3 3 and T = 318 K for one mfp thick annular region. (Volume = 1. 3 x 10 cm ). The number of adducts that enter and leave the illuminated core (per sec) 20 14 -3 is 2. 8 x 10 , when the molecule density is 10 cm (3 mtorr) [derived from -5 NcA/4]. Their mean exposure time (met) to the radiation is 4 x 10 sec. 1 fl During one such period 1. 12 x 10 D BPF enter and leave; concurrently, 15 3 3 1.37 x 10 photons are absorbed; i.e. on the average one out of 8 adducts picks up a single photon. Goodman calculated that 17 of these pick up two, and one out of 79 picks up three in a coherent sequential process . Thus, 13 < 3i y s» 1. 8 x 10 D BPF enter the annular region (1 mfp =1.5 cm) per met . These either get de-excited or react with a D B prior to entering the re- maining volume. The total moles of B D. produced during a 12 min. run, at the 60 watt level, via ,* D 3 BPF 3 ♦ BD 3 -+ B 2 D 6 + PF 3 ,13 ;xlo / 720 \ 1.8xl0 13 1 r „ „„11, IT ,„ , m = ST—*) 23 X300 [2Xl ° ' [BD 3 ! s S < 5) * -7 If we accept the above estimate for [BD ] , TR = 9.9 x 10 moles, compared -6 SS with 3. 9 x 10 moles of the adduct initially present; that is, a conversion of « 25%, close to that observed. C4-4 Infrared Laser Isotope Separation by John L. Lyman, L-8 and Stephen D. Rockwood, L-7 Los Alamos Scientific Laboratory P. 0. Box 1663 Los Alamos, New Mexico 87545 Extended Abstract 1-5 It has been established in previous work that when SF is irra- diated by intense C0_ laser radiation isotopically selective dissocia- tion occurs. The following points have been established: 1) The reaction is selective when the pressures is less than about 10 torr. 1 ' 2 ' 5 2) Both reaction yield and selectivity decrease with increasing 1,2,3,5 pressure. 3) A fluorine atom scavenger such as H ? enhances reaction yield and 1 2 ■? S selectivity. ' ' ' 4) The reaction yield increased more rapidly than linear with pulse 4 5 energy and there appears to be an intensity threshold. ' 5) The reaction yield depends on the laser frequency, but it does not 4 follow the infrared band contour. 3 6) Other molecules also undergo isotopically selective reactions. In a typical experiment a mixture of SF and hL is irradiated by focussing the C0 ? laser pulses into the reaction cell with a short 2 1 focal length BaF lens. ' If it is assumed that 40 laser photons are required for rapid dissociation (the bond strenth is 28 photons), then C5-1 -4 the photon utilization efficiency is 2.5 x 10 for a sample of 0.2 torr of 25% SF in H_ and 1.0 Joules/pulse. It has been suggested that this efficiency could be improved by irradiating the gas in reflective cylinders. Several cylindrical cells have been constructed and tested. (See Fig. 1 for cell construction and dimensions.) For any meaningful evaluation of laser-induced chemical reactions occurring in these cells the distribution of laser flux (or intensity) must be known. If, it is assumed that geometrical optics is applicable to this problem then the intensity in a cylinder at a point x centi- meters from the entrance end and r centimeters from the cylinder axis is approximately given by h 2 n $(x,r) = $ ~- R (1) o 2ra where $ is the flux of the laser light at the focussing lens and the laser spot is assumed to be of uniform flux with radius b. The cylinder has radius a and wall reflectivity R and n is the mean number of times a ray strikes the wall before reaching the point x. fi is given by n=l/3f (2) where f is the lens focal length. The following table lists some preliminary results for the three cells listed in Fig. 1. All experiments were performed with 25% SF in H irradiated with CO laser pulses at 10.6 urn (P(20) line). The equivalent reaction volume, v, is the fraction of the SF that reacts per laser pulse times the total cell volume and is therefore a C5-2 a measure of the reaction yield. The selectivity factor, a, is defined by 3 32 32 d[^SF ] [""SF ] —34 = a — (3) d[^SF 6 ] r 4 SF 6 ] or as the ratio of equivalent reaction volumes for the two isotopic species. 3 The isotopic enrichment factor for the residual gas, 3, will be given by In 3 = (1-1/cO Nv/V (4) where N is the number of laser pulses and V is the cell volume. Equa- tion 3 was used to calculate a from 3, v, and N. In experiments where the amount reacted was not determined, a was assumed to be five and Eq. 3 was used to calculate v. T; able 1 Relative Cell P(torr) E(J) a v (cm ) Yield 1 n. 1 3.5 5* 6.0 14.2 1 0.5 2.5 6.4 3.1 13.0 1 1.0 3.5 4.5 1.4 6.7 1 2.0 3.9 1.7 2.3 15.6 2 0.1 2.1 5* 16.1 ^50 2 0.2 3.4 5* 8.9 13.2 2 1.0 2.8 5* 1.3 3.8 3 0.5 2.7 5* 0.4 1.6 Assumed; see text C5-3 Note from the table that enhancements in reaction yield as high as 50 were obtained, this corresponds to a photon utilization efficency of 0.5%. This is certainly an improvement over the single focus effi- ciency. The highest efficiency was obtained with cell 2 which has the highest wall reflectivity (> 0.97). The reflectivity is of cells 1 and 3 are respectively 0.87 and 0.70. Experiments are continuing with these cells. The pulsed CO TEA lasers normally give mode locked pulses. That is the 200 ns laser pulse is really a train of closely spaced subnano- second spikes. This mode locking can be supressed by adding a cw gain cell to the laser cavity. Experiments performed with and without mode locking indicated that the selectivity of the SF laser induced disso- ciation is not significantly altered by mode locking. However, the reaction yield is greater by 25 to 50% for a mode locked pulse than a non-mode locked pulse of the same total energy. References 1. R. V. Ambartzumian, Y. A. Gorokhov, V. S. Letokhov, JETP Lett. 21_, 171 (1975). 2. J. L. Lyman, R. J. Jensen, J. P. Rink, C. P. Robinson, and S. D. Rockwood, Appl. Phys. Lett. 27_, 87 (1975). 3. J. L. Lyman and S. D. Rockwood, J. Appl. Phys. 4_7_, 595 (1976). 4. R. V. Ambartzumian, Y. A. Gorokhov, V. S. Letokhov, G. N. Makarov, and A. A. Puretskii, Zh. ETF. Pis. Red., 23, 26 (1976). 5. J. L. Lyman and S. D. Rockwood, Proceedings of the Laser-Optics International Laser '75 Conference Nov 11-13, 1975, Anaheim, CA (to be published) . C5-4 6. T. P. Cotter (private communication) Figure 1 - light trap cell 4> (r)-r = constant # (r) -- energy/area METAL TUBING NaCI WINDOW CELL (1) 5 cm I.D. x 20 cm LONG, Au CELL (2) 0.95 cm I.D. x 61 cm LONG, Cu CELL (3) 0,23 cm I.D. x 61 cm LONG, SS C5-5 CO TE Laser Induced Photochemical Enrichment of Carbon Isotopes Joseph J. Ritter and Samuel M. Freundf National Bureau of Standards Washington, D. C. 20234 Several groups have previously reported on the CO„ laser photolysis 12 3 3 gaseous CI CF ' ' . In one case , carbon isotopic perturbations were noted in samples subjected to high energy pulses from a CO^ transverse excitation (TE) laser. We have studied the CO TE laser-induced chemical reactions of Cl„CF with several reagents and have noted substantial carbon isotopic enrichments along with evidence for the participation of a reactive intermediate, dif luorocarbene . The systems studied along with the principal products are given below: Reaction Reactants Principal Products I C1 2 CF 2 + °2 C0F 2' (C1CF 2 ) 2' C1 2 II C1 2 CF 2 + N ° C 2 F 4' C0F 2' C1CF0 ' N0C1 ' (C1CF 2 ) 2 III CI CF + (H C) C=CH C o F a + com P lex array of other products IV C1 9 CF ? + HC1 HCF CI, (ClCF ) CI ' ^ 2 2 2 2 The reaction mixtures were contained in a 1 I stainless-steel cell and subjected to focussed 929 cm (P-36) radiation from a C0„ TE laser delivering 300 ns wide pulses, with 0.20 J/pulse at 2 pps. Chlorine (identified by color and mass spectrometry) was removed by treatment of the crude reaction mixtures with Hg at 25°. The resid- ual gases were then subjected to GLC (10 m, 20% squalane on firebrick) . Carbonyl fluoride-containing mixtures were passed over uncoated fire- t Present address, Los Alamos Scientific Laboratories, Los Alamos, New Mexico. C6-1 brick to effect a quantiative COF -> CO conversion, in order to avoid this conversion by the firebrick support within the GLC column. Sepa- rated components were identified by ir and isotope ratios determined by mass spectrometry with a typical relative standard deviation of < 3%. Estimates of halocarbon product recoveries were made from GLC peak area measurements, assuming equivalent TC detector response for all compounds. Other, generally less volatile materials, in amounts too small for unequivocal identification, were noted in all of these systems. Mass spectrometric examination of the products of reaction I, 13 indicated a depletion of C while the analysis of recovered, unreacted 13 CI CF indicated its C content to have changed from an initial 1% to 13 the 16-20% range. This alteration in C content is sufficient that it is manifested in the ir spectrum of the recovered material with the appearance of two new bands (1131 and 1077 cm ) as shown in Figure a. Additional experiments performed under identical conditions but with the R-18 (1077 cm ) line of the TE laser resulted in recovered Cl-CF which was depleted in 13 C Thus the results of the isotopically specific reactions were used to confirm the assignment of the 1077 and -1 13 12 929 cm absorptions to C and C species of the CI CF respectively. 13 For II and III unreacted Cl_CF , enriched in C was recovered. 12 12 The appearance of C enriched COF from reaction I and C enriched C„F from reactions II and III constitutes evidence for the laser induced, isotopically specific formation of dif luorocarbene (:CF ) Dif luorocarbene is also implicated in the formation of HCF CI from 4 reactions with HCl . The fact that C F is not also noted in the HC1 and O reactions is probably due to the facility of the + :CF„ ■> C6-2 COF + O, Cl + :CF -*■ CI CF and HC1 + :CF ■> HCF Cl reactions as com- pared to the dimerization of :CF . On the other hand, our results indi- cate that when Cl CF is laser-irradiated in the presence of compounds very reactive toward chlorine such as NO or an olefin, the Cl + :CF ■> Cl CF reaction is accordingly less favored and the probability for 3 dimerization of :CF thus improved. Other workers have observed C F from the laser photodissociation of Cl CF . Further studies designed to affirm the role of laser-produced di- f luorocarbene and an assessment of its potential as a reagent in iso- topically specific syntheses are currently in progress. 1200 noo 1000 CM" 900 800 Fig. a. TOP: C^CF- prior to laser photolysis. Pressure approximately 10 torr. BOTTOM: CI2CF2 recovered after laser photolysis in the presence of O2. Pressure approximately 10 torr. C6-3 REFERENCES 1. M. P. Freeman, and D. N. Travis, J. Chem. Phys. 60, 231 (1974). 2. R. N. Zitter, R. A. Lau and K. S. Wills, J. Amer. Chem. Soc . 97_, 2578 (1975) . 3. J. L. Lyman and S. D. Rockwood, J. Appl. Phys. 4_7, 595 (1976). 4. Walter Mahler, Inorg. Chem. 2, 230 (1962). C6-4 Measurement Of Mean Lives In Atomic Uranium Jules Z. Klose National Bureau of Standards Washington, D.C. 2023^ During the past several years the availability of tunable dye lasers has led to attempts to develop inexpensive methods of producing isotopically enriched uranium to meet the needs of the expanding nuclear power industry. One of these methods involves isotope sep- aration through stepwise photoionization processes in which selected levels of the isotope of interest are excited with tuned laser light , and the final step of ionization is achieved with a laser or an ultra- violet lamp. The isotope ions are then separated electrostatically from the neutral atoms. Using this method, groups at the Lawrence (1 2) (3) Livermore Laboratory ' and the Avco Everett Research Laboratory 235 238 have done experiments demonstrating the separation of U from U on a small scale. 235 238 In selecting a transition to be used in separating U from U, the overriding requirement is that the isotope shift must be large enough to avoid overlapping the unshifted line with the hyper fine structure of the isotope line. Once this requirement has been met, the further criteria of strong absorption and suitable level lifetime must be applied. Thus, selecting transitions requires knowledge of both the absorption cross sections, i.e., transition probabilities, and the excited state lifetimes. Although these two quantities are related, their relationship is usually not a simple one since most excited atomic states have several decay modes. However, in some cases decay C7-1 branching ratios, i.e., relative transition probabilities, are avail- able from other sources such as arc measurements . In such cases the absolute transition probabilities can be derived from the level life- times. To help provide some of this needed lifetime data, the National Bureau of Standards has been carrying on a program of measurements of (h) mean lives of excited levels in the uranium atom. In a comple- mentary program at NBS, relative transition probabilities in uranium are being determined using a stabilized arc. At NBS uranium atomic lifetime determinations have been made using low-energy electrons for excitation and a method of delayed coincidence for detection. In order to provide free atoms for the measurements, a Knudsen effusion device especially designed to produce diffuse beams (h) of atoms of heavy elements was constructed and placed in operation. At this writing a result of 7-3 ± 1.1 ns has been obtained for the mean life of the upper level of the 358^.9 A resonance transition in natural uranium. For U this level has an energy given in wavenumbers of 27886.99 cm • In addition efforts have been made to measure the mean life of the 16900.39 cm " level using the 5915-^ A resonance transi- tion. These efforts were largely thwarted by the very high-intensity scattered light from the source at ~ 2000°C, which gave rise to very poor signal-to-background ratios. In an attempt to overcome this difficulty of low signal-to- background ratios , a tunable dye laser pumped by a pulsed nitrogen laser is being installed as the source of excitation. The tunable laser will provide such strong decay signals that much more efficient means of detecting the decays can be used. Also, the problem of C7-2 cascading into the level of interest from simultaneously excited higher levels will be eliminated by selective level excitation with the tunable laser. (3) The groups at the Avco Everett Research Laboratory ' and the (7) Lawrence Livermore Laboratory have reported measuring mean lives in U I using their experimental isotope separation setups. In their methods the lifetime of an intermediate level is obtained by observing the yield of isotope ions as a function of the delay between the exciting and ionizing laser pulses. References 1. S. A. Tuccio, J. W. Dubrin, 0. G. Peterson, and B. B. Snavely, IEEE J. Quantum Electron. QE-10, 790 (19T 1 +). 2. S. A. Tuccio, R. J. Foley, J. W. Dubrin, and 0. Krikorian, IEEE J. Quantum Electron. QE-11 , 101D (1975). 3. G. S. Janes, I. Itzkan, C. T. Pike, R. H. Levy, and L. Levin, IEEE J. Quantum Electron. QE-11 , 101D (1975). k. J. Z. Klose, Proceedings of the SPIE/SPSE Technical Symposium East, March 22-25, 1976, Vol. 76 (in press). 5. P. A. Voigt, Phys. Rev. All, I8U5 (1975). 6. J. Z. Klose, Phys. Rev. A 11, l8Uo (1975). 7. R. W. Solarz, J. A. Paisner, L. R. Carlson, C. A. May, and S. A. Johnson, Lawrence Livermore Laboratory Report UCRL-77590, 1975 (unpublished). 8. L. J. Radziemski, Jr., S. Gerstenkorn, and P. Luc, Opt. Commun. 1_5_, 273 (1975). C7-3 Cm" 1 50000 r IONIZATION LIMIT 49250 ± 160 Cm- 1 40000 30000 20000 10000 - .-2 2 000 28650 « 16000 34659 , i- r _ Jr7 I I I® I 0) 17J361 h 3 3 Si" i i i J .6 116900 Jr7 620 V; V c ' Fig. 1. Examples of two- and three-stage ionization processes for separating U from U. The energies of the levels and transitions are given in wavenumbers (cm ) , and the wavelengths of „the transitions are given in angstroms (A) . The values given are^f or U,„but differences in energies and wavelengths between U and U are generally not discernible within the precision of the quantities given in the figure. This figure was reproduced with changes from reference (8). C7-4 HIGH -TEMPERATURE SOURCE PHOTO- MULTIPLIER AMPLIFIER DISCRIMINATOR START TIME TO PULSE- HEIGHT CONVERTER PULSER STOP PULSE-HEIGHT ANALYZER Fig. 2. Block diagram of the delayed coincidence experimental setup used for measuring mean lives in heavy elements. C7-5 HIGH-TEMPERATURE SOURCE MONOCHROMATOR SAMPLING SCOPE SIGNAL AVERAGER PHOTO- ULTIPLIER PHOTODIODE TUNABLE DYE LASER NPUT TRIGGER PULSE □ NITROGEI LASER SWEEP OUTPUT □ ADDRESS ADVANCE PULSE Fig. 3. Block diagram of a system for measuring atomic mean lives utilizing the strong decay signal produced "by a pulsed tunable dye laser as the source of excitation. C7-6 Principles of Photochemical Separation of Isomeric Nuclides - G.C. Baldwin , H.M. Clark, D. Hakala, R.R. Reeves, Rensselaer Polytechnic Institute Abstract Aside from intrinsic scientific interest, photochemical separa- tion of isomeric states of a single nuclide offers one possibility of achieving a nuclear population inversion, essential to the de- (1 2) velopment of gamma ray lasers v ' . Just as for isotope separation, photochemical methods of isomer separation depend upon the existence and magnitude of spectral differences between the species to be separated exceeding the resolution of tunable optical sources. Mercury has already been shown to be capable of isotopi cally- (3) selective photochemical reaction using 254nm radiation^ . These studies have been performed on the even mass number stable isotopes (i.e. 198 and 202) which have zero nuclear spin and one hyperfine component per isotope within the hyperfine structure of the 254nm line. Mercury 197m, the excited nuclear state, has a half-life of 24 hours, with 3 hyperfine components due to the nuclear spin (4 5) 197a splitting effect v ' . The ground state, Hg 3 , is also radioactive with a half-life of 64 hours, but the nuclear spin splitting results in a doublet^ 4 ' 5 ). Mass, spin, quadrupole, and nuclear volume effects give rise to shifts and splittings of the optical energy levels of atoms and mole- cules. In the case of mercury, its optical spectrum has been shown to contain several lines having well-resolved hyperfine structures characteristic of the different isotopic species. The hyperfine C8-1 structure of the 254nm line has been measured for Hg and Hg 9 . Isotopes have been separated by chemical reaction of mercury atoms selectively excited at 254nm by means of a mercury lamp incorporating a single isotope (or enriched in a single isotope). The excited atoms can react with oxygen to produce isotopically enriched HgO (3) which may be recovered v . In a similar manner it should be possible to excite and separate "I Q"7 m the Hg ' . However, a different source of radiation is needed. The specific wavelength required may be obtained by means of a nitrogen laser coupled to a tunable dye laser. The 254nm line can then be scanned using a nonlinear potassium pentaborate (KPB) doubling crystal. Resolution is in the order of .0008nm using an incavity etalon with such a laser. One absorption line of the triplet of the Hg ' m at 254nm is almost 0.002nm from the nearest Hg g component and about 0.0026nm from the nearest hyperfine component of any of the (5) stable isotopes v . Therefore it should be possible to recover the Hg enriched with respect to Hg * and the other isotopes. The degree of isomer enrichment may be determined by measurement of the characteristic radiations emitted in the decay of the Hg and Hg 9 . It is noted also that the use of recoil-based radiochemical isotope enrichment (The Szilard-Chalmers process) during the prepa- ration of the isomeric species would facilitate the subsequent photo- chemical enrichment of the isomeric state. Another element which has been studied for photochemical isotope enrichment^ ' and has isomeric states of reasonable lifetime is bromine. Selective excitation of the molecule, Br2» results in the C8-2 formation of bromine atoms. Reaction of these atoms with HI can produce isotopic enrichment in the product formed initially. Bromine-80 has a metastable state with a half-life of 4.4 hours and a ground state with a half-life of 17.6 minutes. It should be feasible to separate these isomers by a photochemical process if the molecules can be selectively electronically excited. The absorption spectra for molecules containing the isomers has not been measured, but isomer shifts should exist for these species compared to the normal bromine. Partially supported by The National Science Foundation On leave to Los Alamos Scientific Laboratory, 1975-76 1. G.C. Baldwin and R.V. KhokhlovPhysics Today 28 (2), 32 (1975). 2. V.S. Letokhov, Science 180, 451 (1973), J90, 344 (1975) 3. H.E. Gunning and O.P. Strausz, in "Advances in Photochemistry" ed by Noyes, Hammond, and Pitts, Vol I, p209 (1963) 4. CM. Lederer, J.M. Hollander, I. Perlman, Table of Isotopes, John Wiley (1967) 5. A.C. Mellissinos, S.P. Davis, Phys. Rev. 115, 130 (1959) 6. S.R. Leone, C.B. Moore, Phys. Rev. Letters, 33, 269 (1974) C8-3 The Early Years of Photochemistry in the Vacuum Ultraviolet. Wilhelm Groth Institut fur Physikalische Chemie der Universitat Bonn. The wavelength region between 1850 A and about 1250 A was opened to spectroscopy around 1900 through the pioneer- ing work of V. Schumann. Photochemistry, which is the young- er sister of spectroscopy, reached the Schumann region only after 1930. After shortly describing the importance of pho- tochemistry for the kinetics of thermal reactions, especially for the mechanism of chain reactions, the first spectroscopic investigations in the Schumann ultraviolet with the light of condensed sparks between metal electrodes and with quartz optic are mentioned. Difficulties originated in the trans- parency of the window materials, the absolute energy measur- ments and above all in the lack of suitable monochromatic light sources. Starting with the development of the high-current low- voltage xenon lamp of Harteck and Oppenheimer the technical improvements of the rare gas lamps are described: the application of low-current glow discharges, micro wave dis- charges, and the use of other rare gases than xenon, f. i. krypton and argon. Dl-1 In the beginning the absorption and photochemical primary and total reactions of simple gases were investigat- ed : oxygen, nitrogen, hydrogen, ammonia, carbon dioxide, carbon monoxide, water vapour, and some organic compounds like methane, ethane, and propane. Remarkable are some experiments carried out by the author and Hans Suess in 1937 who illuminated mixtures of water vapour and carbon dioxide by the xenon resonance wave lengths. The following primary reactions occur: (1) C0 2 + hy — * CO + (2) H 2 + h V * OH + H H atoms react with CO to give formaldehyde or glyoxal according to (3) H + CO(+M) — ► HCO (+M) H 9 C0 + CO (4) 2 HCO * (HCO) 2 This process was assumed to have played an important role in the earth 's atmosphere. It would give an explanation for the first appearance of free oxygen in the primitive atmo- sphere and for the formation of certain carbon compounds that were probably the prerequisite for the evolution of organic life. Later on the hypotheses of Oparin, Urey, Kuhn, and Bernal assumed that the earth's primitive atmosphere con- tained hydrogen, methane, and water vapour. Since the Dl-2 greater part of the energy available in the earth's atmo- sphere would be contributed by the ultraviolet radiation of the sun, the earlier photochemical experiments were resumed in 1957 by the author and Hanns von Weyssenhoff, using mixtures of methane of ethane with ammonia and water vapour in mercury-sensitized as well as direct photolysis experi- ments. For irradiation the resonance wave lengths of mercury at 2537 8 and 1849 8, of xenon at 1469 A* and 1296 8, and of krypton at 1236 A and 1165 A were used. Aminoacids (glycine, 6l*-alinine, and fi^-aminobutyric acid), and fatty acids (formic acid, acetic acid, propionic acid), and hydrocar- bons could be detected in several experiments. These results were confirmed in 1959 by Terenin, in 1961 by Dodonova and Siderova, and in 1966 by Ponnamperuna and Flores. Special photochemical reactions in the Schumann ultra- violet which were carried out in the 1930 's with a xenon lamp were concerned with the reactions of atoms. It could be shown that even at extremely small oxygen partial press- ures in mixtures with hydrogen and carbon oxide the ozone formation in a triple collision predominates the primary attack on hydrogen or carbon oxide molecules. Itwas assumed that the primarily formed energy rich ozone at low atom concentrations reacts with the molecules before it loses its energy by collisions. Also the nitrogen oxidation was investigated in the light of the xenon lamp. N 2 could be detected mass-spectro- Dl-3 metrically; about 10~ 4 of the atoms formed NO. The small amount of N„0 which is found in the earth's atmosphere can at least partially be produced by this photochemical reaction, Dl-4 Fluorescence and Decomposition of Tertiary Amines in a Glow Discharge Francoise Lahmani* and R. Srinivasan IBM Thomas J. Watson Research Center Yorktown Heights, New York 10598 Introduction Although electric discharges in gases have been extensively used to produce the excited states of many atoms and small molecules which were subsequently studied by spectroscopy, little attention has been given to large polyatomic, fluorescent molecules (with the exception of benzene and toluene) . Tertiary aliphatic amines which have been shown to be strong emitters in the near ultraviolet should offer an inter- esting class of compounds for study. We have conducted an investiga- tion of the fluorescence and decomposition of such tertiary amines in a continuous electric discharge in a flow system. In this communica- tion, we shall confine our remarks to diazabicyclo [2 . 2 .2] octane (DAECO) which seems to be typical of these systems. Experimental The flow system and discharge that were used were unexceptional. The flow rates were about 0.02 mole/hr at the vapor pressure of DABCO which ranged from 0.03 to 0.3 torr. The light emitted from the dis- charge was analyzed by a 1-meter, Jarrell-Ash spectrometer and de- tected by a EMI photomultiplier (S-20 response) . Results The following is a summary of the results that were obtained: * on leave from Laboratoire de Photophysique Moleculaire University de Paris Sud, Orsay, France. D2-1 1) In pure DABCO vapor (0.3 torr) , at the lowest voltage at which a discharge could be sustained (60V/cm) a purple glow was observed which on spectroscopic analysis was found to be identical to the fluorescence of DABCO vapor that has been reported . 2) With an increase in the energy of the discharge, the intensity of the light that was emitted increased. A new emission which con- 2 sisted of the CN* violet system at 421.6, 388.3, and 359.0 nm was observed. With an increase in voltage, the intensity of CN* emission increased rapidly while the fluorescence of DABCO remained steady or decreased slightly. 3) Addition of helium sharply increased the current as well as the emission from CN*. Argon produced qualitatively the same effect as helium. At a given voltage and a constant ratio of helium to DABCO, the intensity of CN* emission was roughly proportional to the current up to a 20-fold increase. At the same time, the fluorescence of DABCO was only slightly increased or not at all. 2 2 A) The CN* B Z ■+ X Z violet bands Av=o Av=-1 were studied under high resolution in order to obtain information on the vibrational and rotational intensity distributions. The weakness or absence of anoma- lous distributions in the intensities indicated that the first excited 2 2 electronic state A II of CN was of minor importance relative to B E in this study. 5) The intensity distribution pattern was the same (within exper- imental uncertainty) in the presence and absence of Helium or Argon, with one exception. 6) Under conditions in which only the fluorescence of DABCO was D2-2 observed, no decomposition products were detected. When CN* emission was also present, the products that were detected were C„ hydrocarbons, HCN, and a reactive small molecule which may be C ? H_N. Discussion Present observations suggest that there are two independent routes to the excitation of DABCO in an electric discharge. These consist of (i) one that leads to excitation to the fluorescent electronic state which has been tentatively identified by Halpern as a low-lying optically forbidden singlet state with a life'time of 1040 nsec. (ii) a second path which induces the dissociation of the mole- cule. The mechanism by which DABCO dissociates and gives rise to CN* 2 B E cannot be deduced from the present experiments , but it is obvious that CN*cannot be a direct dissociation product of DABCO. The observation that the distribution of vibrational and rotation- 2 al energy in the B E state appears to be normal (without any population inversion) is consistent with other studies on electric discharges. It suggests that either a thermally equilibrated ground state of CN is the 2 precursor to CN* B E, the excitation being due to collisions between CN and electrons, or that CN*is produced in a dissociation process from unstable intermediates without much change in the equilibrium C-N bond 2 distance. This kind of intensity distribution in CN* B E has been 3 obtained in electron impact induced dissociation of cyanides and in 4 the photodissociation of HCN . It is markedly different from results obtained in the reaction of active nitrogen with hydrocarbons or dissociative excitation of halogen cyanides by collision with metastable argon atoms . D2-3 The enhancement of the decomposition of DABCO relative to its fluorescence on the addition of helium or argon confirms the existence of two parallel pathways for the excitation of DABCO. References (1) A. M. Halpern, Mol. Photochem. 1 517 (1973). (2) R. W. B. Pearse and A. G. Gaydon, The Identification of Molecular Spectra, John Wiley and Sons, Inc., New York (1963). (3) I. Tokue, T. Urisu and K. Kuchitsu, J. of Photochem. 3, 273 (1974- 1975). (4) A. Mele and H. Okabe, J. Chem. Phys . 51, 4798 (1969). (5) N. H. Kiess and H. P. Broida, 7th Symposium on Combustion, 1958, Bulterworth (1959) p. 207. (6) J. A. Coxon, D. W. Setser W. H. Duewer, J. Chem. Phys. 58, 2244, (1973). D2-4 Photochemistry arid Flash Photolysis of 5-Nitroquinoline A. C. Testa and A. Cu, Department of Chemistry, St. John's University, Jamaica, New York 11439 As an extension of our photochemical studies on aro- matic nitro compounds into heterocyclics we have investi- gated the photochemistry of 5-nitroquinoline, which in the presence of $0% isopropyl alcohol-water solutions contain- ing HG1 results in photoreduction to a chlorinated 5-amino- quinoline. The quantum yield of 313 nm photolysis, which proceeds via the triplet state, increases with HC1 concen- tration and levels off at a maximum value of 5«4 x 10 when (HC1) - 1 M, Photochemical disappearance is attenu- ated in sulfuric acid solutions as wel 1 as in neat iso- propyl alcohol, which indicates that the chloride ion serves as an electron donor to the triplet state of the aro- matic nitro compound. Measured triplet yields for 1- and 2-nitronaphthalene are 0.63 and O.83, respectively, and it is reasonable to expect a comparable value for 5-nitro- quinoline . In order to evaluate the importance of the chloride ion as an electron donor we have carried out a flash photol- ysis study of this molecule in a variety of solvents with the aim of trying to identify triplets, anions and neutral radicals that may account for its photochemical behavior. Although the absorption spectrum of neither the triplet nor the radical anion is known we felt that a comparison of P3-1 photochemical and flash photolysis data might provide the needed correlation to assign any observed transients. Flash photolysis of 5 -nit ro qui no line does not result in any triplet -triplet absorption for any of the solutions studied, which suggests that the triplet is short lived ( < i|.0 usee). That the triplet is populated can readily be discerned from the phosphorescence yield of 5-nitroquinoline in EPA at 77 K, which was measured and found to be 0.27. The distinct 0-0 band, vibrational structure and similarity to naphthalene and nitronaphthalene phosphorescence provide evidence that the lowest triplet is tt,tt . The photochemi- cally active medium where reduction occurs with 313 r™ photolysis is $0% isopropyl alcohol-water containing HC1, and flash photolysis of these solutions do result in the production of two species absorbing at 550 and I4.IO nm with lifetimes in the millisecond time range and exhibiting first order decay kinetics. The optical density of the 550 tran- sient increases up to (HCD'vO^ M, and then decreases rap- idly for increasing acidic solutions. In contrast, the op- tical density of the I4.IO transient in 50% isopropyl alcohol- water increases continuously from 0.02 M to 6 M HC1, and is also observed in 12 M HC1. In view of the necessity of HC1 and isopropyl alcohol for formation of the 550 transient, we assign its absorption spectrum to the radical anion generated by electron trans- fer from the chloride ion. We have already reported a D3-2 similar event for nitrobenzene and [(.-nitropyridine. Since the first order decay constants for the 550 and i+,10 tran- sients each remain constant in the acid range 0.5-3 N HG1, we believe that an equilibrium is operating between the an- ion and its protonated form. Thus, we consider it an at- tractive possibility to consider the i\.10 transient as the protonated anion. Despite the uncertain assignment of the L\.l0 transient flash photolysis data support the view that the photoreduc- tion of 5-nitroquinoline occurs via two transients. It ap- pears, however, that alcohol is not necessary for the photo- reduction of 5-nitroquinoline since the quantum yield in 12 M HG1 has been determined to be 5.7 x 10 -2 A summary of the photochemical quantum yields, optical density of the Pig. 1. -Absorption data for the 14.10(A) and the 550 nm (a) .6 transient as a function of HC1 >- concentration. t (o) =313 nm quantum yields. o X ►-GM 2 (HCL),M. 4 6 550 and I4.IO transients, as a function of HC1 concentration are summarized in Pig. 1. It is seen that the maximum opti' D3-3 cal density for the 550 transient occurs where the disap- pearance quantum yield has been observed to achieve its up- per limit of S-k x 10 . That this transient is not pres- ent in neat isopropyl alcohol or 50$ isopropyl alcohol with 6 N sulfuric acid underlines the importance of the chloride ion as an electron donor. This transient is absent in air- saturated solutions and further supports the view that it is the radical anion generated via electron transfer, i.e., ArN0 2 * 3 + CI" — > ArN0 2 "+ CI- . The lifetime of this tran- sient is 5»3 msec. Since the I4.IO absorption can be ob- served in 12 M HG1 as well, as in 50% isopropyl alcohol-wa- ter with 6 N HC1, it appears that the alcohol may not be the predominant species responsible for its formation. The lifetime of the lj.10 transient is 1.1 msec. It is evident that for solutions in which (HC1)>0.5 M, the quantum yield remains constant, but the optical density of the 550 nm transient decreases while that of the I4.IO in- creases. In the acid range less than 0.5 M the intensity of both species is increasing as is the photochemical quan- tum yield. It thus seems reasonable to conclude that both transients contribute to the observed photoreduction. The relatively large phosphorescence yield for 5-nitroquinoline suggests that its triplet-triplet absorption should be readily observed in laser flash photolysis. P3-4 PHOTOCHEMISTRY AND PHOTOPHYSICS OF NITROGEN HETEROCYCLES Marian S. Henry and Morton Z. Hoffman Department of Chemistry, Boston University, Boston, Massachusetts 02215 Aromatic N-heterocycle compounds, such as 2,2'-bipyridine (bipy) , provide an opportunity to examine the effects of varying molecular structure on the energetics and reactivities of the excited states. The nitrogen heteroatom is an intimate part of the aromatic structure and, being a Lewis base, is subject to protonation and coordination to metal ions. These latter processes, as well as substitution on the aromatic system are expected to have a significant effect on the excited states and their observed photochemistry and photophysics. This re- search is directed specifically to the establishment of structural- energetic-reactivity relationships for the excited states of these molecules. The initial phase of this research has involved the characteriza- tion of the photophysics and photochemistry of bipy. The ground state and lowest triplet state absorption spectra, and fluorescence excitation and emission spectra have been examined as a function of pH. Acid dis- sociation constants have been calculated for the ground state S and the Si, Tj., and T 2 excited states. The results are given in Table I. Re- sults for So are in agreement with those reported by McBryde Relative fluorescence intensities are 5:1: 1000 for bipy, bipyH , and bipyH| respectively. Fluorescence excitation and emission spectra show a good mirror-image relationship for bipy and bipyHi , but not for D4-1 Table I. pK values for bipy Equilibrium S a Sj^ Tj*_ T 2 C bipyH| + t bipyH + + H + -0.5 -7.76 -0.5 -1.4 bipyH + t bipy + H + 4.5 3.8 5.0 6.7 a. Calculated from pH dependence of absorption spectra b. Calculated from pH dependence of emission spectra.. „. c. Estimated from Ti absorption and Forster cycle ' Table II. State of Protonation of Excited States pH range Sjl < (-8) bipyHi (-7) - (- -2) bipyH - 3.5 bipyH ^ 4.8 bipy > 6 bipy 1l bipyH§ + bipyHf bipyH bipyH bipy D4-2 bipyH . In the ground state, bipy is known to be either cis- planar + +(2 3) (bipyH ) or trans- planar (bipyH* and bipy) ' . The mirror-image relationship between absorption and fluorescence spectra and the small Stokes' shift found for bipyH! (3.5 kK) and bipy (2.1 kK) suggest a small geometry change upon excitation. Inasmuch as the ground state (4) extinction coefficients are similar , the radiative rate constants are expected to be similar. Therefore, the decreased fluorescence intensity of bipy, relative to bipyHf , is attributed to much faster nonradiative processes, either intersystem crossing to T x or internal conversion to S . A significantly larger Stokes 1 shift (5.8 kK) and the lack of a mirror-image relationship for bipyH point to a change in geometry in the S -Si process. The effect may be due to twisting of the two pyridine rings . The pK values of Ti and T 2 are similar to those of S while those of Si, particularly for bipyHl , are highly shifted to less positive values. The sensitivity of the S x equilibrium bipyH Z bipy + H towards buffer concentration indicates short fluorescence life- times ' . Thus the position of the equilibrium is set by pH and the extent to which equilibrium is reached before fluorescence occurs is set by the buffer concentration. It is therefore possible to select the state of protonation of the two excited states, S x and Ti, which most determine the photochemistry (Table II) . By measurement of the rate of hydrogen atom abstraction from isopropyl alcohol the lowest free base triplet state has been found * (7) to be of it, it character, in agreement with Gondo and Kanda . Its D4-3 absorption spectrum is very similar to that of the biphenyl triplet . The decay of triplet bipy is second-order with no ionic strength de- pendence. The decay of triplet bipyH is slower but also second-order with an ionic strength dependence consistent with the excited state 2+ having a +1 charge. Triplet bipyH 2 decays via first-order kinetics with k = 3 x 10 3 sec x . It is apparent that a competition exists between uni- and bimolecular decay under the relatively high concentration of triplet produced by the flash. Triplet bipy does not react with halide ions but triplet bipyH reacts with I and Br , but not CI . Finally, we have found that triplet bipyH is quenched by Mn 2 but not by Zn 2 . 1. W.A.E. McBryde, Can. J. Chem . , 43, 3472 (1965). 2. K. Nakamoto, J. Phys. Chem ., 64, 1420 (1960). 3. T. McL. Spotswood and C.I. Tanzer, Aust. J. Chem ., 20 , 1227 (1967). 4. R.H. Linnell and A. Kaczmarczyk, J. Phys. Chem ., 65 , 1196 (1961). 5. N. Lasser and J. Feitelson, J. Phys. Chem ., 77, 1011 (1973); 79, 1334 (1975). 6. S.G. Schulman and A.C. Capomacchia, J. Phys. Chem ., 79 , 1337 (1975). 7. Y. Gondo and Y. Kanda, Bull. Chem. Soc. Jap ., 38, 1187 (1965). 8. G. Porter and M.W. Windsor, Proc. Roy, Soc , 245A, 238 (1958). 9. J.F. Ireland and P.A.H. Wyatt, J. C. S. Faraday I , 68, 1053 (1972). 10. H.H. Jaffe and H. L. Jones, J. Org. Chem ., 30, 964 (1965). D4-4 The Photophysics of Several Condensed Ring Heteroaromatic Compounds F. Sheldon Wettack , Robert Klapthor, Allan Shedd, Mary Koeppe, Ken Janda, Patti Dwyer, and Kathleen Stratton Department of Chemistry Hope College Holland, Michigan 49423 As a part of our interest in the photophysics of aromatic com- pounds, we have examined the series of condensed-ring heteroaromatic compounds, indene, indole, benzofuran and benzothiophene. The ring system of these compounds can be considered to be isoelectronic with naphthalene, but with less symmetry and a variable heteroatom. Pre- vious work on absorption and fluorescence and the variation in such properties as weight, aromaticity, and vibrational frequencies combine to yield an ideal series for a photophysical investigation. In particular we have been interested in learning of the role of the heteroatom in radiative and non-radiative processes from both the sing- let and triplet states. Although work directed toward this end has been carried out in both the condensed and gas phases, the results re- ported in this paper will be confined to the condensed phase. Results at 298 K In the liquid phase at room temperature the fluorescence of the series varies significantly. Quantum yields are shown in Table I. Also shown are fluorescence lifetimes measured for the same solutions and calculated values of the radiative (k~) and nonradiative (k ) rate constants from the excited singlet state. D5-1 Tabl e I Compound f x f (nsec) 0.073 15 . ISC 0.67 k £ (xl0" .49 ■ 7 ) k (xlO' nr 6.2 • 7 ) k. (xlO -7 ) ISC ' Indene 4.5 Indole 0.41 9.6 0.07 4.3 6.1 0.7 Benzofuran 0.23 7.0 0.07 3.3 11 1 Benzothiop? lene 0.014 <0.5 0.73 1.4 100 91 We have also measured the intersystem crossing yield (cf>. ) for these compounds using the sensitized 1 ,3-pentadiene isomerization tech- (31 nique of Lamola and Hammond. The results are shown in Table I. The triplet yield measurements were carried out with benzene and cyclo- -2 -3 hexane solutions at concentrations of 10 -10 M whereas the fluores- cence measurements were performed at 10 -10 M in cyclohexane. Results at 77 K Phosphorescence yields at 77 K in 2,2,4-trimethylpentane (isooc- tane) solvent and lifetimes in EPA are shown in Table II. Table II Compound 4 x (sec) k <$. (sec ) _ _j)_ p p isc Indole 0.40 5.6 7.1 x 10 2 Benzofuran 0.34 1.2 28 x 10~ 2 -2 Benzothiophene 0.23 0.44 52 x 10 It was found that the most reproducibile yields were obtained with isooctane solvent. Cyclohexane, methanol and EPA were also used but problems with solubility, fractured cells and nonreducibility were encountered. The yields shown in Table II are based on naphthalene as a standard ( = 0.051) . Experimental In all of the work, spectra corrected for emission detector sensi- D5-2 tivity were used to calculate the yields. In addition, the effective optical density over the excitation bandwidth was used to account for the relative absorption by standard and unknown. The phosphorescence spectra and lifetimes were obtained using a Farrand spectrofluorometer modified to accept cylindrical cells at liquid nitrogen temperatures. The fluorescence spectra were taken on the same instrument and all absorption measurements were made on a Cary 14. Fluorescent lifetimes were determined by the time-correlated single photon counting technique. Each quantum yield was determined several times in independent experi- ments. Reproducibilities in the fluorescence yields are better than 10% and the phosphorescence yields are reproducible to 10%. Discussion At 298 K sufficient data are available to calculate k. , the rate isc constant for intersystem crossing, using d>. = k. t^. The results / 6) 6 T 1SC 1SC f are included in Table I. Clearly the sulfur heteroatom influences the rate of intersystem crossing in benzothiophene. Otherwise k. goes approximately as the degree of aromaticity in the compounds. It should be noted that the . values were determined at higher concentrations T isc & than the fluorescence yields and lifetimes. Preliminary work with indole indicates that t~ is concentration dependent and hence the inter- system crossing yield is a minimum value. It is possible that k. for indole and, perhaps benzofuran, may be as much as an order of magnitude higher than the values shown in Table I. At 77 K the relationship = k t . holds. If one assumes that P P P isc 3 k =0, the relationship d> /& c = k. /k_, known as Kasha's intersystem nr ' r Y p' T f isc f J (4) crossing ratio, is obtained. If ~ is taken to be 0.6 for indole , P5~3 7 -1 k. is calculated to be 3 x 10 sec . If this same value for k. isc isc exists in the liquid phase at 298 K, , =0.3. The relationship /t = k d>. leads to the values of k 4. P p p isc p isc shown in Table II. These values imply that k and/or d>. are increas- r ' P isc ing in going from indole to benzothiophene at 77 K. Since the phospho- rescence yields indicate that a significant increase in . is unlike- J 6 T isc ly, one concludes that k is increasing. We find no evidence of phosphorescence from indene at 77 K in isooctane. Phosphorescence has been reported for EPA solutions. Whether this is an indication that the singlet state photophysics of indene at 77 K is isooctane does not include intersystem crossing or whether all of the energy dissipation from the triplet state is of a radiationless nature is yet unknown. An exhaustive search for phos- phorescence in isooctane is continuing. Benzothiophene, indole, and benzofuran show relatively high phos- phorescence yields compared to other hydrocarbon aromatic compounds and the values are consistent with emission yields found for dibenzo con- densed-ring heteroaromatic molecules. The fact that indene shows no phosphorescence is consistent with the low yields found in compounds which strictly involve it-it* excited triplet states (e.g., naphthalene, benzene) . (1) J.M. Hollas, Spectrochimica Acta, 1_9, 753 (1963) (2) B.L. Van Duuren, Anal. Chem. , 32, 1436 (1960) (3) A.M. Lamola and G.S. Hammond, J. Chem. Phys., 43_, 2129 (1965) (4) J. Eisenger and G. Navon, J. Chem. Phys., 5£, 2069 (1969) (5) J.B. Birks, "Photophysics of Aromatic Molecules" ' 3 Wiley-Inter- science, New York, 1970 D5-4 "Molecular Weight Dependence of Triplet -Triplet Processes in Poly- (2-Vinylnaphthalene)".* Nicholas F. Pasch and SJF\ _Wehher. Depart- ment of Chemistry, University of Texas, Austin, Texas 78712 In 1969 Cozzens and Fox demonstrated the occurence of triplet - triplet annihilation leading to delayed fluorescence in polymers of 1- vinylnaphthalene (P1VN). Since that time a number of observations of delayed fluorescence in polymer solutions and films for both homopoly- 2~5 5 mers and copolymers have been reported. Kloffer and Fischer have shown previously that the intensity of delayed fluorescence in poly(vinylcarbazole) increases with P, the degree of polymerization. It is our purpose to report some similar experimental observations of delayed fluorescence for solutions of poly(2-vinylnaphthalene) (=P2VN) at 77°K with P values from ^100 to ^3000. All polymers were synthesized using sealed ampules containing freshly sublimed 2-vinylnaphthalene in "spectro" grade benzene (**1.5 M) and initiated with AIBN (2 x 10~ 3 to 1 x 10" 2 M). The highest MW samples were obtained by eliminating the solvent from this method. All MWs were determined by viscometry using the results of L. Utrecki, 6 R. Simha, and N. Eliezer. After the initial set of experiments described herein was completed we received from Professor R. Simha two samples of anionically poly- merized poly(2~vinylnaphthalene) used in reference 7. These samples were used as received and gave no evidence of any spectroscopically significant contamination. P6XL For all experiments we used a 1:1 THF:Et 2 (tetrahydrofuran:diethyl ether) glass at 77°K. The THF was freshly distilled from lithium alum- inum hydride with triphenylmethyl radical as indicator. Diethyl ether usually was found to give a sufficiently weak phosphorescence (near 500 nm) to be used without further purification. For more recent experiments the diethyl ether was freshly distilled from lithium alum- inum hydride. A solution approximately 10" 3 M in naphthalene groups was made up with the 1:1 THF:Et 2 mixture using quartzware that had been recently pyrolyzed with a Mekker burner to remove any organic residues. Suc- cessive dilutions were made in several cases and the spectra of the 10"? 10 and 10~5 m solutions were compared. For the highest molecular weight polymers there was a significant (a*15%) increase of the delayed fluorescence intensity relative to the naphthalene phosphorescence in the 10 M solution compared to the 10" ^ M. No significant change occurred in the next 10-fold dilution except for a diminished signal to noise ratio. Hence the results we report are for ^10 M solutions in naphthalene chromophores. A phosphorimeter of our own design was operated in the traditional way. A single chopper alternately exposed the sample to excitation and then exposed a McPherson model 218 monochrometer fitted with an EMI 6255S photomultiplier to the delayed emission. Excitation and observa- tion times were approximately 2. 35 msec. For long time decay curve D6-2 experiments a pair of Vincent Associates Uni-blitz shutters was used to cut off excitation and expose the photomultiplier to the long lived emis- sion at a given wavelength, selected on the monochrometer. For all decay curve measurements the photomultiplier current was amplified by a Keithley model 427 current to voltage converter and signal averaged by a Fabri-Tek model 1060 instrument computer (manufactured by the Nicolet Instrument Co.)- Sample excitation was effected by a 200 watt high pressure Hg lamp filtered by either a Corning 7-54 filter or a 7-54 plus an Oriel 313.0 nm interference filter. of The delayed emission spectrum/these polymers consists of two fea- tures: (1) a region to the long wavelength side of 480 nm, typical of naphthalene phosphorescence, and (2) a peak around 340 nm, typical of naphthalene fluorescence. Hence triplet-triplet annihilation leading to delayed fluorescence is present in P2VN. The relative intensity of the delayed fluorescence to the phosphorescence is dependent on the molec- ular weight (MW) of the polymer. In general the higher the MW the larger the ratio IrjpAp (intensity of delayed fluorescence to phosphor- escence). We have plotted L-vp/Ip vs.V P* in Fig. 1. The choice of ■fP for the abscissa is for convenience of display. While the trend of the results is clear, individual differences be- tween polymers of approximately the same MW prevent an unambiguous determination of the functional dependence of Iop/Ip on P» Two features worth noting are: (1) the apparent extrapolation of Irjp/Ip to zero for D6-3 P f 0, and (2) the apparent "saturation" of the IrjpAp ratio f° r the four polymers with the highest MW. Observation (1) follows if one treats the number of triplets on the polymer as obeying a Poisson distribution where the average number of triplets per polymer chain is small. This treatment yields the solid line in Fig. 1. Observation (2) is expected when P exceeds the average diffusion length of the triplet exciton. How- ever, since two of the high molecular weight polymers were bulk poly- merizations, it is not clear if they should be regarded as typical of the solution synthesized polymers. Certain general features of the decay curves for all polymer samples can be summarized: (1) the delayed fluorescence decays much faster than the phosphorescence for the first 0.5 sec. After the delayed fluor- escence has decayed to approximately 0. 1 to 0.05 of its initial value, the rate of decay of the delayed fluorescence is approximately twice that of the phosphorescence for polymers with P greater than 700. This does not seem to be the case for shorter polymers but poorer sig- nal to noise ratios for delayed fluorescence in these samples prevents a firm statement from being made. (2) Both delayed fluorescence and phosphorescence display highly nonexponential decay for approximately 0.5 to 1.5 sec after cessation of excitation, at which point the former is very weak and the latter appears to be nearly exponential. The non- exponent iality for phosphorescence decay is much more pronounced for longer polymers than short polymers. (3) Only a slight dependence of P6-4 the decay curves on I is observed. This implies that the kinetic be- cX havior of the triplets is dominated by uniexcitonic processes (emission, trapping, and radiationless transitions). (4) In general the decay of delayed fluorescence is lower as the MW of a polymer increases, al- though differences in individual samples preclude a clear-cut relation from being derived. *Supported by the Robert A. Welch Foundation References 1. R.F. Cozzens and R.B. Fox, J. Chem. Phys. 50, 1532 (1969). 2. (a) R.B. Fox, T. R. Price, R.F. Cozzens and W. H. Echols, Macromolecules 7, 937 (1974); (b) R.B. Fox, T.R. Price, R.F. Cozzens, and J.R. McDonald, J. Chem. Phys. 57, 2284 (1972); (c) R.B. Fox, T.R. Price and R.F. Cozzens, Ibid 54 , 79 (1971). 3. C Helene and J.W. Longworth, J. Chem. Phys. 57_, 399 (1972). 4. C. David, M. Lempereur and G. Geuskens, Europ. R)lymer J. _8, 417 (1972) and references therein to earlier work. 5. W. Kloffer and Dieter Fischer, J. Polymer Sci., Symposium No. 40, 43 (1973). 6. L. Utracki, R. Simha, and N. Eliezer, Polymer 10, 43 (1969). D6-5 Fig. 1. The ratio I DF /Ip vs. P. The solid line is from a Poisson distribution of excitons. The closed circles are for our original samples the open circles for newer samples of our poly- mers, and the open triangles are the samples provided by Professor Simha. D6-6 Singlet And Triplet Precursors of A B. Stevens and J. A. Ors Department of Chemistry, University of South Florida, Tampa, Florida 33620. The quantum yield y of photosensitised addition of molecular oxygen to an organic acceptor M is the product of the quantum yield y of 0„ A formation and the efficiency d>.„- of o A addition to the ac- 2 MO 2 ceptor (process 1) where M + A ■> MO 1. °2 1a "*" °2 3e 2 ' d>,,_ = k., [Ml /(k, [Ml + k„) . Measurements of y„_ as a function of dis- M0„ 11 2 M0„ 1 solved oxygen concentration have been analysed to identify the opera- tive oxygen quenching process of singlet (S..) and triplet (T ) states of the sensitiser of which processes 3-6 are spin-allowed and exothermic S l + °2 3z "*" T l + °2 1a 3 ' S l + °2 3e "" T l + °2 3e 4 ' T n + oA -> S + O^A 5. 12 o 2 T n + o 3 S ■> S + o 3 Z 6. 12 o 2 if both the singlet-triplet splitting AE (process 3) and triplet energies E (process 5) exceed the excitation energy of A at 8000 cm . In terms of the parameters a = k„/(k„ + k.) and e = k c /(k c + k, ) the overall reaction quantum yield is given by equation I where y tc; is Y = a {ey + P \a + e(l-Y )1) I Y M0 V M0 l IS L V Y IS' J; the sensitiser intersystem crossing yield and the probability that S 3 is quenched by Z, P = K[0„]/(1 + K[0 9 ]) is available from indepen- dent measurements of the Stern-Volmer fluorescence quenching constant K. D7-1 Linear plots of the data y (P ) provide the quotients (Table 1) %0 2 (P 2 = ° )/Y M0 2

Wrt and equation I in the form A 'MO ? MO Y A = Y MQ (1 + 3/[M]) = ey is + P Q [a + e(l - Y Ig )] HI is used to present the experimental data Y A (P n )• Quantum yields Y Mn of self-sensitised photoperoxidation of rubrene, tetracene, 9,10-di- methylanthracene (DMA) and 9 ,10-dimethyl-l,2-benzanthracene (DMBA) were estimated from recordings of the optical density as a function of time at the actinic wavelength and the source intensity at this wavelength determined by ferrioxalate actinometry. In the case of perylene, this was used to sensitise the photoperoxidation of DMA and DMBA at 436 nm where these acceptors are transparent at the concentrations used. The data presented in Figs. 1 and 2 illustrate two behavioral patterns, viz a) Y A (P n =l)=a+e^l.O for those sensitisers with high inter- system crossing yield. Since Y A ( p n = 0) = £Y T c % Y Tq for these com- pounds (tetracene and DMBA) , it must be concluded that e ^ 1 and a ^ consistent with 4 and 5 as the dominant oxygen quenching process. D7-2 b) Y A (P n =l)=a+e^2.0 for rubrene, perylene and DMA which exhibit high fluorescence quantum yields (^0.9) and low intersystem crossing efficiencies. Since neither a nor e as defined can exceed unity, it must be concluded that a ^ e ^ 1 or that processes 3 and 5 are largely responsible for oxygen quenching of the singlet and triplet states of these sensitisers. The reverse of process 3 represents the 4 energy pooling process responsible for the oxygen enhancement of fluo- rescence under conditions where triplet state relaxation is relatively slow. Since AE exceeds 8000 cm "" for the sensitisers examined here, the nature of the singlet quenching process is essentially determined by the relative rates of non-radiative transitions from the complex formed 3 3 3 1 initially r. (S E) to lower energy complex states T (T A) and 3 (T E). If processes 3 and 4 operate by an exchange mechanism these g rates will largely depend on the relative energies of these states viz 3 r (t 3 e) > V (s 3 e) > V (t/a) k » k. gn ll fl 34 3 r. (s 3 E) > 3 r (T 3 E) > 3 r (t/a) w » k. ll gn fl 3 4 which reflect the ordering of molecular states T n > S l > T l Y IS "• ° S l > T n > T l Y IS ** ° expected for sensitisers of high and low fluorescence yields and inter- system crossing rate constants k which differ by an order of magnitude J, O (Table 1) . The quenching processes 3 and 4 should not therefore be mutually exclusive particularly for sensitisers where AE ^ 0. D7-3 References 1. B. Stevens and B.E. Algar, Ann. New York Acad. Sci., 171 50 (1970) 2. A. Kearvell and F. Wilkinson, Chem. Phys. Letters, JL1 472 (1972). 3. B. Stevens, S.R. Perez and J. A. Ors, J. Amer. Chem. Soc, 96^ 6846 (1974). 4. R.D. Kenner and A.U. Khan, Chem. Phys. Letters, %. 6 ^3 (1975). 5. B. Stevens and B.E. Algar, J. Phys. Chem., T2 2582 (1968). Table 1. Photoperoxidation Parameters in Benzene at 25°C Sensitiser 10 3 3(M) a Y IS 10 -\ s (s-h a + e 0.98 ^ 0.02 ■4 3.0 1.9 ± 0.4 0.90 •£ 0.10 < 1.5 2.0 ± 0.4 0.89 < 0.11 < 7.1 2.2 ± 0.3 0.19 0.68 b 120 1.2 ± 0.2 0.36 .£ 0.64 ^ 27 1.0 ± 0.2 Rubrene DMA Perylene Tetracene DMBA ref. 5 2.0 1.0 1.2 2.4 1.6 b ref. 2 • Rubrene O Tetracene 1.0 The Heavy Atom Effect On The Photochemical Cycloaddition Processes Of Acenaphthylene B. F. Plummer and Lou Jean Scott, Trinity University, San Antonio, Texas, 78281| Modification of photochemical processes through the influence of 1 heavy atom perturbation continues to be a fruitful area of study. Quantum mechanical interpretations satisfactorily rationalize the ef- fect of heavy atom perturbations. However, predictions of specific molecular compounds that will show a measureable change in photochemis- 2 try under heavy atom perturbation have not yet materialized. The discovery by Cowan and Drisko that the photodimerization of 3 acenaphthylene (A) showed a dramatic heavy atom effect (HAE) stimulated us to explore the potential of this perturbation when applied to the cycloaddition reactions of other molecules to A. Compounds such as acrylonitrile,oi -chloroacrylonitrile, cyclopentadiene(CP)and cis and trans - 1,3 -pentadiene were found to cycloadd photochemically to A when HA perturbation in the form of brominated solvents was used. The triplet nature of these direct irradiation processes was verified through S-V quenching studies using cyclooctatetraene and ferrocene as quenchers. The triplet energy of these quenchers lies in the range of 30-UO kcal/mole and thus establishes a lower limit for the triplet energy of excited triplet A as near ho kcal/mol. The pro- duct ratios from direct and triplet sensitized radiations were compared for A reacting with CP and found to be similar. The sensitizer used for these studies was the dicyclohexyl l8-crown-6-ether complex of di- s odium Rose Bengal. D8-1 The influence of heavy atom solvents such as bromoethane, 1,2-dibro- moethane, and bromobenzene as compared to cyclohexane and acetonitrile was studied. The quantum yield of the reaction was found to vary in a predictable way with a change in the concentration of CP. The mech- anism of this reaction was found to conform to the general kinetic expression in equation 1. 1 x ( k *\ _L (1) « W. ot , *r ** *isc ** r isc Cd] = [cp] The quantum yield of reaction is , Qi = the fraction of intermediate diradicals forming product, k,= the rate constant of excited triplet state decay of A and k = the rate constant of bimolecular formation of — r intermediate from A and CP . Using the reasonable assumption that 0. for A is unity in heavy atom solvents, a value f or of, of 0.2 was obtained from the kinetic data. Thus, about 20$ of the diradicals produced continue on to form products, the remainder apparently dissociating to form A and CP. A comparison of the external perturbation of bromine was made to internal perturbation by studying the photocycloaddition of 5-bromoace- naphthylene (ABr) to CP. Similar kinetic expressions were obeyed for A and ABr reacting with CP. However, it was learned that the internal heavy atom exerts a larger effect upon k, than does the external heavy atom. The unique capability of A to magnify heavy atom perturbation in conjunction with its relatively poor fluorescence quantum yield ( 0f< 0.001) and phosphorescence quantum yield (0p^O) has led us to D8-2 explore the photochemistry of the acetyl derivative of A. The new mole- cule £-acetoacenaphthylene (aA) was synthesized by acetylating acenaph- thene with acetic anhydride and magnesium perchlorate and then subse- quently dehydrogenating the saturated precursor with DDQ. Single pho- ton emission studies of aA showed only a very weak fluorescence near UOO nm and no detectable phosphorescence. The absorption spectrum of aA resembled that of A except for a larger absorptivity coefficient at similar regions of maximum absorption. The yellow color of the com- pound suggests that the onset of the Tr-^ / Tt transition is of lower energy than the n-»1r and thus explains our inability to detect a n-*Tf transition in its absorption spectrum. We have studied the excited state cycloaddition of aA to CP in regular and heavy atom solvents. The photoproducts formed are analog- ous to those formed between A and C_P, as verified by gc product distributions, nmr and ir measurements, and mass spectroscopic analysis. We have studied the direct and sensitized product distributions of the reaction of aA with CP and find that they are similar. Addition- ally, the direct reaction can be quenched with ferrocene. The kinetic studies of aA reacting with OP have been investigated and data will be reported that suggest an intermediate ro2.e of the carbonyl group in aA in effecting spin-orbital perturbation of the reactions of aA. References 1. (a) B.F. Plummer and W. W. Schloman, Jr., J. Amer . Chem . Soc , in press, (b) D. 0. Cowan and J. C. Koziar, ibid ., ?8, 1001, (1976). (c) A. R. Gutierre and D. G. Whitten, ibid., %j 7128 (197h). 2. G. A. Giachino and D. R. Kearns, J. Chem . Phys ., $1? 2 ?6U (1970). D8-3 3. D. 0. Cowan and R. L. Drisko, Tetrahedron Lett . 12#? (1967). U. B. F. Flummer, W. I. Ferree, Jr., and W. W. Schloman, Jr., J. Amer . Chem. Soc , 96, 77Ul(197U) and references therein. P8-4 Laser Induced Fluorescence Emission Spectroscopy of H CO(A, A Kenneth Y. Tang and Edward K.C. Lee Department of Chemistry University of California Irvine, California 92717 The single vibronic level (SVL) radiative lifetimes (r„) of the K ,~1 N 1 first excited singlet state of H CO(A A ) have been studied recently. The most interesting finding was a large variation of the radiative lifetimes showing a general trend of increasing radiative lifetime with increasing vibrational energy (E . ) and expectedly shorter values of t for the SVL levels with one quantum excitation of the asymmetric C-H R stretch mode (v ) . More specifically, the radiative lifetime of the 5 level at E . = 2968 cm , T (5 ) was found to be~0.7 x 10 sec, several vi b R times shorter thanr_(2 k ) = 6.2 x 10 sec, at E . = 2U71 cm '" and R vib p Q (i -| t_(2 k ) = 11 x 10 sec at E .. = 33^3 cm . Since the non-totally R vib symmetric, out-of-plane puckering mode ( v. ) is much more active than v in the absorption and in the emission spectra for electric dipole for- bidden but vibronically allowed transition ( A - A ) , this behavior of the radiative lifetime is very surprising. In order to establish the extent of the Herzberg-Teller type vibronic coupling in the radiative transitions from the 5 level, we have decided to determine quantita- tively the activity of the v mode in the molecular resonance fluores- cence spectra from several single vibronic levels at low pressures (nearly collision free). The most interesting results have been obtain- -1 m ed with the 5 and the 1 h levels which are nearly degenerate (2.7 c and perturbed by Coriolis coupling. * This research has been supported by the National Science Foundation and the Office of Naval Research. El-1 For the purpose of illustration, the spectra taken at 320T.0 S(vac) and 3215.1 A(vac) excitation with 0.5 torr of H CO at 23 C are shown in Figure 1. The collisionally relaxed spectrum obtained at X (vac) = ex 3207-0 A and 5-0 torr of added N is shown in the middle for comparison. The pumping was achieved with a flashlamp-pumped dye laser equipped with frequency doubling crystals and fine tuning etalons, and the fluores- cence was detected with an optical multichannel analyzer. Despite our attempt to obtain an SVL spectrum from the "selective" excitation of the ro-vibronic level of either 5 or 1 U , we were unable to obtain a "level" purity better than 80% as you can see from Figure 1. 1 1 R We beleive that the 3207.0 S excitation corresponds to the 1 k Q (J'= 13. K' = h <4- J" = 13, K" = 3) transition and 3215.1 8 excitation cor- responds to the 5 1 ? Q^ (J' = 7, K' = 2 «- J" = 7, K" = 3) transition and the overlapping 5 1 P P-, (J 1 = 2, K' = 2 «- J" = 3, K" = 3) transition. In examining the SVL emission spectrum we find that the intense f bands in the emission involve the odd quantum number change in the out- of-plane bending mode (*v) and the progressions are formed by the C-0 stretch mode ( v ) . Hence, the emission spectrum from the 5 level is similar to the one from the 1* level, showing always the 5-, contribu- tion as if v is the "frozen" mode. Likewise, the emission spectrum from the 1 h level is also similar to the one from the h level, again showing the 1 contribution as if K. is the "frozen" mode. The most intense bands from the 5 level are k 5, , 2 k 5, and ^ 5 » where as the most intense bands from the 1 k level are 1,U and 1 2 h . Fluorescence gives spectra whose distinctive features characterize the emission from the levels 5 and 1 h . This makes us to believe that Ei-2 the intramolecular energy flow between these two vibronic levels is slow Q compared to the time scale (10 sec) of the unimolecular decay processes (mostly non-radiative), although these two states are nearly degenerate with 2.7 cm of separation. The fluorescence intensity I from an SVL is proportional to the radiative decay rate constant (k ) and the population (N). Since the 5 11 -1 and the 1 h levels are ~ 2.7cm " appart , the Boltzmannized population 1 11 ratio at room temperature should be very close to unity: N(5 )/N(l h = 1.013. In the presence of 10~100 torr of N , the 3207.0 8 (5 1 ) and the 3215.1 A (l h ) excitations give an identical "equilibrated" spectrum consisting of the emission from the two SVL's as shown in Figure 1 (middle). Considering the lifetimes (~15 nsec ) and the pres- sures used, the collision induced vibrational relaxations to other v ibronic levels are negligible and that the 5 *WAr¥± h relaxation is considerably more efficient as in rotational relaxation . The intensity ratios, I„(5 )/I_(l h ), obtained from the two excitation wavelengths as r r a function of N pressure gives a limiting value of ~1.8 at high pres- sures. Therefore, k (5 ) is 1.8 times greater than k_(l k ). Using the K r\ value of k (5 ) = 1.3 x 10 sec "'" obtained previously we calculate k K ' K (l h ) to be ~ 0.7 x 10 sec . This value is greater than k (h ) = R 0.5 x 10 sec obtained previously. In conclusion, the present data clearly indicate that the radiative transitions (even from the 5 level) show insignificant involvement of the v mode for the Herzberg-Teller type vibronic coupling but yet the overall radiative rate from the 5 level is 1.8 times greater than that from the 1 k level. Also, it appears that k (l k ) is also l.k times El -3 as great as kp(^ )• We believe that further theoretical advance should shed more light on the significance of the "frozen" mode in enhancing the radiative transition rates. References : 1. R.G. Miller and E.K.C. Lee, Chem. Phys. Lett., 33, 10U (1975). 2. K.Y. Tang and E,K,C, Lee, submitted for publication to Chem, Phys, Lett. El -4 5 level X = 3215-1 fi (vac) ex H 2 C0 =0.5 torr X ex = 320T *° ? ( vac ) H 2 CO =0.5 torr N = 5.0 torr » 1 U level X = 3207. £ (vac) ex H CO =0.5 torr ■> — r t r t r 370 U00 U30 Figure 1 . Emission Spectrum U60 (nm) El-5 "Laser Excited NO2 Fluorescence Lifetime Studies in the 600 nm Region" V. M. Donnelly and F. Kaufman, Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260. A partial analysis of the absorption and fluorescence spectra of N0 2 2 indicates that the B.. state is responsible for much of the absorption -11 -12 above 23,000 cm , while below 19,000 cm ' the B. state carries vir- 2 3 tually all the oscillator strength. ' The visible absorption spectrum is extremely complex due mainly to extensive perturbations of the ex- cited states by high vibrational levels of the ground state. This fac- tor also accounts, at least in part, for the fact that both excited states exhibit fluorescence lifetimes much longer than the 0.3 ysec order of magnitude value calculated from the integrated absorption co- efficient. Broad band modulated excitation studies have reported fluorescence lifetimes of 50 to 90 ysec in the 400-600 nm region. Nar- row bandwidth pulsed laser experiments have shown that in the 451-461 nm region , decays are slightly non-exponential with average lifetimes of about 80 ysec, while at 593 nm, decays are very non-exponential due 3 to at least two components with lifetimes of 30 and 115 ysec. We re- port here the results of laser excited fluorescence lifetime experiments in the 578-612 nm region. At pressures as low as 10 torr decays are highly non-exponential for some excitation wavelengths , and nearly exponential for others. These results can be explained in terms of a 2 single electronic state ( B_) vibronically coupled to high vibrational levels of the ground state. Fluorescence was excited by a Nd-YAG pumped tunable dye laser with a bandwidth of either 1.5 or 0.15 cm . The fluorescence was detected through cutoff or interference filters by a photomultiplier tube which E2-1 observed the central portion of a 72<£ spherical fluorescence cell. 4 5 Fluorescence decays were summed over 10 to 10 laser pulses on a multi- channel analyzer with 10 ysec per channel time resolution. Nonexponen- tial decays were fit to the expression I(t) = Re" t/T S + e" t/T L (1) Very good fits were obtained in all cases. However this does not prove that decays are simple biexponentials , so the parameters x , x and R should be interpreted with caution. Decays were recorded as a function of excitation energy and pressure (P < 25 mtorr) . For excitation coin- 2 2 cident with the strong B 2 ■*• A absorption at 585, 593, and 612 nm, the decays are highly non-exponential. Plots of x and x vs . P extrapolated to zero pressure give "lifetimes" in the 22-75 ysec range for x and 170-260 ysec for x . The slopes of the x vs . P and x O Li o Li -10 -10 3 -1 vs. P plots are typically 5 x 10 and 0.75 x 10 cm sec , respec- 2 tively. When the laser wavelength does not correspond to strong B features (578, 594, 603 nm) the decays are nearly exponential at all pressures. The slopes and intercepts of the x vs. P plots at these wavelengths are in the same range as the x vs . P plot slopes and in- tercepts at 585, 593, and 612 nm. R in equation (1) varies from ~ to 3 over the 578-612 nm region. Other experiments show that x and espe- cially R vary strongly for small changes (0.15 cm ) in the excitation energy. The largest R and the shortest x correspond to strong features in the fluorescence excitation spectrum. 2 3 The overwhelming spectroscopic evidence J against the existence of more than one strongly absorbing electronic state in this wavelength region leads us to believe that these results are due to variable vi- E2-2 2 bronic coupling between the B„ state and upper vibrational levels of the ground state as suggested by Douglas. Excitation at 578, 594, and 603 nm is due to high levels of the ground state which lie close enough 2 to levels of the B~ state so that they can borrow oscillator strength and make transitions possible. Absorption to and fluorescence from 2 such levels are weak since they are mostly A in character and the transition is forbidden; hence the fluorescence lifetime will be much longer than that calculated from the integrated absorption coefficient. 2 At 585, 593, and 612 nm the upper state has more B^ character due to 2 the coincidence of these wavelengths with positions of zero-order B ? 2 levels. Even so, the state populated is still mostly A because of vibronic coupling. So at these latter wavelengths the average lifetime would be expected to be shorter than at 578, 594, and 603 nm, but still much longer than the integrated absorption coefficient value. Previous attempts to explain the lifetime anomaly by electronic state mixing were not successful because the ratio of the density of ground state to excited state vibrational levels is only about a factor of ten (at ~ 600 nm) leaving an unexplained factor of about 30. However this cal- culation was based on a single excited state, whereas it has been dem- onstrated that over the entire absorption region, two electronic states are excited. In order to get a more reasonable estimate we have assum- 2 ed, first, that the B state is responsible for all absorption at energies above 25,000 cm ; second, that all absorption at energies -1 2 less than 18,000 cm is due to the B„ state; and third, that absorp- 2 -1 tion by the B„ state is symmetric about a maximum at 20,000 cm J where it is responsible for 70% of the total absorption. We thereby calculate £2-3 a lifetime of roughly 3 ysec for the 2 B ? state. When this is multi- 2 plied by a level density ratio of 10 the lifetime anomaly for the B_ state is essentially removed. The radiationless transition theory of Bixon and Jortner predicts non-exponential fluorescence decays of molecules such as NO- due to vi- bronic coupling. The present results will be discussed in terms of this theory. A. E. Douglas, and K. P. Huber, Can. J. Phys . 43, 74 (65); J. L. Hard- wick and J. C. D. Brand, Chem. Phys. Lett. 21, 458 (73); G. I. Senum, and S. E. Schwartz, Chem. Phys. Lett. 32, 569 (75). 2 K. Abe, F. Myers, T. K. McCubbin, Jr., and S. R. Polo, J. Mol. Spec- trosc. 50, 413 (74); J. C. D. Brand, J. L. Hardwick, R. J. Pirkle, and C. J. Seliskar, Can. J. Phys. 51, 2184 (73); R. E. Smalley, L. Wharton, and D. H. Levy, J. Chem. Phys. 63, 4977 (75). 3 C. G. Stevens, M. W. Swagel, R. Wallace, and R. N. Zare, Chem. Phys. Lett. 18, 465 (73). 4 L. F. Keyser, F. Kaufman, and E. C. Zipf, Chem. Phys. Lett. 2_, 523 (68); S. E. Schwartz and H. S. Johnston, J. Chem. Phys. 51, 1286 (69); L. F. Keyser, S. Z. Levine, and F. Kaufman, J. Chem. Phys. 54_, 355 (71), P. B. Sackett and Y. T. Yardley, J. Chem. Phys. 57, 152 (72). A. E. Douglas, J. Chem. Phys. 45, 1007 (66). M. Bixon and J. Jortner, J. Chem. Phys. 50, 3284 (69). E2.-4 o + Laser Fluorescence Studies, Including The B 3 II(0 ) States Of BrF, IF And ICl . By: Michael A. A. Clyne , Alan H. Curran and I. Stuart McDermid. Department of Chemistry, Queen Mary College, Mile End Road, London El 4NS, England. A number of new laser excitation spectra, mostly of transient molecules, are reported. These include:- (a) the B J II(0 ) - X 1 E trans- itions of BrF, IF and ICl (440-630 nm) ; and (b) the A 3 II - X X Z trans- ition of SO (259-261 nm) . We also report that laser excitation of CIO radicals is unobservably weak in the most favourable (263-280 nm) region of its A 2 II - X 2 II system; fluorescence is therefore unlikely to be use- ful as a monitor for stratospheric ClO. In this paper, we emphasize results of spectroscopic studies of the Q + 1 + B J II(0 ) - X X E visible transitions of BrF, IF and ICl, using laser fluor- escence. Rotational constants for the excited B state of BrF (8^v'$3) have been measured for the first time (in collaboration with J.A.Coxon) . o + An interesting predissociation at v'=8, J' -29 of BrF B 3 II(0 ) has been identified. Results on this predissociation, and hopefully also on the lifetimes of the excited state, will be presented. Strong fluorescence by the transient molecule IF also was observed over a wide range of excited B state vibrational levels, up to v'=ll, J'=45, where the onset of predissociation is confirmed. The source of the transient BrF and IF molecules were the rapid reactions (1) and (2) carried out in a discharge-flow system [e.H. Appelman and M.A.A Clyne, JCS Faraday I, 1975, 71_ , 2072^:- E3-1 F + Br 2 -> BrF + Br . . . (1) , F + IC1 ■*■ IF .+ CI .. . (2) . Observation of fluorescence from the B state of ICl, which is very weakly bound, was surprising. Because of the unexpectedly short life- o + time of the B°II(0 ) state, fluorescence to the ground state from the normally dominant, but long-lived (^ 100 ysec) A 3 II(1) state of ICl, could be quenched out almost entirely by using ICl pressures near 5 Torr, In this way, the overlapping strong A-X fluorescence could be virtually completely suppressed. Energy levels up to v'=2, J' ^70, of I 35 Cl o + B°II(0 ) were observed in fluorescence in the B-X transition. The dominant transition at 300 K were the 2,3; 1,3; 2,4 and 1.4 bands. Absence of fluorescence from v'=3, which is sharp in absorption, indicates that the B state of ICl is completely predissociated above v'=2. RKGC potential energy curves and Franck -Condon factors for the B and X states of IF, ICl and IBr will be presented. The absence of o + detectable fluorescence from the B 3 n(0 ) state of IBr will be dis- cussed in the light of these results. The non-collisional lifetimes of the B°II(0 ) states of the halogens and interhalogens will be discussed, in relation to the corresponding B-X absorption spectra. E3-2 E3-3 Laser- induced Fluorescence Of CN Radicals Produced By Photodissocia- tion Of RCN Regina J. Cody and Michael J. Sabety-Dzvonik*, Astrochemistry Branch, Laboratory for Extraterrestrial Physics, NASA/Goddard Space Flight Center, Greenbelt, Maryland 20771 * NAS/NRC Research Associate Laser-induced fluorescence spectroscopy has become a useful tool in probing free radical processes in photodissociation, flames, elec- tric discharges, etc. With this technique measurements have been made of radical properties and formation processes and of the internal energy content of product radicals. In addition to laboratory dis- sociative processes, radical emissions are also widely observed in astronomical spectra, including those originating from comets. Gen- erally, the first radical emission observed as a comet comes toward the sun is the Violet Band system of CN. The source mechanisms for many of these cometary radicals are unknown. It has been postulated that photodissociation of appropriate parent molecules by solar radia- tion could be a source for these radicals. In this work we will re- port our observations on the CN radicals formed by the VUV photodis- sociation of cyanoacetylene (C^HCN) , which has been seen in interstel- lar space, and of methylcyanide (CH-CN) and HCN, both of which were observed in the atmosphere of Comet Kohoutek. A schematic of the flash photolysis-laser induced fluorescence apparatus, which has been described previously in detail, is shown in Fig. 1. Two flashlamps with different spectral outputs are coupled E4-1 with various UV filters to dissociate the parent compounds in different absorption regions above 135 nm. One flashlamp is a commercial xenon lamp with a spectral output above 190 nm, while the other is an argon lamp (~ 0.6 /is FWHM, 0.8 - 1.0 joule, 50 pps, pressure < 1 atm) with a spectral output above 135 nm (short wavelength cut-off of SrF_). The spectral outputs of both lamps were measured with aim normal incidence spectrometer using a sodium salicylate coated photomultiplier tube. The time delay between the firing of the flashlamp and the triggering of the N„ laser pumped dye laser can be varied from 2 jus to 20 ms. The 2 + CN(X E ) fragments produced by the photolysis are excited by the dye 2 + laser output to the B 2 state via the Violet Band system, Au = sequence. The fluorescence radiated by these CN(B) radicals is de- tected by a photomultiplier tube after passage through an holographic grating monochromator which acts essentially as a broad band inter- ference filter (FWHM ~ 10 nm) . The number of collisions which the radical undergoes between formation and detection is controlled by the time delay and the total gas pressure. Both pure parent gas samples and mixtures of RCN in N- or an inert gas are photolyzed. Pure samples of CLHCN were photolyzed at a pressure of 0.04 torr and a time delay of 3 jus, which correspond to a collision number of < 2 for a kinetic temperature of 300 K. The spectrum of the CN radicals showed predominantly the 0-0, 1-1, and 2-2 bands with strong bandheads for the 0-0 and 1-1 bands. After correction for the variation in laser intensity with wavelength, the intensities of the rotational lines in the R branches of the 0-0 and 1-1 bands were computer fitted to the Maxwell -Bo ltzmann equation. Preliminary results yield temperatures of E4-2 ~ 1800°K and ~ 1400°K for the u = and u = 1 levels, respectively. The areas under the rotational distribution curves were integrated and ratioed to give the relative vibrational population distribution of N Q :N :N 2 = 1.0:0.50:0.32. 2 Previous work has shown that CN formed in the A( II.) state, u = level, can be collisionally induced to transfer energy into the X state, U = 4 level, through near energy resonances between certain rotational states in the A and X states. This energy transfer results in an en- hancement of the intensity of the 4-4 band at high collision number. Photolysis of high pressure (P = 50 torr) mixtures of » -s z; J L 7 P" H ft> c l-i h- / \ H- H l- 1 o o 3 03 CO \ t. -J o -f- K3 fD — -■ '— rt ft) V / O i- 1 Ln a^ 3 O ft) LO CTv rt H rr to 0) h- 1 3 3 3 3 W 3 i! 03 H 03 ss > ? JL N 1 ro rt |; ! l x d en ►d H Ui • N f C 03 n (D < > ^__J to a \ / _4 i Kj > IT" 1 03 CO ft) H E5^-4 CO U cO H T3 0) O Xf C CXM-I CJ co a rt cu 0J -a 43 » CU g"g O CJ M <4-l •H cu O LO 4J S-- H •H u U co O CO o o "d CM g CJ 4-> cO vO CO B £ a 4-1 .-< CU to to cO O- XI 4-1 CO CO •H CU i-i iH •H o- n 0) £ CU O 0) • -H e • • O 0) 4J pj }-< C K -H p 01 CD CO • IS o J^ C e LO o co U CO c • a, a) o s-i m j^ 4-1 4J •vt CO j-i o © CU 3 o- JOO CO v£> £ A^Tsua^ui souaosaaonxa E5-5 Spectroscopic and Photochemical Studies of Gaseous Ions Using Tunable Dye Lasers. John R. Eyler , Department of Chemistry, University of Florida, Gainesville, Florida 32611. In recent years tunable laser sources have been used in ever-increas- ing numbers for both spectroscopic and photochemical studies of neutral species. However, there have been few applications of these powerful, essentially monochromatic light sources to the study of ionic species. Our work involves a wide ranging study of the irradiation of gaseous ions by tunable organic dye lasers, with the belief that results of importance to both ionic spectroscopy and ionic photochemistry will be forthcoming. Investigations to date have been carried out in the areas of photodetachment of negative ions (A + hv ->■ A» + e ) and photo- dissociation of positive ions (A + hv ■> B + C) . These studies use a pulsed ion cyclotron resonance (icr) mass spec- 1-2 trometer to generate either positive or negative ions by several mechanisms and to trap them for periods as long as several seconds in duration. During this time the ions are subjected to tunable irradia- tion from a flashlamp-pumped dye laser. The effect of this irradiation upon the trapped ions is greatly enhanced by use of an intracavity 3 technique which places the icr cell inside the laser cavity. Such a technique gives improved sensitivity when examining processes of low cross-section and threshold behavior. Recent experimental work in the area of negative ion photodetachment has been carried out by Brauman and co-workers using an icr spectro- 4 meter and a conventional arc lamp-monochromator illumination source. E6-1 Species such as NH , PH , AsH , SeH , C H , C.H.N , CHO , and C,H_S have been investigated and improved values of electron affinities 6 5 and related thermochemical quantities for these species and associated neutrals have been obtained. A value for the spin-orbit coupling con- stant of SeH* was determined from an analysis of the relative cross- section vs. wavelength curve for SeH photodetachment . Tunable laser irradiation has been used to photodetach negative ions in a beam ex- periment by Lineberger and co-workers. Use of laser irradiation allows much higher precision in electron affinity determinations, and makes possible observation of two-photon effects. Our initial research using the laser-icr technique involved the photo- — ft — — detachment of SH ions (SH + hv -> SH« + e ). This work demonstrated that fine structure unobserved in previous SH photodetachment studies could be obtained when using this technique which combines production of ions with relatively low rotational "temperatures" with the in- herently high resolution of the tunable dye laser. Current work in- volves study of photodetachment of species such as SD , Cl 9 , and C 2 H". A majority of the ionic photodissociation studies reported in the past several years have been carried out by Dunbar and co-workers using an icr spectrometer and arc lamp illumination. These have pro- duced much valuable information about the thermochemistry, energy levels, and structures of various gaseous ions such as NO , C.H , H o C, n H , , and CH_C1 . In addition, these workers have studied the chemistry of photo-produced ions in selected systems. Recently, other 8,9 groups have also reported photodissociation results. E6-2 Our first experiments in the area of dye laser-induced photodissoc- iation involved C-,H (parent) ions in gaseous toluene (C^H + hv ■* / o / o C 7 H + H»). The observed onset and general shape of the dissocia- tion spectrum were in quite good agreement with an earlier study of much lower resolution. The ease of studying both negative ion photo- detachment and positive ion photodissociation with the laser-icr technique allowed a comparison of the cross-section for C H photo- / o dissociation versus that for SH photodetachment under quite similar experimental conditions. This resulted in a value of (490 nm) = — 1 ft 7 + 1.01 ±0.24 x 10 cm for the absolute cross-section for C H photo- / o dissociation at 490 nm. No significant fine structure was seen in the photodissociation spectrum of C.H in spite of the high energy resolution of dye laser irradiation. We are currently investigating a number of positive ions of simpler structure, in hopes that fine structure related to vibra- tional energy levels in the pertinent electronic states of the ions involved will be observed. Two of the species being studied are C H Cl + and CH0CH0 + . References 1. Recent reviews include J.D. Baldeschwieler and S.S. Woodgate, Accounts Chem. Res. 4, 114 (1971), J.L. Beauchamp, Ann. Rev. Phys. Chem. 22, 527 (1971), and J. M.S. Henis in "Ion-Molecule Reactions, Vol. 2", Ed. by J.L. Franklin, (Plenum Press, New York, 1972), Ch. 9. 2. R.T. Mclver, Jr., and R.C. Dunbar, Int. J. Mass. Spectrom. Ion Phys. I, 471 (1971). 3. J.R. Eyler, Rev. Sci. Ins t rum. , ^5, 1154 (1974). E5-3 4. J.H. Richardson, L.M. Stephenson, and J.I. Brauman, J. Amer. Chem. Soc. 97, 296 (1975), and references contained therein. 5. W.C. Lineberger in "Chemical and Biochemical Applications of Lasers, Vol. 1" Ed. by C.B. Moore (Academic Press, New York, 1974), Ch. 3. 6. J.R. Eyler and G.H. Atkinson, Chem. Phys. Lett., 28(2), 217 (1974). 7. R.C. Dunbar in "Interactions Between Ions and Molecules" Ed. by Pierre Ausloos (Plenum Press, New York, 1975), p. 579. 8. B.S. Freiser and J.L. Beauchamp, J. Amer. Chem. Soc. 96^, 6260 (1975). 9. M.L. Vestal and J.H. Futrell, Chem. Phys. Lett. 2% (4), 559 (1974). 10. J.R. Eyler, J. Amer. Chem. Soc, in press. 11. R.C. Dunbar, ibid., 95, 472 (1973). E6-4 Photodissociation Spectra Of Positive Ions With Time-Of-Flight Analysis Timothy F. Thomas* and John F. Paulson Air Force Geophysics Laboratory (LKB) Hanscom AFB, Bedford, MA 01731 One of the many new opportunities provided photochemists by the development of tunable lasers in the visible and UV regions is the study of very dilute systems such as well-defined ion beams. Recent applica- tions include measurement of photo-detachment spectra of numerous nega- tive ions, photo-destruction spectra of polyatomic negative ions such - (2) as CO and , and determination of kinetic energy spectra upon + +(3 4) photodissociation of H and D ' . We have assembled an apparatus capable of determining both absolute photodissociation cross-sections of ions as a function of wavelength (the photodissociation spectrum) and the kinetic energy spectrum of product ions by time-of-f light analysis. Figure 1 shows the apparatus used. A pulsed laser beam, ^2 1/2 mm in diameter, crosses a mass-selected ion beam whose energy can be varied from 10 to 200 eV by the decelerating plates shown. The quadrupole mass analyzer is set to pass only the product ion formed during a phot- olysis, or to monitor the primary beam intensity intermittently. A photo-diode triggers the time-of-f light logic at the start of each laser pulse so that the time between the pulse and arrival of a product ion at the electron multiplier can be measured to an accuracy of 10 nanoseconds , if desired. The polarization rotator shown in the laser beam permits determination of the direction of the electronic transition moment, as well as checking whether product ions ejected at large angles to the flight axis are accepted by the quadrupole mass analyzer. E7-1 Partial results obtained for three reactant ions are given below. D . The only transition accessible at the available wavelengths is lsa -> 2PO . Figure 2 confirms that the dipole moment of the transition g u lies along the internuclear axis. Figure 3 shows the time-of -flight + spectrum obtained at primary ion energies so low that only D ions ejected in the forward direction are accepted by the quadrupole mass analyzer. Product ions produced from different vibrational levels of the lsa state possess different kinetic energies and thus show dif- ferent flight times as indicated on the figure. To obtain absolute cross sections the primary ion energy was in- creased to 173 V so that all product ions were accepted by the quadru- pole analyzer. Average values obtained are shown in Table I, along with the photodestruction results of von Busch and Dunn who used a o o band width of 200 A versus <^ 1 A used here. Table I. Cross Sections for D + hv ■*■ D + D A (nm) = 445 585 590 595 600 a x 10 19 (cm 2 ) = 9.5+0.7 5.3+0.3 5.1+0.2 5.7+0.3 5.2+0.3 von Busch S Dunn: = 9.9 5.8 5.7 5.6 5.5 (interpolated) HD + . Following the same electronic transition as in V> 2 ., two dissociative processes are possible: + + ~ ~ ~-19 2 HD + hv -*■ H + D , a = 2.0 x 10 cm + -19 2 +H+D , a= 2.0x 10 cm The cross sections shown are average of several results at 590 nm, using a source pressure of 0.62 Torr, ionizing voltage of 200 V, and primary ion energy of 173 eV. F.7-2 E. I. Ion Source = *-Sample Pyroelectric Energy Meter Decelerating Lens System n,, \ inn inn 6" Sector Magnetic Analyzer Polarization Rotator Pulse S Trigger Generator Start Time of Flight Logic Event-'' Photodiode Multi- Channel Analyzer Amplifier Discriminator J3-T Electron Multiplier i "-HFaraday Cup 50 cm Quadrupole rJj Mass Analyzer Tunable Dye Laser Vibrating Reed Electrometer Figure 1. Apparatus for photo-dissociation of ion with time-of -flight analysis of product ions. 30 60 £polor td, ' rM,) Figure 2. Effect of direction of laser beam polarization on cross section for photo-dissociation. 9-1 = corresponds to hori- zontal polarization, ion energy: 23 V. Primary E7-3 36 r 30 » 24 c =J o o - 18 E 12 Laser Pulse Dj hv - 14 II I I D + + D 8 6 5 I • • • • • • • • •• • • 1 1 • • m 20 25 30 Time (microseconds) 35 w 40 Figure 3. Time-of-f light spectrum for D + . X. = 580.0 nra, primary ion energy = 17.5 V, source pressure =0.64 Torr, ionizing voltage = 200 V. Quad^upole rods were biased at ^ + 8.5 V. E7-4 NO . Table II summarizes the data obtained for the process: NO + hv -*■ NO + N. Table II. Photodissociation Cross Sections for NO 2 X (nm) = 640 445 322.0 300.2 a x 10 (cm 2 ) = 0.013 0.05 4.3+0.8 0.5+0.2 a) Source Pressure = 0.82 Torr, ionizing voltage = 70 V, primary ion energy =173 eV. The result at 322.0 nm was found to be at a very distinct peak, half- width of ^ 2 nm, in the photodissociation spectrum. Photoelectron spectroscopy of NO suggests that this wavelength corresponds to the transition X 2 tt(0,0,0) -»■ A 2 E + (0,0,1) in which one vibrational quantum is gained by the anti-symmetric stretching mode (2460 cm ) . Recent photoelectron-photoion coincidence spectroscopy of NO indicates predissociation in the A state begins when one vibrational quantum is in the symmetric stretching mode (1350 cm )• Approximately 40% of the ions in the (0,0,1) level are reported to dissociate, and the remainder presumably fluoresce. The low cross sections at the two longer wavelengths and the failure to observe a significant change in a (322.0) upon addition of N to increase the ion source pressure to 1.9 Torr both indicate that most of the NO ions enter the photolysis region in their ground vibrational states. * NRE Senior Research Associate - On leave from University of Missouri- Kansas City, 1975-6. E7-5 References 1.) W.C.Lineberger, et al. , J. Chem. Phys. 61, 1300 (1974), and ref- erences therein. 2.) J.R. Peterson, et al., J. Chem. Phys. 62_, 4826; 63_; 1612 (1975). 3.) N.P.F.B. van Asselt, J. G. Maas , and J. Los, Chemical Physics 5_, 429 (1974). 4.) J.B. Ozenne, D. Pham, and J. Dump, Chem. Phys. Lett. 17_, 422 (1972) . 5.) F.V. Busch and G.H. Dunn, Phys. Rev. A5_, 1726 (1972). 6.) C.R. Brundle and D.W. Turner, Int. J. Mass Spectrom. Ion Phys. 2_, 195 (1969). 7.) J.H.D. Eland, Int. J. Mass Spectrom. Ion Phys., 12_, 389 (1973); 13, 251 (1974). E7-6 Vibrational States Of Molecules In The Visible Region By Thermo-optical Spectroscopy and the Local Mode Model A. C. Albrecht , Department of Chemistry Cornell University, Ithaca, New York 1U853 When a very small fraction of an incident laser beam is absorbed by a liquid, a steady state temperature gradient on the order of millj degrees can appear transverse to the propagation direction of the incident light. This produces an index of refraction gradient causing the transitting beam to diverge. When a tunable laser is used for the heating beam, this thermal lens effect provides a basis 1 for an extremely sensitive new spectroscopy. We have explored the absorption characteristics in the red (Rhodamine 6-G) region of the visible spectrum of several characteristic organic molecules. These include several aromatic hydrocarbons, methanol, acetone, methylene- dichloride, and several alkanes. Isotope studies show that absorption by these molecules is derived from stretching excitations of bonds containing the hydrogen atom. The excitation energy is more than fifty percent of that needed for bond dissociation. Thus this spectro- scopy offers an opportunity to explore molecular potential energy surfaces along reaction coordinates and at energies approaching those required for breaking bonds. The spectroscopy itself has been developed both in a single beam transient mode as well as in a dual beam, automatic scanning mode. In the latter case a fixed wavelength laser is used to probe the lens E8-1 which is formed by a chopped,tunable, heating beam. Synchronous detection is used. In the single beam mode the time course of the developing lens is followed from the first moment the sample is illum- inated to the point where the stationary temperature gradient has appeared. An absolute measure of the extinction coefficient can be made in this case. Extinction coefficients nine orders of magnitude weaker than those of typical allowed electronic transitions can easily be measured. Samples having optical thicknesses of kilometers give easily detected thermal lensing effects in centimeter path lengths. Singlet-triplet transitions are candidates for such spectro- scopy. One such transition has been identified in anthracene. The theoretical understanding of vibrational molecular states responsive to light in the visible region appears to be greatly facilitated by the local-mode model (IM). Here the zeroth order description of an excited state is not given by the usual overtone or combination (normal mode) approach but rather by expressing the occupation number of local mode anharmonic vibrators. Thus in benzene 2 for example , the 1M model pictures the C-H stretching states to be built up of six identical, uncoupled, anharmonic C-H oscillators. The IM model can incorporate any degree of anharmonicity in its zeroth order framework. Since it is an uncoupled system model, and since the perturbation due to the light wave is also separable, the Ui model predicts that multiple quantal excitation can be one -photon allowed only when they take place within one bond. The model predicts the very simple spectroscopy of a one dimensional anharmonic oscilla- tor for all degrees of excitation even though the state density rises E8^ 2 abruptly. The number of states for j identical oscillators containing altogether v quanta is just the binomial coefficient ((d+v-l)/(j-l)). Thus for the six dimensional C-H stretching space in benzene (3=6) there are 1*62 states at v=6. The observed line however is that of a single state - that (of E-^ symmetry) which carries all six quanta in one bond. This is by far the most anharmonic of the h6Z states. The JM model can enumerate all levels appearing at any v once a two para- meter fit is made to the observed 'overtone' spectrum. The correspond- ing energy level diagram for the deuterium isotope is automatically determined by the mass effect. Although only two parameters are needed for the energy fit, an effort to predict intensities requires an expan- sion of the potential energy function at least to the fifth power in displacement. Thus the attempt to match intensities of these very weak transitions provides considerable analytical insight into the potential energy surface. As even higher transitions are observed it is expected that a serious challenge to theory may emerge, for now, as dissociation is approached, the zeroth order Born-Oppenheimer approx- imation must be questioned. The LM model predicts that in a polyatomic molecule having several equivalent oscillators there is a remarkable harmonic multi- photon channel that can provide a rapid route for the deposit of intense monochromatic infrared laser radiation in' a manner having photochemical significance. The fact that pulsed infrared light can bring about very energetic molecular excitation, as seen by visible luminescences as well as by photodecomposition,has been the subject of intense interest. The basic question concerns the mechanism of how E8-3 a large number of photons of a single frequency can build up in a single molecule, especially when its vibrational levels are known to be anharmonic . We will show how in benzene, for example, the LM model predicts a highly allowed six photon, sequential, transition among vibrational states which are harmonic. (At v ■ 6 this harmonic state is the one where one excitation appears in each of the six uncoupled C-H oscillators). The implications of this to examples in the liter- ature will be touched upon. Supported in part by a grant from the national Science Founda- tion and the Materials Science Center of Cornell University. This paper includes work found in references 1 and 2 as well as more recent studies carried out in our laboratory by Dr. R. Swofford and Mr. M. Burberry. 1. M. E. Long, Robert L. Swofford and A. C. Albrecht, Science 191, 183 (1976). 2. M. E. Long, R. L. Swofford and A. C. Albrecht, J. Chem. Phys. M. is. Long, K. l 65, 000 (1976). Eb"4 The In-Situ Measurement of Atoms And Radicals In the Upper Atmosphere James G. Anderson Department of Atmospheric and Oceanic Science Space Physics Research Laboratory University of Michigan, Ann Arbor, Michigan 48109 Although upper atmospheric photochemistry has become rapidly more complicated in the past five years as our appreciation for the multitude of gases present in the stratosphere has grown, the basic reaction pattern displayed by each of the chemical cycles involving oxygen, hydrogen, nitrogen and chlorine remains simple. Figure 1 generalizes this pattern into three common parts: (1) Chemical source terms, which represent the upward flow of stable polyatomic molecules from the earth's surface and troposphere; (2) Linking radicals or molecular fragments which are formed from the chemical source terms either by photolysis or by chemical reaction; (3) Reservoir or sink terms which are formed by the recombination of the radicals and are recycled into the radical system by photolysis and chemical reaction and removed by downward and meridional transport. The source and sink terms are relatively stable chemically, having chemical lifetimes on the order of weeks to months so that their global distribution is, in general, governed by transport processes rather than by details of the local chemical environment. In contrast, the radicals have chemical lifetimes on the order of minutes and they thus reflect, with considerable alacrity, the chemical conditions in their vicinity. A composite of the hydrogen, nitrogen and chlorine systems, including only the major reaction paths, is shown in Figure 2 which can Fl-1 be subdivided by inspection into three cycles similar in form to Figure 1. now DOWNWARD DIFFUSION ~~~T" UPWARD DIFFUSION SYSTEM OVERVIEW ^SSBSSRSS COMPOSITE OXYGEN, HYDROGEN, NITROGEN, AND CHLORINE PHOTOCHEMISTRY Significant progress has been made over the past five years in the measurement of stratospheric source and sink terms most notably by infrared absorption and emission techniques, whole air samples and filter trapment. These techniques have proven to be powerful for the determination of absolute concentrations in the part per million to part per billion range where extended regions of spatial integration were acceptable. Because of the long chemical lifetime of these species, details of their local structure were not relevant and there developed a very direct and beneficial exchange between field measure- ments and theoretical calculations. In order to hope for such a exchange in the case of radicals, an experiment must (a) have a detection sensitivity in the part per Fl-2 trillion range in a defined volume element in which other interacting radicals are simultaneously measured, (b) possess a means for accurate absolute calibration and (c) be capable of measuring the diurnal behavior of several species. A technique which quite naturally satisfies these requirements is atomic and molecular resonance fluorescence in combination with a high velocity sample flow which eliminates heterogeneous interaction between the atmospheric sample and the instrument. This geometry is familiar to laboratory gas phase kinetics studies and its adaptation for this work is shown in Figure 3. A beam of photons, resonant with a preselected electronic transition of an atom or molecule is passed across the flowing gas sample confined to the interior of a pod or "nacelle" which establishes laminar flow around and through the device as it is lowered through the atmosphere on a stabilized parachute platform. Photons scattered out of the lamp beam are collected and counted by a photomultiplier with associated optics which observes in a direction perpendicular to both the incident photon beam and the gas flow direction. Absolute calibration is accomplished using a laboratory system shown in Figure 4 which provides a known quantity of the atom or radical in a flow at the appropriate total pressure, temperature and flow velocity. As Figure 4 shows, the flight instrument occupies the central portion of the reaction zone so that a known concentration of the atom or radical, formed either by discharge alone or by chemical reaction in the upstream portion of the reaction zone, flows through the instrument's interior. After calibration, the instrument is F1^3 DESCENT t^ VELOCITY VECTOR PHOTOMULTIPLIER TUBE PULSE COUNTING ELECTRONICS COLLECTION OPTICS SCATTERING VOLUME removed from the system, fitted with airfoil sections and flown optically unaltered from its configuration during calibration. Helium filled research balloons capable of lofting the instrument package to 45 kilometers from where the measurement commences upon deployment onto the stabilized parachute platform have thus far been used exclusively although rocket or aircraft deployment is equally possible. Results of two flights for the measurement of ground state 3 atomic oxygen 0( P) are shown in Figure 5 and two flights investigating the vertical profile of the OH radical are shown in Figure 6. These results will be discussed as well as the development of instruments for the measurement of Cl, CIO, NO, NO and HO. Fl-4 50- 5 * - 40 L±J O 3 i i 1 1 1 i 1 i i i I 1 1 1 1 1 I i I in I 0( 3 P)INTHE EARTH's STRATOSPHERE " ' J • 25 NOVEMBER , 1974 ■ ^ x = 56° 1030 AM. CST , ■ • 7 FEBRUARY, 1975 x= 51 ° K>20A.li CST MODEL CALCULATION 30 V HIGH OH * HO z k.2110 10 cmJ/MC I LOW OH 20H) . The source of 0( 1 D) in turn is the photolysis of ozone in the Hartley continuum ( <3000A) and possibly in the Huggins bands between 3000A and 3500A. 1 i 3 + hv (A<3000A) + 0( D) + , ( Ag) (1) 3 + hv (3000A[0 3 ] relative to the values below 3000A which is assumed to be unity. Thus it is evident that a careful study is needed to determine the quantum yield of 0(D) from the photolysis of ozone. Experimental : l ? 13 Moorgat x and Martin have both studied the photolysis of ozone with CW arc lamps using the reaction of O(-^D) with N2O to subsequently produce excited state NO2 . This technique is used in the present investigation and involves the reaction of 0(D) with N according to the following scheme: 3 + hv -> 0( D) + 2 0( 1 D) + N 2 -+ 2N0 * N 2 + 2 NO + 3 -> N0 2 *+ 2 + N0 2 + 2 N0 2 * + N0 2 + hv Thus, the amount of N0 2 * formed is directly related to the amount of 0(D) produced. The quantum yield of 0(D) can now be determined by measuring the fluorescence of NO * as a function of wavelength for the photodissociation of ozone. N 2 does not absorb in the photodissociation region of approximately 3000A, and the N0 2 fluorescence occurs at F3-3 - 4 - wavelengths greater than 6000A so that conditions exist for an inherently clean experiment. The experimental arrange- ment consists of a flashlamp pumped dye laser which is frequency doubled to provide the photodissociation energy in the region of 2900A to 3200A (See figure 1) . The use of a tunable dye laser allows a high photon flux and a high spectral resolution to give better definition to the nature of the quantum yield. Doubling is accomplished by the use of an ADA and an RDP temperature tuned crystal and the NO2 fluorescence is monitored by standard photon counting techniques. Results of this investigation will be discussed in light of previous experiments and newly acquired data. F3-4 Flash Lamp Pumped Tele Dye Laser BS SHG a. Trigger Trigger SHG Second Harmonic Generation BS Beam Splitter PD Photodiode PM Photo Multiplier MCA Multi-Channel Analyzer Tele Beam Reducing Telescope F3-5 REFERENCES 1. K. Wanatabe, E.C.V. Inn, M. Zelikoff, J. Chem. Phys . , 21, 1026(1953). 2. E.C.Y. Inn and Y. Lanaka, J. Opt. Soc. Am., 4_3. 870 (1953). 3. M. Vigroix, "Contribution a 1' Etude Experimentals de l'ozone", Marson et cie, Paris (1953). 4. M. Ackerman in "Mesopheric Models and Related Experi- ments", D. Reidel, Ed., Dordrecht (1971), p. 149. 5. M. Griggs, J. Chem. Phys., 4_9, 857 (1968) 6. R.F. Hampson, ed. , J. Phys. Chem., Ref. Data 3 (1973). 7. I.T.N. Jones and R.P. Wayne, Proc. Roy. Soc. A., 319 273 (1970). 8. E. Castellano and H.J. Schumacher, Z. Physik. Chem., 65, 62 (1969). 9. R. Simonaito, S. Braslonsky, J. Heicklen, M. Nicolet, Chem. Phys. Lett. 19, 601 (1973). 10. 0. Kajimoto and R.J. Cvetanovic, Chem. Phys. Lett., 37, 533 (1976). 11. C.L. Lin and W.B. DeMore, J. Photochem. 2, 161 (1973- 74). 12. G. K. Moorgat and P. Warneck, Z. Natruforsek 30A , 835 (1975) . 13. D. Martin, J. Girnan, & H.S. Johnston, 167 ACS National Meeting, March 31-Apr. 5, Los Angeles, 1974. F3-6 Absence of NO Photolysis in the Troposphere D.H. Stedman , R.J. Cicerone W.L. Chameides, R.B. Harvey Department of Atmospheric and Oceanic Science University of Michigan Ann Arbor, Michigan 48109 Since the early studies of Bates and Hays [1967] it has been accepted that N_0 photolyzes in the troposphere. Thus, it was thought that not all of the N„0 released at ground level (by what- ever natural source) could be transported intact to the stratosphere. Once in the stratosphere a small fraction of the N„0 is attacked by OC^D) to yield NO molecules, and this process furnishes the NO to conterbalance much of the natural production of 0_ [ Crutzen , 1970; McElroy and McConnell , 1971] . From theoretical numerical models (e.g., Crutzen , 1974; McElroy et al . , 1974), an N~0 source of about 7 3.5 x 10 metric tons per year was needed to balance the altitude- integrated photochemical loss of N^O, almost half of which loss was tropo spheric. Using the photodissociation cross section shown by Bates and Hays , the photodissociation rate, j (N„0) N„0 + hv ■> 0( 1 D) + N 2 2 when Rayleigh scattering is included, is calculated to be about 3 x 10 sec in the lower troposphere. Recently, however, Johnston and Selwyn [1975] have shown that the absorption cross section is negligible in the A > 260 nm spectral region important for tropospheric photochemistry. Their work would imply a zero value for j (N„0) in the troposphere. F4-1 While their work was in progress we decided to use an actinometer, previously used to measure j (N0„) ( Jackson et al . , 1975) , to directly measure j (N_0) at ground level. In the experiment for j (N_0) , pure N„0 at latm was used as the flowing gas in the quartz tube [Jackson et al., 1975] with a flow just enough for the downstream NO 3 -1 detector (^30 cm min ) . The residence time in the tube was about 100 sec. If the NO were photolyzed then the product 0( D) reacts according to: 0( D) + N 2 -> N + 2 -> 2N0 at equal rates [see data survey edited by Hampson and Garvin, 1975] . Thu s a quantum yield of one NO per N~0 photolyzed is expected. If j (NO) ^ 3 x 10~ sec" then the yield of NO would be 3 x 10~ —8 -9 x 100 = 3 x 10 atm. Since the NO detector has a noise level ^-10 atm. then quantitive detection of j (N„0) was possible if Bates and Hays' cross sections applied. When the experiment was first performed, exposure of the N„0- containing tube to sunlight generated significant quantities of NO. At this point, the calibrated NO detector was switched to the NO x (NO + NO„) mode, and an impurity ^0.2 ppm NO_ in the N„0 was deter- mined. The photolysis of this impurity was generating the NO, and when the impurity was removed with an ascarite trap the NO generated by sunlight on NO became undetectable. Figure 1 shows the trace from such an experiment, performed at a time when j (N0„) was also being measured. Results of twenty-seven measurements around solar noon gave an F4-2 upper limit for j (NO) determined experimentally as <1 x 10 "" sec Thus, tropospheric photolysis of N 2 has been shown to be negligible by direct experiment, as predicted by the concurrent spectroscopic study of Johnston and Selwyn. With the Bates and Hays [1967] cross sections a j (N 2 0) = 10~ sec could only be observed with an 3 — 19 —2 column of >3 x 10 cm , three or four times larger than the normal ozone content. 0.007 o ~ 0.005 z 0.003 PURE N 2 A rJ N 2 + Olppm N0 2 PURE N 2 r\ 20 10 } I W J \ / Oct 21, 1975 1=00 p.m. EDT 20 TIME min * 20 10 Si a. a. O Z 10 Figure 1: The lower trace and right hand axis shows a record of NO formation in the solar photolysis of N2O, with and without a small impurity of NO2. Three traces are shown with sunlight intercepted by a black cloth at the beginning and end of each trace (dotted lines) . The upper trace shows the concurrent measured value of the solar photolysis rate of NO», j (NO2) , on the left hand axis. NO production in the absence of im- purities is not distinguishable from the background. F4-3 References Bates, D.R. and P.B. Hays, Atmospheric nitrous oxide, Planet. Space Sci., 15, 189-197, 1967. Crutzen, P.J., The influence of nitrogen oxides on the atmospheric ozone content, Quart. J. Roy. Met. Soc , 96 , 320-327, 1970. Crutzen, P.J., A review of upper atmospheric photochemistry, Can. J. Chem . , 52_, 1569-1581, 1974. Hampson, R.F. , and D. Garvin, (Editors), Chemical kinetic and photo- chemical data for modelling atmospheric chemistry, NBS Technical Note 866, National Bureau of Standards, Washington, D.C. , 113 p. , 1975. Hunten, D.M., Residence time of aerosols and gases in the stratosphere, Geophys. Res. Lett ., 2_, 26-28, 1975. Jackson, J.O., D.H. Stedman, R.G. Smith, L.H. Hecker, and P.O. Warner, Direct N0_ photolysis rate monitor," Rev. Sci. In strum . , 46_ (4) , 376-378, 1975. Johnston, H.S., and G. Selwyn, New cross sections for the absorption of near lutraviolet radiation by nitrous oxide (NO) , submitted to Geophys. Res. Lett . , 1975. McElroy, M.B., and J.C. McConnell, Nitrous oxide: a natural source source of stratospheric NO, J. Atmos. Sci ., 28 , 1095-1098, 1971. McElroy, M.B., S.C. Wofsy, J.E. Penner, and J.C. McConnell, Atmospheric ozone: possible impact of stratospheric aviation, J. Atmos. Sci . , 31, 287-303, 1974. F4-4 Temperature Dependence of 0( D) Reactions of Atmospheric Importance J. A. Davidson and H. I . Sen iff York University, Toronto, Ontario, Canada and G. E. Streit, A. L. Schmel tekopf , and Carleton J. Howard Aeronomy Laboratory NOAA Environmental Research Laboratories Boulder, Colorado 80302 The reactions of 0(D) with NpO, H 2 0, H 2 » and CH 4 are now well established as the principal sources of the radical species (NO, OH) responsible for the natural destruction of stratospheric ozone. Because of the importance of these reactions, there have been many determinations of the relative i rates of the reactions of 0( D) with these and other 1 -2 species. On the other hand, systematic investigations of absolute reaction rates have been limited to those conducted O _ C 7 in Husain's " lab and our own. 1 In order to assess the relative importance of 0( D) de- activation and reaction to produce radical species under stratospheric conditions we have undertaken an investigation of the temperature dependence of the reactions of 0( D) of stratospheric importance. The apparatus used in these studies has been described in detail previously and hence will only be discussed F5-1 briefly here. A frequency quadrupled Nd-YAG laser, yielding 1-4 mj at 266 nm in a 10 nsec pulse, was employed to photo- lyze 0~ to produce 0( D) in a temperature regulated cell. The time evolution of the 0( D) concentration was then followed by means of the emission at 630 nm. Because 0( D) has a long radiative lifetime (t = 100 sec) and reacts at near collision frequency with most gases, signal averaging techniques were required to obtain a statistically meaning- ful decay trace. Our treatment of this data in order to extract rate constants has been previously reported. In Table I our room temperature rate constants are compared with the results from other laboratories. With the exception of 0~, the results obtained in Husain's lab using the time resolved attenuation of resonance absorption are significantly higher than our results. We have previously suggested that this discrepancy may be due to the use of a rather large correction to Beer's law employed by Husain. 1 3 Subsequently, Phillips has calculated that in the case of the 0- rate determination this correction was appropriate, while in all other cases the correction should have been much smaller, hence leading to a reduction of the reported rates by about a factor of 2. We will be presenting the most recent results obtained in our investigation of the temperature dependence of the reactions of 0( D). These will include the rates of 0( D) w ith H 2 0, CH 4 , H 2 , N 2 0, Og, Ng. C0 2> 3 , HC1, NH 3 and with F5-2 trace atmospheric halocarbons. This research was conducted in the NOAA Aeronomy Labor- atory and was supported in part by NASA contract MGR 52-134- 005. 1. R. J. Cvetanovic, Can. J. Chem. 53_, 1452 (1974). 2. D. Garvin and R. F. Hampson, Editors, NBSIR 74-430 (1974). 3. R. F. Heidner, III, D. Husain, and J. R. Wiesenfeld, Chem. Phys. Lett. 1_6, 530 (1972). 4. R. F. Heidner, III and D. Husain, Nature Phys. Sci. 241 , 10 (1973). 5. R. F. Heidner, III and D. Husain, Int. J. Chem. Kinet. 5, 819 (1973). 6. R. F. Heidner, III and D. Husain, Int. J. Chem. Kinet. 6, 77 (1974). 7. J. A. Davidson, C. M. Sadowski, H. I. Schiff, G. E. Streit, C. J. Howard, D. A. Jennings, and A. L. Schmelt ekopf, J. Chem. Phys. 64, 57 (1976). 8. D. Biedenkapp, L. G. Hartshorn, and E. J. Bair, Chem. Phys. Lett. 5, 379 (1970) . 9. D. R. Snelling and E. J. Bair, J. Chem. Phys. 47_, 228 (1967). 10. D. R. Snelling and E. J. Bair, J. Chem. Phys. 48_, 5737 (1968). F5-3 11. J. F. Noxon, J. Chem. Phys. 52, 1852 (1970). 12. R. Gilpin, H. I. Schiff, and K. H. Welge, J. Chem. Phys 55_, 1087 (1971). 13. L. F. Phillips, Chem. Phys. Lett. 37, 421 (1976). F5-4 -M CO H s o ch VO OO OJ CM o°° CM + 1 -4- LPs d ON CO •\ * ir\ to ft rH CM •\ ■H Jh H rH ft- m i 1 O CM 41 H CO O © m •>— <• w ■p a Q CM s o d) ^ 3 oj cO ■S3 IPs s CM + | ■p W C3A £ CO *- s CM +J CJ w Cl) co On o ft- i o H S 1 ' CM o OJ Ch O ft- !>s CM fn O + 1 CO o co + 1 CM CM d • oo OO + 1 O -4 o + 1 C— • VO • o O d vo CM + 1 + 1 oo ft CM oo no + 1 -d- H* LTN o OJ • • + 1 + 1 OO CO • • H H l.O oo O + 1 + 1 c— oo CM rH ft OO + 1 + 1 H OO • • OO H o OO • • H + 1 + 1 O H oo CM H + 1 CM + 1 -4- • • CM H CM ON o + 1 H + 1 CM CM H vo o ft o + 1 o\ VO + 1 o oo d d LTN O LTN o + 1 o c— + 1 H -4" o O o CM 1\ + 'l H + 1 -4- -«£. l/-\ + 1 CD r^ O -P nd CD P CD P CO >> CJ CO CJ o < >> H ft ca -P ■H g •H ft 5-i O u ft CM CM CM F5-5 Modeling Stratospheric Photochemistry and Kinetics John McAfee and Paul J. Crutzen National Center for Atmospheric Research Upper Atmospheric Research Branch Boulder, Colorado 80303 An analysis will be presented of our present understanding of stratospheric photochemistry. The effect of uncertainties in rate coefficients will be discussed with emphasis on the impact of human activity on the ozone layer. F6-1 THE RATE CONSTANT FOR + NO + M FROM 217-500 K IN FIVE HEAT BATH GASES 1 2 J. V, Michael , W. A. Payne and D. A. Whytock, Astro chemistry Branch, Laboratory for Extraterrestrial Physics, NASA/Goddard Space Flight Center, Greenbelt, Maryland The absolute rate constant for + NO + M has been measured pre- 3 viously by a variety of techniques and is of interest in atmospheric 4 modelling. Most workers have used the discharge flow method with chemiluminescent detection of [o]; however mass spectrometric and e.s.r. detection have also been employed. Values relative to a known reaction have also been reported from steady state photochemical experiments. There are six real time studies. Three utilize phase sensitive detec- 3 4 tion and two make direct measurements from chemiluminescence. There has been only one direct measurement with the atom resonance fluores- cence technique, at room temperature in Ar, The currently accepted value for the temperature dependent rate constant is based primarily on the work of Klein and Herron and Clyne and Thrush and is given from 200-500 K in Arrhenius form as k - 3.0x10" exp(940/T) cm molecule"' sec" 1 (M = , Ar). 3,4 The apparatus used in the present study has been described pre- fi -L. -? viously. Binary mixtures of NO in heat bath (2.5x10 ^ X^ ^10 ) were repetitively flash dissociated (X ^ 142 nm, t = 2.5 /isec) in a 4 3 flowing system yielding N( S) and 0( P) after the method of Stuhl and 7 3 4 Niki. Further 0( P) is produced by the fast reaction, N( S) + NO -* 3 N~ + 0( P). This reaction is effectively complete before the first data are colletected (delay time ^ 300 jusec). Flash energies, and therefore concentrations, are sufficiently low so that the secondary reaction, + N0„ -> 0_ + NO, is totally negligible. F7-1 Fluorescent photons are repetitively counted and stored with a multichannel analyzer under a given set of experimental conditions. The resulting decay constant can be expressed as, "observed * k t M M + V < X > where k. is the termolecular rate constant for + NO + M, and k, is t a the diffusional loss rate constant out of the viewing zone. At a given pressure and temperature, at least three decay constants are obtained as a function of [NO]. Plots of k , against [NO] yield k, = k [m] as the linear least squares slope, and the intercept gives k,. Plots of a k, . against [m] then yield k as the linear least squares slope for a given temperature. In nearly all cases the least squares standard de- viation in the intercept is within one standard deviation of zero. Determinations at six temperatures (217 K, 252 K, 298 K, 355 K, 418 K, and 500 K) are then carried out for any given heat bath gas. The heat bath gases used are N 9 , He, Ne, Ar, and Kr. The results for M = NL are shown in Table I and can be expressed in Arrhenius form as k = (15.5+2.0)xl0 exp (582+37) /T) cm molecule sec . This result can be directly compared to that of Klein and Herron who give (3.97+0.55)xl0 exp ( (971+50) /T) cm molecule sec (M=flL). The room temperature values are in good agreement, but dis- crepancies occur for all other temperatures above room temperature, the largest being at 500 K where the present results are 1.8 times larger. The temperature dependence of k with the other heat bath gases appears to be similar. F7-2 TABLE I: Rate Data for + NO + N -> NO, , + N T/K k /10 cm molecule sec 217 20.8 + 1.2 a 252 17.2 + 2.3 298 11.5 + 0.9 355 7.40+ 0.68 418 6.24+ 0.23 500 4.92+ 0.25 a. quoted error limits are the standard deviation 1. NAS/NRC Senior Resident Research Associate. 2. Catholic University, on leave from the Chemistry Department, The University of Essex, Colchester, England. 3. Reviewed by D.L. Baulch, D.D. Drysdale, D.G. Home, and A.C. Lloyd, "Evaluated Kinetic Data for High Temperature Reactions", Vol. 4, Butterworths, London, 1972. 4. R. F. Hampson, Jr. and D. Garvin, editors, "Chemical Kinetic and Photochemical Data for Modelling Atmospheric Chemistry", NBS Technical Note 866, U.S. Government Printing Office, Washington, 1975. 5. T.G. Slanger and G. Black, J. Chem. Phys. 53, 3717 (1970). 6. R.B. Klemm and L.J. Stief, J. Chem. Phys. 61., 4900 (1974). 7. F. Stuhl and H. Niki, J. Chem. Phys. 55, 3943 (1971). 8. F.S. Klein and J.T. Herron, J. Chem. Phys. 41, 1285 (1964). F7-3 Determination By The Phase Shift Technique Of The 3 Temperature Dependence Of The Reactions of 0( P) with HC1, HBr, and HI. D . L. Singleton and R.J. Cvetanovic Division of Chemistry, National Research Council of Canada Ottawa, Canada. A few studies have been reported recently on some specific aspects of the reaction of ground state oxygen atoms with hydrogen halides, O + HX ■*■ OH + X, namely, the effect of HX vibrational energy on the reaction rate and 2 ) the relative yields of excited products , OH(v=0,l) and o X( P 3 /, |). The temperature dependence of the bulk rate constants has received little attention, and, in the case of HC1, where more than one determination has been reported, the results are in poor agreement. We have used a phase shift technique to determine the rate constants for the reactions of 0( P) with HC1, HBr, and HI between 298-55^K. Also, as an incidental part 3 of this 'work, the rate constant of the reaction 0( P) + NO + M -> NO + M (M = HCl) has been determined between 298- 398K. Experiment al 3) As described previously ground state oxygen atoms are formed by modulated photosensitized decomposition of nitrous oxide in a flowing mixture of NO, NO, HX, and Hg F8-1 according to the simple reaction scheme: Hg( 1 So^25iLJH$ H g( 3 Pl ) Hg( P!) + N 2 0( 3 P) + HX 0( 3 P) + NO + M _^ Hg( 1 So) + N 2 + 0( 3 P) OH + X N0„ + M -> 2 (1) (2) (3) The expression for the phase shift, (J), between the incident 25^ nm light, modulated at frequency V, and the chemilumi- nescence from reaction (3) i fSf* ■ k 2 [HC1 l + S [N ° ][M] (I) - The phase angles are measured with appropriate filter - phot omult ipli er combinations and a lo-ck-in amplifier Two methods of modulation were used. The first, for the HC1 reaction below 500K where frequencies of 30-90 Hz were required, involved mechanically chopping the light from a rf powered low pressure mercury lamp. In the second method, for all the other reactions, which required frequen- cies of 1.3-6 kHz, the amplitude of the rf that powered the lamp was modulated. Total pressures were between 30-70 Torr and were varied by a factor of two at each temperature except for HC1 at 298K where the total pressure was 16-18 Torr. Appro- ximate relative concentrations [N 0]:[N0]:[HX] were: HC1 (T<500K), 1:0. 001-0. 01:0. 1-1;HC1(T>500K) , 1:0.1:0.1; HBr, 1:0.1: 0.01-0.1; HI, 1:0.1: 0.001-0.01. The residence time of the gases in the reaction cell was <0.k sec for the HI experiments; <2-5 sec for HBr and HC1 (T>500K); and <0.2 sec F8-2 for HC1 (T<500K) . The total flow rates were between 10 -k - 3 -1 to 10 .mole sec The rate of production of 0-atoms 12 -3-1 was about 1x10 atoms cm sec as determined by gas chro- matographic analysis of the products of the + 1-butene reaction. Results For the reactions with HCl(T>500K), HBr , and HI, the rate constants k and k were determined from inter- cepts and slopes of plots of 2ttv/ [HX] t an<£ vs. [NO] [M]/ [HX] . The values of k (M=N 0) were within about 10$ of the previously determined values. The standard deviation of each value of k was 7-9$ for HC1 and 2-5$ for HBr and HI. The values of k are shown in Figure 1. The rate constants plotted in Figure 1 cover a range of reactivity of four to five orders of magnitude. The reactions of oxygen atoms with HC1 (T>500K) and with HBr and HI are sufficiently fast to be measured by the proce- 3) dure described previously. The reaction with HC1 at temperatures smaller than about 500K is much slower and required a modified procedure. The reaction with HCl(T<500K) required lower frequencies, but at these longer time scales, diffusional loss of 0( P) occurred from the viewing zone of the photo- multiplier that detected the chemilumines cence . In this case, Equation I becomes 27rv/tan4> = k [N0][M] + k /[M] F8~3 in the absence of HC1 and assuming a first order diffusional loss of oxygen atoms. The first order diffusion rate constant, k , was obtained in experiments without HC1 pre- sent from the intercept of plots of 2ttv[M] /t ant}) vs. [NO] [M] at each temperature below 500K. The slopes gave values for k (M=N 0) which were in good agreement with previously- determined values. An additional complication arises because the higher HC1 concentrations (which were even as high as the NO concentrations at 298K and 330K) meant that HC1 was a significant third body in reaction 3. Incorpora- ting k„ and a, (the third body efficiency for HCl), into the simple mechanism, we obtain, in place of Equation I, 2ttv/ tan<|) = k 2 [HCl]+k [N0]([N 0] + a[HCl])+k D /[M]. Plots of 2irv/tan(J)-k /[M] vs. [NO], in experiments where [NO] was varied while all other concentrations were held constant, gave k [HCl] as the intercept. From the slope, the value of a was obtained at each temperature (298-UU3K) using the known concentrations of NO and HCl, and the previously determined value of k (M=N 0). The results are given in Figure 1. The Arrhenius expres- sion for the 0+HBr reaction is ( 8 . 09 ±0 . 86 ) xlO 9 exp(-3. 59±0. Ookcal mole -1 /RT), and for the 0+HI reaction^ 2 . 82±0 . 27 ) xlO ° exp(-1.99± 0. OTkcalmole" /RT) 1. mole" sec" . The plot of Ink vs. /T for 0+HC1 is not linear. A one parameter empirical expression for the results over the temperature interval of these experiments is k = ( 3 . 1±2 . 1 ) xlO -13 T ^ T • H± • 09 ) 1>raole _1 F8-4 sec , which was obtained from a least squares treatment of Ink vs. InT. All the indicated uncertainties are one standard deviation. At room temperature, HC1 is 1.7 times as efficient as NO as a third body in reaction 3. The Arrhenius ex- pression for reaction 3, with M=HC1, is (3.3±1.0)xl0 exp (0.6±0.2 kcalmole" /RT ) between 298-398K, which compares with the previously determined expression for M=N , 6.12xl0 9 exp(l.23kcalmole -1 /RT) 1 . 2 mole~ 2 s e c" 1 . A comparison is made with some literature values for k in Figure 1. The results for HBr are in good agree- k) ment with Smith's values and with Glass's room tempera- 2 ) 7-1-1 ture value, 2.7x10 l.mole sec . For HI, the only avai- lable comparison is with the approximate room temperature 2 ) 9 -1-1 value reported by Glass (^1x10 l.mole sec ), which is in close agreement. For HC1 , the present values agree well at higher temperatures with Smith's values but are somewhat larger at lower temperatures. The results of references 5 and 6 for the HC1 reaction are also shown in Figure 1. Further work is in progress to establish the validity of the present values for the 0+HC1 reaction. References 1) Z. Karny, B. Katz, A. Sz8ke, Chem. Phys. Lett., ^5_, 100 (1975); D. Arnoldi and J. Wolfrum, Chem. Phys. Lett., 2U , 23U (197*0; G.P. Quigley and G.J. Wolga, J. Chem. Phys. , 63, 5263 (1975) . 2) G.A. Takacs and G.P. Glass, J. Phys. Chem., 77 , 1182 (1973). 3) D.L. Singleton, S. Furuyama, R.J. C vet anovi c , and R.S. F8-5 I0'° F _0> O E m 7 10' r- CM 10* 10'r- 0+HBr *^ + ^ + + HCI N £s + +* 1 1.5 2.0 2.5 3.0 3.5 lOOO/KK"') Figure 1. Arrhenius plots of k : • this work; + Ref. k ; Ref. 5; -•-• RefT 6. Irwin, J. Chem. Phys . , 6_3, 1003 (1975). k) R.D.H. Brown and I.W.M. Smith, Int. J. Chem. Kinet., 7, 301 (1975). 5) E.L. Wong and F.E. Belles, NASA Technical Note D-6U95 (1971). 6) V.P. Balakhnin, V.I. Egorov, and E.I. Intezarova, Kinetics and Catalysis, 12, 258 (1971). F8-6 High Energy Pulsed Laser Photolysis of Some Chromium(III) and Cobalt(III) Complexes. Arthur W. Adamson , A.R.Gutierrez, R.E.Wright, and R.T.Walters, Department of Chemistry, University of Southern California, Los Angeles, California 90007. A current interest of this Laboratory is in the study of excited state processes through the use of a high energy pulsed laser system. Results of three types of experiments will be reported. Figure 1 allows illustration of the first two. The photochemical literature for Cr(III) complexes so far consists just of quantum yield measurements (1). The yield for photochemistry from the first thermally equili- brated excited (thexi) quartet state (^Lj ) is given by ^ = $ k x where is the yield for producing Hj , k is its chemical rate con- q cq stant, and 1/t = (k„„ + k „ + k ) , k „ and k being the non- q cq nrq eq nrq eq 3 radiative and radiative deactivation rate constants respectively. Alternatively, photochemistry may derive from reaction of the first doublet thexi state ( 2 D in 0.), with 2 = 4> d k ,t., where k . is the chemical rate constant and 1/t. = (k . + k . + k ,). In addition, it d cd nrd ed has been postulated that the 2 D and ^L^ states may thermally intercon- vert (2). Clearly, a full elucidation of such systems requires determi- nation of the separate rate constants. It is certainly insufficient to find <\>i + 2 or even the separate yields. These last have been inferred from quenching experiments (See Refs. 1,2). The first type of laser study is that of the determination of ligand and solvent effects on k . through studies of emission life- times. Weak room temperature emission has been reported for several Gl-l Cr(III) complexes (3,4) and additional results are included in Tables I and II. These demonstrate that emission lifetimes of various com- plexes vary greatly with the type of ligation and, for a given complex, with solvent medium. A correlating observation is that water either as a ligand or as solvent catalyzes nonradiative deactivation. That is, we suppose k . to be negligibly affected by solvent environment and, as suggested by low temperature studies (5), also negligibly by such changes as replacing coordinated NH 3 by H 2 0. Thus variations in x . should be due mainly to variations in k .. It may be that ligand-solvent interactions that provide good vibrational communication lead to large nonradiative deactivation rates. If so, the process may be less intramolecular than due to interaction between the excited state and the phonon system of the medium. It may be wery similar in nature to the activation process in thermal reaction kinetics. The second and third types of study are ones for which the laser system was designed. The system comprises a Q-switched, gated, amplified, and frequency doubled Nd oscillator; maximum output is 1 to 1.5 J at 530 nm with 20 nsec halftime. If the beam is collimated to 3 mm diameter, the nominally absorbed number of photons (i.e. assuming no ground state bleaching) can be made greatly to exceed the number of molecules of compound. This capability makes possible the investigation of (a) two (successive) photon processes and (b) single pulse photolyses in which primary product formation is monitored. As an example of (a), we obtain t. (for trans-Cr(NH 3 ) 2 (NCS)i f " in acetonitrile both from emission decay and from excited state absorption (for which the spectrum has Gl-2 been reported (6)). For several systems excited state absorption has also been observed as an excess attentuation of the transmitted pulse, the pulse shape being skewed if the lifetime of the absorbing excited state is greater than that of the pulse. A third experiment of type (a) is that of two-photon processes involving Co(III) ammines. As illustrated in Figure 2, absorption of a second 530 nm pulse should populate a charge transfer excited state, the evidence being the appearance of Co(II) as photolysis product. The effect has been re- ported for Co(NH 3 ) 6 3+ using a highly focussed, multiply pulsed N 2 laser (337 nm) (7). We find it for Co(NH 3 ) 5 Br 2+ after a few unfocussed pulses at 530 nm. Type (b) experiments are in progress with various Cr(III) ammines and acidoammines. We observe the rise time of primary photo- product formation by means of a suitably monochromated monitoring beam. The 2 process should lead to product with an appearance time x ., which typically is in the short microsecond range. The ^ process may produce product much more quickly; the rise time in this case should yield the so far inaccessible x . q (1) E.Zinato, "Concepts of Inorganic Photochemistry," A.W.Adamson and P.F.Fleischauer, Wiley, 1975. (2) R.Ballardine, G.Varani, H.F. Wasgestian, L.Morri, and V.Balzani, J.Phys.Chem. , 77, 2947 (1973). (3) N.A.P.Kane-Maguire and C.H.Langford, Chem.Comm., 895 (1971). A.W.Adamson, C.Geosling, R.Pribush, and R.Wright, Inorg.Chim.Acta, 1_6, L5 (1976). K.K.Chatterjee and L.S.Forster, Spectrochim.Acta, 20, 1603 (1964). (6) T.Ohno and S.Kato, Bull .Chem.Soc. Japan, 43, 8 (1970). Gl-3 Cr(III) complexes (3,4) and additional results are included in Tables I and II. These demonstrate that emission lifetimes of various com- plexes vary greatly with the type of ligation and, for a given complex, with solvent medium. A correlating observation is that water either as a ligand or as solvent catalyzes nonradiative deactivation. That is, we suppose k . to be negligibly affected by solvent environment and, as suggested by low temperature studies (5), also negligibly by such changes as replacing coordinated NH 3 by H 2 0. Thus variations in t. should be due mainly to variations in k .. It may be that ligand-solvent interactions that provide good vibrational communication lead to large nonradiative deactivation rates. If so, the process may be less intramolecular than due to interaction between the excited state and the phonon system of the medium. It may be very similar in nature to the activation process in thermal reaction kinetics. The second and third types of study are ones for which the laser system was designed. The system comprises a Q-switched, gated, amplified, and frequency doubled Nd oscillator; maximum output is 1 to 1.5 J at 530 nm with 20 nsec halftime. If the beam is collimated to 3 mm diameter, the nominally absorbed number of photons (i.e. assuming no ground state bleaching) can be made greatly to exceed the number of molecules of compound. This capability makes possible the investigation of (a) two (successive) photon processes and (b) single pulse photolyses in which primary product formation is monitored. As an example of (a), we obtain x . (for trans_-Cr(NH 3 ) 2 (NCS)i f " in acetonitrile both from emission decay and from excited state absorption (for which the spectrum has Gl-2 been reported (6)). For several systems excited state absorption has also been observed as an excess attentuation of the transmitted pulse, the pulse shape being skewed if the lifetime of the absorbing excited state is greater than that of the pulse. A third experiment of type (a) is that of two-photon processes involving Co(III) ammines. As illustrated in Figure 2, absorption of a second 530 nm pulse should populate a charge transfer excited state, the evidence being the appearance of Co(II) as photolysis product. The effect has been re- ported for Co(NH 3 ) 6 3+ using a highly focussed, multiply pulsed N 2 laser (337 nm) (7). We find it for Co(NH 3 ) 5 Br 2+ after a few unfocussed pulses at 530 nm. Type (b) experiments are in progress with various Cr(III) ammines and acidoammines. We observe the rise time of primary photo- product formation by means of a suitably monochromated monitoring beam. The 2 process should lead to product with an appearance time t ., which typically is in the short microsecond range. The <$>i process may produce product much more quickly; the rise time in this case should yield the so far inaccessible x . q (1) E.Zinato, "Concepts of Inorganic Photochemistry," A.W.Adamson and P.F.Fleischauer, Wiley, 1975. (2) R.Ballardine, G.Varani, H.F. Wasgestian, L.Morri, and V.Balzani, J.Phys.Chem. , 77_, 2947 (1973). (3) N.A.P.Kane-Maguire and C.H.Langford, Chem.Comm., 895 (1971). A.W.Adamson, C.Geosling, R.Pribush, and R.Wright, Inorg.Chim.Acta, 16^, L5 (1976). K.K.Chatterjee and L.S.Forster, S pec troc him. Acta, 20, 1603 (1964). (6) T.Ohno and S.Kato, Bull .Chem.Soc. Japan, 43, 8 (1970). Gl-3 TABLE 1 Emission Lifetimes for Cr(III) Complexes at 25°C Complex Cr(bipyr) 3 3+ Cr(NH 3 ) 6 3+ t-Cr(en) 2 (NCS) 2 Cr(en) 3 3+ T,ns (in H 2 0) 45,000 1,800 1,490 1,300 Complex Cr(NH 3 ) 5 NCS 2+ t-Cr(en) 2 F 2 + t-Cr(NH 3 ) 2 (NCS) l+ ' Cr(NCS) 6 3 - x,ns (in H 2 0) 140 (148 60% AN) 139 <20 (120 AN) <20 (30 AN) Cr(en) 2 (H 2 0)F 2+ , Cr(NH 3 ) 5 (H 2 0) 3+ , Cr(NH 3 ) 5 (N0 2 ) 2+ , cis-Cr(L)Cl ? + , cis- a-Cr(trien)C1? + : <20 ns. Cr(urea) 6 3+ , Cr(acac) 3 : no observable emission. L = 2,3,2-tet TABLE 2 Solvent Effects on the Emission Lifetime of trans-Cr(NH 3 ) : >(NCS) 4 at 25°C Solvent x,ns Solvent T,ns water <20 DMF 78 methanol 23 DMS0 89 acetic acid 26 nitromethane 105 ethanol 31 acetonitrile 120 THF 39 (deaerated) 177 acetone 158 Gi-4 L'gand h«ld iWti (or O^ symmetry I T Chemical reaction Radiationlet* deactivation by medium I Energy — *» 0»»IOr Figure 1 100 80- UJ 60 40 20 POSSIBLE TWO PHOTON PROCESS FOR A Coim ) COMPLEX 310 nm 7 s J _^~- 530 nm ~^-r— 'CT . • c — coin) ~* 3 CT ~0.3 i (B) — aquation 530 nm -j^- ~G£jal Figure 2 Gl-5 STUDIES USING A COMBINATION OF FLASH PHOTOLYSIS AND PULSED MAGNETIC INDUCTION: APPLICATION TO THE N0 3 RADICAL IN AQUEOUS ACID SOLUTION AT 25 °C. T . W . Mart in (Department of Chemistry, Vanderbilt University, Box 1506, Station B, Nashville, Tennessee 37235) and M. V. Stevens (Duck- worth Pathology Laboratory, Methodist Hospital, 1265 Union Avenue, Memphis, Tennessee 3810*1 ). Introduction . The photolysis of (NHi f ) 2 Ce(N03) 6 in acid solution is a convenient means of studying the chemistry of the N0 3 radical, one of the strongest oxidants known. Over the years we have learned a few 1 2 things about its absorption spectrum, decay kinetics, and competitive electron transfer with various Ce(lll) species. More recently, by using a pulsed laser and kinetic spectroscopy, we have obtained deeper insight into the comprehensive mechanism of the generation and decay of the NO3 radical in nitric and perchloric acid solutions at room temperature. It was found that the 20 ns pulse from a 4-joule fre- quency-doubled ruby laser was virtually instantaneous in comparison to the thermal rate processes involved. Hence it was possible to make a kinetic analysis over the complete lifetime of NO3 in these solvents. These studies revealed that an excited-state metastable complex, # [Ce(lll) • • «N03] , is a precursor to the NO3 radical. Thus the earlier 5 published mechanism must be enlarged as follows to account for this complex: (1) Ce(lV)N0 3 " + hv > [Ce(lll) •••N0 3 ] (2) [Ce(lll)»--N0 3 ] k2 > Ce(lV)N0 3 " (3) [Ce(lll)«--N0 3 ] 3 > Ce(lll) + N0 3 G2-1 (k) Ce(lll) + N0 3 klt ) Ce(lV)N0 3 " (5) N0 3 + N0 3 k 5 > N 2 6 (6) N 2 6 + 2 Ce(lV) > 2 N0 2 + + 2 + 2 Ce(lll) (T) N0 2 + + N0 3 ~ + H 2 > 2 H0N0 2 2 H + + 2N0 3 ~ The measured rate constants and quantum yields derived from our anal- yses, together with a new value of the extinction coefficient for NO 3 are summarized in Table 1. From these data, with regard to the gener- ation of NO3, we infer that k 2 increases on dilution of H0N0 2 because of increased coordination of H 2 with Ce(lV); whereas the opposite is true for IC3 which increases with increased coordination by NO3 . From these trends and since $ = k3/(k 2 +k3), it is not surprising that IMU3 $ becomes constant and equal to unity when [H0N0 2 ] - 6M. As to the JNO3 decay modes, k^ follows similar trends to k3, but ks is essentially constant and independent of [H0N0 2 ]. Magnetic Studies . Concurrent with our NO 3 work, we developed an appar- atus which can combine either conventional flash photolysis or pulsed laser techniques with the simultaneous use of a high field pulsed magnet capable of generating inductions to 200 kilogauss (kG). The purpose was to test if magnetic inductions could in any way truly influence either the excited or the ground states of species in solu- tion at 25°C. So far we have measured two bona fide magnetic effects. One had to do with the classic photoisomerization of stilbene where the inductions stimulated the S -*■ T intersystem crossing mode in the * excited state as efficiently as benzophenone ; the other is the reverse effect of stimulating the T ■*■ S rate determining step monitored by kj+ in the above mechanism. G2-2 Because of the novelty and lack of theory for this experimental approach, we have hesitated to publish our results. But this second effect involving NO3 has convinced us of the possible general appli- cability of this method for probing mechanism in both the ground and excited state of organic as well as inorganic systems. In Table 2, where the [HONO2] is relatively high, we note that the size of k^. is only slightly enhanced, about 6%, which is probably due only to a small magneto-striction effect on the concentration of paramagnetic species in the ground state. However, in perchloric acid (Table 3), where the coordination sphere of Ce(lll) is richer in H 2 than NO3 , we observe a large specific rate enhancement that is quite non-linear particularly at B > 75 kG. We suspect this enhancement at B > 150 kG will approach an upper limit of 3, because we believe k^ normally monitors the electron transfer process involving only one of the levels of the ground triplet recombination complex, [Ce(lll)»»* Tn NO3] °, which is active in converting to the ground singlet state, o [Ce(lV)NC>3 ] °. We postulate that this non-linear magnetic effect either tends to make all three levels of the triplet complex kineti- cally active in undergoing Tq ■*■ Sq crossing or it acts to mix those levels in such a way as to enhance the population of the one active level so as to increase k^ with respect to the unperturbed specific rate k^ . If either interpretation is correct, this technique repre- sents a new way independent of magnetic resonance for demonstrating and studying CIDEP effects by monitoring short-lived intermediates in both organic and inorganic systems . Its sensitivity to the structure of the ligand coordination sphere is also remarkable. G2-3 u CO PJ CO < S o LO LO PJ «vj o o 3C vO , — * X o 1— 1 ■p o 1— 1 bo • 1— ( 0) T-H d> 1—1 o u •M i i—( 3 o a) a* o > a> • •H o +J 10 rH O o3 o M-l .—1 +-> <4-l i— 1 e rt f-4 o e v ' T— I o | to CJ s •H 1 00 *3- CO o 2 o +l o o • o bO tH 00 c • •H •p rj- 6 3 3 o It to XI to rt a TS T3 0) LO c c to 03 •H ^ 03 /■ — \ i — \ ■M CO CO c o AS o 2 + o * » •M ■(-> c O .— i m o c o3 i— i •H 3 T3 E 4-> o c oj cd 03 f— 1 u PQ Q 3 X O T3 G2-4 TABLE 2. EFFECT OF A MAGNETIC INDUCTION OF 150 KILOGAUSS ON THE DECAY OF N0 3 IN AQUEOUS NITRIC ACID AT 25 ± 1°C [HONO2 ] (U B ± a) x 10" 6 M" 1 s" 1 at B = 150 kG (k k ± of xHfW 1 at B = kG kit /ki» .599 ± .089 1.06 ± .18 .996 ± .036 1 . 09 ± .05 1.78 ± .10 1.06± .10 3.96 ± .26 1.03 ± .08 1.00 M .630 + .075 3.00 M 1.08 + .039 6.00 M 1.86 + .14 12.00 M 4.08 + .16 Control samples run under Identical conditions except B = kG, TABLE 3. EFFECT OF MAGNETIC INDUCTION ON THE DECAY OF NO 3 IN 3M AQUEOUS PERCHLORIC ACID AT 25 ± 1°C Magnetic Induction B(kG) (U B ± 0) x 10" 6 M~ ■1 -1 s (k 4 ± a) a x 10" 6 M" at B = kG •1 - s -1 k 4 /kit 150 .896 ± .150 .372 ± .052 2.31 ± .52 150 .771 ± .055 .352 ± .071 2.18 ± .47 75 .425 ± .052 .356 ± .022 1.20 ± .16 75 .428 ± .078 .356 ± .022 1.21 ± .20 Control samples run under identical conditions except B = kG. References (T. W. Martin et al . ) : l) J. Am. Chem . Soc , 85., 113 (1963) ; 2) ibid ., 86, 2595 (196U); 3) ibid . , 9_2_, 5075 (19T0); k) Manuscript in preparation; 5) J. Am. Chem . Soc . , 508U (1970) ; 6) Manuscript in pre- paration. Acknowledgements : We are grateful to the National Science Foundation, the U.S. Atomic Energy Commission and Vanderbilt University for sponsoring these studies. G2-5 Luminescence of Some Metal Trisdithioacetylacetonate Complexes by M. K. DeArmond and J. Merrill Department of Chemistry, North Carolina State University Luminescence of metal complexes can within the limit of small mix- ing of ligand and metal d-orbitals be classified as "d-d" emission (localized orbital), charge transfer (delocalized orbital) or ligand localized orbital (delocalized orbital) . Since emission typically occurs only from the lowest manifold of excited states only localized "d-d" orbital emission is observed for first transition series complexes while either localized and delocalized orbital emission can be observed for heavy d transition metal complexes dependant upon the ligand used. For example, tris acetylacetonato Cr(III) exhibits lowest energy "d-d 2 4 absorption and a characteristic narrow structured E ■> A„ emission while the bidentate sulfur chelates, tris ethylxanthatochromium(III) (Cr(exan)~) and trisdithiocarbamato chromium (III) (Cr(dtc)~) exhibit lowest energy (intense) d-d absorption and "d-d" emission. Emission of c. d Rh(III) and Ir(III) complexes can be categorized as "metal localized 3+ orbital" (i.e. Rh(dtc) 3 ) or "delocalized orbital", ([Rh(dip)„] where dip = 2,2'-dipyridyl) . The former type of emission is broad, structure- less and exhibits a large Stokes Shift while the latter emission exhib- its vibrational structure and a relatively small Stokes Shift. The 3-diketone complexes of Rh(III) and Ir(III) have been report- 2 ed to exhibit d-d emission at 11.8 and 11.3 kK although the spectra was so weak that no lifetime could be measured. Such an emission is unique since no d-d absorption can be identified at energies lower than the intense delocalized orbital transitions. However large Stokes Shifts of emission are characteristic of d localized orbital emitters, there- G3-1 fore the d-d absorption may be covered by the intense low energy bands. To better understand the nature of the interaction between d orbi- tals and high energy ligand Il-orbitals and the resultant energy trans- fer processes, neutral tris dithioacetylacetonate complexes, M(sacsac)^ of Cr(III), Rh(III) and Ir (III), were synthesized absorption and emiss- ion spectra determined and lifetimes measured. In the limit of small mixing of ligand and metal orbitals, the low lying d-d absorption although hidden by the more intense deloca- lized orbital transitions, may still permit typical d-d emission to be observed. But in the limit of substantial mixing of metal and ligand orbital the emission should be of the characteristic "delocalized orbital" type for these weak field ligands. All sacsac complexes (Cr(III), Rh(III), and Ir(III)) were synthe- 3 sized according to literature methods, purified by recrystallization and preparative thin layer chromatography. All compounds were analyzed for C, H, and S by Atlantic Microlab, Inc., Atlanta, Georgia and gave proper stoichiometry. Absorption spectra were determined at room and liquid nitrogen temperature using a Cary 14 spectrophotometer. Emission and excitation spectra were determined with an Aminco-Bowman Spectrophotof luorometer at 77°K in EPA solvent with both a Hamamatsu R-136 and RCA 7102 red sensitive tube. Emission decay was determined using a Laser Energy Incorporated (LEI) 20&/att pulsed KL laser (A % 337 nm, pulse width = 5 nsec) both with the single shot technique and oscilloscope detection and using pulse averaging technique with a Princeton Applied Research (PAR) G3-2 Boxcar Averager Model 160 with an X-Y recorder. Lifetime values were determined from decay curves using a least squared program. The Cr(sacsac)~ complex gave no emission out to 1.0 Um. The Rh(sacsac)~ and Ir (sacsac) „ gave broad (v . = 1840 cm , V , = 2090 cm ) emission bands with maxima at 12.9 kK and 12.3 kK. The lifetime values measured for the Rh(sacsac)« and Ir(sacsac)~ were at 77°K 23.8 usee and 1.13 usee. The experimental Stokes Shift is for these sacsac's much smaller than that for the d-d emission of the Rh and Ir complexes of 4 dtp (diethydithiophosphate) and dmtc (dimethyldithiocarbamate) . More- over the intensity of the lowest absorption band for these sacsac's is greater than that of a typical d-d band. On the basis of the energy, shape and lifetime of the emission, the Ir(sacsac)- and Rh(sacsac)„ emission are best assigned as "d-d" localized orbital emission with the corresponding absorption transition covered by an intense charge transfer transition. The character of the lowest excited states for Cr(sacsac)- cannot be determined since no emission could be observed. The absence of emission may result from either a very low quantum efficiency and/or it may be beyond 1 um. Since, within the crystal field theory, the 2 position of the E state is relatively independant of crystal field strength and dependant only upon the Racah parameters, B and C, tris aca, dmtc, and exan (ethylxanthate) and dtp complexes can permit an 9 L approximation to the position of the E -> A band in the Cr(sacsac).. 2 4 These compounds (with the exception of Cr(dtp)^) show E -*■ A phos- phorescence in the 12,000 cm region. Therefore, absence of emission G3-3 for Cr(sacsac)^ can be interpreted as for Cr(dtp)~ to indicate that the 4 2 T» minimum lies lower than the E state, therefore, that a very weak red fluorescence (beyond 1 ym) occurs. Alternatively, a strong mixing of d and IT orbitals may occur such that the emission (if observable in the infrared) cannot be classified within the small mixing limit. Attempts to synthesize additional S analogs of the 3-diketone com- III III III plexes of Cr , Rh and Ir are continuing. Synthesis of a filled shell complex of the dithioketone, In(sacsac) „ , has not been success- ful. Such a compound if available may permit location of the lowest IT — IT* triplet for the ligand. References 1. M. K. DeArmond, Accounts Chem. Res., 7_, 309 (1974). 2. G. A. Crosby, R. J. Watts and S. J. Westlake, J. Chem. Phys., 55_, 4663 (1971). 3. G. A. Heath, R. L. Martin and A. F. Masters, Aus . J. Chem., 25 , 2547 (1972). 4. J. E. Hillis and M. K. DeArmond, Chem. Phys. Letters, 10, 325 (197]) 5. W. J. Mitchell and M. K. DeArmond, J. Luminescence, 4, 137 (1971). G3-4 Luminescent Transition Metal Complex Photosensitizers J. N. Demas , J. W. Addington, R. P. McBride, E. W. Harris, and S. Peterson, Chemistry Department, University of Virginia, Charlottesville, Virginia 22901, USA. Ruthenium ( II ) , osmium(ll) and iridium(lll) complexes with diimine ligands possess nearly ideal properties as photosensitizers. They absorb intensely, emit strongly in fluid solution, have long lives, are generally highly photo- stable and exhibit wavelength independent efficiencies of population of the emitting state. Table I shows properties of thirteen complexes which have been studied as photosensitizers. The zero point energies of the excited state, E 's, the mean lifetimes, "^o ' s , the bimolecular quenching constants for oxygen deactivation, k 2 ' s and the limiting quantum yields, 0o's, for photooxygenation of tetramethylethylene (TME) at in- finite concentrations of O2 and TME for each complex are indicated; all values are for methanol at ~ 21°C. The 0o ' s are all less than unity. The Rose Bengal yield of 0.80 equals its intersystem crossing yield of 0.84 within experimental error. Thus, the subunity yields for the complexes do not arise from inefficient reaction of 1 02 with TME. Since the intersystem crossing/ internal G4-1 conversion efficiencies for complexes of this type has been shown to be unity, the inefficiencies of photooxygenations are caused by an inefficient energy transfer step. We attribute this inefficiency to formation of an exciplex which undergoes radiationless deactivation before dissocia- tion to form 1 2 - From a combination of intensity and decay time quenching data, it has been found that Cu 2 , Co 2 and Ni 2 quench Ru(phen ) 2 (CN) 2 and Ru(bipy ) 2 (CN) 2 by both static and dynamic processes. The association constant estimated from these data are given in Table II. Since other neutral complexes without CN groups do not give static quenching, and since CN can be a bridging ligand, the association appears to arise from the RuL 2 (CN) 2 complex acting as a ligand for the metal ion to yield a Ru-CN-M species. Ru(bipy ) 2 (CN) 2 and Ru(phen) 2 (CN) 2 are also protonated, presumably at the CN. In the case of Ru(bipy ) 2 (CN) 2 we find evidence for reversible acid-base equilibrium in the excited state. Details will be presented. 1) J. N. Demas and J. W. Adding ton, J. Amer . Chem. Soc, in press . 2) J. N. Demas, E. W. Harris, C. M. Flynn, Jr., and D. Diemente, J. Amer. Chem. Soc, 97_, 3838 (1975). G4-2 3) J. N. Demas and J. W. Addington, J. Amer. Chem. Soc . , 96_, 3663 (1974). Table I Properties of Transition Metal Complexes a b Photosensitizers ' in Methanol at ~ 21°C Complex E ?o k 2 xlO~ 9 O (kK) (usee) (M^s"" 1 ) (TME) [Ru(bipy) 3 ] 2 + 18.0 0.765 1.8 0.86 [Ru(phen) 3 ] 2+ 18.4 0.313 3.3 0.75 [Ru(Clphen)(phen) 2 ] 2+ 0.947 2.3 0.81 [Ru(Brphen) (phen) 2 ] 2 0.989 2.3 0.80 Ru[(SO 3 2 )phen] (phen) 2 17.9 3.98 1.8 0.82 Ru(bipy) 2 (CN) 2 18.3 0.40 2.8 0.79 Ru(phen) 2 (CN) 2 18.3 1.58 5.0 0.68 [0s(bipy) 3 ] 2 + 14.9 0.049 4.5 — [0s(phen) 3 ] 2+ 15.3 0.183 5.7 0.76 [Os(0 2 phen) (phen) 2 ] 2 15.0 0.212 4.6 0.78 0s[ (SO 3 0) 2 phen] (phen) 2 14.8 0.093 6.8 0.74 [Ir(bipy) 3 ] 3+ 22.7 2.4 0.34 — [Ir(phen) 3 ] 3+ 22.5 2.6 0.28 — Rose Bengal — — — 0.80 G4-3 Table I - continued a) bipy = 2,2 ' -bipyridine, phen = 1, 10-phenanthroline, Xphen = 5-halo-l, 10-phenanthroline, 2 phen = 4,7- diphenyl-1, 10-phenanthroline, (SO 3 0) 2 phen = disulphonated-4,7-diphenyl-l,l0-phenthroline. b) Data taken in part from References 1-3. Table II Association Constants for Metal Ions with Ruthenium ( II) -Cyanide Complexes in Water at ~ 22°C K eq (M _1 ) Ru (b ipy) 2 (CN) 2 Ru (phen) 2 (CN) 2 Cu* + 220 270 Ni 2+ 10 20 Co 2+ 11 17 G4-4 Picosecond Studies of Transition Metal Complexes Patrick E. Hoggard Department of Chemistry, Polytechnic Institute of New York Alexander D. Kirk Department of Chemistry, University of Victoria, B.C., Canada Gerald B. Porter Department of Chemistry, University of British Columbia Mark G. Rockley Department of Chemistry, Oklahoma State University Maurice W. Windsor Department of Chemistry, Washington State University Lifetimes of excited electronic states of a number of transition metal complexes have been measured in solution at room temperature with picosecond time resolution [1]. Two experimentally distinct lifetimes can be retrieved with this technique. The ground state bleaching (GSB) lifetime is that for the total repopulation of the ground state, measured by the regrowth in intensity of absorption bands, initially bleached because of depopulation by the pumping pulse. The excited state absorption (ESA) lifetime is that for depopulation of a particular excited state, measured by the decay of the electronic absorption spectrum from that state. Both of these processes are observed only under somewhat special conditions, and in addition we have thus far been limited by the necessity that the sample absorb significantly at 530 nm, 3+ the double frequency of a Nd laser. To observe GSB, the molar extinction coefficient at 530 nm must be higher than about 5000. A similar requirement applies to ESA - the extinction coefficient G5-1 must be high somewhere within the range of the instrumentation and the continuum pulse used for analysis. Additionally, observation is impossible if the excited state absorption is masked by the ground state absorption. In spite of these restrictions, a number of transition metal complexes have been found for which we can observe GSB and/or ESA. A summary of some of these results is shown in the Table. In the case of the Fe(II) complexes, there are two absorption bands in the visible region, an intense band above 500 nm, represen- ting a CTTL transition, and another much weaker band below 900 nm, 3 1 assigned to a T. ■+■ A- ligand field transition. In addition, 3 1 transitions from CT and T- states are probably masked by the spin allowed charge transfer band. An estimate of the radiative life- 1 time of the CT state from the absorption spectrum plus the absence of observable luminescence allows us to rule out this state as too short-lived to contribute to the observed GSB lifetime. The 3 830 ps lifetime is consistent with estimates for the T_ lifetime, lg 3 but could represent decay into the ground state from any of T , 1 3 T. , or CT. lg Chromium (III) complexes exhibit small extinction coefficients in general, so that GSB is not expected to be observed. Excited 4 state absorption from the lowest ligand field states, T~ and 2 E would also be expected to exhibit small extinction coefficients g within the manifold of ligand field states. Observation of ESA in the visible range is thus to be expected only with ligands which induce charge transfer states in the region of 30-40 kK above G5-2 the ground state. Thiocyanate and acetylacetonate complexes exhibit the required CT bands, and ESA was observed for these complexes. 4 Although the T_ state is reached by absorption, the ESA 2 spectra of all 3 complexes match the ESA of the E state measured at 77°K by Ohno and Kato [2]. Unless the ESA spectra of 4 2 T_ and E states are fortuitously the same, it appears that 2g g 4 2 the T_ state undergoes intersystem crossing to the E state in a time less than the duration of the pump pulse ( ca . 5 ps) . References 1) A. D. Kirk, P. E. Hoggard, G. B. Porter, M. G. Rockley, and M. W. Windsor, Chem. Phys. Lett ., 37, 199 (1976). 2) T. Ohno and S. Kato, Bull. Chem. Soc. Japan , 43 , 8 (1970). G5-3 CO H c <+-( o o •H •U 01 •H a 0) >•. a H cfl l-i H ■u 1= CO o ■u •H en +J P- 13 u CU o 4J to •H 43 O < X w CU 4-1 CO 60 4-> C C/D -H 43 T3 O C CO 3 Oi O rH j-i m o c CO 00 •H rH n CO CD U U-l 4J CO e £3 M CO M ^ H rH 43 a> CO 00 43 M o CO rl 43 PM a CO CO p- TJ a o 0) o CO > CM 00 at VD II en 43 II H O H V) n CU CD »w M-t CO CO c C CO CO M hJ i-i H + H 0) CU 00 S 00 n M CO CO 42 43 c_> c_> S a a e c c fi o o o m CN o CO ■d- + + W cz> X CN CN CJ O 1 CU r""i r— i N— ' z I-— » CO H CO CO CN v— ' CO a <»~N ^N /""s CN •"> V B >> C >> /-v o •-s o D- CU P. CO CO Crt CJ> •H 4= •H 3* O c_> 43 a. 43 § CO 525 v— / N—' v—/ v—' v~> V— ' CU cu 3 u M M Pn Pn rt u U O >£) G5-A Electron Transfer Properties of Excited States of Transition Metal Complexes by V.Balzani , F.Bolletta, M. Maestri, A. Juris, and N.Serpone Istituto Chimico "G.Ciamician" dell'Universita, Bologna, Italy. Until a few years ago, the attention of inorganic photo- chemists was entirely focussed on the intramolecular photo- reactions and on the photosolvation reactions of transition 2 1 metal complexes. ,J In the last few years, however, there has been an increasing interest in the intermolecular redox properties of transition metal complexes. The protagonist of this new trend in inorganic photochem- 2+ istry is the Ru(bipy).. complex, whose lowest excited state, commonly indicated by (- > CT)Ru(bipy)^ since it has triplet multiplicity and metal-to-ligand charge transfer character, is able to act as an energy donor, electron donor, and 11 12 electron acceptor * depending on the specific nature of the reaction partner: ( 3 CT)Ru(bipy)2 + + A ****** > Ru(bipy)!: + + * A (1) 1 ( 3 CT)Ru(bipy)^ + + B red * > -Ru(bipy) 3+ + B" (2) ( 3 CT)Ru(bipy)^ + + C kox * > Ru(bipy)+ + C + (3) The ( J CT)Ru(bipyK excited state is luminescent in fluid solution at room temperature, a very precious property be- cause it provides a simple means of monitoring the interac- tions between the excited state and the quencher. The life- 's p . time of (- > CT)Ru(bipy)f is 680 nsec in deaerated aqueous so- G6-1 lution. The absorption spectrum of this excited state has recently been obtained by means of laser flash spectroscopy and shows a maximum at about 360 nm. -* The quantum yield of triplet formation, which has been measured with different 1 1 techniques, is certainly higher than 0.5 and probably near unity, • ' It should also be recalled that the Ru(bipy)? + complex absorbs strongly in the visible, with a maximum at 450 nm ( £=15,000). The "thermodynamic" barriers for reactions 1,2, and 3 may be evaluated knowing that the energy difference between ( 3 CT)Ru(bipy)^ + and ground state Ru(bipy)^ 1 " is 17.1 kK (2.12 eV),and that the reduction potentials of the (^CT)Ru(bipy)r + - -Ru(bipy)* and Ru(bipy)^ + -(^CT)Ru(bipy)^ + couples are +0.84 V 11 and -0.83 V 11 ' 16 , respectively, vs. the NHE. Kinetic barriers have also been found to play an important role in 6 Q 11 both energy and electron ' transfer reactions; unfortu- nately, systematic studies are still lacking. It is inter- esting to note that reactions 2 and 3 provide for the con- version of light energy into chemical energy. The products of these reactions, however, are difficult to separate and generally undergo a fast back electron transfer reaction which dissipates the converted light energy as heat in the reaction medium. Indeed, when the two products of reaction 2 (or 3) have opposite electric charge the back electron transfer reaction is too fast to be seen with conventional flash techniques. ,17 However, relatively high stationary G6-2 concentrations of the products can be obtained in more favor- able cases, and this can be used to produce a photogalvanic effect. 18 Another very interesting complex from the point of view of electron transfer reactions is Cr(bipyH + . Its lowest excited state, which has doublet multiplicity and metal-centered character, is relatively long lived ( about 50/isec) and emits luminescence in fluid solution. As the Cr(bipy)^ 4 "- Cr(bipyK reduction potential is -0.25 V and the spectro- scopic excitation energy of ( MC)Cr(bipy)< + is 13.8 kK (1.72 eV), the reduction potential of the ( MC)Cr(bipyH + -Cr(bipy)? + 1 Q 2 couple is expected to be about +1.45 V. y Thus, the ( MC)- Cr(bipy)^ excited state is expected to act as a very strong oxidant. This expectation seems to be verified since ( MC)- Cr(bipyH + is quenched at nearly diffusion controlled rates by a number of species which do not possess excited states lower than 13.8 kK but have E°(Q-Q") lower than 1.45 V. 20 For example: ( 2 MC)Cr(bipy)^ + + I" 22Ll-> Cr(bipy) 2+ + I (4) The quenching of ( MC)Cr(bipy)x by electron transfer to the quencher is presently under investigation in our laboratory. One of the most interesting systems we have examined thus far is that given by a solution containing Ru(bipy)-, and Cr(bipy)^ + . '» 19 The excited Ru complex is able to reduce the ground state Cr complex, and the excited Cr complex is able to oxidize the ground state Ru complex. In both cases G6-3 one obtains the Ru(bipy)^ and Cr(bipy)-. complexes which then undergo a back electron transfer reaction to reach their equilibrium situation: Ru(bipy)^*" + ( 2 MC)Cr(bipy)^ + t.. J Nft k V Ru(bipy)^ + + Cr(bipy)^ + < c _ Ru(bipy)^ + Cr(bipy)^ + J hv ! y k b ( 3 CT)Ru(bipy) 2+ + Cr(bipy) 3+ We have here a system in which light absorption leads to the same products, regardless of the absorbing species. The rate q constants for the two excited state reactions are k =3.3x10 a 11 ft 1 1 mol" s~ and k, =4.0x10 mol" s~" . The reason for k„>k, is d a b probably due to kinetic factors which are related to the dif- ferent electronic distribution in the two excited states. The redox properties of the excited states of other Ru(Il) and Cr(lll) complexes as well as of other transition metal complexes are currently under investigation. Acknowledgment . - Pinantial support from the Italian Na- tional Research Council is gratefully acknowledged. References and Notes 1. On leave from the Department of Chemistry, Sir George Williams Campus, Concordia University, Montreal, Canada. 2. V.Balzani and V.Carassiti, "Photochemistry of Coordination Compounds", Academic Press, London, 1970. 3. A.W.Adamson and P.D.Pleischauer, eds. , "Concepts in Inor- ganic Photochemistry", Wiley, New York, 1975. G6-4 4. Strictly speaking, the spin label is meaningless because of the dominant role played by spin-orbit coupling. 5. G.A.Crosby, K.W. Hipps, and W.H.Elfring Jr., J.Am.Chem. Soc, 96, 629 (1974). 6. V.Balzani, L.Moggi, M.P.Manfrin, F.Bolletta, and G.S. Laurence, Coordination Chem.Rev. , V5» 321 (1975). 7. H.D.Gafney and A.W.Adamson,J.Am.Chem.Soc. ,94,8238 (1972). 8. C.R.Bock, T.J.Meyer, and D.G.Whitten, J.Am.Chem. Soc. , 96, 4710 (1974). 9. G.Navon and N.Sutin, Inorg.Chem. , 13., 2159 (1974). 10. G.S.Laurence and V.Balzani, Inorg.Chem., J^3, 2976 (1974) 11. C.Creutz and N.Sutin, Inorg.Chem., in press. 12. A. Juris, M.T.Gandolfi, M.F.Manfrin, and V.Balzani, J.Am. Chem.Soc, 98, 1047 (1976). 13. R.Bensasson, C.Salet, and V.Balzani, submitted. 14. J.N.Demas and G.A.Crosby, J. Am. Chem.Soc. ,93., 2841 (1971). 15. F.Bolletta, M. Maestri, and V.Balzani, submitted. 16. C.R.Bock, T.J.Meyer, and D.G.Whitten, J. Am. Chem.Soc. , 97, 2909 (1975). 17. R.Ballardini, G.Varani, and F.Scandola, to be published. 18. C.T.Lin and N.Sutin, J.Phys.Chem. , 80, 97 (1976). 19. F.Bolletta, M.Maestri, L.Moggi, and V.Balzani, J.C.S. Chem.Commun. , 901 (1975). 20. M.Maestri, N.Serpone, F.Bolletta, L.Moggi and V.Balzani, manuscript in preparation. G6-5 A Comparison of the Excited-State Electron-Transfer Reactions Of Ru(bipy)^ + And Os(bipy)^ + Peter Fisher, Edward Finkenberg, And Harry D. Gafney City University of New York Department of Chemistry Queens College Flushing, N.Y. 11367 2+ The mechanism of quenching of Ru(bipy) (bipy denotes bipyridine) luminescence is an area of active investigation. Recent studies with organic and inorganic substrates have established that quenching can occur by an electron transfer or energy transfer mechanism. For the quenching of the luminescent charge transfer state, LCT, of * Ru(bipy) by the acidopentaamminecobalt(III) complexes, Co(NH ) X , the evidence, at present, supports an electron-transfer mechanism. A detailed study of the reduction of 13 Co(NH ) X complexes 2+ by * Ru(bipy) offers further support for an excited-state electron- 3 2 2+, transfer reaction. The reduction potential of * Ru(bipy) , 0.8v, Ices 1 and small values, 2.0 - — t— ■ * of the energies of activation suggest ' mole to 6& the reactions are not thermodynamically controlled. As indicated by the data in Table I, some degree of correlation exists between 6 • and the thermal rates of reduction of the cobalt(III) com- plexes. This correlation suggests the reactions are kinetically controlled, that is, the rate of electron transfer must be competitive with the rate of relaxation of the excited state of the electron donor. The correlation is not perfect, indicating other factors are also 2+ important. Since * Ru(bipy) , a strong reductant, is converted on 3+ electron transfer to a strong oxidant, Ru(bipy)» , a necessary cri- G7-1 teria for a net electron transfer to occur is that the oxidant undergo an irreversible reduction. For these cobalt (III) complexes, an irre- versible reduction is accompanied by dissociation of the coordinated ligands. The differences in reactivities of Co(NH3)5H20^ + , Co(NH3)5NCS 2+ , Co(NH3)5Ng + , and CoCNt^^SoJ, only the latter two show a measurable ^CoClI)* ^ s attributed to differences in the rate of ligand dissociation and/or intersystems crossing within the reduced cobalt substrate. The latter rate must be competitive with the rate of reverse 3+ electron transfer from the reduced cobalt substrate to Ru(bipy)3 in order for a net reaction to occur. To further test these conclusions, a study of the reduction of the same cobalt (III) complexes by *0s(bipy)3 under identical conditions has been undertaken. Like *Ru(bipy) 2+ , *Os(bipy)? + is of sufficient potential to reduce all cobalt (III) complexes studied. The potential of the reaction *Os(bipy) 2+ + 0s(bipy)3 + + e" has recently been calculated to be 0.96vf" consequently Os(bipy) 2+ is a better reducing agent than *Ru(bipy) 2+ , E° = .84v. Although a stronger reducing agent, the lifetime of *Os(bipy) 2+ , 19.2 nsec, is considerably shorter than that of Ru(bipy)^ , 600 nsec. A priori it was expected that Os(bipy)^ would reduce only those cobalt (III) com- > 7 1 —1 plexes which had thermal rates of reduction _ 10' M~ x sec , ±^. e_. , Co(NH 3 ) 5 Cl 2+ , Co(NH 3 ) 5 Br 2+ , and Co(NH 3 ) 5 I 2+ . As indicated by Figure I, the available data indicates Os(bipy)^ reacts with cobalt (III) com- plexes which have thermal rates of reduction < 10' M -1 sec"-'-. This sug- gests that the rate of electron transfer is much faster than the rate G7-2 rate of relaxation. Furthermore, the limiting yields of $r (-rT) obtain- ed with Os(bipy)„ , Figure I, are essentially unity and much larger than those obtained with Ru(bipy) , Table I. The increased limiting & 9-4- efficiency of Os(bipy)^ is though to reflect a decrease in the rate of reverse electron transfer due to the lower oxidation potential of Os(bipy)^ + . References 1. C.T. Lin and N. Sutin, J. Phys. Chem. , 80, 97 (1976) and references therein. 2. P.K. Lam, A.W. Adamson, and H.D. Gafney, submitted to J. Amer. Chem. Soc. . G7-3 TABLE I Comparison of the Quantum Yields of Reduction of Various Co(NHo)cX Com- plexes by RuCbipy)^"** with the Thermal Rates of Reduction by Cr(bipy) 2+ . „ . .00 —1 _1 \\ Complex* 0^^ 0^^ k(M sec i ) b RNH 3+ <.001 - 6.9 x 10 2 RCN 2+ <.001 RH 2 3+ <.001 - 5.0 x 10 4 RF 2+ <.0003 - 1.8 x 10 3 R00CCH 2+ <.001 - 1.2 x 10 3 RNCS 2+ <.0001 - 1.0 x 10 4 RCot 0.014 3 '3 RN 2+ 0.051 RSol" 0.043 4 RC1 2+ 0.063 RBr 2+ 0.104 2+ RI 0.141 r'c.o"!' 0.097 2 4 R=Co(NH n , R =Co(NH ) 3 ; 5 3 4 0.1 4.1 x 10 4 0.1 4.5 x 10 0.6 8.0 x 10 5 0.7 5.0 x 10 6 1.0 _ G7-4 Figure 1 Dependence ef ^(. / TI \ en the ceneentratien ef Ce(NH 3 ) JC • CoCNHj-Br* O Ce(NH 3 ) 5 Cl 2+ ■ Ce(NH 3 ) 5 N^ n+ D Ce(NH 3 ) 5 C0 3 250 l/(Ge(NH 3 ) 5 X n+ ), M -1 G7-5 Photodissociation Of Simple Polyatomic Molecules J. P. Simons Chemistry Department, The University, Birmingham B15 2TT, England Ten years ago, the majority of photochemists were satisfied if they were able to identify the primary products of a molecular photo- dissociation. Now, with the advent of the new heavy artillery of techniques for studying energy disposal and angular distributions they are much more demanding, and as the new results have flowed in, (2) the theoretical models have burgeoned to codify and rationalise the new data. Until recently, all the dynamical models for treating the problem of vibrational energy disposal adopted the quasi-diatomic approximation, an artificial device that resolves the full intra- molecular potential into two components but has the virtue of allowing the analysis of a complex process in a relatively simple way. Each of the publications follows the format - introduction to the physical model, mathematical analysis, comparison with current experimental data. The last section is essential, not only because it is the experimental horse that pulls the theoretical cart, but also because none of the models is predictive. This has often led to a 'Comedy of Errors' because the current data chosen for comparison, (commonly the near u.-v. 3 photodissociation of ICN or the quenching of Hg(6 P ) by CO or NO) , proved to be inaccurate. The most recent dynamical models have (3) chosen the vacuum u.-v. photolysis of the halogen cyanides for comparison, but here too, experimental re-appraisal of the data (4) suggests that some of the results may have been in error. Nevertheless there is now a general consensus on the importance of geometry and force Hl-1 constant changes in the transfer from the bound to the repulsive electronic surface during photodissociation, and the utility of a Franck-Condon description for 'predicting' vibrational energy distributions in the separating fragments. The new data for vibrational energy disposal in the vacuum u.-v. photodissociation of the halogen cyanides will be presented and compared with alternative theoretical models. The new results suggest that vibrational populations calculated from observations of the 2 + fluorescence of CN in the N E state can be modified by collisionally 2 induced transfer from fragments in the much longer-lived A n state, which has a lifetime in the region of (l-10)ys. Thus collisional effects can be important, even at pressures as low as 0.1 torr. Collisional interchange between the A and B states ( 6") has previously been suggested by Luk and Bersohm to account for the pressure dependence of the fluorescence decay of CN(B) , produced in the vacuum u.-v. flash photolysis of ICN. Results obtained using a new variation of the photofluorescence technique, polarised photofluorescence excitation spectroscopy, will be presented. The technique depends on the formation of rotationally excited fluorescent fragments which retain their memory of the orientation of the photo-selected parent molecule: the requirement is not restrictive since rotational excitation following photodissociation in the vacuum u.-v. appears to be the rule rather than the exception. Measurement of the polarisation of the fluorescence as a function of the absorbed photon energy provides information relating to the symmetry of the vibronic state Hl-2 initially populated and its lifetime prior to dissociation. The technique is a member of the family of anisotropic photo- dissociation experiments developed by Zare , Bersohn, Wilson and their co-workers. ' Polarised spectra obtained from the (7 9) halogen cyanides, HCN, HO and D ? will be discussed. Finally, a time of flight photofragment spectrometer will be described, which incorporates a tunable, frequency-doubled, dye laser as the polarised photolysis beam and a resonance fluorescence detection system. References (1) See, for example, (a) anisotropic photodissociation: R. Bersohn and S.-C. Yang, J. Chem. Phys. 61, 440 (1974); R. Bersohn, M. Dzvonik and S.-C. Yang, J. Chem. Phys. 61, 4408 (1974). (b) photofragment spectroscopy: G. E. Busch and K. R. Wilson, J. Chem. Phys. 56, 3626, 3638, 3655 (1972). (c) photodissociation laser spectroscopy: G. A. West and M. J. Berry, J. Chem. Phys. 61, 4700 (1974); M. J. Berry, J. Chem. Phys. 61, 3114 (1974). (2) For leading references see (a) Y. B. Band and K. F. Freed, J. Chem. Phys. 62, 4479 (1975), (b) S. Mukamel and J. Jortner, J. Chem. Phys. 60_, 4760 (1974). (c) M. J. Berry, Chem. Phys. Letters, 29, 323, 329 (1974): also ref. 1(c). Hl-3 (d) J. P. Simons and P. W. Tasker, Mol. Phys. 26, 1267 (1973); 27, 1691 (1974). (e) M. Quack and J. Troe, Ber. Bunsenges. 79, 469 (1975). (3) A. Mele and H. Okabe, J. Chem. Phys. 51, 4798 (1969). (4) M. R. N. Ashfold and J. P. Simons, unpublished work. (5) T. J. Cook and D. H. Levy, J. Chem. Phys. 57, 5059 (1972). (6) C. K. Luk and R. Bersohn, J. Chem. Phys. 58, 2153 (1973). (7) G. A. Chamberlain and J. P. Simons, J. Chem. Soc. , Faraday Trans, II, 71, 2043 (1975). (8) R. J. Van Brunt and R. N. Zafe, J. Chem. Phys. 48, 4304 (1968). (9) M. T. Macpherson and J. P. Simons, unpublished work. (10) M. R. Levy and J. P. Simons, unpublished work. Hl-4 Energy Distribution in the Photodissociation of Methylketene at 215 nm M. E. Umstead , R. G. Shortridge, and M. C. Lin Physical Chemistry Branch, Chemistry Division Naval Research Laboratory, Washington, D.C. 20375 The photolysis of methylketene has been studied by Kistiakowsky and co-workers. 1 ' 2 They concluded that the reaction proceeded prin- cipally by means of the following paths: CH3CHCO + hv -• CHCH 3 + CO (1) CHCH 3 •* C 2 H^ (2) c 2 nS "• C 2 H 2 + H 2 (3) t C 2 H 4 ' + M - C 2 H 4 + M (4) CHCH 3 + CH3CHCO - C 4 H 8 + CO (5) a mechanism which is consistent with that proposed by Frey for the photodissociation of diazoe thane. 3 We have investigated the photodissociation of methylketene (CH 3 CH=C=0) in a quartz flash system (X ^ 200 nm) at 300°K. The pro- ducts of the photodecomposition were analyzed by gas chromatography, and the initial vibrational excitation of the CO formed was measured by means of a laser resonance absorption method. 4 The results of the pro- duct analysis indicated the presence of CO, C 2 H 2 , and C 2 H 4 , with lesser amounts of CH 4 , C 2 H 6 , C 3 H 6 , C 3 H 4 (propadiene and propyne) and various C 4 hydrocarbons, and are generally in agreement with those of Chong and Kistiakowsky. 2 In order to measure the initial vibrational population distrib- ution of the CO, a stabilized cw CO laser preset at the various vibrat- ional-rotational CO lines was directed along the axis of a quartz flash H2-1 tube. Mixtures of methylketene in He were flash photolyzed, and the population distribution determined from time-resolved absorption curves measured for all vibrational levels populated by the photodissociation. The CO laser resonance absorption measurements showed that the CO formed in the photodissociation of methylketene in the 200-230 nm region was vibrationally excited to v=9 and had a Boltzmann vibrational temper- ature of 3800 ± 500°K. The results of two sets of C 3 H 4 data are plot- ted in Fig. 1. One set was obtained from a 1% mixture of C 3 H 4 in He, flashed at an energy of 0.5 kJ, and the other from a 1.5% mixture flash- ed at 1.0 kJ. The fact that the two sets of data led to the same pop- ulation distribution, even though the C 3 H 4 concentrations and the flash energies differed considerably, provides evidence that the excited CO is indeed produced in the primary step, and not by secondary reactions. This is also supported by an experiment in which 2 was added as a free radical scavenger. A mixture containing 1% C3H4O and 2% 2 in He was flashed at an energy of 1.0 kJ. The results of the CO excitation measurements were the same as those from the 2 -free mixtures, within experimental error. The amount of vibrational energy channelled into CO was calculated from the expression = S f E V v*0 V V where f = N /£ „_ N is the normalized vibrational population distri- v v v^0 v bution and E is the vibrational energy of CO at the vth level with the v zero-point energy excluded. The data presented in Fig. 1 give rise to E = 3.7 ± 0.34 kcal/mole, which indicates that only about 3% of the H2-2 total available energy, E = hv - AHo + RT a= 126 kcal/mole, is carried by the CO as vibrational energy. The results obtained for methylketene are listed in the Table and are compared with those obtained for the isomeric compound, acrolein, 5 and the chemically activated systems, 0( 3 P) + propadiene and propyne. In the propadiene and propyne reactions, it has been shown, based upon calculations in terms of simple statistical models, that the reaction energies have been randomized in the activated complexes before disso- ciation takes place. 6 Randomization of E t from the propadiene reac- r tot tion in an activated cyclopropanone complex followed by its direct dissociation into CO and ethylene provided good agreement between the calculated and the experimental values of (E ). In the case of propyne, however, an excited methylketene complex was assumed, along with its dissociation into CO and CH 3 CH which subsequently isomerized to C 2 H 4 . Thus the heat of isomerization of CH 3 CH, AH? so = - 68 kcal/mole, is not available for CO excitation, and randomization of E - AHicr, among the tot ■ LSO modes of the complex provided a calculated value for (E ) which agreed with the experimental. Also shown in Fig. 1 are the experimental CO population distribu- tion from the photodissociation of acrolein, and calculated curves for methylketene, Curves I and II. Curve I is the distribution obtained by randomizing the full E^ in the statistical model and clearly predicts ° tot J c too hot a CO vibrational distribution. Curve II assumes dissociation of the complex into CH 3 CH and CO*, and thus is based upon randomization of E - AHi so . It provides better agreement with the experimental values, but indicates that energy randomization is not complete before H2-3 dissociation takes place. The results obtained from methylketene and acrolein are very similar to each other. At this time, it is not known with certainty whether this indicates that the photodissociation of both compounds proceeds via similar complexes, or is merely fortuitous. Table Average Vibrational Energy of CO Formed in the Decomposition of Various C3H4O Molecules# Reaction C 3 H 4 E tot <\> -==0 -==0* ~126 3.7*0.3 =-=0 =-=0* -131 3.3* .2 04== i f 123 6.8* .6 0+-s -- ot 122 2.3* .3 #A11 energies in kcal/mole. References 1. G. B. Kistiakowsky and B. H. Mahan, J. Am. Chem. Soc, J79, 2412 (1957) 2. D. P. Chong and G. B. Kistiakowsky, J. Phys. Chem., 68, 1793 (1964) 3. H. M. Frey, J. Chem. Soc, 1 962, 2293 4. M. C. Lin and R. G. Shortridge, Chem. Phys. Lett., 29, 42 (1974) 5. R. G. Shortridge, M. E. Umstead and M. C. Lin, Presented at VIII International Conference on Photochemistry, Aug 1975 6. M. C. Lin, R. G. Shortridge and M. E. Umstead, Chem. Phys. Lett., 37, 279 (1976) H2-4 k* |.o ^-==0, .5 kj 1.55s- ==o, |.o kj — - 2.0%=-=o, |.o kj — STAT. CALC. (sec the text) 10 20 30 40 50 60 Ev/kcal-mole Figure 1. Vibrational energy distributions of the CO formed in the photodissociation of methylketene and acrolein. H2-5 The Production And Reactions Of Vibrationally Excited 1 ,1,2, 2-Tetrachloroe thane By M.H.J. Wijnen Chemistry Department, Hunter College Of The City University of New York 695 Park Avenue, New York, N.Y. 10021 Decomposition reactions of vibrationally excited halo- ethanes have been investigated extensively by Setser and 1) 2) coworkers and by Pritchard ' and coworkers. In the past vibrationally excited chloroethanes were produced by the photolysis of ketene in the presence of chlorinated methanes or by photochemical reactions of mixtures of chlorinated methanes. Both techniques yield two different radicals and thus three vibrationally excited ethanes (the direct combi- nation and the cross-combination products of the radicals involved) . It would, of course, be advantageous to study a system in which only one type of excited molecule is produced. Such a system is provided by the photolysis of phosgene in the presence of chlorinated methanes. To illustrate this method we have studied the photolysis of phosgene in the presence of CH 2 C1 2 at various pressures and in the presence of octafluorocyclobutane (OFCB) as an inert deactivator. The important reactions occurring in this system are H3-1 given byt C0C1 2 + hv CI + CH 2 C1 2 2 CHC1 2 TCE* + CH 2 C1 2 TCE* + C0C1 2 TCE* + OFCB TCE* -> 2 CI + CO -> CHC1 2 + HC1 -> TCE* (1,1,2,2-C 2 H 2 C1*) -> TCE + CH 2 C1 2 la -> TCE + C0C1 2 lb > TCE + OFCB lc — > C 2 HC1 3 + HC1 2 As clearly shown in the mechanism only one type of vibra- tionally excited molecule is produced, thus simplifying the analysis and the interpretation of the results. We have carefully looked for decomposition products other than HC1 elimination products. We have not observed any evidence for the occurrence of H 2 or Cl 2 eliminations from 1,1,2, 2-tetrachloroethane . The following experiments were carried outi A) C0C1 2 was photolysed in the presence of CH 2 C1 2 (ratio COClVCHgClg * 1) at total pressure between 0.3 and 5*1 torr. B) C0C1 2 was photolysed at a constant pressure of 0.85 torr, the pressure of CH 2 C1 2 was varied from 0.85 to k.6 torr. C) C0C1 2 (part, press. 0.85 torr) was photolysed in the presence of CH 2 C1 2 (0.85 torr), while the pressure of OFCB was varied from zero to 6.2 torr. The data were plotted according to the following H3-2 equations t R TCe/ R C 2 HC1 3 = (k la + k lb*/ k 2 P part. (A) k lb (C0Cl 2 )A 2 + k la /k 2 CH 2 C1 2 (B) k la + k lt/ k 2 P part. + k lc/ k 2 (0FCB > < C > The plot of Rm CE /R c u C i versus the partial pressures of C0C1 2 ( s CHgClp) should give a straight line without inter- cept. The ratio Rm CE /R c Hcl varies linearly with pressure hut a small intercept is observed, indicating that perhaps some other minor reactions may play a part in the mechanism. The following quantitative data were obtained i < k la + k lb>A 2 = 3.8 torr" 1 (Eq. A) k la/ k 2 ss 1 . 5 torr" (Eq. B) k lt/ k 2 = 4.1 torr (Eq. B) (k la + k lb )A 2 sr 5.9 torr" (Eq. C) k lo/ k 2 s 2.2 torr" (Eq. C) t is clear that the data are quite consistent since three different plots (obtained from different series of experiments) yield a value of 3*8 t 5*6 and 5*9 torr" for the sum of the deactivation steps la and lb over the decomposition step 2 thus confirming the simplicity and reliability of our method. A comparison of our data with those available in the literature will be made. H3-3 1) Setser, D.W. and coworkers, J.Chem.Phys. £6, 2 ^3 (1972) and earlier publications. 2) Pritchard, G.O. et al, J.Am.Chem. Soc. 2£» W*2 (1968) and earlier publications. H3-4 Measurement of Branching Ratios for the O+CS — >OCS+S Reaction Ronald E. Graham and David Gutman Department of Chemistry Illinois Institute of Technology Chicago, Illinois 60616 The reaction between oxygen atoms and carbon disulfide has 1 2 been shown to proceed by all three possible exothermic channels, ' — k >CS+S0 + 31 kcal (1) o+cs 2 k 2 >0CS+S + 55 kcal (2) -k 3 >C0+S 2 + 83 kcal (3) Highly vibrationally excited CO is also produced in this system by a rapid secondary reaction involving the CS produced in Route (1) , 0+CS >C0* + S. (4) This vibrationally excited CO is capable of producing either pulsed and continuous laser emission in a variety of experimental arrange- ments. The kinetics of this system has been studied to better 3 4 understand the performance of this chemical laser. ' It has been shown that the products of Routes (2) and (3) can significantly effect the laser output. The OCS produced by Route (2) selectively relaxes the lower excited vibrational states of CO enhancing the laser output from higher levels and decreasing it from lower levels. The CO produced in Route (3) is itself vibrationally excited and is produced with a different vibrational population distribution than that produced in Reaction (4). H4-1 Quantitative information on all three routes is still needed to better understand the factors which govern the performance of the chemical laser driven by this reaction. We report here measurements of the branching ratio for Route (2), R~ = k~/k, as a function of temperature. Slagle, Gilbert, and Gutman have 2 measured R at 302 °K and found it to be 0.093. Nielsen and Bauer have used this value in their modeling study of the CS +0 9 chemical laser and report that the OCS produced by Route (2) has a significant effect on laser output and is in fact the most important * 3 relaxer of CO near v = 5. This room temperature value of R_ was used since there have been no determinations of this branching ratio near the laser temperatures, 400-600°K, and there is no knowledge of its temperature dependence. Because of the demonstra- ted importance of Route (2) in this chemical laser, we have measured R_ at seven temperatures between 249 and 500°K. Oxygen atoms (produced by the N+NO — »N +0) reaction and CS were mixed in a fast flow reactor. The rate of CS loss and OCS production were simultaneously monitored by mass spectrometric analyses of gas sampled through a 0.033 cm diam hole in the end of the reactor. The experimental procedure and data analysis were the same as described before. The branching ratio for 2 Route (2) was obtained from the relation, R 2 = k 2 /k = A[0CS] t /A[CS 2 ] t , (I) where A[0CS] is the concentration of OCS at time t, and A[CS ] = [CS-] -[CS_] , is the decrease in [CS n ] from its initial value. L o 2. t 2. H4-2 At each of the seven temperatures between 249 and 500 °K, R was measured in 6-9 experiments and found to be independent of [0] , [CS„] , [M] , flow velocity, and extent of reaction. In o z o experiments in which [0] was in great excess, [CS ] vs. t profiles were used to also obtain k, the overall rate constant for the 0+CS 2 reaction. The results of these experiments are given in Table I. The branching ratio for Route (2) decreases only slightly in importance between 249 and 500°K dropping from 0.098 + 0.004 to 0.081 + 0.007. The small monatonic decrease in R_ , although statistically evident, cannot be used for lengthy extrapolations to other temperatures due to the large uncertainty of each determina- tion (about + 20%). We would suggest a value of R_ = 0.085 as probably being most appropriate in the 400-600°K range. The overall rate constants measured between 249 and 500°K are consistent with other determinations obtained from flow reactor studies in which [0] was in excess, but lie somewhat above those from studies in which [CS ] was in excess and in which a stoichiometric factor had to be determined and used. Calculations of "Prior" branching ratios predict lower values with virtually no temperature dependence. References * The authors gratefully acknowledge support for this research from the National Science Foundation. 1. I. W. M. Smith, Trans. Faraday Soc. 64, 378 (1968). 2. I. R. Slagle, J. R. Gilbert, and D. Gutman, J. Chem. Phys. 61, 704 (1974). H4-3 3. N. Nielsen, "Development of a Computer Model for the (CS-+CL) Chemical Laser", Department of Chemistry, Cornell University, 1974. 4. H. V. Lilenfeld, R. F. Webbink, W. Q. Jeffers, and J. D. Kelly, "Modeling of the CW CO Chemical Laser", I.E.E.E., J. Quantum Electr. (In Press). 5. J. W. Hudgens, J. T. Gleaves, and J. D. McDonald, J. Chem. Phys. 64, 2528 (1976). H4-4 TABLE I RESULTS OF EXPERIMENTS TO MEASURE O+CS RATE CONSTANTS AND BRANCHING RATIOS TRMri'KATijPE [::..< \>r kxlO 12 (°K) (k2/k) (cm 3 raolec" 1 sec" 1 ) 249 9.8 (+ 0.4)* 2.9 (+ 0.2)* 273 9.8 (+ 0.5) 3.6 (+ 0.3) 295 9.6 (+ 0.6) 4.1 (+ 0.2) 335 9.4 (+ 0.5) 5.1 (+ 0.6) 376 8.7 (+ 0.5) 6.6 (+ 0.3) 431 8.2 (+ 0.1) 8.5 (+ 0.6) 500 8.1 (+ 0.7) 11.2 (+ 0.8) * Average of at least 6 experiments, with one standard deviation given in parenthesis. Estimated accuracy + 20%. H4-5 Photodissociation of Molecular Beams of Metallic Iodides M. Kawasaki , H. Litvak, S.-J. Lee, and R. Bersohn Department of Chemistry, Columbia University, New York, New York 10027 Photodissociation processes are among the most important in photochemistry, but are difficult to study by conventional absorption methods; molecular beam photofragment spectroscopy, on the other hand, has proven to be a very effective technique for elucidating photodis- sociation dynamics. We report here photofragment spectroscopy studies of the elementary photodissociation processes of Til and Cdl2 at wave- lengths ^300 nm. The photochemical and reaction dynamics of metallic halides have long been subjects of interest, while photodissociation of iodo-compounds is particularly relevant to the development of new laser systems. Our apparatus has been largely described elsewhere [1] and con- sists basically of mutually perpendicular molecular beam, intersecting polarized photon beam, and ultrahigh vacuum (^10 torr) mass spectro- meter photofragment detector. The molecular beam effuses from a high temperature (^600K) oven in a separate vacuum chamber (^10 torr back- ground pressure) ; the photon beam is from either a Chromatix CMX-4 frequency-doubled tunable dye laser or a 1000W Hg-Xe lamp with appro- priate polarizer and filters (laser polarization is controlled by means of a Fresnel rhomb % wave plate). With this apparatus, we are able to observe (1) the photofragment mass spectrum, (2) the angular distribu- tion of recoiling fragments with respect to the electric vector of the polarized light, (3) the time-of-f light (TOF) distribution of photo- fragments arriving at the detector following a laser pulse, and (4) H5-1 the variation of the above quantities with photon energy. From the angular distribution f(6) one can obtain the symmetry of the dissoci- ative transition by use of the relation f(6) = -,— {1+$ [■? cos 2 (0-0 J- k ]} , where 9 is the angle be- tween the photon electric vector and the fragment recoil direction (which is necessarily into the mass spectrometer for detected frag- ments!), is an angular offset produced by the transformation from center-of-mass (cm.) to laboratory (lab) coordinates, and 8 is the anisotropy parameter (in the lab system; the anisotropy in the cm. system is usually nearly the same [2]); for purely parallel transi- tions, 8=2 and f(6) peaks at (0-6 )=0°, while for purely perpendicular transitions, g=-l and f(0) peaks at (0-0 Q )=9O . (This is true for direct dissociations, which seem to be the case in the present study; for indirect transitions, one needs to take into account the lifetime of the predissociative state [3].) From the TOF distribution, (also processed via a lab->-cm. transformation), one can obtain the transla- tional energy of the recoiling fragments and thus, through energy con- servation, their internal excitation as well. Results and Discussion 1. Til. The TOF distribution of I fragments from laser dissoci- ation at 300 nm shows that either or both of the following processes must occur: Til ^ V T1*( 2 P , 22 kcal mol _1 excitation) + I( 2 P ) 3/2' 3/2 b,v Tl ^2 p ^ + I *(2 p 21 kcal moi-iexcitation) . 1/2 1/2 the energy difference between these two processes is too small to re- solve in our experiments. (The first process has been observed pre- H5-2 viously using Tl atomic resonance absorption [4].) There is no evi- dence for dissociation into two ground state atoms. The angular distribution of I fragments at ^310 nm obtained with the Hg-Xe lamp is nearly isotropic ( 3= 0.07±0.02, e Q = -6°±10°) and is consistent with photodissociation via two separate transitions of opposite symmetry or via a single transition of mixed (parallel plus perpendicular) symmetry, as would be the case if either (or both) the ground or excited states contains even a slight (^5%) mixing of ionic and co- valent character [5,6] . 2. Cdl 2 . The TOF distribution of I fragments (and to a much lesser extent of cdl fragments) from laser dissociation at 300nm shows two partially resolved peaks having different angular distribu- tions. The angular distribution of the slower peak, has a maximum at (6-6 )-0° > corresponding to a parallel transition, while the transla- tional energy of this peak corresponds to 33 kcal mol~-'-(-80% of the total available energy) internal excitation of the fragments. The symmetry parallel of the transition suggests that the photodissociation process A hv o 2 _i is Cdl 9 "-»■ CdI(X z Z) + I*( Pi,, 21 kcal mol L excitation), the re- * '2 maining 12 kcal mol going into Cdl vibration and rotation. The an- gular distribution of the fast peak has a maximum at (0-A o )-9O°, cor- responding to a perpendicular transition, while only - 44% of the total available energy (18 kcal mol - -'-) appears as fragment internal excitation. This is too little energy for production of I*, and the perpendicular symmetry of the transition suggest the production of ground state I and vibrationally/rotationally excited (the full 18 kcal mol -1 ) Cdl (X 2 E). H5-3 The existence of two transitions of opposite symmetry (and of unequal probabilities) is consistent with the relatively low anisotropy (6=0. 46*0. 08, 6 =30°±3°) observed in the I fragment angular distribu- tion at -310 nm obtained with the Hg-Xe lamp (but without TOF dis- crimination). Additional experiments at different photon energies are in progress. References 1. M.J. Dzvonik and S.C. Yang, Rev. Sci. Instrum. 45, 750 (1974). 2. G.E. Busch and K.R. Wilson, J. Chem. Phys. 56, 3638 (1972). 3. M. Dzvonik, S. Yang, and R. Bersohn, J. Chem. Phys. 61, 4408 (1974) 4. P. Davidovits and J. A. Bellisio, J. Chem. Phys. 50, 3560 (1969). 5. R.C. Ormerod, T.R. Powers, and T.L. Rose, J. Chem. Phys. 60, 5109 (1974). 6. R.N. Zare and D.R. Herschbach, J. Mole. Spectrosc. 15, 462 (1965). *This work is supported by NSF grant MPS-74-22100. H5-4 The Photochemistry of 2-Furaldehyde in the Gas Phase A. Gandini, P. A. Hackett , J. M. Parsons and R. A. Back Division of Chemistry, National Research Council of Canada Ottawa, Canada The photochemistry and photophysics of 2 -fur aldehyde vapor following excitation in its first two absorption bands have been studied. Excitation in the tt*-HT transition (220-275 nm) led to photo- decomposition but yielded no detectable emission of light. Photolysis at 253. T nm at 65 and furaldehyde pressures between 0.2 and 7 Torr gave CO, furan, propyne, allene and cyclopropene as major products, with acetylene and COp in much smaller yields. A study of the variation in product quantum yields with furaldehyde pressure showed that c|)p n approaches 2 at low pressure and decreases towards at high pressure or with added COp. The yields of the other major products follow a similar trend but with more complex pressure dependence, particularly in the case of furan. The mercury-photosensitized decomposition at 253.7 nm gave essentially the same results as the direct photolysis. The mechanism invoked to explain these observations at 253.7 nm postulates the rapid initial sequence, F + hv -> X F ■*> 3 F + F # 1 3 i where F and F are the tttt* excited singlet and triplet states, and F' is the vibrationally excited ground state of furaldehyde, from which decomposition occurs. The reactions following this fast double inter- system crossing involve F and excited furan and C_h\ intermediates H6-1 which can decompose, isomerize, polymerize, or be deactivated "by collision. It is suggested that the mercury photosensitized reaction 3 directly populates the F state, which then undergoes the same reactions as in the direct photolysis. Evidence for the proposed mechanism is discussed, in particular the lack of light emission and the lack of efficient quenching by oxygen, both of which point to very short life- 1 3 times for F and F. A comparison of the present results with the earlier ones obtained by Hiraoka and Srinivasan show some significant differences which are tentatively explained. The behaviour of 2-furaldehyde excited in the TT*-«-n transition (290-370 nm) is quite different. Although the products of the photolysis at 313 and 336 nm were the same as the major ones found at 253.7 nm, quantum yields were much smaller, falling in the range -2 -3 10 -10 . Resinification on the cell walls was more important, with a quantum yield of about 0.05, which made quantitative studies of the photolysis virtually impossible, particularly at lower pressures. The large difference in the yields of decomposition from the two excited states may probably be attributed to the difference in their energies. Photophysical experiments showed that excitation in the TT*-*-n band of furaldehyde gave no detectable fluorescence, but modest phosphorescence was observed. The dependence on pressure, temperature and excitation wavelength of both the lifetime and the quantum yield of this emission from the 7T*«ti triplet state were investigated, having established that the quantum yield of intersystem crossing is close to unity at pressures above about 0.3 Torr. The results have been interpreted in terms of the following mechanism: H6-2 1 3 F + hv ► F ► F 3 k l F ► F + hv (phosphorescence) (l) 3 F + F — > 2F (2) 3 F _J— F (3) 3 1 where F and F are now the 7T*-<-n triplet and singlet states of —1 7 —1 —1 fur aldehyde . Values of L = 1.1 x , k ? = 6.3 x 10 M s , and 2 -1 k. = 1.2 x 10 s were obtained at room temperature, and k„ showed a negative temperature dependence. A long-lived emission from a quenching impurity present in trace amounts in all samples of 2- furaldehyde tested was also observed. This was minimized by rigorous purification and was taken into account in the treatment of the phosphorescence results. H6-3 Photoinitiated Decomposition Of Monosilane E. R. Austin and F. W. Lampe The Pennsylvania State University University Park, Pennsylvania 16802 A mass-spectrometric study of the hydrogen-atom initiated decompo- sition of monosilane has been carried out. Silyl radicals were gener- ated by mercury photosensitization of hydrogen-monosilane mixtures which consisted of about 95% hydrogen. Under these conditions more than 94% 3 of the Hg( P ) atoms that are quenched by collision react with hydrogen to form hydrogen atoms and all the hydrogen atoms formed react with 1 2 monosilane via (1) to form silyl radicals. Our attention is centered on the further reactions of the silyl radicals. H + SiH, + SiH + H 2 (1) The reactions were carried out in a photolysis cell containing a pin-hole leak leading into the ionization region of a time-of-f light mass spectrometer. In the photolysis of such a flow system the con- centrations of all substances will approach steady-states that are reached when the rates of introduction into the cell become equal to rates of loss from the cell. Illumination of the photolysis cell containing the H 9 -SiH, mixture o with 2537 A radiation results mainly in the formation of Si H, and 2 6 depletion of SiH.. Small amounts of Si H and Si.H _ appear after long H J O 4 1(J illumination times. In addition to the volatile products a solid film is deposited on the walls and quartz window of the cell. In Figure 1 is shown the time dependence of the concentrations of SiH. and Si_H, (both relative to the initial concentration of SiH. ) for 4 2 6 4 H7-1 two cells, one with a pyrex wall and one with a stainless steel wall. For irradiation times greater than 50 seconds the behavior in the two cells is different. However, the initial slopes of both the depletion of SiH, and the formation of Si„H, are independent of the cell wall. When the stainless steel cell was packed with quartz wool, identical results for the initial slopes were obtained. A fourth cell prepared by presilanation of the pyrex wall with (CH~) ^SiCl^ gave completely dif- ferent results; the depletion rate of SiH, was reduced by a factor of 2, while the formation rate of Si„H, was decreased by a factor of 7. These data indicate that in a clean cell, with walls which do not contain adsorbed silanes, the initial rates of depletion of SiH, and formation of Si H, are independent of the nature and area of the sur- z b face. The ratio of the initial rate of depletion of SiH, to the initial rate of formation of Si„H, is 2 as the simple stoichiometry of (2) re- quires, 2 SiH 4 ->■ Si 2 H 6 + H 2 (2) When the cell walls are silanated, most of the SiH, consumed does not produce Si„H, or any other volatile product but results in an increased formation of film on the walls of the cell. The protium-deuterium isotopic composition of the disilane formed 3 by the Hg ( P ) photosensitized reaction in a (96:2:2) H~-SiH,-SiD, mix- ture has been studied. The method used was to determine the mass spec- tra of the 56-68 region at a series of ionizing electron energies near the ionization potential of Si„H,. Ion-currents of all parent ions and of m/e 60 relative to the ion current at m/e 68 were plotted as a func- tion of nominal electron energy. A linear extrapolation of the i,„/i ratio yields a nominal electron energy of 11.0 eV for the disappearance H7-2 of all fragment ions from the spectrum. The relative amounts of the various deuterated disilanes are given by extrapolating each of the plots in Figure 2 to 11.0 eV and making isotope corrections. These results are shown below. 28 Si-Mass Spectrum of Disilane Products at 11 eV m/e Ion Relative Abundance 62 Si H, + 90±26 2 6 63 Si 2 H 5 D + 43±11 64 Si 2 H 4 D 2 + 37±15 65 Si 2 H 3 D 3 + 3±13 66 Si H D. + 100±16 2 2 4 67 Si 2 HD + 66±11 68 Si D, + 74±16 2 6 It is clear that Si«H«D„ is virtually absent. This is in agreement 3 with the proposal of Strausz et al that association of silyl radicals does not occur and that the formation of disilane occurs by the sequence (3) and (4) . The distribution of disilanes suggests that the H-atom SiH + SiH„ -* SiH, + SiH 2 (3) SiH + SiH. -> Si H, (4) 2 4 2 6 attacks SiH. to form SiH„ radicals about twice as fast as it attacks 4 3 SiD. to form SiD. to form SiD» radicals. This is in accord with the 4 4 3 corresponding rate constant ratio of 2-2.5 as reported by Potzinger et al 4 . H7-3 References 1. P. Potzinger, L. C. Glasgow and B. Reimann, Z. Naturforsch. 29a , 493 (1974). 2. D. Mihelcic, P. Potzinger and R. N. Schindler, Ber. Bunsenges. Phys. Chem. 78, 82 (1974). 3. T. L. Pollack, H. S. Sandhu, A. Jodhan and 0. P. Strausz, J. Am. Chem. Soc. 95, 1017 (1973). 4. R. Laupert, P. Potzinger, D. Mihelcic and V. Schubert, To be published in J. Phys. Chem. H7-4 1.00 0.80 CSiH 4 ] CSiH 4 3 o 0.60 t 1 1 i r o stainless steel walls □ pyrex walls Irradiation Time in Seconds 100 200 300 400 Figure 1. Depletion of Sill, and formation of Si H, during photolysis 12 13 14 65 ^68 J I I 12 13 14 Figure 2. Ion current ratios versus nominal electron energy, H7-5 Excitation of HNO by ( A ) T. Ishiwata, H. Akimoto and I . Tanaka Department of Chemistry, Tokyo Institute of Technology, Ohokayama, Meguroku, Tokyo, JAPAN *National Institute for Environmental Studies Reaction system (I) Active oxygen + C^ + NO Fig. 1 H + NO (010H000) nm c u l/> HN0( 3 A") + (V) 2 g l g HN0( 3 A") + 2 ( X A ) > HN0( 1 A M ) + 2 (V) . The internal relaxation in the triplet state of HNO would be responsible to the intensity alternation at 42 kcal mol in the chemiluminescent spectra. Emission intensity of HNO at 795 nm ( 1 ) - 10 / o - 7 Slope s - 5 2.0 / - i. /° " / / 1 / / . 1 C 2 H 4 ♦ ♦ NO ♦ 2 ('& g )/ 2 Fig. 4 I i i Emission intensity of 0, ( A ) at 1270 nm 2 9 In the reaction system (I) the formation of the highest 3 4 5 Emission intensity ot Oj ( Aq) «l 1270 nm H8-3 rotational level with K' = 7 of (010) would due to the two step energy transfer processes without relaxation in the triplet state, since two 0«( A ) molecules possess the energy of 45 kcal ,-i mol In the reaction system (II) , the spectral information and the nearly first order dependency of OJ A ) differed from the reaction o system (I) probably indicates the direct formation of HNO in the triplet state from the reaction of H + NO and the followed excitation step by 0«( A ) . O This stepwise mechanism suggests the triplet state energy of HNO at 0.8 eV above the ground state. Recent theoretical values are presented in Table together with our value. Table calc. (eV) exptl. (eV) State a b c d e ■""A' 0.0 0.0 0.0 0.0 0.0 3 A" 0.73 0.68 0.71 0.8 1 A" 1.45 2.04 1.61 1.63 a) A. W. Salloto et al., J. Chem. Phys., 52 2936 (1970) b) G. R. Williams, Chem. Phys. Lett., 30 495 (1975) c) A. A. Wu et al., ibid., 35_ 316 (1975) d) Our value, ibid., 27 260 (1974) e) M. J. Y. Clement et al., Can. J. Phys., 39 205 (1961) H8-4 Collision And Photo induced Dissociation of NH And HO J. Masanet , J. Fournier and C. Vermeil Equipe de Recherche associee " a l'ESPCI - 10 rue Vauquelin 75231 PARIS Cedex 05 - France. Comparable amount of electronic energy may he given to a mole- . . . 3 cule either by collision with argon metastable atoms( P , 11.5 and 11.7 eV) , by direct photoexcitation at the argon resonance lines 13 ( ' P ; 11.8 and 11.6 eV) or by argon photosensitized excitation (- \% of the molecule in argon). Depending upon the potential energy sur- faces of the molecule in the considered energy range and of the nature of metastable atom-molecule interaction different excited (or ionized) states may be reached leading eventually to dissociation and (or) ioni- zation. Understanding energy transfer from metastable argon atoms is important for chemical lasers; some insigth on the mechanism may be gai- ned by comparison with photoabsorption processes. Light emissions from excited state photofragments of the NH and HO molecules have been recorded in our laboratory for the direct and argon sensitized photolysis; for metastable atom studies, the expe- riments with NH, have been performed at Boulder Colorado USA, those with (1 ) HO are taken from Clyne and coll. (1969) ' • H 2 The following table summarizes the results 2 + . OH A E is produced in the three cases with the same rotationnai population of the v' = level; the v' = 1 level is seen also in the photodissociation case, but too weak for a rotationnai analysis to be performed. The distribution is different from that given by Carrington (2) at 121.6 nm or by us at the krypton resonance lines. The fact that identical results are obtained whatever is the excitation mode above J J eV favors a predissociation of the excited state(s) of HO leading (not exclusively) to that exit channel. NH„ With a ionization potential of 10. 1 U eV for NH , NH may be H9-1 Direct and Argon sensitized photolyse Collision with metastable argon Atoms Fragment HO ' emission 2 + 2 OH A I + X 7T 2 + OH A Z X 2 7T NH, Ionization Fragments emission 1 1 NH c it •> a A - J = 16 1 - J = 13 NH b 1 E + * X 3 E~ - J = 6 no NH 2 A 2 A^X 2 B 1 NH c 1 tt + a^ - J = 16 1 - J = 13 NH A 3 tt ■* X 3 Z~ - J = 12 1 - 1 J = 8 ^2 ^2 no NH 2 A A^X B ] • • • • (3) formed and is indeed found in the three cases. The direct and sen- sitized photoionization quantum yield are about the same (- 0.5). One of the dissociative channel of the super excited state is found in every case and leads to NH(c)+ H with the same internal energy distri- bution. By analogy with H_0 it is proposed that it correlates to a pre- dissociated state of NH . NH(b) is only formed by photoexcitation of NH and is thought to arise from a discret Frank-London state of NH disso- (U-5) ^ ciating into NH(b) + 2H in agreement with previous work . This super excited state could belong to a Rydberg series converging to the second ionization potential of NH . 3 With the metastable argon atoms, mostly Ar( P p )j a loosely bound complex may be formed with a lifetime sufficient for electron exchange to occur and to allow the following dissociation : Ar( 3 P ) + NH ■* (ArNH ) + NH(A 3 tt) + H + Ar (1) This spin forbidden dissociation does not occur either in direct or argon sensitized phot odissociat ion. This reaction path is shown to be in competition with at least two other exit channels H9-2 Ar 3 P Q 2 + NH 3 -» NH* + e +Ar( 1 S Q ) (2) ■> NH(c'tt) + H 2 + Ar ( 1 S Q ) (3) As the two last channels are also observed in the photoinduced dissociation it is tempting to suggest that the electronic state which dissociates into NH(b) gives by collision with metastable atoms the molecular complex of reaction (1). This would be in agreement with the discrete nature of this state as postulated above. REFERENCES 1/ M.A.A.CLYNE, J.A.COXON, D.W.SETSER and D.H.STEDMAN, Trans. Far. Soc . 65 (1969) 1177. 2/ T. CARRINGTON, J. Chem. Phys . h]_ (196*0 2012. 3/ R.L.LILLY, R.E.REBBERT and P. AUSL00S, J. Phot. 2 (l 973/7*0 h9 . h/ J. MASANET, A. GILLES and C. VERMEIL, J. Phot. 3 (197V75) UlT. 5/ C. VERMEIL, J. MASANET and A. GILLES, Int. J. Radiat . Phys. Chem. 7 (1975) 275. H9-3 Photogalvanic Cells M.D. Archer and M.I.C. Ferreira, The Royal Institution, 21 Albemarle St., London W1X 4BS, U.K., and W.J. Albery and W.R. Bowen, Physical Chemistry Laboratory, South Parks Road, Oxford 0X1 3QZ, U.K. We consider the performance of photogalvanic cells of the type shown in Fig. 1 and Scheme 1 , with particular reference to the much (1) studied iron - thionine cell. Scheme 1 In dark solution : A+Z*-^B+Y Photochemistry : A+Z — *B+Y At electrodes : A+e — »B Y+e^Z The electrodes are identical, and the Fig. 1 The photogalvanic cell V. / — f ' 1— Dark Illuminated solution solution Light device is a concentration cell, working by virtue of the different chemical compositions of the dark and photos tationary states. Cell performance is characterized by three parameters: open circuit potential, short circuit current and maximum power. The open circuit potential change on illumination, from a thermodynamic dark potential to a mixed potential, depends upon the (2) elctrode kinetics of A,B and Y,Z. ' If both are highly reversible, then the electrode restores thermodynamic equilibrium at its surface, and there is no potential change on illumination, and no power can be obtained from such a cell. If one couple is more reversible than the other, then the electrode potential does change on illumination, to an extent which depends on the balance between the electrode kinetics, and the rates of back reaction in solution and of transport to the electrode. 11-1 Maximum potential shifts are obtained if one couple (A,B say) is completely reversible and the other (Y,Z) completely irreversible. In this case, the electrode potential follows the Nernst expression for the A,B couple. In the iron-thionine cell, although both couples, thionine/leuco- thionine (T/l) and Fe /Fe are separately reversible at clean metallic electrodes, the iron couple is rather strongly repressed by the presence of thionine in solution, as illustrated by Figures 2 and 3> a *id so the potential shift on illumination is negative. Full Nernstian behaviour with respect to the T,L couple is not observed on platinum, as shown in Figure 4« Chronocoulometric measurements can reveal the extent to which thionine is adsorbed on the electrode surface. Figure 2 ^ A Cyclic voltammagrams on Pt spherical microelectrode ^0 a) 1CT 2 M FeSO in 0.5M H 2 S0 4 a') Same + 3 * 10~ 5 M thionine -100 Figure 3 DC voltammagrams on platinum rotating disc electrode b) 10" 2 M FeSO in 0.1M K 2 S0 4 4 b') Same + 3 * 10" 5 M thionine c) 3-5* 10" 5 M Fe 2 (S0 4 ) 5 -2 in 0.1M K 2 S0 4 c') Same + 3 * 10" 5 M thionine 0.4 E(NHE) Figure 4 Steady state potential of illuminated electrode as a function of the composition of the photos tationary state 0.6 . 0.3 0.2 Composition of dark solution 3 x 10-5M thionine, 10~ 2 M FeS04, 0.1M K 2 S0 4 , pH 2.5 None A 5 1CT 5m □ 1 10'% Fe 2 (S0 4 ) 5 Q \ \ O^Cfc E = E° + RT In [T] (Nemst equation ' 2F [LJ for' T,L couple) 0.2 0.4 0.6 0.8 1.0 [L]/([T] + [L] ) in photostationary state When the photogalvanic cell is short circuited, then the current obtained is a maximum if one couple is completely reversible at the electrode and the other completely irreversible.^' (The iron- thionine system on platinum is therefore not ideal.) In this case, the electrode processes are the reduction of A at the dark electrode, and the oxidation of B at the illuminated electrode. If the kinetic lengths 11-3 of A and B (the distances over which they diffuse before they react with Z and Y respectively) are shorter than the diffusion layer thickness at the electrode surface, then there may be an important catalytic contribution to the short circuit current from the following sequence of reactions: /Solution : Y + B — >A + Z Dark side .Electrode : e + A — >B X 'Solution : Z + A — > B + Y Illuminated side (Electrode : B — » A + e X The homogeneous reactions regenerate the species that is reacting at the electrode. Thus by appropriate manipulation of the thermal and photochemical reaction rates in bulk solution, it may be possible to increase the present low efficiencies of photogalvanic cells considerably. References 1. M.D. Archer, J. Appl. Electrochem. , _5, 1975, 17 gives pre-1975 references. See also R. Gomer, Electrochim. Acta, 20, 1975» 13 and W.D.K. Clark and J. A. Eckert, Solar Energy, r7» 1975, 147- 2. W.J. Albery and M.D. Archer, Electrochim. Acta, in press. 3. W.J. Albery and M.D. Archer, submitted to J. Electrochem. Soc. U"4 The Importance of Intermediate Partitioning In Energy Storing Photoreactions G. Jones, II , W.R. Bergmark, and M. Santhanam Department of Chemistry, Boston University Boston, Massachusetts 02215 The problem which is central to the observation of significant photoreactivity using low energy (single) photons (practically speaking, the visible portion of solar radiation) is illustrated in the figure below. The objective of one energy conversion scheme is to drive a starting material A to photoproduct B, which is energy- rich, reasonably kinetically stable, and revertible to A. Quantum efficiency will depend largely on the extent to which A* avoids a sizeable ground state potential energy barrier. In selecting materials which absorb increasingly toward the red, a point is reached at which A* has insufficient energy to mount the thermal surface, and photo- reactivity will cease unless special mechanisms are employed P.E. Rx Coordinate 12-1 (e.g, significant thermal activation of A*, tunneling). We have exam- ined several model systems for which the excitation energy (A - A*) exceeds the thermal barrier (A - T) by no more than 20 kcal/mol. The data permit the generalizations that kinetically identifiable interme- diates are important for endoergic photoisomerizations to stable pro- ducts, and that it is the partitioning of these intermediates ( and not competitive decays of the initial excited species) which determine quantum efficiencies. The case for intermediate partitioning is illustrated for dienone 1. Quenching studies show that the lifetime of reactive trip- ^ let 1 is extremely short and temperature dependent (k. „ =1.1 - r\, decay 3.5 x 10 sec" , 4 - 51°C). Normal (unreactive) radiationless decay O _"! for cyclic enones is no faster than 10 sec" . A fast reactive decay channel for triplet 1 appears responsible for the short lifetime, but direct formation of 2 without competition is ruled out by the observed quantum efficiency (0 = 0.36, 320 - 380 nm). The required mechanism in- volves one or more intermediates, formation of which from triplet 1 is lifetime limiting and partitioning of which (to ] and to 2) determines the quantum yip 1H . % % hV 12-2 h\> *0 "o o o CHpCHp CH«CH^ eye 0.36 'fir <0.001 T flr (nsec) CH/jCHrt H 0.26 0.16 1.7 CH^CHp CH 3 0.05 0.14 2.8 CH 2 H 0.15 0.06 <0.5 CHOH H 0.29 0.02 <0.5 (^0.01) A similar pattern of kinetics and quantum yields is found for linked anthracenes 3 (see table). The linking of anthracenes leads to f\j enhanced quantum efficiencies for photocyloaddition (3 + 4). (The con- % a. centration dependent quantum yield for 9-methyl anthracene photodimer- 2 ization reaches a maximum of 0.14.) Fluorescence quantum yields are low and fluorescence lifetimes short in a way highly dependent on struc- ture (note table divisions) and not related to quantum efficiency for photoisomerization. Neither emission nor radiation! ess deactivation 8 -1 expected for anthracenes (normally, intersytem crossing, k % 10 sec" , followed by triplet decay) account significantly for the rapid decay of excited singlets 3. The results are consistent with efficient forma- tion of an intermediate which partitions to 3 and 4. 12-3 The photocycloaddition of biacetyl (5) and olefins has been investi- gated. The interaction of excited biacetyl and cis-1 ,2-dimethoxyethene 9 -1 -1 (6) in benzene leads to quenching; k (fir) ■ 2.5 x 10 M sec and k (phos) = 6.2 x 10 M" sec" . The quantun yield for photocycloaddition (5 + 6 ■* 7) is 0.11 + 0.02 at 0.001 - 0.01 M 6 (a concentration range in which virtually all biacetyl triplets, but negligible singlets, are quenched). This limiting value of the quantum yield indicates that trip- let quenching encounters not only lead to product (7) but also provide a path for decay back to 5 and 6. A clue to the nature of an intermediate species which would account for the results (but not exclusively) is found in the reaction stereochemistry (i.e., 5 + 6 + trans oxetane as r\j 'Xj shown) . O O /OMe r\ /=\ — *> MaH HMO / MeO OMe ' I OMe 5 6 7 Should this pattern of intermediate partitioning be even more general? Current theory regarding reactive radiationless decay affirms 3 the trend. The systems selected, for which kinetic stability and thermo- dynamic instability in photoproduct are compatible, exhibit weakly avoided surface crossings (normally S Q and S 2 ). At some near mid-point along the reaction coordinate, excited and ground surfaces will be close lying for facile radiationless transition and minima in S, and T, will be provided. From this di radical or biradicaloid geometry (corresponding, 12-4 in the cycloadditions presented, to structures with one bond made and the other left incomplete) excited molecules funnel to product and back to reactant. Anticipation of electronic factors which control parti- tioning should be valuable in designing energy storing systems which can be driven with high quantum efficiency. Acknowledgements . This work was supported by the Office of Naval Research, the Advanced Research Projects Agency, and the National Science Foundation (through a fellowship to W.R.B.) References (1) G. Jones, II and B.R. Ramachandran, J. Photochem . , in press. (2) T.M. Vember, T.V. Veselova, I.E. Obyknovennaya, A.S. Cherkasov, and V.I. Shirokov, Izvestiya Akademii Nauk SSSR. Seriya Fizi- cheskaya , 3£, 837 (1973). (3) J. Michl, Fortschritte der Chem. Forschung , 46, 1 (1974). 12-5 Optimization Of The Iron-Thionine Photogalvanic Cell; Photochemical Aspects P.P. Wildes , N.N. Lichtin and M.Z. Hoffman Dept. of Chemistry, Boston University Boston, MA 02159 The iron-thionine photogalvanic cell was first investigated by E. Rabinowitch (1) and has since been the subject of a number of investigations (2-6). Studies of this cell can be regarded as prototypical efforts directed towards ultimate realization of a photogalvanic device which is a practical and economically signifi- cant means of converting the solar flux to electrical power. One requirement which must be met to achieve this goal is conversion of -2 incident solar power (^100 mW cm at ground level) to electrical power with an efficiency of ^5% (i.e., a 5% engineering efficiency). If the quantum efficiency of the iron-thionine photogalvanic cell for conversion of absorbed photons to electrons delivered to an external circuit were 100%, its engineering efficiency could be at best only ^3%. This follows because no more than 15% of solar power falls within the absorption band of thionine and the potential of the cell is no more than 20% of the potential of the incident photons. Sensitization to the blue with a quantum efficiency of 100% would afford the possibility of increasing maximum engineering efficiency to ^9%. Optimization of photochemical and photophysical processes is thus a necessary (but not sufficient) condition for achievement of a practical sunlight engineering efficiency. Photochemical aspects with which we have been concerned include efficiency of photooxida- 3 2+ tion of Fe(II) by triplet thionine ( TH ) (7), sensitization to the blue and composition as well as kinetics of decay of the 13-1 photochemical steady state at high concentrations of thionine. Table 1 summarizes information on the dependence of rate constants upon solvent and anions for thionine and methylene blue. Table 1 Kinetics of Excited States of Thionine and Methylene Blue First Order Decay, k, , and Reduction by Fe(II), k„, of Triplet States IONIC k STRENGTH 1 l - DYE SOLVENT pH ANION M 10 sec 10 M sec TH+ H o 2.0 HSO ~/S0 = — 1.3 600 2 4 4 TH+ H 2 1.7 F 3 CS0 3 — 1.5 65 TH+ 50 v/v % "2.0" HSO "/S0, = — 1.2 5,000 aq. MeCN TH+ 50 v/v % "1.7" F CSO " -- 1.2 55 aq. MeCN MB+ H 2 1 HSO 4 "/S0, = — 2.4 38 MB+ HO 1 F 3 CS0 3~ "' 2 * 2 9 MB+ H o 2.0 HS0~/S0. = 0.2 2.2 55 2 4 4_ MB+ HO 2.1 HSO ~/S0 ~ 1.0 2.2 80 2 4_ 4_ MB+ H o 3.1 HSO. /SO, 1.0 2.2 80 2 4_ 4_ MB+ 50 v/v % "2" HSO /SO 0.2 2.0 700 aq. MeCN MB+ 50 v/v % "2" F CSO ~ 0.2 2.1 10 aq. MeCN 2+ 2+ a. At the pH of these measurements, triplet TH and MBH b. Values in quotation marks are pH values of solutions of identical composition in H-0. 3 2+ The rate constant for the first order decay of TH„ and 3 2+ MBH is not significantly affected by a change in anion from CF SO to HSO, /SO, or by a change in solvent from water to 50 v/v 3 2+ per cent aqueous acetonitrile. The rate of reaction of TH ? and 3 2+ MBH with Fe(II) is, however, strongly affected by changes in the anion present and in some cases by changes in solvent. The faster rate of reaction of the triplet excited dyes in the presence of 13-2 sulfate anions compared to solutions with CF-SO anions may result from reduction in electrostatic repulsion by ion pairing. In any 3 2+ case, the efficiency of photooxidation of Fe(II) by TH_ is higher in the presence of sulfate anions than with CF SO anions. We are also studying factors which affect the rates of other reactions occurring in the cell solutions. The maximum current obtained from the cell is related to the concentration of reduced thionine present at the anode when the cell is irradiated with sun- light. Under constant illumination, the cell solution attains a photostationary state composition in which some of the dye has been reduced. The photostationary state composition of the solution depends on several parameters including the initial concentrations of dye, Fe(II) and Fe(III), the intensity of light and the length of the light path through the cell solution. Table 2 summarizes some data showing the dependence of photostationary state solution composition on reactant concentrations and cell thickness. The details of the solution dynamics which result in the photostationary state compositions shown and the implications of these values in maximizing the cell engineering efficiency are currently being investigated. Acknowledgment: This work was supported by NSF RANN Grant SE/AER/72- 03597. We wish to acknowledge valuable conversations with Dr. E. Berman (Boston Univ.), J. A. Eckert (Exxon Research and Engineering Co.) and Prof. H. Linschitz (Brandeis Univ.) and the use of laser flash equipment in the laboratories of Prof. Linschitz and Dr. E. Hayon (U. S. Army Natick Laboratories). 13-3 Table 2 Photoreduction of Thionine in 50 v/v% Aqueous Acetonitrile Solution containing 0.01N H„S0, % Reduction of 'o [Fe(II)]^ [Fe(III)l b M x 10~ 4 Thionine at [Thionine' Cell Thickness Photos tationary M x 10 -4 O M ° cm x 10 _df Stated 7.5 0.01 0.9 80 40 7.5 0.025 2.25 80 41 7.5 0.05 4.5 80 25 7.5 0.01 0.9 25 51 7.5 0.025 2.25 25 56 7.5 0.05 4.5 25 45 3.75 0.01 0.9 80 59 3.75 0.025 2.25 80 74 3.75 0.05 4.5 80 42 3.75 o.'oi 0.9 25 73 3.75 0.025 2.25 25 66 3.75 0.05 4.5 25 47 a. Under illumination with ^50mW/cm of light from a Xenon lamp. b. Initial dark concentrations. c. Probable errors are -10%. References 1. E. Rabinowitch, J. Chem. Phys. , 8_, 560 (1940). 2. L. J. Miller, U. S. Dept. Commerce, Off. Tech. Serv., Report AD 282, 878 (1962). 3. A. E. Potter and L. H. Thaler, Sol. Energy , 2> 1 (1957). 4. K. G. Mathai and E. Rabinowitch, J. Phys. Chem. , 66 , 663 (1962). 5. N. Kamiya and M. Okawara, Denki Kagaku , 38 , 273 (1970). 6. W. D. K. Clark and J. A. Eckert, Sol. Energy , 17 , 147 (1975). 7. J. Faure, R. Bonneau and J. Joussot-Dubien, Photochem. Photobiol., 6, 331 (1967). 13-4 •k A Bioirtimetic Approach to Solar Energy Conversion. Thomas R. Janson , and Joseph J. Katz, Chemistry Division, Argonne National Laboratory, Argonne, Illinois 60439. A generation ago, R. Emerson and W. Arnold developed the concept of the photosynthetic unit as the basic device used by green plants in photosynthesis for light energy conversion. Chlorophyll (chloro- phylls a and b_ in green plants, bacteriochlorophyll in photosynthetic bacteria) acts as the primary photoacceptor . A great many chlorophyll molecules (300 hundred or more) act cooperatively in the light conversion process. The great majority of the chlorophyll molecules in the photosynthetic unit are passive, and act as antennas to collect light quanta whose energy is then funneled to a few special chlorophyll molecules in a photo-reaction center where energy conversion occurs. Antenna and photo-reactive chloro- phyll, together with electron transfer proteins that serve as conduits, constitute the photosynthetic unit in which light energy is made available for chemical purposes. It has long been known that the visible absorption spectra of both antenna and photo-reaction center chlorophyll are red-shifted relative to a solution of chlorophyll in, for example, ether or acetone solution prepared in the laboratory. It has also long been established that a characteristic free radical signal is produced in photo-reaction center chlorophyll in the process of light conversion. Laboratory investiga- tions, mainly by infrared and magnetic resonance spectroscopy, now make it possible to provide plausible interpretations on the molecule ■k Work performed under the auspices of USERDA. 14-1 for the anomalous visible absorption spectra of in vivo chlorophyll, and for the unusual features of the free radical signal. Thus, new models for both antenna and reaction center chlorophyll have emerged and these have become the basis for a serious effort to produce new devices for solar energy conversion. The information so far developed on plant photosynthesis can be used in a biomimetic approach to solar energy conversion. Essentially, solar energy in a biomimetic context uses light for chemical purposes, i.e., for the synthesis of compounds, rather than as a low-grade heat source. From the standpoint of the efficient utilization of the energy of light, such an approach has much to recommend it. First steps have been taken in the direction of a device intended to mimic the light conversion step in plants, and in which light- mediated electron transfer can be investigated. The device studied consists of a cell composed of two glass compartments, each with its own platinum electrode. External leads are gold soldered to each platinum electrode and connected to an electrometer to monitor current, emf, and resistance changes. When assembled the cell compartments are separated by a membrane (a plastic film or metal foil) carrying the photo-active chlorophyll- water adduct. In the case of metal foils the photo-active chlorophyll species is deposited on one side (generally the side facing the elec- tron donor) by evaporation of an octane suspension of the chlorophyll- water adduct. One compartment is then filled (under nitrogen) with 1-2 ml of a suitable degassed aqueous solution of an electron donor such as ascorbate and/or (tetramethylphenylenediamine) (TMPD) . The 14-2 other compartment is then filled with a solution of an electron accep- tor such as potassium f erricyanide, dichlorophenol indophenol (DCPIP) or the sodium salt of anthraquinone sulfonic acid. After thermal equilibration in the dark the supported photo-active chlorophyll film is irradiated with visible (red) light and the electrical consequences observed. Although white light is generally used in these experiments, it should be noted that when red light 0W4O nm, which corresponds to the absorption maximum of the photo-active chlorophyll film) is used photo-induced electrical changes are still observed. A change in potential (emf) across the separating membrane upon irradiation with visible light was the most commonly measured elec- trical parameter. Under our experimental conditions, we observed increases in emf, decreases in resistance, increases in conductivity, and increases in current flow. In our best cell to date, using an aluminum support film, a potential difference of 422 mV was recorded associated with a current flow of 2.36 x 10 amp. A rough estimate of the efficiency of the photo-activity of this cell is 0.0024%, comparable to the best organic photo-voltaic devices in the literature. The magnitude of the photo-induced effects has been shown to depend on the type of film support used, the nature of the donor and acceptors, and on the light flux. In conclusion, it has been shown that Chi ji (740) as a thin film on a suitable support material in contact with aqueous solution of various donors and acceptors indeeds mediates and produces photo-electric responses. 14-3 Light Energy Conversion Via Photoredox Processes in Mi cellar Systems Michael Gratzel, Hahn-Meitner-Institut fur Kernforschung Berlin GmbH Bereich Strahlenchemie, D 1000 Berlin 39 This paper describes electron transfer reactions involving ex- cited states of organic molecules by which exploitation of solar light energy may be achieved. Three different types of processes are dis- cussed: i) Photoejection of electrons from a photoactive probe P into a po- lar liquid such as alcohol or water: -^ hv + 1 P >P +e 1 solv where P stands for a cation radical and e ., for the solvated solv electron T ii) Electron transfer from the triplet state of the molecule (P ) to various acceptors: (2) P T +A >P + + A" T iii) Electron transfer from various donors to an acceptor triplet A (3) A T + D >D + +A" These reactions are examined by conventional photolysis and pulsed laser techniques. A frequency doubled, Q-switched ruby laser (pulse- width 15 ns, 347. 1 nm) is used to excite the photoactive species. Tran- sitory species are identified and their kinetics monitored by fast kine- tic spectroscopy or conductivity technique. 15-1 2 A prominent feature of this work is the comparison of the nature of photoreactions 1-3 and subsequent dark reactions in homogeneous polar media such as alcohol or water and aqueous micellar solutions. It is attempted to illustrate the important role surfactant micelles could play in the search for photochemical systems suitable for conversion of light energy into chemical or electrical energy. Reaction (1) has attracted attention as a possible pathway for pho- tochemical production of hydrogen from water since hydrated electrons produced in the photoionisation event are known to undergo the diffus- ion controlled reaction (3) e" + e" * H + 2 OH" aq aq 2 The two major problems which heterofore have prevented successful application of such a system are: (a) The majority of organic sensiti- zers with ionization potentials low enough to permit electron ejection by visible or near u.v. light are sparsely soluble or insoluble in water; (b) Rapid recombination of e /P ion pairs leads to annihilation of electrons before hydrogen can be formed. These obstacles may be overcome by usage of anionic micellar solutions. A schematic illustrat- ion of a photoionization event in such a system is given in Fig. 1. The sensitizer P is solubilized in the hydrophobic interior of the micellar aggregate whose negative surface charge prevents reentry of + the hydrated electron and recombination with the parent ion P . These 15-2 Fig.l - Schematic presentat- ion of photoionization and electron transfer processes in solutions of surfactant mi- celles containing a solubiliz- ed photoactive probe P. The electron acceptor M n is lo- cated in the Stern layer of the micelle. phenomena are illustrated by laser photolysis results obtained from solutions of N,N,N',N' -tetramethylbenzidine (TMB)in sodiumlauryl- sulfate micelles. Comparison of photoionization cross sections in al- coholic and micellar solution also reveals the favourable result that the micelles enhance strongly the yield of electrons. This effect is discussed in terms of an electron tunneling. T The triplet states of phenothiazine (PTH ) and TMB were found to have strongly reducing properties and readily transfer electrons to acceptors such as methyl viologen, duroquinone, europium(III) and chromium(III) ions. Some of these reactions have particular import- ance for the photo reduction of water as the reduced form of the accept- or is capable of evolving hydrogen from water according to (4) A + H 2 2 H 2 +A+OH ' This reaction has to compete with back transfer of electrons from A to P which in homogeneous solution occurs at a diffusion controlled rate. In eluded in Fig. 1 is an illustration of a forward electron trans- fer process in a micellar system. These intramicellar processes oc- 15^3 cur at a very fast rate. The back reaction, however, may be strongly inhibited by the micellar systems. The last type of photo redox processes discussed in this paper T are electron abstraction reactions of triplet duroquinone (DQ ). Inmix- T tures of water/ ethanol DQ abstracts electrons from a variety of sub- 2+ 4- 2- strates such as Fe , Fe(CN) , CO , diphenylamin and trimethoxy D O benzene. The rate constants for these redox processes are close to the 9 10 - 1 - 1 2 - diffusion controlled limit (10 -10 M s ) except for CO where k 7 -1 -1 equals 7 x 10 M s ). Anionic micelles were found to have a pronounc- 2+ T ed catalytic effect on the reaction of Fe with DQ . A schematic il- lustration of such an intramicellar electron transfer process is given in Fig. 2 Fig. 2 -Sche- matic illlus- tration of an intramicellar electron trans fer process from Fe to DOjsolubiliz- ed in anionic micelles Converseley cationic micelles drastically enhance the rate of electron 2- transfer from CO~ to DQ. The latter process is of particular interest as CO is known to form oxygen in its subsequent reactions. Hence O evolution from water by visible light becomes feasible via excited qui- none redox reactions. 15-4 Photoelectrolysis of Water by Solar Energy* D. I. Tchernev Lincoln Laboratory, Massachusetts Institute of Technology Lexington, Massachusetts 02173 Although the free energy required for the decomposition of water into gaseous H and O is only 1.23 eV, while the peak of the solar spectrum occurs at a photon energy of about 2. 4 eV, solar energy cannot be utilized for the production of H fuel by the direct photodecomposition of water, because the threshold energy for this direct reaction is about 6.5 eV. Fujishima and Honda have recently shown, however, that this threshold can be greatly re- duced if decomposition is accomplished by means of photoelectrolysis, a process in which a semiconductor is used as a catalyst. This is a promising process for the large-scale utilization of solar energy to produce H fuel. Ci We have been investigating the physics and electrochemistry of photoelectroly- sis by experiments on cells with polycrystalline as well as single-crystal TiO anodes. This work has led to studies with SrTiO anodes, and we report some of these results as well. The basic photoelectrolysis cell consists of an n-type semiconduct- ing working electrode, an electrolyte, and a Pt counterelectrode. Because of the difference in work functions between the semiconductor and the electrolyte, the energy bands of the semiconductor are bent at the surface, so that the analog of a Schottky barrier exists at the semiconductor-electrolyte interface. *This work was sponsored by the Department of the Air Force. 16-1 The semiconductor surface is irradiated by photons of energy at least equal to its energy gap. The photon-generated hole -electron pairs do not recombine but are separated by the electric field of the barrier; the electrons move away from the surface into the bulk of the semiconductor and then through the external circuit to the Pt cathode, where they discharge H (2e" +2H -» H ), while the holes remain at the surface of the semiconductor anode, where they can interact with the electrolyte to produce O (2p +H.O-* 1/2 O +2H"*). m £k d* The overall chemical reaction is 2hv+H 2 0-l/2 2 +H 2 (1) provided the semiconductor is chemically inert, serving only to absorb the light and to produce the holes and electrons that make the reaction possible. Two conditions are necessary for efficient photoelectrolysis: (1) The energy bands at the semiconductor-electrolyte interface must be bent in order to separate the holes and electrons excited by the light and (2) the relevant electronic levels must line up, i.e. , the hole states of the anode with e(H O/O ) and the Fermi level of the cathode with e(H + /H ). The anode material used in most of our experiments was the rutile form of TiO . Some TiO anodes were fabricated from single crystals, but most were polycrystalline disks prepared by hot-pressing powdered rutile at 750 C under a pressure of 10,000 psi. To make the disks conducting (p ~ 10 Q-cm), they were heated in vacuum for three hours at~ 600 C. In initial experiments the external quantum efficiency, r\, defined as the ratio of the number of electrons flowing in the external circuit (N ) to 16-2 the number of photons incident on the cell (N ), was measured as a function of photon energy hv. The maximum values of r\, 82% for the single crystal and 60% for the polycrystalline disk, occur at hv » 4 eV. Quantum efficiency measurements were also made with TiO films on Ti foil that had been therm- ally oxidized. The efficiencies were comparable to those obtained with single crystal anodes. When the maximum measured r\ values are corrected by taking account of estimated reflection and absorption losses, we find that the internal quantum efficiency was close to 100% for both single-crystal and polycrystal- line TiO anodes. This high quantum efficiency shows that the band bending in TiO was sufficient to separate the electron-hole pairs generated by photon absorption and in addition that the hole states of the anodes lined up with the (HO/O) level of the electrolyte. As stated above, however, for efficient photoelectrolysis it is also necessary for the Fermi level of the cathode, e , F to line up with the electrochemical potential of the H /H level in the solution, e(H + /H 2 ). The quantum efficiency experiments just described were performed with the photoelectrolysis cell open to air, so that the electrolyte contained dissolved oxygen. In this case the energetically favorable process at the cathode is the transfer of electrons from the cathode (at e ) to the H o 0/0 causing the reduction of oxygen: |o 2 +H 2 0+2e"-»20H" . (2) The cell now functions in the galvanic mode, and no H gas is evolved. Under 16-3 these conditions, we have observed that the O bubbles formed at the anode Li migrate to the cathode, where they disappear. This obstacle to the production of H gas can be eliminated by re- moving the dissolved oxygen from the cathode compartment of the cell. However, the cell current then decreases considerably, so that r| becomes only 1-2%, even for the best TiO anodes. This occurs because the band bending under photoelectrolysis conditions is very small for TiO . However, Li better results are obtained for SrTiO , for which the band bending is ~ 0. 2 V larger because of its lower electron affinity. This increase in band bending results in a quantum efficiency for photoelectrolysis of 10% at hv = 3. 8 eV, about one order of magnitude larger than that for TiO . The SrTiO experi- ments were performed with single -crystal and polycrystalline anodes. When dissolved O was removed from the solution, the amounts of H evolved at the cathode and O evolved at the anode were in good agreement with the predic- tions of Eq. (1), showing that photoelectrolysis was the only reaction occurr- ing. Ultimately, the efficiency with which solar energy can be used to generate H by photoelectrolysis will depend on the semiconductors that are used as electrode materials. Only about 10% of the solar energy reaching the earth is supplied by photons with energies higher than 3. eV, the energy gap of TiO or SrTiO„. Therefore the efficiency of solar utilization will be Ci O limited to a few percent unless it is possible to find a suitable electrode material with a significantly lower energy gap. 16-4 RECENT WORK IN EXCIPLEX PHOTOPHYSICS, by Desmond V. O'Connor and William R. Ware, Photochemistry Unit, Department of Chemistry, Univer- sity of Western Ontario, London, Ontario, Canada. Introduction While there have been numerous studies reported in the literature aimed at revealing the photophysical and photochemical properties of exciplexes, there have been relatively few that attempted to obtain simultaneously (a) a test of mechanism (b) all the pertinent rate con- stants (c) the temperature coefficients of these rate constants and thus the activation parameters as well as the thermodynamic properties of the exciplex equilibrium. In fact, what is needed are such studies carried out with a number of donor-acceptor pairs of different pro- perties in a number of solvents ranging from non-polar to polar. The results of such studies cannot fail to enhance our knowledge of the exciplex. Such a program is currently underway and this paper consti- tutes a preliminary report. Experimental Exciplex photophysics have been investigated by observing the fluorescence decay of both the monomer and exciplex, by observing the relative yields of monomer and exciplex emission as a function of quencher concentration and by measuring the steady-state monomer quenching. The fluorescence decay measurements were done with the single photon technique and deconvolution accomplished by iterative convolution. Results The principal system to be discussed is a-cyanonaphthalene quenched by dimethyl cyclopentene. This has been investigated in hexane and di- ethyl ether (DEE) over a wide temperature range. The experimental observations are in qualitative agreement with the well-known excimer or exciplex mechanism. That is prod. Jl-l The appropriate equations are I M (t) = C.e'^ + Cae"^* -Xit -\zt- I E (t) = C 3 (e" AlL - e- A2L ) (1) (2) Xi, 2 = %[ki + k 2 + k 3 [Q] + ki, + k p ± {(ki + k 2 + k 3 [Q] - U - k p ) k- + Akak^CQ]^] (3) k p = k 5 + k 6 One also has (4) \i + X 2 = ki + k 2 + k 3 [Q] + ki, + k r XiX 2 = (ki + k 2 ) (k, + k p ) + k 3 k p [Q] Another useful equation is the indicial equation used to derive Eq. 1 and 2. This is X 2 - [k, + k 2 + k 3 [Q] + (k, + k p )]X + (k, + k 2 ) (k- + k p ) + kakp [Q] = Finally, the two steady-state equations are k 3 k (5) (6) (7) K. SV (ki + k 2 )(k„ + kp) TO -iji-l 1 IQT 71 (8) (9) These equations were utilized to calculate individual rate con- stants on the assumption that the mechanism is valid. It will be shown that quantitative aspects of the experimental results are con- sistent with this model. The results are summarized in the table. The following conclusions appear consistent with the data. (a) The forward rate of quenching is diffusion controlled. (b) The low-temperature results in hexane are consistent with the Jl-2 Hexane — — > Diethyl ether T°K 273 262 251 242 233 303 293 283 273 k 3 xlO" 10 FPsec" 1 0.98 0.88 0.78 0.67 0.60 1.23 1.21 1.08 0.93 kalO" 8 sec" ! 1.14 0.49 0.20 0.10 0.04C 0.55 0.25 0.17 0.05^ k xlO" 8 sec" * 0.98 0.77 0.62 0.52 0.46 0.79 0.65 0.54 0.44 ^ % 1.94 2.25 2.58 2.82 3.18 2.35 2.69 2.81 3.24 AS 3 ^ -4.78 -4.68 -4.58 -4.58 -4.44 -4.42 -4.18 -4.11 -4.10 ASi/ 18.97 18.94 19.00 19.25 19.18 14.27 13.97 14.56 13.75 AS -23.15 -23.62 -23.58 -23.83 -23.62 -18.68 -18.15 -18.67 17.85 AG° -2.42 -2.70 -2.97 -3.13 -3.39 -3.26 -3.60 -3.64 -4.05 AG 3 ^ 2.36 2.31 2. 25 2.23 2.17 2.56 2.46 2.42 2.42 AG,/ 4,78 5.02 5.23 5.36 5.55 5.81 6.06 6.06 6.44 Jl-3 high-temperature results in hexane where the rapid equilibrium approximation is valid (see J.Am.Chem.Soc. , 96, 7853 (1974)). Thus the kinetics are now resolved over a range of 85°. (c) The quenching is inefficient and this appears entirely due to the feedback step (U). (d) The thermodynamic parameters and activation parameters are similar to those obtained for other exciplexes and the exci- plex binding energy is consistent with electrochemical data. (e) The steady-state studies confirm the fluorescence decay measurement. (f) Changing from hexane to DEE has a profound effect on k k but leaves AE? and AH C unchanged. While k k and k decrease as one goes to DEE, K~ v increases! (g) In hexane k 5 f f(T). Jl-4 Charge Transfer and Hydrogen Atom Transfer Reactions of Excited Aromatic Hydrocarbon - Amine Systems N. Mataga , T. Okada, and T. Mori Department of Chemistry, Faculty of Engineering Science, Osaka University, Toyonaka, Osaka 560, Japan It is well known that excited aromatic hydrocarbon forms exciplexes with N ,N-dimethylaniline (DMA) or N,N- diethylaniline (DEA). Contrary to this, when amines with free N-H groups such as aniline and N-methylaniline are used as electron donors for the exciplex formation with excited aromatic hydrocarbons in solution, no exciplex fluorescence can be observed although the fluorescence of aromatic hydrocarbon is quenched. On the basis of pre- liminary studies upon pyrene-aniline system by means of laser photolysis as well as the measurements of the effects of solvent polarity upon the fluorescence quenching rate, hydrogen atom transfer assisted by charge transfer inter- action in the encounter complex has been proposed as a 2 ) mechanism of the quenching of exciplex fluorescence. Since this is an important problem in the exciplex chemis- try, we have made a more detailed and extensive studies by means of fluorescence and transient absorption measurements with laser photolysis as well as conventional flash photo- J2-1 lysis methods . The "bimolecular rate constants (k ) of the quenching of pyrene fluorescence in n-hexane "by various primary and secondary amines were measured. The k values in the case of aliphatic amines are considerably smaller than those of the aromatic amines, which seems to indicate that an addi- tional activation energy over that for the aromatic amine system appears necessary for the quenching "by aliphatic amines. The difference of the quenching rate constant bet- ween aromatic and aliphatic amines was also observed in the case of the fluorescent exciplex formation of pyrene-DMA 3 ) and pyrene-tr ibutylamine (TBA) systems. Presumably, the additional activation energy may be necessary for the change of the structure of TBA from the pyramidal to the 3) planar one in the course of the charge transfer. It may be possible to assume similar structural change in case of the present systems. It has been confirmed that the k values increase re- el markably with decrease of the ionization potential (I ) of g the amines. This result suggest the fluorescence quenching mechanism of the charge transfer in the encounter complex. Although the values of k as well as I for N-methylaniline 0. P (NMA) or N-ethylaniline (NEA) are rather close to those of DMA or DEA, respectively, no exciplex fluorescence can be observed in the case of the formers. This result suggests the following mechanism of the quenching. J2-2 k k k, * q p a. . A + HD -*■ (A"....HD ) + (AH....D) ■* AH + D (l) yia. kV r A + HD and A + HD V_^ rf other reaction product Assuming that the rate constants of the radiative and ra- diationless transitions {k' and k! ) of (A . . . . HD ) are equal to those of pyrene-DMA or pyrene-DEA exciplex in hexane, respectively, (k + k! ) y k has "been estimated to 9 * P 1 'v p be larger than 2 x 10 s If AH and D radicals are formed by the process k , their absorption spectra may be observed by means of flash photolysis. Transient absorption bands were observed in the 380 *\j U30nm region for many pyr ene-pr imary and secon- dary amine systems in hexane. The observed spectra can be explained as a superposition of absorption bands of pyrene triplet state and those of some other transient species. At the delay time of 50us, the T-T absorption band becomes much weaker while the absorption band of the other tran- sient remains fairly strong. In order to identify the long lived transient, 1 , 2-dihydropyrene (AH p ) in methylcyclo- hexane-isopentane (3 : l) mixture at T7K was irradiated with a low pressure mercury lamp, and absorption spectra of photoproduct were measured. It seems probable that the radical AH is formed by irradiation. The obtained spectrum of the photoproduct is rather close to that of the long lived transient species observed by the flash photolysis J2-3 method. Furthermore, the spectrum of AH was measured by means of pulse radiolysis of solutions of pyrene in ethanol containing lvol . % sulfuric acid, where AH may be produced by the following reactions. C 2 H 5 OH --VVVW e solv , H, C^OH^ C^O, and other neutral radicals e . + H ■*■ H soiv (2) H + A •*■ AH The spectrum shows peaks at ^05nm and 395nm and is essen- tially identical with the spectrum obtained by subtracting the T-T absorption band of pyrene from the observed tran- sient spectra in pyrene-pr imary amine and secondary amine systems . The confirmation of the formation of D by means of the transient spectral measurement is rather difficult since the absorption spectra of anilino radicals are observed in the wavelength region of about l+00nm and their intensities are rather weak. However, since the absorption spectrum of diphenyl amino radical shows maximum at ca. TTOnm, we have examined pyrene-diphenyl amine (DPA) system in hexane solu- tion and have observed clearly the absorption band of diphenyl amino radical in addition to the absorption band of AH. Thus, the formation of both AH and D has been con- firmed . Although it is assumed in eq.(l) that the radicals are formed from the singlet excited state, it might be possible J2-4 that they are formed by the reaction in the triplet state. We have examined this problem by using a dilute solution of anthracene-DPA system in hexane, and confirmed that no re- action occurs in the triplet state since the decay process of anthracene triplet state is not affected by the addition of DPA. If the charge transfer mechanism of the fluorescence quenching is valid, k will depend upon the solvent polari- ty. Actually, we have confirmed in several cases where the additional activation energy over that for the diffusional motion is necessary for the charge transfer, that the k q value decreases with increase of the solvent polarity. This result indicates that not only the I value but also P the solvent polarity affects the additional activation en- ergy and it is decreased with increasing solvent polarity. For example, in the case of pyrene-aniline system, k is 5.7 x 10 M s in hexane, 2.6 x 10 M s in isopropanol, 9 -1 -1 and 5«3 x 10 M s in acetonitr ile , and for the pyrene- T -1 -1 dibutylamine system, it is 3.0 x 10 M s in hexane, Q _ -] -| fill 2.0 x io Ms in ethylether, Q.k x 10 Ms in acetone, Pi — 1 1 and 8.7 x 10 M s in acetonitr ile . Roughly speaking, approximately linear relation can be observed between logk and the reciprocal of the solvent dielectric constant e. This result can be explained satis- 2) factorily on the basis of the reasoning given elsewhere. Namely, we assume that the activation energy for the charge J2-5 transfer is determined by the Horiuchi-Polanyi ' s relation. The free energy of the final state of the charge transfer (the solvated ion pairs) may he written as, F = (A/e) + B (3) where A and B are constants irrelevant of e. The change of F. with e (from e° to e) is, AF. = A(l/e° - 1/e) (h) Then, the change of activation energy AE is given by, AE = aAF . , a : constant 1 (5) according to Horiuchi-Polanyi ' s relation. The ratio of k ' s in solvents with dielectric constant e and e° can be q written as , k (e)/k (e°) = exp(aAF./RT) q q. i (6) Thus, logk (e) is inversely propartional to e. Of course, in sufficiently polar solvents such as acetone and aceto- nitrile, ionic dissociation into solvated radical ions, A and HD , will complete with the proton transfer in the en- counter complex. Actually, transient spectra observed by the laser photolysis of pyrene-aniline system in aceto- nitrile show strong absorption bands of solvated radical ions . 1) N. Mataga, T. Okada and H. Oohari, Bull. Chem. Soc . Jpn. , 39, 2563 (1966). 2) N. Mataga, in "Exciplex," ed. by M. Gordon and W. R. Ware, Academic Press, New York, 1975, p. 113. 3) N. Nakashima, N. Mataga, and C. Yamanaka, Int. J. Chem. Kinetics, 5, 833 (1973). J2-6 Protolysis Equilibria Of Lumichrome (6 ,7-Dimethyl-Alloxazine) In She Lowest Excited Singlet State By Bertil Holmstrom , Sture Bergstrom and Hans-Bertil Larson, Insti- tute of Physical Chemistry, Chalmers Institute of Technology and the University of Gothenburg, Gothenburg, Sweden Fluorescence is a standard assay method for riboflavin (Vitamin Bp). This substance is very unstable to light, two major decompo- sition products being lumiflavin (6 ,7, 9-trimethyl-iso-alloxazine) and lumichrome (6 ,7-dimethyl-alloxazine) . A thorough study of the fluorescence properties of these compounds may, therefore, be of some proctical importance. Lumiflavin has a fluorescence spectrum very similar to that of the parent compound (an emission peak at 520 nm) ; the emission disappears in acid medium (pH = 1.5) and in basic medium (pH = 10). lumichrome = H ? Lc In contrast to this, the pH variation of lumichrome fluorescence presents a very complicated picture containing several features of more general interest. In the acidity/basicity range of this study, from concentrated sulphuric acid to strong sodium hydroxide solu- tion, at least three different protolysis forms are evident in emission and four in absorption (cf Pig 1 and 2). The various pK* values were estimated by two principally different methods: (i) direct observation of the change in fluorescence spectrum with pH, J3-1 600 SOO 400 300 600 £00 400 ZOO 600 5"00 400 ZOO Fig 1. Absorption spectra (broken lines; Bausch & Lomb Spectronic 505) and emission spectra (full lines; Aminco Bowman, uncorrected), (l) Strong sulphuric acid; (2) Weakly acid solution; (3) Basic so- lution. and (ii) from observed absorption and emission spectra of the "pu- re" protolytic forms using the Forster cycle. - Measurements of fluorescence lifetimes (following pulse excitation) was employed to get additional information of the dynamics of the system in weakly basic media. Changes in emission spectra with pH . - In the acid region, the change in emission spectrum occurs at a slightly lower pH than the change in absorption spectrum, indicating that the exciting form of the neutral dye is slightly stronger as a base than the ground state molecule (pK = - 1, pX* = o). In strong sulphuric acid medium, there is another change in absorption spectrum (pK„ = = - 8) without any corresponding change in the emission spectrum. Both absorption and emission spectra exhibit marked red shifts when going to stronger acid. - In basic medium, there is a change in the absorption spectrum corresponding to pK ? = 12. The shape of J3-2 to Fig 2. Summary of absorp- tion and fluorescence H3LC "a. "LC data for lumic hr ome , ta- ken from curves like /?, ("a\ ,"s those shown in Fig 1. w A ^ 0_; o — ^ ® ted) The ordinate gives the absorption or (uncorrec- emission maxima. Abscissa values below -i refer to the Harmnet -5" O € to 1$ r t ^0 n function |_1J. the emission spectrum is, in neutral and moderately basic media, strongly dependent on the buffer concentration (Fig 3), with the ratio I (470)/l (51 o) changing from 0,65 to 3,6 when the buffer concentration is increased from 0,2 M to 9 M. This indicates a competition between direct fluorescence and the establishement of the new protolytic equilibrium, according to "Scheme I": / B "" 1 H 2 .tc-hcv c HLc + Kcv e The Forster cycle . If the entropy change on protonation is assumed to be the same in the excited state as in the ground state (_i e_ O* 0\ AS = A S_ ; , the difference between the two states in standard free / O"* o \ energy change on protonation (AG - A_G ; will be the same as the difference between the two states in the enthaply change on proto- / o* o -\ nation (AH - AH J, If the latter can be determined from the ob- served absorption and emission spectra, it is possible to estimate J3-3 the difference in pK between the two protolysis forms (c_f Pig 4) : where h £ v and h c_ v represent the energy differences between the first excited singlet state and the ground state (the energy of the 0-0 band) in the two protolysis forms. In the ideal case of mirror image symmetry between absorption and emission spectra, the energy of the 0-0 transition is quite well defined. The best thing that can be done in the lumichrome case is to take mean of the wave number of the absorption maximum, v , and the emission a maximum, v . This leads to the predictions pK* = pK_. + 0.0 ^ and pK* = pKp - 6.7 ~ 5, in rather good agreement with the values obtained by the previous method. rro 5bo 0,2M / 1.0 M 3.4 M 9.4 w nrn ^5"0 400 H>L? © hey! ® hcv' UnU AH"/L HU -*» =£ C^K) ir=± Pig 3. Fluorescence spectra of weakly acidic lumichrome solutions, with different buffer concentrations, as given. Pig 4. The Porster cycle. L = Avogadro's constant. J3-4 ! ■ i; ' ! ' i ? f\\\ -;.. 1 ,1.1 ' f / \\ \ -(•* f i : U i (1/ \\\ ii 1/ '1/ 1 1 .A \ Pig 5. Smoothed pulse fluorometry curves corresponding to the time chapes of processes a, b, and c in Pig 4. TRW Nanosecond lamp; sampling technique [2 J . (a) Scatt- ered light, (b) Direct fluor- escence (X = 450 nm) . (c) Trans- formed fluorescence (X = 510 nm) , -- ; - :r ..._ rt _i_.j_,.., rt .., _. .4. 4. ..... . t . j... .. | ..... Pulse fluorometry . - In the broad pH region where the stable pro- tolytic forms are H„Lc and HLc , additional information about the molecular dynamics (cf_ "Scheme I" above) can be obtained from pulse fluorometry. Pig 5 shows (smoothed) curves from one experi- ment, showing the shape of the lamp pulse (a), the direct fluor- escence (b), and the transformed fluorescence (c). Assuming the decays to be exponential, the lifetimes of the two fluorescence components can be determined by Laplace transform deconvolution [3]. As expected from "Scheme I", the lifetime of the direct fluorescence is strongly decreased on increasing buffer concen- tration, (it is hoped that single photon counting equipment, pre- sently being installed, will allow more precise measurements than the intensity measurement technique now used) . References 1. Michaelis,L. , and Granick,S., J.A.C.S. 1942, 64, 1861. 2. Lytle,P.E., Photochem. Photobiol. 1973, J_7, 75. 3* Almgren,M. , Chemica Scripta 1972, _3, 145. J3-5 Photochemistry of Pheophytins and Porphyrins D. C. Brune, J. Fajer and S. P. Van Division of Molecular Sciences Department of Applied Science Brookhaven National Laboratory Upton, New York 11973 Recent pulsed laser spectroscopic results have revealed two short lived intermediates involved in the primary photochemistry of photo- synthetic bacteria. The first of these, which appears in less than ten picoseconds is postulated to consist of a dimeric bacteriochloro- phyll cation radical and of an anion radical of a demetallated chloro- phyll, bacteriopheophytin. When further chemistry is blocked, this primary charge separation is annihilated (in nanoseconds) by recombina- tion of the radicals to yield fluorescence and a triplet state . We describe here the _in vitro photochemical generation as well as the optical and magnetic characterization of the anion radicals of bacte- riopheophytins a and b, the pigments found in Rhodopseudomonas spheroides and viridis, respectively. The radicals undergo isotopic substitution at the proton positions which carry high unpaired spin densities. The occurrence of chlorophyll-derived anion and cation radicals in photosynthesis has logically led to attempts at mimicking the natural act with artificial porphyrin systems. One of these is a heterogeneous system consisting of multiple monolayers of porphyrins deposited on an aluminum electrode which is immersed in a ferro-ferricyanide aqueous c. solution. This photoelectrochemical cell, originally described by Wang was reported to generate ~1 V versus a platinum electrode (open circuit) on illumination of the porphyrin-coated electrode. Attempts to J4-1 reproduce this cell using a number of porphyrins which exhibit a range of oxidation potentials have been unsuccessful, the photovoltages actu- ally observed are an order of magnitude less than previously reported, notwithstanding efforts to remove or introduce impurities or charge carriers. A. Chlorophyll radicals Photolysis of bacteriopheophytin (BPh) a and b in pyridine contain- ing lOmM Na_S and 1 M H„0 yields radicals which, on the basis of their optical and electron spin resonance spectra are similar to the one electron reduction products obtained by electrolysis in aprotic sol- vents. The reduction can be induced by irradiation in each of the ab- sorption bands of the pheophytins. For BPh a (5 x 10 M, room tem- _2 perature) the quantum yield at 750 nm is 5 x 10 Analysis of the esr spectra is facilitated by electron-nuclear double resonance which reveals some of the proton interactions predicted by self«*consistent field molecular orbital calculations and by selec- tive deuteration experiments. Examples of the latter are provided in Figs. 1-3 which display the esr spectra of deuterated BPh _a in deu- terated pyridine and water, and computer simulations of the experimental spectra using the hyperfine splitting constants shown. If photolysis is performed in the presence of H„0 instead of D~0, exchange of hydro- gen for deuterium occurs first at the a, (3, 6 methine positions (Fig. 2), and eventually the methyl groups also undergo partial exchange (Fig. 3). (The assignments have been verified by nuclear magnetic resonance.) The sites which exchange are those which carry high unpaired spin densities and a significant isotope kinetic effect is observed as well. J4-2 Protonated BPh _a will substitute deuterium approximately an order of magnitude faster than the deuterated BPh will exchange with hydrogen. B. Photoelectrochemical cells Cells patterned on the Wang design shown were tested for photo- effects using copper, zinc, the |-l-oxo-iron (III) dimer of me so tetraphenyl-porphyrins, zinc porphine and-phthalocyanine. The cell consisted of 50-70 layers of porphyrin, deposited on a clean aluminum surface in a Langmuir trough, and immersed in an aqueous solution of potassium ferri-ferrocyanide. 50-70 monolayers Al porphyrin 0.1 M K Fe(CN) 6 K.Fe(CN), Pt Photovoltages (open circuit) of approximately 100 mv can be induced using white light with ultraviolet and infrared cut offs or by direct excitation of the porphyrin Soret bands (~400 nm). Attempts to improve the photovoltages by changing the electrode surface, or by including free base porphyrins, or a porphyrin cation radical as possible impurities into the porphyrin coated electrode yielded no significant effects. References 1. Rockley, M. G. , Windsor, M. W. , Cogdell, R. J. and Parson, W. W. (1975), Proc. Nat. Acad. Sci. USA 2251. 2. Kaufmann, K. J., Dutton, P. L., Netzel, T. L. , Leigh, J. S. and Rentzepis, P. M. (1975), Science JL88, 1301. 3. Dutton, P, L. , Kaufmann, K. J., Chance, B. and Rentzepis, P. M. , (1975) FEBS Lett. 60, 275. J4-3 4. Thurnauer, M. C, Katz, J. and Norris, J. R. (1975), Proc. Nat. Acad. Sci. USA _72, 3270. 5. Fajer, J., Brune, D. C. , Davis, M. S., Forman, A. and Spaulding, L. D. (1975), Proc. Nat. Acad. Sci. USA 72, 4956. 6. Wang, J. H. (1969), Proc. Nat. Acad. Sci. USA _62, 653. simulation BPh^/C 5 D 5 N/D 2 OG Fig. 1 Second derivative esr spectrum of the perdeuterated BPh anion generated photolytically in CrD-N/D-O and a computer simulation which assumes the splitting constants shown. Fig. 2 BPh generated in C-D N/H.O, in which the methine deuterons have exchanged. Fig. 3 BPh" in CJJ/HO, in which the methine and ~50% of the 1 and 5 methyl deuterons have exchanged. J4-4 Fig. 2 2 nd deriv simulation BPht/C-D-N/HLO U 5 5 2 Fig. 3 2 nd deriv simulation BPhp/C 5 D 5 N/H 2 A Kinetic Study of CH 3 2 and (CH 3 ) 3 C0 2 Radical Reactions by Kinetic Flash Spectroscopy Michael R. Whitbeck, Jan W. Bottenheim, Stuart Z. Levine, and Jack G. Calvert Department of Chemistry, The Ohio State University , Columbus, Ohio The reactions of the hydroperoxy (H0 2 ), alkyl peroxy (R0 2 ), and the acylperoxy (RC00 2 ) radicals are believed to be essential components of the reaction mechanism which describes photochemical smog generation. For example, these radicals are believed to be largely responsible for the oxidation of NO to N0 2 in the NO -hydrocarbon-polluted atmospheres : -X. H0 2 + NO -♦ HO + N0 2 (l) R0 2 + NO -♦ RO + N0 2 (2) RC00 2 + NO -> RC0 2 + N0 2 (3) It is postulated that HO, RO, and RC0 2 radical products of these re- actions react in the atmosphere with 2 , hydrocarbons, aldehydes, CO, etc., to regenerate other H0 2 , R0 2 , and RC00 2 radicals. In theory the NO -» N0 2 conversion continues to occur by way of these chain reactions, and the increasing ratio of [N0 2 ]/[NO] causes the ozone to rise through the rapid chemical interactions of the sequence k-6: N0 2 + hv -» 0( 3 P) + NO (h) 0( 3 P) + 2 + M -* 3 + M (5) 3 + NO -» 2 + N0 2 (6) In view of the great significance attached to the reactions 1-3 by atmospheric scientists today, it is disappointing to observe that the rate constant for reaction 1 has been determined only in indirect Kl-.l 2 experiments, while those for reactions 2 and 3 have not been determined experimentally by any method to our knowledge; present estimates were derived from computer simulations of very complex smog chamber experi- ments and are of very limited accuracy. In theory the alkyl peroxy radicals may also be important reactants in the atmospheric oxidation of S0 2 : R0 2 + S0 2 -> RO + S0 3 (7) H0 2 + S0 2 -> HO + S0 3 (8) 2a Davis and coworkers have studied reaction 8 for which they estimated k 8 = (5-2 ± l.l) x 10 5 i. mole~ 1 sec~ 1 . This rate constant coupled with the computer simulation of the complex chemistry of the polluted atmos- phere suggest that reaction 8 is a major removal path for S0 2 in the "I "K polluted atmosphere. In view of the fact that reaction 7 is about 8 kcal/mole more exothermic than reaction 8, the rate constant for reaction 7 could in theory be somewhat larger than k 8 . If this is the case, then the significance of reaction 7 in the atmospheric conversion of S0 2 is probable. In this work we attempted to study the reactions of several alkyl peroxy radicals using kinetic flash spectroscopy. We generated CH 3 2 and (CH 3 ) 3 C0 2 radicals by flash photolysis of (CH 3 ) 2 N 2 -0 2 and [ (CH 3 ) 3 C] 2 N 2 -0 2 mixtures, respectively. Previous work of Bjellqvist and Reit- 3 k 5 berger, Parkes and co-workers, and Burggraf and Firestone, have shown that the alkylperoxy radicals exhibit an ultraviolet absorption band (2400 -3000 A) of moderate intensity which is convenient for kinetic studies of these radicals. Kl-2 The xenon flash lamp employed in these experiments operated at 20 kJ per flash with a half -width at half -maximum intensity of about 50 ^xsec. The Vycor tube of the flash lamp paralleled the quartz re- action cell (220 cm in .length, 6.3 cm in internal diameter) but was separated from it by a Pyrex plate filter (0.5 cm thickness) which transmitted only light of X > 2900 A. A white surfaced reflector sur- rounded the flash lamp and cell. The cell housed a multiple reflection White optical system which was set for a path length of 800 cm in most of this work. Actinometric estimaces of the extent of azoalkane photolysis, and hence the number of alkyl (and R0 2 ) radicals formed per flash, were made by two independent methods : (l) measurements of the pressure increase which resulted in the flash photolysis of the pure azoalkane (1-5 Torr); (2) measurements of the amount of nitrogen product formed following the flash photolysis using mass spectrometric analysis. Both methods gave essentially the same results. Since very small amounts of decomposition of the azoalkane occurred in these ex- periments (less than .k-Qfo of the initial amount), and the pressure of added oxygen was high (about 50 Torr), the temperature rise during an experiment was no more than a few degrees. The ultraviolet and visible spectrum of the flash products was recorded photographically using a O.65 m Hilger spectrograph and a small argon flash lamp which was fired with delay times varied from 0.6 to 2.0 msec following the flash. In kinetic runs a continuous I5O w point source xenon arc was employed together with a . 5 m Jarrell Ash monochrometer equipped with an RCA IP- 28 photomultiplier-oscilloscope combination. Both sets of data define an absorption band with a maximum near 255O A which we believe Kl-3 results from the R0 2 radicals. In most kinetic runs absorbance at 265O A was followed to obtain the rate data. The kinetics of the CH 3 2 and (CH 3 ) 3 C0 2 radical disappearance was found to be second order when the appropriate azoalkane was photolyzed in mixtures with pure 2 . The rate constants for the radical decay k are in reasonable accord with those Parkes and co-workers who used a modulated beam photolysis experiment. However this agreement may be somewhat fortuitous in that the absorption spectrum observed by these workers and attributed to the CH 3 2 radical is somewhat different in band position and extinction than we observe. Perhaps interferences from the H0 2 radical absorption may have been important in the previous modulated beam photolysis studies. The build-up of CH 2 product and its time dependence observed in the previous study would insure a major source of H0 2 species. Experiments with flash photolyzed mixtures of azoalkane, 2 , and S0 2 were made in an attempt to derive the constant for reaction of CH 3 2 with S0 2 . The addition of S0 2 to this system at pressures in excess of O.5 Torr resulted in aerosol formation following the flash, and so the useful range of S0 2 pressure was very limited. The present data suggest that k 7 < 2 x 10 6 I. mole -1 sec -1 . We are continuing further work with the flash system and with a steady state photochemical system with FTS infrared analysis in an attempt to further define esti- mates of k 2 and k 7 . References l) For a general review of the very extensive literature and a more detailed discussion of the current ideas concerning the mechanism of photochemical smog formation, see: a) K. L. Demerjian, J. A. Kerr, and J. G. Calvert, Adv. Environ. Sci. Technol., k, 1 (197*0; Kl-4 b) J. G. Calvert and R. D. McQuigg, Int. J. Chem. Kinet., Symp. No. 1, 113 (1975); c) H. Niki, E. E. Daby, and B. Weinstock, Chapter 2, in "Photochemical Smog and Ozone Reactions", F. F. Gould, Editor, Advances in Chemistry Series 113, Amer. Chem. Soc, Washington, D.C., 1972, p. 16; d) A. P. Altshuller and J. J. Bufalini, Environ. Sci. Technol., 5, 39 (1971). 2) a) W. A. Payne. L. J. Stief, and D. D. Davis, J. Amer. Chem. Soc, 95, 76lk (I973r, b) R. Simonaitis and J. Heicklen, J. Phys. Chem., 77, 1096 (1973). 3) B. Bjellqvist and T. Reitberger, INIS, 1104 (1971). k) D. A. Parkes, D. M. Paul, C. P. Quinn, R. C. Robson, Chem. Phys. Letters, 23, ^25 (1973). 5) L. W. Burggraf and R. F. Firestone, J. Phys. Chem., 78, 5O8 (197*0. K1^5 The Reaction of OH Radicals with C H\- and C H, Ralph Overend and George Paraskevopoulos Division of Chemistry National Research Council of Canada Ottawa, Canada K1A0R9 Introduction There is considerable interest in the mechanism and the rates of reaction of olefins with radicals of atmospheric interest. In the case of hydroxyl radicals with ethylene and propylene there is little rate data at atmospheric pressure; the majority of the data having been obtained at low pressures. The reported rates vary over a range of 3-5 and in the case of ethylene a systematic effect of pressure has been la) identified . In the present paper we report measurements of hydroxyl radical reactions over the pressure range 20-^+00 torr and with various third bodies, as measured by the technique of flash photolysis resonance absorption. The results for ethylene show an effect due to pressure and nature of the third body while those of propylene are pressure indepen- dent in the range of our measurements. Experimental The apparatus and technique have been described in detail else- 2) where . Hydroxyl radicals were created by the Vacuum U.V (A>l60 nm) photolysis of water in the presence of the other reactant; their concen- tration was monitored by following the time resolved attenuation of the OH resonance radiation (Q or Q rotational lines of the (o,o) band of the A I ->X II transition) produced by a microwave discharge in a low K2-1 pressure Ar/H p O mixture. Signal averaging was used to improve the sig- nal to noise ratio. The concentration of the olefin was such that the 3 5 -1 measured pseudo-first order rates were between h x 10 - 10 s To minimize the photolysis of the olefin while still permit- ting the generation of sufficient hydroxyl radicals from the water vapor, a filter 15 mm deep of NH at 10 torr was used to reduce photolysis at wavelengths longer than 190 nm. Results and Discussion Propylene . Initial experiments with the propylene/OH system showed that the apparent rate decreased with repeated flashes on the same mixture and that the decrease was larger with high (~600 J) flash energy and high absorption by OH, i.e. (high hydroxyl concentration). Gas chroma- tographic analyses of C EV/He mixtures corresponding to our experimental range showed that the direct photolysis of CX at 600 J was approxi- mately 5-10$ per flash, while with water present ~30-U0$ of C-EV was consumed per flash. All experimental values reported were obtained in experiments of a single flash <100 J and low HO pressure (i.e. absorp- tion by OH of <5%) . We calculate that under the worst conditions of low CX pressure the ratio C H^/0H is >20 and consumption of olefins should not significantly affect our measurement. The second order rate con- 13 3 -1 -1 stant obtained is, k_ = 2.5 ± 0.5 x 10 cm mol s and was not C3H5 affected by the pressure or the nature of the third body (M = He, 50 torr - M = SF^, U00 torr.). A fivefold variation of the ratio C Hg/0H did not produce a change in the measured rate within the experimental error; 2) this, in accordance with a numerical simulation would indicate no in- fluence of secondary radical reactions on the rate constant. K2-2 Ethylene . The conditions for minimum OH, described above were also used in the experiments with C H, . The reaction was studied over a range of pressures and with He, SF^, CF, and HO as third bodies. The rate was found to depend on the pressure and the nature of the third body in la) agreement with another study . HO and SF^- were found to be very efficient third bodies while He was very inefficient. The limiting sec- 12 3 ond order value of k T _ was found to be k^ = 6.5 ± 0.5 x 10 '' cm I loo IIoo mol s . It was established that the limiting value depends only on the pressure and the nature of the third body and not on the OH concen- tration. References 1. a) D.D. Davis, S. Fischer, R. Schiff, R.T. Watson and W. Bolinger. J. Chem. Phys. 63, 1707 (1975). b) R. Atkinson and J.N. Pitts, ibid. 63_, 3591 (1975). c) A.V. Pastrana and R.W. Carr, Jr. J. Phys. Chem. 79, 765 (1975) and references therein. 2. R.P. Over end, G. Paraskevopoulos and R.J. Cvetanovic. Can. J. Chem. 53, 337^ (1975). K2-3 On the Triplet State (s) of Sulphur Dioxide Jan W. Bottenheim , Fu Su, David L. Thorsell, Jack G. Calvert and Edward Damon, The Ohio State University, Columbus, Ohio ^3210. Introduction Theoretical work indicates that the first three triplet-states of S0 2 ( 3 B 2 , 3 A 2 , 3 Bx) are very close in energy and can be expected in the wavelength range of 3^00-3900 A. High resolution spectroscopic work has shown perturbations in 3 Bx bands which give indirect evidence of the proximity of the 3 A 2 and 3 B 2 states in this region, although the major features of the optical spectrum are undoubtedly due to the o 2 Bx-state. Phosphorescence-sens itization studies with S0 2 -biacetyl mixtures form other evidence for the presence of at least one other triplet state besides the 3 B!-state in this region. Moreover, photochemical reaction studies seem to imply the participation of more than the two optically accessible singlet states ( 1 A 2 and 1 B 1 ) and the 3 Bi- k state only. Direct evidence for the presence of the other triplet states is missing as of today, however. The theoretically calculated zero-pressure lifetime of the emit- ting 3 Bi -state of S0 2 , based on measurements of the integrated absorp- tion spectrum indicates a reciprocal lifetime of 79±5 sec -1 . The reported experimental values of this lifetime, based on phosphores- cence or lifetime studies, are rather dramatically at odds with this number, ranging from 375 to 2000 sec -1 . If these data are correct a large part of the 3 B;l -state molecules should decay in a unimolecular non-radiative fashion, which is not expected within the present K3-1 theories of radiationless transitions, and a modification of this ■■■ theory might he necessary. * We have initiated research to establish more clarity into these matters and "will report some of the results here. Two different approaches were followed. The direct singlet-triplet absorption '* spectrum was probed with a much longer pressure -path than before in an attempt to identify new absorption-bands that might be described to the other triplet-states. The major part of this work concerns phosphorescence lifetime studies, in which we used several different lasers to excite S0 2 -molecules at a range of important wavelengths. Experimental The direct singlet-triplet absorption study was performed with a 2 m long multiple-pass White-cell. With S0 2 pressures of 1+00 to 600 Torr a pressure-pathlength of 28,800 to h^>,200 Torr-meter was obtained. The absorption-spectrum was probed with a 150 Watt Xe-high-pressure lamp with a soft-glass filter in front to minimize output below 3200 A, and monitored with a 0.5 m Jarrel-Ash monochromator . In the phosphorescence experiments the reaction cell was of the 22 i variety (diameter 35 cm) to allow the measurement of long-living •$ emitting species to low pressure regions without having major wall- 7 deactivation effects to correct for. The cell has two side-arms with fused-on Brewster angle windows (Suprasil) for entrance and exit of the laser-beam, and a third side-arm at 90° with a 2 inch diameter Suprasil-window to which as close as possible was positioned an RCA 7265 photomultiplier. Between the window and the PM-tube a Kodak Wratten 2B filter and an Optics Industries cutoff filter nr. WG-395 K3-2 1 ;i were inserted to minimize scattered laserlight and fluorescence of singlet-excited S0 2 -molecules . All lasers used in this study were KDP-doubled dye-lasers of moderately high power in the uv (3 to 10 Kwatt). An echelle grating in the 100$ mirror in the dye-laser cavity narrowed all laserlight to a 2.5 A "bandwidth. For excitation at 36 31 A a Ruby laser pumped solution of dye nr 16 (Candela) was employed, resulting in an approx. 1 Mwatt 25 nsec laserpulse at 7262 A. All other wavelengths were obtained by flashlamp-excitation of appropriate dye-solutions in a coaxial flashlamp-dyeflow cell (Candela, model CL-66), resulting in approx. O.I5O Mwatt, I5O nsec laser pulses. The lasers could not be rapidly pulsed, so essentially single- shot experiments were done. The phosphorescence signals were moni- tored via a Tektronix 770^- oscilloscope, and digitized with a Biomation 610 transient digitizer for storage on paper tape. The traces of several shots at a particular S0 2 pressure were computer averaged and then displayed in a semi-log form on the CRT-screen of a PDP-7 com- puter, after which appropriate least-square treatment was performed to establish the decay-parameters . Results and Discussion a. Direct singlet-triplet absorption study . Several new bands were observed in the absorption spectrum between 38OO and UlOO A. A close analysis shows that all these bands are hot bands of the a 3 Bx- X Ai spectrum, however. Table 1 shows the identification of these bands, whereas Table 2 compares some relative integrated absorption- intensities with calculated relative number densities of ground state K3-3 S0 2 , assuming a Boltzraann distribution only. A conclusive test was next performed, where the cell was cooled to -15°C. A dramatic decrease of intensity of all new bands was observed. Table 1 Identification Band-nr v cm 1 24651 2 24752 3 24989 1st hot band 25258 4 25*166 0-0 band 25776 5 25991 0,0,0 0,0,0 0,1,0 0,0,0 0,2,0 0,0,0 0,2,0 1,1,0 )«-(i,o,o Mo, 2,0 )«-(l,0,0 Mo, 1,0 Mo, 2,0 Mo, 0,0 Mo, 1,0 Mo, 2,0 Av exp ^calc 1145 1151 1042 1035.4 1149 1151 518 518 1029 1055.4 (b) (b) 5O5 518 1045 IO55.I+ b iff indicate the experimental and calculated distance to the indicated triplet band. This band was used as reference. All other bands are determined from their distance to this band. (o,o,o)«-(o,i,o) Calc 1 Exp. 1 Table 2 (o,o,oMo,2,o) 0.084 0.101 (0,0,0)4.(1,0,0) 0.048 0.056 From these experiments it follows that the integrated extinction coefficient of transitions to other triplet-states, if present in the studied wavelength- area should be at least a factor 50 less than to the 3 Bx -state. It seems therefore unlikely that such other states will be identifiable at all in absorption. b. Phosphorescence studies . Our data-analysis was initially directed to detect similar non- exponential decay-curves as seen in 8 fluorescence-studies by Brus and McDonald Only single -exponential K3r-4 decay was observed to pressures as low as 2 m Torr. Some curvature below this pressure becomes noticeable, but this is readily shown to be due to residual fluorescence under these conditions. (Only with excitation in the singlet manifold was it possible to study the phos- phorescence down to these low pressures.) Figure 1 shows a Stern-Volmer plot of the measured decay-rates of the phosphorescence signal as function of pressure for some different wavelengths of excitation. Since all data points fall very well within the error limits of the least squares line, we conclude that no matter how S0 2 is excited, the same phosphorescing state is monitored. This applies equally well for excitation at 3631 A, at which wavelength strong perturbations are observed in the absorption-spectrum, tenta- tively ascribed to the influence of other triplet-states, and for exci- tation at 3130 A at which wavelength two singlet states are populated ( A2 and 1 B 1 ) as possible precursors for the triplet-state. The latter observation is of interest, since it indicates that the very efficient collisional deactivation of the 1 A 2 -state should lead at least in part to triplet-population. In fact, simple spin- orbit coupling should favor the """Ag-state as the precursor to the 3 Bi- o o state; and for excitations at 3235 A and 3251 A, where we estimate that the 1 A 2 -state is initially populated to the extent of 2k and 58% at 3235 A and 3251 A respectively. All these results point to the same conclusion that only one triple state is seen in phosphorescence. Although cascading from another triplet-state has been suggested is not excluded, it is unlikely in view of the fact that our conclusion holds to pressures K3-5 down to at least 10 m Torr, where even complete vibrational relaxation within the 3 B!-state is not complete Our phosphorescence data permit a new estimate of the zero -pressure lifetime of the 3 Bx -state. It has been suggested that the much higher experimentally determined reciprocal zero pressure lifetimes are in error due to too long extrapolation to zero pressure . Our data "based on sever- al excitation-wavelengths in the pressure-region of 10 to 100 m Torr, see Figure 1, all tend to indicate a value of 130 to I5O sec" 1 . Taking the inherent experimental errors in our experiments and in the determin- ation of the integrated absorption-coefficient into account, we feel this value is the same order of magnitude and therefore suggest with some caution that the true zero pressure quantum yield of phosphorescence of the 3 B;l -state may be near unity in agreement with the existing theory. Acknowledgements Many stimulating discussions with Professor J. D. Meadors during the construction of the different lasers is gratefully acknowledged. Professor C. Weldon Mathews, Dr. K. J. Chung, Mrs. R. Damon and B. Hessler made valuable contributions to parts of this work, and computer time was contributed by the ElectroScience Laboratory of The Ohio State University. Literature cited : (1) T. F. Hayes, G. V. Pfeiffer. J. Am. Chem. Soc, 90, 4773 (1968); I. H. Hillier, V. R. Saunders, Molec. Phys., 22, 193 (1971); K. J. Chung, Ph.D. Thesis, Ohio State University, 1974. (2) J. C. D. Brand, V. T. Jones, C. diLauro, J. Mol. Spectr., V?, 404 (1973). K3~6 (3) A. M. Fatta, E. Mathias, J. Heicklen, L. Stockburger III, S. Bras- lavski, J. Fhotochem., 2, 119 (1973/74). (4) cf. ref. 3; also K. J. Chung, J. G. Calvert, J. W. Bottenheim, Int. J. Chem. Kinet., 7, l6l (1975). (5) S. J. Strickler, J. P. Vikesland, H. D. Bier, J. Chem. Phys., 60, 664 (1974). (6) J. P. Briggs, R. B. Caton, M. J. Smith, Can. J. Chem., £3, 2133 (1975). (7) P. B. Sackett, J. T. Yardley, J. Chem. Phys., £7, 152 (1972). (8) L. E. Brus, J. R. McDonalds, J. Chem. Phys., 6l, 97 (1974). (9) H. D. Mettee, J. Chem. Phys., 49, 1784 (1968). (10) F. C. James, J. A. Kerr, J. P. Simons, Chem. Phys. Lett., 2£, 431 (1974). (11) R. B. Caton, A. R. Gangadharan, Can. J. Chem., £2, 2389 (1974). K3-7 Fig-1. Stern-Volmer Plot for the Phosphorescence of SO2 3 - ■ 30^6 X rt"N ' 3130 1 • 3235 X | a 3238 . A 3631 O

Products (including PAN) The chain terminating step involves the formation of stable compounds. Peroxyacetyl nitrate (PAN) is one such compound formed from the reac- tion of nitrogen dioxide with peroxyacetyl radical. PAN, a strong oxidizing compound, is one of the principal nitrogen- containing products of photochemical smog; it and its homologs have been shown to be strong eye irritants and can cause severe plant K5-1 damage. The infrared absorption bands of PAN were first observed in lab- oratory irradiations of auto exhaust by Stephens et. al_. in 1956^ , (2) and shown present in the Los Angeles atmosphere in 1 957 v . It was first called "Compound X" because of the unidentified IR absorption band in the spectra of irradiated auto exhaust. In 1961 the formula of "Compound X" was established as that of peroxyacetyl nitrate. A simple preparation of PAN-type compounds that produce high yields in a short time was needed to prepare samples for research, testing and calibration of PAN detection instruments. Four methods have been employed to prepare PAN-type compounds. These are: Photolysis of nitrogen dioxide in the presence of an olefin. Photolysis of di-acyl compounds and nitrogen dioxide. Photolysis of low concentrations of alkyl nitrite in oxygen. The dark reaction of an aldehyde with nitrogen dioxide and ozone. A new PAN synthesis suggested itself during a study of reactions of chlorine atoms with organic material. Chlorine atoms are highly selective in the abstraction of hydrogen atoms. The weakest bonded hydrogen is preferentially abstracted to form HC1 . The chlorine atom abstraction of an aldehydic hydrogen from an aldehyde yields the acyl- type free radicals that are precursors of PAN-type compounds. In the presence of oxygen and nitrogen dioxide PAN-type compounds are formed. Any of the PAN homologs can be produced from its parent aldehyde cleanly and at \/ery high yields. In the case of acetaldehyde the reactions are: K5-2 CI + hv + 2C1 CI + CH 3 CH0 -»• HC1 + CH 3 CO CH 3 C0 + 2 - CH 3 CO(0 2 )- CH 3 C0(0 2 ). + N0 2 -► CH 3 CO(0 2 )N0 2 (PAN) The reactions were carried out in a 690 liter glass photochemical reaction cell surrounded with forty-eight 40 watt black lamps in which molecular chlorine photodissociated with a half-life of about 2 minutes. A rapid scan Fourier Transform Infrared Spectrometer was used to analyze reactants and products with an absorption path length of 500 meters. The formation of PAN is shown in Figure I. The conversion of alde- hyde is essentially complete after 20 minutes. Only traces of such side products as CO, HC00H, HN0 3 and CH 2 were observed. When 4 ppm acetaldehyde, and 1.6 ppm N0 2 were photolyzed in the absence of chlorine, PAN formed very slowly, only 0.15 ppm PAN formed after 120 minutes of irradiation. With the addition of 0.8 ppm chlorine to the system, PAN increased from 0.15 to 1.55 ppm after 4 minutes of irradiation. Figure II shows the peroxypropionyl nitrate formed from propion- aldehyde, and Figure III shows the peroxybenzoyl nitrate formed from benzaldehyde. Peroxyformyl nitrate, the simplest homolog of the series, is apparently formed in the irradiation of formaldehyde, NOp and chlorine but is \/ery unstable. The chlorine-aldehyde-N0 2 method of PAN preparation appears to be general in its application. High yields of PAN are produced relatively K5-3 free of organic side products and gives further confirmation to the generally accepted molecular structure of peroxyacyl nitrates. 600 CM Figure I Spectrum at 500-M path length of peroxyacetyl nitrate formed from the irradiation of 4 ppm acetaldehyde, 2 ppm NO2, 2 ppm CI 2 in 760 torr air. 600 CM Figure II. Spectrum at 500-M path length of peroxypropionyl nitrate formed from the irradiation of 3 ppm propionaldehyde, 3.1 ppm NO2, 1.6 ppm CI 2 in 760 torr air. Figure III. Spectrum at 500-M path length of peroxybenzoyl nitrate formed from the irradiation of 3 ppm benzaldehyde, 3.4 ppm NO2, 1.8 ppm CI 2 in 760 torr air. Literature Cited (1) Stephens, E.R., Hanst, P.L., Doerr, R.C., Scott, W.E., Ind. Eng. Chem., 48, 1998 (1956). K5-4 (2) Scott, W.E., Stephens, E.R., Hanst, P.L., Doerr, R.C., Proc. Am. Petrol. Inst. (Ill) 171 (1957). K5-5 Chemical Control of Photochemical Smog. Julian Heicklen , Department of Chemistry and Center for Air Environment Studies, The Pennsylvania State University, University Park, PA 16802. Photochemical smog is generally associated with the atmospheric con- dition resulting from the chemical reactions of hydrocarbons and the oxides of nitrogen in the atmosphere under the influence of sunlight. A principal source of the hydrocarbons and oxides of nitrogen is from automobile exhausts. The automobile exhaust contains unburned and partially oxidized hydrocarbons, as well as nitric oxide, NO. Relatively little nitrogen dioxide, NO2, is emitted directly from automobiles. It, as well as 3 and other oxidants, are produced in the atmosphere and are secondary pollutants. The initially emitted gases are non-toxic at their concentrations in the atmosphere. On the other hand, 03 and other oxidants are a dan- ger in the atmosphere. The photochemical cycle which causes the conversion of NO to NO2 is now generally accepted to be of a free radical nature involving the HO radical and can be summarized as HO + HC -> R(+ H 2 0) R + 2 -> R0 2 R0 2 + NO -> RO + NO 2 R0(+ 2 ) ^ HO or HO 2 + aldehydes or ketones HO 2 + NO ■* HO + NO 2 where HC represents the hydrocarbon and R represents the free radical. The HO radical, which must be produced initially as the result of photo- lysis of some compound in the atmosphere (presumably aldehydes) is re- generated in the chain and can catalyze the oxidation of many NO mole- cules. K6-1 If the conversion of NO to NO2 is effected through a long chain mechanism involving HO radicals, then this conversion could be inhibited by molecules which remove HO radicals in such a manner that the chain is not regenerated. Such molecules would most likely contain an easily abstractable hydrogen atom. However the radical that is produced (either initially or ultimately) when this hydrogen atom is removed must not retain a hydrogen atom on the a-carbon atom, so that the chain will not be regenerated. In our laboratory several such free radical scavengers have been tested, and they all inhibit the conversion of NO to N0 2 in irradiated synthetic HC-NO-0 2 mixtures. As an inhibitor to use in urban atmospheres, we have chosen diethylhydroxylamine (DEHA) C 2 H ^ N-O-H C 2 H5 (DEHA) for two reasons: 1. It is one of the most effective inhibitors found, only surpassed by N-methy laniline . 2. It is a simple non-aromatic molecule composed only of C, N, H, and atoms, and thus it and its reaction products are likely to be non- toxic. Studies in our laboratory have shown that with C 2 H 4 as the hydro- carbon and DEHA present at 25% of the C 2 H k levels, the removal rate of C 2 H lt is reduced by a factor of 5 and the rate of conversion of NO to N0 2 is reduced by a factor of 20. There are two effects operating. One of these is the scavenging of free radicals by DEHA to inhibit the photochemical oxidation processes. The second effect is that any O3 produced reacts rapidly with DEHA (as shown in separate experiments) to K6-2 produce CH3CHO, C2H5OH, and C 2 H 5 N0 2 . Therefore any N0 2 that is produced photodissociates to convert it back to NO, and the oxygen atom produced adds to O2 to produce O3 which is removed by DEHA. As a result photo- chemical smog formation not only is delayed but is eliminated as long as DEHA is present. The C2H5NO2 and C2H5ONO2 are not produced in the initial inhibiting reaction but are formed only when the DEHA is nearly exhausted from our reaction mixtures and NO2 and O3 are produced. The C2H5ONO2 and C2H5NO2 come from the oxidation of DEHA in the presence of NO2 and O3, respec- tively. Under conditions that would prevail in the atmosphere with DEHA present, no NO2 or O3 would be produced and consequently no C2H5ONO2 or C2H5NO2 will be formed. If it is desired to cleanse the atmosphere of DEHA without producing these compounds, this can be done at night when NO2 and O3 do not form. The overall effect of adding DEHA to the atmo- sphere will be to retard hydrocarbon oxidation and to convert some of the NO to N2O. N2O is already present at 0.25 ppm in the air. It diffuses to the stratosphere where 98% of it becomes N2 . The results appear to be the same in the presence or absence of H2O vapor. Changing the initial concentrations of reactants by a factor of 15 also appears to have little if any effect, though all of our mix- tures were of the same relative composition, i.e. [olefin]/ [NO]/ [DEHA] = 4/2/1. If these results can be extrapolated to urban atmospheres then we can expect that 30-50 ppb of DEHA will have a sufficient retarding effect on the oxidation of NO to NO2 (i.e. about a factor of 3 or more) to effectively prevent photochemical smog formation (i.e. keep oxidant levels below 0.08 ppm). Presumably DEHA will be needed for 6-8 hours a K6-3 day for ^ 200 days /year in Los Angeles and 50-100 days /year in another 10 U.S. cities. The advantages of adding DEHA to urban atmospheres rather than using automobile control devices are numereous: 1. The DEHA method controls stationery as well as mobile source emissions. 2. The DEHA method is projected to cost about $200 million annually, which is at least a factor of 100 cheaper than automobile control devices. 3. The DEHA method can provide the immediate elimination of photochemical smog, whereas it will take at least 10 years to equip all cars with emission control devices. 4. No government regulation of the individual is involved with chemical control, thus minimizing government interference into the lives of the individual citizen. 5. The DEHA method is believed to be safe, unlike the automobile catalytic control devices which introduce significant and harmful amounts of ^SOi* into the atmosphere. 6. In fact not only does DEHA not introduce HaSOl, into the air, but it should reduce HaSOit concentrations because it will inhibit the free- radical oxidation of S0 2 . Furthermore DEHA is a mild base and has the potential for neutralizing H 2 S0i t . One obvious problem that must be faced is how DEHA is to be disper- sed into the air. Since the reaction times in the air are of the order of several hours, mixing does not have to be rapid. DEHA could be intro- duced by evaporation (vapor pressure is 3-4 Torr at 25 °C) or spraying Kfr-4 from moving or stationery sources on the freeways . DEHA is a colorless liquid and its vapor does not absorb radiation o above 2700A, so that it is not photochemically reactive (no radiation o < 2900A reaches the troposphere) . The odor threshold of DEHA is 0.5 ppm, > 10 times the level that would be used. Thus it will cause no annoyance to the general popula- tion. It has a faint odor at 1.4 ppm and a moderate odor at 7.5 ppm, so that the odor is a built-in safety device in case high concentrations develop in some region. Detailed toxicologicul studies on rats have been conducted in our laboratory. The animals were exposed for over 1300 hours to 9 ± 1 ppm ,° DEHA, 10 ± 1 ppm C2H5NO2, and the vapor of (C 2 H 5 ) 2 N0S-— H (the adduct of x DEHA and SO2). Animals have been periodically sacrificed and subjected to hematological and blood chemistry evaluation and complete post mortem examinations. Only two deviations have occurred between the control and test animals. 1. After 855 hours of exposure one male rat was found to have a malig- nant tumor in the skin on a rear leg. No other tumors have appeared. 2. After 1227 hours exposure the two exposed females that were sacri- ficed had fluid in the uterus. Subsequent matings of 9 other ex- posed females with 3 exposed males (3 females per male) resulted in normal births with normal offspring. K6-5 Photochemically Induced Free Radical Reactions in Nitrogen Dioxide-Acetaldehyde Mixtures* E. R. Allen Atmospheric Sciences Research Center State University of New York at Albany Albany, New York 12222 Abstract Photochemical studies have been made of the room temperature rates of formation of products and decay of reactants during irradiation at 366 nm of binary mixtures of nitrogen dioxide and acetaldehyde. Pre- liminary results from computerized numerical modelling incorporating currently accepted kinetic parameters of the elementary processes occurring in these systems has enabled a viable overall mechanism to be established and provided additional reliable information on less well known rate constants for hydrocarbon radical-nitrogen oxide reactions. Knowledge of the latter is of importance in chemical models describing the propagation of photochemical air pollution and in estimating urban concentrations of potentially reactive or environmentally hazardous nitrogenous products. ■k -k * * Introduction Static photolysis studies of nitrogen oxide-hydrocarbon mixtures were undertaken at the NCAR laboratories to complement and augment parallel investigations of related atomic oxygen-hydrocarbon reactions utilizing dynamic discharge flow techniques. The latter studies pro- vided kinetic data on the initial (primary thermal process) O-atom *Work performed at the National Center for Atmospheric Research (NCAR) , Boulder, Colorado. NCAR is sponsored by the National Science Foundation, K7-1 attack on hydrocarbons, whereas the former investigations were employed to indicate the mode of O-atom attack on the hydrocarbons and allowed us to follow the subsequent reactions (secondary thermal processes) in- volving product free radicals and the various molecular species present. In addition, the static photochemical experiments were designed to pro- vide information concerning relative and absolute rate coefficients for the simultaneous and consecutive secondary processes which, due to their competitive nature, are not easily isolated for study by other techniques. The basic system employed, consisting of a photochemical reactor used in conjunction with periodic sampling and gas chromato- graphic analysis of irradiated mixtures, has been described elsewhere. Since these early reports the system has been substantially upgraded and refined to the extent that additional details, concerning modifica- tions and improvements incorporated more recently , warrant inclusion here. Early work consisted mainly of preliminary investigations designed to: test the feasibility of using gas chromatographic analysis to follow the course of reaction during photolysis, isolate and identify the products of reaction, calibrate the analytical system for known products and reactants, and obtain relative rate coefficients of elementary processes from initial rates of product formation assuming the steady state approximation method applied for the experimental conditions selected. A similar experimental approach has been used by other investigators to study reactions in the nitrogen oxide- acetaldehyde system. However, these investigators examined different aspects of initial O-atom attack and end-product formation due to differing experimental procedures and priorities. Here we are mainly K7-2 concerned with defining the overall mechanism and assigning rate con- stants to the elementary processes involved. These studies were part of a laboratory program to quantitatively identify selected chemical transformations occurring in polluted air and to provide kinetic data on elementary processes suitable for in- corporation into atmospheric chemical models, which are used to describe the urban and industrial environment. Experimental Experiments were conducted in a conventional high vacuum, mercury free static photolysis system. All glass stopcocks were lubricated with Dow Corning silicone grease, which proved to be more stable to attack by nitrogen oxides and organic vapors than other commercially available lubricants. Radiation at 366 nm was obtained from the filtered (Corning CS-7-83) emission of a PEK110 high pressure mercury arc lamp. Intensities incident on the photochemical reactor were measured using a potassium ferrioxalate actinometer 8 and transmitted radiation monitored by recording the output of an RCA935 phototube coupled to a Pacific Photometric Photometer. Periodically small aliquots (0.1 cm 3 ) of the reaction mixture in the cell (300 cm 3 ) were extracted using a leak-free rotary valve developed from a design reported by Alperstein and Bradow. These samples were immediately injected either on to parallel Poropak Q columns at room temperature and -78 C for the separation and detection of CH, , C0 2 , N„0 and N„, 2 , NO, respectively, using helium ionization; or on to a 100' capillary column at 50 C coated with 3, 3' oxydipropionitrile for the separation and detection of organic compounds using flame ionization. Gas pressures K7-3 were monitored using a modified Wallace and Tiernan absolute gauge and a Statham PA707 TC-5-350 transducer. Eastman Organic Chemicals acetaldehyde and organic products were purified by trap to trap distillation at reduced temperatures. Matheson nitrogen dioxide (99.5% min.) was oxidized by bubbling oxygen through the liquid at C, dried with P?0,- and distilled into a blackened storage reservoir. Permanent gases of high purity were obtained directly from Matheson cylinders. Results and Discussion It is not possible to present all of the data obtained to date because of space limitations. However, the following thirteen step mechanism accounts for more than 95% of the products observed. N0 2 + hv ■+ NO + (1) fabs I o + N0 2 ■> 2 + NO (2) k 2 + CH 3 CHO ■> OH + CH 3 CO (3) k 3 OH + CH 3 CHO -»■ H 2 + CH 3 CO OH + NO 5 HNO- OH + NO 5 HN0 2 (A) \ (5) k 5 (6) k 6 CH 3 CO + N0 2 -> CH 3 + C0 2 + NO (7) k 7 CH 3 CO + CH 3 + CO (8) k 8 CH 3 + N0 2 -> CH 3 N0 2 (9) k 9 CH 3 + N0 2 ■* CH + NO (10) k 10 CH 3 + N0 2 -*■ CH 3 ON0 2 (ID k ll CH~0 + NO ■> CH 0N0 (12) k 12 M CH 3 + NO 5 CH 3 NO (13) k 13 Relative and absolute rate coefficients for these reactions will be discussed in terms of initial product formation rates and the matching of experimental data to computerized simulations of product formation and reactant decay. The latter procedure involved using the K7-A Runge-Kutta technique for solving a set of differential equations in- corporating coefficients for reactions (1) through (13) and the initial reactant concentrations, over one minute time intervals. Acknowledgement The author is greatly indebted to T. R. Englert who performed the large number of necessary experiments, and M. Sutterlin, who developed and ran the computer program. References 1. R. D. Cadle and E. R. Allen, "Reactions of ( 3 P) Atoms with Aldehydes in Photochemical Smog." In "Chemical Reactions in Urban Atmospheres," C. S. Tuesday (ed.), Elsevier Publishing Co., 63-87 (1971). 2. E. R. Allen and K. W. Bagley, "The Reactions of Methyl and Acetyl Radicals with the Oxides of Nitrogen," Berichte Bunseges. Physik. Chemie., 72, 227-233 (1968). 3. E. R. Allen, "Further Studies of Reactions of Radicals with the Oxides of Nitrogen." Paper presented at 8th Conference on Photo- chemistry, Ottawa, Ontario, Canada, June 1968, 4. K. W. Watkins, P. J. Ogren and E. R. Allen, "The Effects of Col- lisional Moderator Gases on Photochemically- Induced Reactions in Nitrogen Oxide-Propylene Mixtures." Paper presented at 8th International Conference on Photochemistry, Edmonton, Alberta, Canada, August 1975. 5. H. E. Avery and R. J. Cvetanovic , "Reactions of Oxygen Atoms with Acetaldehyde," J. Chem. Phys. , 43, 3727-3733 (1965). 6. B. M. Collins and M. I. Christie, "Reactions of Oxygen Atoms with Acetaldehyde," Nature, 218, 1245-1246 (1968). 7. H. E. Avery, D. M. Hayes and L. Phillips, "Reactions of Acetyl and Methyl Radicals with Nitric Oxide," J. Phys. Chem., 73, 3498-3499 (1969). 8. C. G. Hatchard and D. A. Parker, Proc. Roy. Soc. (London), A235 , 518 (1956). 9. M. Alperstein and R. L. Bradow, Anal. Chem., 38, 366 (1966). K7-5 PHOTO-OXIDATION OF TOLUENE-NO^ O-N SYSTEM IN GAS PHASE H. Akimoto, M. Hoshino, G. Inoue, M. Okuda and N. Washida Div. of Atomospheric Environment National Institute for Environmental Studies P.O. Yatabe, Ibaraki 300-21, Japan Introduction A number of studies have been made on the photo-oxidation of olefin- N0_-air system as a model reaction of photochemical air pollution in 1 2) urban atomosphere. Although similar studies ' are also available on aromatic hydrocarbon-N0 9 -air system in less detail, most of the studies concern with the so called "photochemical reactivity", and studies on the reaction products characteristic to aromatic hydrocarbons have rare- ly been carried out. Studies aiming at the elucidation of reaction mechanism of oxidation of aromatic hydrocarbons in gas phase are also very scarce. 3) 4) Jones and Cvetanovic , and Grovenstein and Mosher have studied reactions between oxygen atom and toluene. The main products of these reaction systems were phenol and cresols. On the other hand, Kopeczynsky studied the photo-oxidation of alkyl benzenes-NO -air system using a long path infrared spectrometer and identified aldehydes, CO., CO, formic acid and peroxyacetyl nitrate ( PAN ) among the reaction products. The purpose of this study is the elucidation of the reaction mech- anism of photo-oxidation of toluene-N0_-0 -N system in the concentration range of 34 ppm of toluene, 10 - 200 ppm of NO in N and /or at 1 atm. Experimental The reaction chamber was a Pyrex cylinder with the dimensions of 240 mm i.d., 1660 mm long and 67 1 volume, and was evacuable to less than 1 x 10 torr. Metal parts containing gas inlet and outlet were attached R8-1 to the both ends of the Pyrex cylinder with viton O-rings . Inside surface of the metal parts were lined with teflon so as to avoid the catalytic decompositon of certain kinds of reaction products. Each end of the re- action chamber' sealed with a pyrex window of 20 mm thick, through which photolyzing light beem was transmitted into the gas sample. Light source was a 500 W xenon short arc lamp and a parallel light beam of 200 mm was obtained with an elliptic mirror, a lens and an off-axis parabolic mirror. These optical arrangement was adopted for the purpose of accom- plishing the uniform irradiation of the gas sample and also of avoiding the illumination of the side wall of the reaction chamber. The reaction mixture was first sampled into a constant volume glass sampling bulb ( about 700 cc ) and then concentrated in a GC sampling tube. The sampling tube was a pyrex spiral with 2 mm i.d. and about 4 m extended length. The concentration of the reaction mixture was achieved by evacuating the air and NO while cooling the tube with ethanol - liq. N at -60 ^ -80 °C. After the concentration, the sampling tube was heated with hot water and the sample was fed into a gaschromatograph directly. This procedure was found to minimize thermal reaction of N0_ with toluene and cresols in the sampling tube. Product identification was performed with a gaschromatograph mass spectrometer (JEOL, JMS-D100 ) using a 3 m GC column of SE-30 on shima- lite W. The GC oven temperature was raised from 80 °C to 200 °C at a rate of 4 °C/min. Quantitative GC analysis was carried out using the same column as above. Results and Discussion In the present study, the wavelength of the irradiation light is K8-2 lmited to about 350 nm and longer. In this spectral region, the photo- oxidation of toluene-NO -0 -N system is initiated by the photo-disso- ciation of NO . NO + hv > NO + (1) The radiation intensity gave a photo-dissociation rate coefficient for N0 2 of 1.3 x 10~ 3 sec" 1 . In the reaction system of toluene ( 34 ppm )-N0 ( 50 ^ 310 ppm )- N„( 1 atm.), oxygen atom produced in the reaction (1) reacts with toluene competing mainly with the reaction with N0~. Products observed in the above mentioned 0~ free system are o-cresol ( other isomers are less than a few percent of o-cresol ), a-nitrotoluene and m-nitrotoluene. The relative yield of a-nitrotoluene to cresol increased with the increae of the initial concentration of N0 9 , whereas that of m-nitrotoluene decreased. The production of a-nitrotoluene strongly suggests the presence of benzyl radical and would imply that the hydrogen abstraction from the methyl group of toluene by oxygen atom does occur, although Cvetanovic did not detect a primary hydrogen abstraction. %> /.@ +0H (2) Q) + OH •— ► |£}» (4) s §*-^r « &"+ N0 * - £&- &+"" (5) Formation of m-nitrotoluene may be explained by the reaction of NO with the OH added radical. The competition between the reaction (4) and reactions OH + NO + M > HN0_ + M (6) OH + N Q 2 + M » ™°3 + M (7) would explain the decrease of the relative yield of m-nitrotoluene at KSt-3 higher NO- concentrations. In the presence of initial , for example toluene ( 34 ppm )-N0 ( 11 ^ 207 ppm )-air system, benzaldehyde, benzylnitrate and nitrocresols are formed in addition to cresols and m-nitrotoluene. The production of a-nitrotoluene is completely suppressed. The formation mechanism of benzylnitrate and benzaldehyde is most likely, ^-° CHzOHOz Q) +NQ* >@ (8) CHxO CHO + Z > 0) + H0 Z (9) where benzyloxy radical would yield by reactions o 4- o z - — >£% do) 5**°*' CH t 0- Q + NO >@ + NO* (11) Under our experimental condition given above, the concentration of NO reaches its photo-stationary level in about 15 min. after the irradia- tion is started and stays constant during the irradiation of several hours. This fact togather with the absence of the induction period for the formation rate of any of the products suggests that no chain reaction is occurring in our reaction system. This condition allows us to deduce the reaction mechanism and some of the rate constant ratio. Reference 1) A. P. Altshuller, P. W. Leach, Int. J. Air Water Poll., 8, 37 (1964). 2) J. M. Heuss, W. A. Glasson, Environ. Sci. Tech., 2, 1109 (1968). 3) G. R. H. Jones, R. J. Cvetanovic, Can. J. Chem. , 39, 2444 (1961). 4) E. Grovenstein Jr., A. J. Mosher, J. Am. Chem. Soc, £2, 3810 (1970). 5) S. L. Kopczynsky, Int. J. Air Water Poll., 8, 107 (1964). KS-4 Effects of Solar Energy Distribution on Photochemical Smog Formation by Gary Z. Whitten , James P. Killus, Henry Hogo, Systems Applications, Incorporated, San Rafael, California; and Marcia C. Dodge, Environmental Protection Agency, Research Triangle Park, North Carolina Introduction The most documented photochemical reaction leading to ozone forma- tion is NO photolysis by ultraviolet light. Thus the intensity of ultraviolet light as monitored in urban areas is often related to or even quantified by the rate of NO photolysis (Holmes et al., 1973). Similarly, the light source in smog chambers is often calibrated by measuring NO photolysis (Wu and Niki, 1975) . The time to the NO peak, i.e., the time to highest NO concentration produced by an irradiated mixture of hydrocarbons and nitrogen oxides, is widely used to indicate the reactivity of the mixture (Altshuller and Cohen, 1963) . Even though NO photolysis is central to photochemical oxidant for- mation , more information regarding the UV spectrum of the light source may be necessary to properly estimate the oxidant severity. Indeed, Jaffe et al. (1974) observed significant differences in oxidant forma- tion from light sources that caused similar NO photolysis rates but had different UV spectra. Aldehyde photolysis has recently been shown to be important in smog chemistry (Hecht et al. , 1973) , and the ultraviolet wavelengths that photolyze aldehydes to form radicals are different from those that photolyze NO . This paper discusses the implications of this wavelength difference on the chemistry of smog formation. The quantum yield for NO photolysis is unity at wavelengths below 400 nm, and rapidly decreases to zero at longer wavelengths. NO photolysis in smog formation occurs mainly by absorption of UV light K9-1 Effects of Solar Energy Distribution on Photochemical Smog Formation Introduction The most documented photochemical reaction leading to ozone forma- tion is NO photolysis by ultraviolet light. Thus the intensity of ultraviolet light as monitored in urban areas is often related to or even quantified by the rate of NO photolysis (Holmes et al. , 1973) . Similarly, the light source in smog chambers is often calibrated by measuring. NO photolysis (Wu and Niki, 1975). The time to the N0„ peak, i.e., the time to highest NO concentration produced by an irradiated mixture of hydrocarbons and nitrogen oxides, is widely used to indicate the reactivity of the mixture (Altshuller and Cohen, 1963). Even though NO photolysis is central to photochemical oxidant for mation, more information regarding the UV spectrum of the light source may be necessary to properly estimate the oxidant severity. Indeed, Jaffe et al. (1974) observed significant differences in oxidant forma- tion from light sources that caused similar NO photolysis rates but had different UV spectra. Aldehyde photolysis has recently been shown to be important in smog chemistry (Hecht et al. , 1974), and the ultraviolet wavelengths that photolyze aldehydes to form radicals are different from those that photolyze NO . This paper discusses the implications of this wavelength difference on the chemistry of smog formation. The quantum yield for NO photolysis is unity at wavelengths below 400 nm, and rapidly decreases to zero at longer wavelengths. NO photolysis in smog formation occurs mainly by absorption of UV light K9-2 between 300 and 400 nm (Jones and Bayes, 1973). Formaldehyde is a major fraction of the total aldehyde found in photochemical smog, but its photolysis reactions are not as well understood as NO photolysis. Of its two photolysis reactions, H CO + hv -»■ H- + HCO* (1) H CO + hv -> H + CO (2) the former is much more important in smog chemistry because it produces two radicals. The quantum yields for reactions (1) and (2) are known to vary with wavelength. The quantum yield is nearly unity at short UV wavelengths for reaction (1) and at long UV wavelengths for reaction (2). The quantum yields equal 0.5 for both reactions at about 312 nm. Thus the aldehyde photolysis reaction important in smog photochemistry occurs by absorption of UV light near the short-wavelength edge of the solar spectrum. Results Personnel at the University of California at Riverside Statewide Air Pollution Research Center (SAPRC) performed two smog chamber experi- ments with very similar initial concentrations and the same light source 4 1/2 months apart (Table 1) . During the time between these two experiments the light source was found to have deteriorated by about 8.5 percent as measured by the NO photolysis rate. However, spectral distribution measurements were not obtained for these two experiments. The time to the NO peak differed by a factor of two between these two experiments. In current mechanistic schemes, an 8.5 percent de- crease in NO photolysis cannot explain such a drastic difference in K9-3 Table 1 INITIAL CONDITIONS ; OF SMOG CHAMBER EXPERIMENTS Experiment 1.036 1 Experiment 2* Propylene (ppm) 1.040 NO (ppm) 1.12 1.11 N0 2 (ppm) 0.16 0.15 HCHO (ppb) 16.0 0.0 CH CHO (ppb) 7.0 2.0 * Experiment 2 was carried out in the same smog chamber with the same light source 4 1/2 months after Experiment 1. time to the NO peak. We were interested to see whether a decrease in the intensity of short UV emitted by the light source could account for this decrease in reactivity. Using a chemical kinetic mechanism for propylene/NO systems de- veloped with SAPRC data by Systems Applications, Incorporated, we per- formed computer simulations in which aldehyde photolysis rate constants were varied to obtain best fits to the experimental data for Experiments 1 and 2. The values of these rate constants that produced the best fits are shown in Table 2 . Thus, large differences in aldehyde photolysis rates can explain the major difference in reactivity between these two experiments, assum- ing that our chemical kinetic mechanism for propylene is basically correct. K9-4 Table 2 RATE CONSTANTS FOR BEST FITS OF COMPUTER SIMULATIONS TO EXPERIMENTAL RESULTS Experiment 1 Experiment 2 NO photolysis rate -1 constant (min ) 0.223 0.204 HCHO photolysis rate -1 -4 -5 constant* (min ) 5.2 x 10 5.2 x 10 CH CHO photolysis rate -1 -4 -5 rate constantt (min ) 5 x 10 5 x 10 Time to NO peak Observed 105 210 Calculated 107 200 * By Reaction (1), HCHO + hv -> H« + HCO*. t By the reaction CH CHO + hv -> CH* + HCO* . Some corroboration for the changes in aldehyde rate constants shown in Table 2 is provided by recent measurements of spectra during the replacement of the light source in the SAPRC chamber. It was found that the used light source emitted less UV light at wavelengths than the new source, and that the difference in intensity was largest at the short wavelengths. Conclusions Two smog chamber experiments at SAPRC with similar initial condi- tions showed a difference of a factor of two in reactivity, as measured by the time to the NO peak. To determine whether this difference could be explained by a deterioration of the chamber light source, we simu- lated these experiments with a computer model. Our results indicate K9-5 that a decrease in the intensity of short UV emitted by the light source could have caused the observed difference in reactivity. Calculations and measurements of UV in the real atmosphere show changes in intensity, as measured by NO photolysis, and in spectral distribution. Recent measurements of erythemal radiation have been made and the erythemal radiation spectrum resembles the absorption spec- tra of aldehydes (Machta et al. , 1975) . Hence the technology apparently exists to easily determine the ratio of light flux in the aldehyde- absorbing region to the flux for NO absorption in the atmosphere. The combination of this measurement with the flux for NO absorption should provide more reliable data for relating the concentration of photochemi- cal oxidant to the light intensity. K9-6 References Altshuller, A. P., and I. R. Cohen (1963), Int. J. Air Water Pollut. , Vol. 7, p. 787. Hecht, T. A., M. K. Liu, and D. C. Whitney (1974), "Mathematical Simulations of Smog Chamber Photochemical Experiments," EPA-650/4- 74-040, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Holmes, J. R. , R. J. O'Brien, J. H. Crabtree, T. A. Hecht, and J. H. Seinfeld (1973) , "Measurement of Ultraviolet Radiation Intensity in Photochemical Smog Studies," Environ. Sci. Tech. , Vol. 7, pp. 519-523. Jaffe, R. J., F. C. Smith, Jr., and K. W. Last (1974), "Study of Factors Affecting Reactions in Environmental Chambers," EPA-650/3- 74-004a, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina. Jones, I. T. N. , and K. D. Bayes (1973), "Photolysis of Nitrogen Dioxide," J. Chem. Phys. , Vol. 59, pp. 4836-4844. Machta, L. , G. Cotton, W. Hass, and W. Komhyr (1975), "Erythemal Ultraviolet Solar Radiation and Environmental Factors , " CIAP Report (February 28, 1975). Wu, C. H., and M. Niki (1975), "Methods for Measuring NO Photo- dissociation Rate," Environ. Sci. Tech., Vol. 9, pp. 46-52. K9-7 "Photophysics of Bound and Dissociative States of Small Molecules in Condensed Phases" L . E ♦ Brus and V. E. Bondybey, Bell Laboratories, Murray Hill, New Jersey 0797^ The cage effect on dissociating small molecules in rare gas solids has been investigated via the techniques of time, wave- length, and polarization resolved fluorescence. Strongly predissociated molecules, such as IC1, and directly dissociating molecules ,_ such as CI and CH I, have been studied. In the solid the host prevents permanent dissociation, and the fragments recombine within a few vibrational periods into the various bound electronic states. In the solid there is no quantized vibrational structure at excitation energies corresponding to gas phase dissociation. This strong cage effect occurs in the impulsive limit with respect to the bulk compressibility relaxation time of the rare gas host. However, bound-bound spectra, involving zero phonon lines of levels very close to the dissociation limit, imply only a weak van der Waals cage influence on the molecular potential. This result implies only minor cage effect for processes occurring in the adiabatic limit with respect to host compression. Polarized excitation ("photo- selection" ) studies allow one to unravel the various radiationless transition pathways in the common case where the absorption spectrum contains several overlapping continua. The predissociated B state of IC1 has an effective double minimum shape in the solid. Excitation below the dissociation barrier produces B (v=0) fluorescence with near unity quantum yield, while excitation above the barrier produces principally A II(v=0) emission via radiationless Ll-1 relaxation through the outer minimum of B. In CH I, it is observed that the "repulsive" excited state, which is employed in gas phase 2 2 photodissociation I* P ,_ -*■ P , p lasers, is actually slightly (< 2000 cm ) chemically bound. In CI , the observed emitting bound state is tentatively assigned as the "forbidden" lowest multiplet 3 component IT . Vibrational relaxation rates of high frequency (> 1000 cm ) modes in small molecules are slow because a large number of 50 cm"" " lattice phonons must be simultaneously produced. An energy gap law has been predicted, such that the rate should decrease exponentially with the quantum size. This prediction is observed to be violated in the vibrational relaxation of OH (A E) and NH (A IT), where the hydrides actually relax faster than the deuterides. The electronic spectra show these molecules to be undergoing slightly perturbed free rotation at U.2°K; and the relaxation data are thought to reflect the involvement of this localized motion. When energy relaxation is slow, long range energy transfer processes have time to develop, and may dominate the excited state population evolution. Three near resonant vibrational energy transfer processes from WD (A IT) to CO and CO have been studied. The transfer quantum yield, and the non- exponential time resolved donor population, both follow Forster (dipole-dipole) kinetics as a function of the acceptor concentration. A strong energy gap (transfer exothermicity) law is found of the form k a exp [-AE/(28 cm" )] where AE = ^ (co„ - w ) . d a 3 Transfer from NH (A IT) to CO is anomalously fast in view of the size Ll-2 of AE, again indicating the hydride dissipates energy into the matrix especially efficiently. A number of spin and symmetry forbidden internal radiation- less transitions have been observed in the first row diatomics C^ — 2 + and C_. For example, vibrational relaxation in C (B I ) occurs u_+ by sequential intersystem crossing, to and from, the nearby L state. Similarly in C excitation into A IT produces with high quantum 3 3 yield the lowest triplet a II . Population in v=0-3 of a II subsequently undergoes the spin and symmetry forbidden intersystem 1 + -5 crossing into X Z on a 10 sec time scale. An extremely strong g energy gap law prevails over the low intramolecular matrix elements in these processes. The photodissociation cage effect allows one to study transient species that have very weakly bound ground electronic states. For example, several bound- free and bound-bound fluorescences have recently been observed in the diatomics XeO and XeF. The polarized excitation spectra in solid Ar allow one to unravel the symmetries and energies of the various excited states. In XeO, an avoided curve crossing involving the !> 0.5 eV bound excited state correlating with ( D) + Xe has been detected. This crossing explains the observed highly efficient bimolecular quenching of ( D) by Xe in the gas phase. Ll-3 Spectroscopy and Photochemistry of Matrix Isolated Metal Hexaf luorides William F. Coleman and Robert T. Paine Department of Chemistry University of New Mexico Albuquerque, NM 87131 and R. Burton Lewis, Robin S. McDowell, Lewellyn H. Jones and Larned B. Asprey Los Alamos Scientific Laboratory University of California Los Alamos, NM 87544 Although the uranium hexafluoride (UF fi ) molecule has been of great interest to chemists and physicists for more than 35 years, it is only recently that reliable spectro- scopic data have been available. The photochemistry of UF,, an area of great potential applicability in an isotope sep- aration program, has not been reported prior to this work. We survey here recent work from our laboratories on the vibrational, electronic absorption, emission and excitation spectra and the ultraviolet photochemistry of low tempera- ture thin film and matrix isolated UF,. 6 1) Vibrational Spectra : The six fundamental frequencies of UF, have been recently determined in dilute Ar 6 matrices at 14°K. In addition to the IR and Raman active modes sufficient combination bands were observed to allow assignment of all fundamentals. Shifts of up -1 (2) to 6 cm from the most reliable gas phase data were obtained in the solid and some evidence exists for the formation of low concentrations of UF fi aggregates in the L2-1 matrices. Strong perturbations are observed in differ- ent matrix hosts, the general trend being that stretching frequencies are more susceptible to the max- trix than bending frequencies. Table I gives the values of v -> v^ in the Ar matrix. 1 6 Table I Fundamental Frequencies of Matrix-Isolated UF, (cm ) v n (A, stretch) 666 1 lg v„ (E stretch) 530 2 g v_ (F n stretch) 619.3 3 lu v. (F., bend) 183.5 4 lu v r (F„ bend) 200 5 2g v r (F„ bend) 143.2 6 2u 2) Electronic Spectra : The general features of the UF fi (3) ultraviolet spectrum have been known for sometime. The spectrum consists of two major regions of absorp- tion a) a weak absorption envelope from 340-410 nm (the A band) and b) a strong manifold from 330 nm to higher energies (the B band) . The weak band has previously been reported to exhibit some structure. A recent study has revealed a significant amount of structure in (4) both the A and B bands for low temperatures matrices. The A band has been assigned as arising from a forbid- den charge transfer transition to two excited electronic states , one predominately singlet in character and the T 9-9 second (the lower energy state) predominately triplet in character. The vibronic transitions have been assigned with v (singlet, electronic) = 25,270 cm , v (triplet, electronic) = 24,510 cm , v = 580 cm , v~ = 480 cm and v. = 185 cm . The B band region has been assigned as arising from both allowed charge- transfer transitions and a weak forbidden transition. Comparison of the observed spectra with several recent calculations on the electronic structure lead to the conclusion that the spin-orbit coupling in UF, is intermediate and the best fit is obtained by applying (5) spin-orbit coupling to the Xa calculations. 3) Emission Spectra ; The luminescence of solid UF, was (7) first discussed in detail by Sheremetev in 1957 and has been observed several times since then. Our matrix luminescence spectra represent the first instance in which significant vibronic structure is present. At 14°K in 200/1 Ar/UF. matrix more than 200 lines are b observed in the emission spectrum. Progressions in v fi , v.. = nv_ (n = 1-16) and several other combinations are 6 5 observed. Phonons , apparently arising from the matrix gas, are observed with differing frequencies in the different matrices. The quantum efficiency is quite temperature and concentration dependent and the position of the 0-0 band is strongly concentration dependent. The 0-0 band is much more intense in a UF fi thin film L2-3 supporting a structure change from matrix to thin film. The excitation spectrum verifies the presence of two electronic states in the A band region, the second hav- ing an internal conversion efficiency to the emitting state of less than unity. The internal conversion from the B band to the emitting state is a very inefficient process and is compatible with the observed photo- chemistry (vide infra) . 4) Photochemistry : Photolysis of matrix isolated UF fi in dilute Ar matrices leads to significant and reversible changes in the vibrational and emission spectra. New bands attributed to isolated UF appear in the IR and a weak near IR luminescence is detected. These spectral features disappear rapidly following exclusion of the UV radiation and the spectra of UF, are regenera- ted. In more concentrated matrices (20/1 Ar/UF^) the 6 emission intensity does not return to the initial value indicative of irreversible formation of polymeric UF,.. In CO matrices, which should be efficient fluorine-atom scavengers, the process is irreversible and IR bands appear which can be assigned to FCO, (FCO) 2 and F ? C0. The photolysis is much more efficient for irradiation in the B band than for the A band and a wavelength depen- dence of the photolysis quantum yield is observed within the A band. L2-4 The discussion will focus primarily on recent results in the emission and photochemical studes. References 1. R. T. Paine, R. S. McDowell, L. B. Asprey and L. H. Jones, J. Chem. Phys. (Communication) , 64, 0000 (1976). 2. R. S. McDowell, L. B. Asprey and R. T. Paine, J. Chem . Phys . , 61, 3571 (1974). 3. D. Lipkin and S. I. Weissman, "Photochemistry of Systems Containing Uranium Compounds," University of California, Radiation Laboratory Report A-520, 1942. 4. W. B. Lewis, L. B. Asprey, L. H. Jones, R. S. McDowell, S. W. Rabideau, A. H. Zeltmann, and R. T. Paine, J. Chem. Phys . , 64, 0000 (1976). 5. M. Boring, J. Chem. Phys . , (to be published). 6. D. E. Ellis and D. D. Koelling, unpublished calcula- tions. 7. G. D. Sheremetev, Optica i Spektroskopiya , 2, 99 (1957). L2-5 Mercury Photosensitized Production Of Trapped Radicals In Organic Glasses At '^77 K 1 by N. Bremer, B. J. Brown, G. H. Morine, and J. E. Willard Department of Chemistry, University of Wisconsin Madison, Wisconsin 53706 Solutions of mercury in liquid hydrocarbons (V> x 10~ 6 M at 25°C) 2 may be stabilized in the glassy state by quench cooling in liquid ni- trogen at 77 K (Fig. I) 3 . When a glassy solution in 3-methylpentane (3MP) is exposed to 254 nm radiation absorbed by the Hg, the intensity of the Hg absorption band diminishes while a broad absorption from 200 nm to 300 nm (Fig. 2) attributable, in part, to 3-methylpentyl radicals appears in parallel with growth in the known ESR signal of the radicals. The rate of growth of the radical concentration approaches zero with continued monochromatic 254 nm illumination (Fig. 3) , but increases upon exposure to broad band 200 nm-300 nm light. Exposure of a fresh 3MP-Hg glass sample to the broad band light at 77 K causes continuous growth in the radical concentration with no plateau, many radicals be- ing produced for each Hg atom present (Fig. 4). These observations in- dicate that a species formed in a primary or secondary act following photoactivation of the Hg by 254 nm light (e.g. HgH, HgR or HHgR) ab- sorbs in the broad band region and is decomposed, liberating Hg which again absorbs 254 nm photons and produces more radicals. No ESR signal from the intermediate has been found, indicating that it is not para- magnetic or that the spectrum is too broad to detect. The initial rate of radical production by illumination of a fresh 3MP-Hg sample is directly proportional to the light intensity. At L3-1 plateaus of the type of Fig. 3, the ratio of radicals to Hg atoms which have disappeared may be greater than unity and it increases with the light intensity, but to a power less than first order. This seems to imply that the initially formed intermediate absorbs 254 nm to regen- erate Hg (e.g. HgH rrf Hg + H) in competition with a reaction which forms a species with relatively low extinction coefficient for 254 nm (e.g. HgH + R» -> HHgR) , which is decomposed at other wavelengths of the broad band light (e.g. HHgR -^> Hg + RH) . Illuminations with the broad band light at 5 K do not produce con- tinuing radical growth as at 77 K, but rather lead to a steady state concentration at ^0.2 radicals per Hg atom (Fig. 4). This may be at- tributed to concerted formation within the low temperature parent cage of a compound with the properties attributed above to HHgR, coupled with an ability of the low temperature matrix to deactivate the HHgR without decomposition when it is activated by a photon from the broad band source. At 23 K the plateau is higher than at 5 K, while at 58 K continued growth occurs, though at a lower rate than at 77 K. Thermal decay of the radicals produced by the photosensitization is only a few percent per hour at 77 K and is undetectable over hours at the lower temperatures studied. These studies indicate: 1) energy transfer from Hg(6 3 P!) atoms to molecules in a hard organic glass inducing bond rupture as in the gas 5 and liquid phases 5 ; 2) a new method of producing trapped radicals for study in glassy matrices; 3) a means of trapping intermediates from Hg photosensitization reactions which may in further investigations assist in identifying them and their reactions; 4) a potential source L3-2 of information on the kinetics of radical-radical reactions in radical clusters in solids; 5) the possibility of false interpretation of pho- tochemical results in condensed systems containing unsuspected Hg as a contaminant. All experiments were made with purified degassed reagents; optical measurements used a 5 cm light path; radical concentrations were deter- mined from the double integral of the first derivative ESR spectrum. References 1. This work has been supported in part by the U. S. Energy Research and Development Administration under Contract No. AT (11-1) -1715 and by the W. F. Vilas Trust of the University of Wisconsin. 2. (a) J. N. Spencer and A. F. Voigt, J. Phys. Chem. , 72, 464 (1968); (b) R. R. Kuntz and G. J. Mains, ibid., 68^, 408 (1964). 3. N. Bremer, B. J. Brown, G. H. Morine, and J. E. Willard, J. Phys. Chem., 79, 2187 (1975). 4. E. D. Sprague and J. E. Willard, J. Chem. Phys., 62, 2603 (1975). 5. For reviews of Hg-photosensitization reactions and references, refer to (a) J. G. Calvert and J. N. Pitts, " Photochemistry" , Wiley, New York, N.Y., 1966; (b) R. J. Cvetanovic, Prog. React. Kinet., 2, 39 (1964); (c) A. B. Callear and P. M. Wood, J. Chem. Soc, Faraday Trans. 2, 68, 302 (1972); (d) A. C. Vikis and D. J. LeRoy, Can. J. Chem., 51, 1207 (1973). L3-3 0.8 .TT 240 250 260 270 WAVELENGTH, nm Fig. 1. Spectra of Hg in 3MP glass at 77 K (5-cm pathlength in two 2.5-cm 2 quartz cells): (A) spectrum of cell walls, 3MP is transparent in this region; (B) 3MP saturated with Hg at 25°C and quenched to 77 K; (C) 3MP saturated with Hg at ^55°C and quenched to 77 K. BEFORE PHOTOLYSIS I MIN PHOTOLYSIS J L 220 240 260 280 300 WAVELENGTH, nm Fig. 2. Spectra of 3MP-Hg glass at 77 K at different times of illumination with 200 nm- 300 nm medium pressure Hg lamp. 40 60 00 PHOTOLYSIS TIME, MIN 120 Fig. 3. Concentration of trapped radicals, measured by ESR, as a function of time of illumination of 3MP-Hg glass at 77 K with 254 nm light from Vycor low pressure Hg lamp. Curve A, 4.1 x 10 5 ergs cm -2 sec~ l ; Curve B, 3.7 x 25 50 75 PHOTOLYSIS TIME, MIN Fig. 4. Concentration of trapped radicals, as a function of time of illumination of 3MP-Hg glass at 77 K, 58 K, 23 K and 5 K with 200 nm-300 nm light from medium pres- sure Hg lamp. L3-A "Magnetic Field Effects on Triplet -Triplet Annihilation in Crystals"* S.H. Tedder and S.E. Webber , Department of Chemistry, University of Texas at Austin, Austin, Texas 78712. It has been known since 1967 that external magnetic fields influ- enced the intensity of delayed fluorescence in molecular crystals. The 2 theory of this effect was first presented by Merrifield, and has been refined by Merrifield and others. Excellent reviews are available. The essential idea of the original explanation of Merrifield may be re- presented by the "reaction" ft *tx7 ^A * £•« «., C ^ * *S (1 ) M A +M B where T^ 1 is a triplet exciton in sublevel M, (T^. . .Tg) is an "intermediate" with total M quantum number of M + M R (spin relaxa- tion is ignored), and S A and Gg are excited and ground state singlets, respectively. The Sa state is responsible for the delayed fluorescence. At high fields triplet exciton pair states ( [M, , Mg^ ) will not be degen- erate in general because of zero-field coupling terms in the total Ham- iltonian. For certain orientations of the magnetic field with respect to the crystallographic axis the levels will become degenerate and the states involved in reaction (1) are mixed with a concomitant change in the rate of singlet formation (usually a decrease). Hence the level crossing resonance is observed as a change in the intensity of delayed fluorescence. An example of this type of data for anthracene doped phenanthrene crystals for different orientations of the crystal in the LA-1 plane of the magnetic field is given in Fig. 1(a). For this case the anni- lation is heterogeneous, i.e. , between a phenanthrene triplet exciton and an anthracene trapped triplet molecule. To a good approximation the minimum of the delayed fluorescence corresponds to the angle at which the high field states I + 1^> ^p/ » I 0» »0p Va = anthracene, P = phe- nanthrene) become degenerate. Similar results have been observed for the delayed fluorescence of pure naphthalene or naphthalene crystals lightly doped with anthracene. Heavier doping with anthracene (but still less than 10" " M/M) totally changes the pattern of resonances for naphthalene crystals in a way that is not presently understood. As can be seen from Fig. 1(a) the character of the level crossing resonance changes rather drastically with the orientation of the crystal. Simple application of previous theory does not account for this effect and we have been led to postulate that the triplet exciton -trapped triplet interact with each other during this annihilation via the intermolecular spin- spin interaction. H gs depends on the orientation of the crystallo- graphic axis with respect to the magnetic field, if the spin states are expressed in the high field representation. If the spin states are mixed by the H ss , and the quantum yield of singlet formation is calculated using a kinetic scheme like (1) and (2), we obtain results which have a quali- tative resemblance to our experimental data, see Fig. 1(b). It is our intention to extend these calculations to the naphthalene system. A second theoretical model that has been studied is the use of level L4-2 avoided crossing (LAC) theory, which has previously been applied to magnetic field effects on prompt fluorescence of small molecules in the gas phase. This model is also similar, but not identical to, an approach given by Johnson and Merrifield. In this model it is assumed that each quantum mechanical pair state evolves in time coherently, and inter- ference effects between states play a significant role in the observed line shape. This theoretical model is much more difficult to apply than the kinetic model, but preliminary calculations have demonstrated qualita- tive agreement with experimental results. Our current goals are: (1) to establish if one or the other (or either) of these models is to be pre- ferred, and (2) to establish if these experiments elucidate the nature of the interaction between the triplet state molecules during annihilation. *Supported by the Robert A. Welch Foundation. References 1. R.C Johnson, R. E. Merrifield, P. Avakian and R.B. Flippen, Phys. Rev. Lett. 19, 285 (1967). 2. R.E. Merrifield, J. Chem. Phys. 48, 4318 (1968). 3. R.C Johnson and R.E. Merrifield, Phys. Rev. Bl, 896 (1970). 4. C E. Swenberg and N. E. Geacintov, Organic Molecular Photophysics Volume 1 (ed. J.B. Birks, John Wiley and Sons, New York, 1973), Chapter 10; ibid, Volume 2, p. 395. 5. S.H. Tedder and S.E. Webber, Chem. Phys. \1, 253 (1976). 6. D. Levy, Adv. Magn. Resonance^, 1 (1973). L4-3 \fti\ f 10° 30° 50° 70° 90° 110° 130° 150° *>. 10° 30° 50° 70° 90° 110° 130° 150° Fig. 1(a). The intensity of anthra- cene delayed fluorescence as a func- tion of the angle 0^ °f ^ e applied 8 kG magnetic field. The values ofot are -90° (H // ac*, b up); 30° (H // be*, a down), and 90° (H // ac*, b down). The traces are not all to the same scale. Fig. 1(b) A model calculation of fusing eq. (2) and consider- ing interexciton spin-spin inter action. For these calculations k-i =0.05, k s = 1 . and kp = 0.5 (k + i does not affect the rel- ative changes of Kand only the relative values of the other con- stants are important). L4-4 THE CONTRIBUTION OF THE PHYSICAL AND CHEMICAL DEFECTS TO THE PHOTOCHEMISTRY OF CRYSTALLINE DURENE AT VERY LOW TEMPERATURES A. DESPRES, V. LEJEUNE and E. MIGIRDICYAN Laboratoire de Photophysique Moleculaire du C.N.R.S. Universite Paris-Sud 91405 ORSAY FRANCE It has been known for some time (J.M. Thomas and J.O. Williams, Progr. Solid State Chem.,6 (1971) 119 and references therein) that line defects (dislocations) and point defects (impurities) may play an important role in photochemical reactions developed in single crystals. The disposition of the molecules at crystalline defects is generally different from that of molecules in the regular lattice. These structural imperfections can affect the migration of excitons through the crystal and/or may facilitate the formation of a transition complex leading to a chemical reaction. The objective of the present work is to study the contribution of the physical and chemical defects to the photochemistry of crystal- line dureneat very low temperatures. The solid state photolysis of durene doped with orthomethyl substituted benzaldehydes like 2,4,5-trimethylbenzaldehyde, 2,4- and 2,5-dimethylbenzaldehydes (2,4-DMB and 2,5-DMB) gives rise, between 200 and 300 K, to the production of duryl (2,4,5-trimethyl benzyl ) radi- cals. These radicals are formed by the detachment of a hydrogen from a durene methyl group and identified by their characteristic green fluo- rescence. No duryl radicals are however produced, between 200 and 300 K under similar photolysis conditions, in durene doped with 3,4-dimethyl benzaldehyde (3,4-DMB) or with 3,4-dimethylacetophenone (3,4-DMA), both guests having no methyl substituent in ortho position with respect to the carbonyl group. The electronic spectra of all these guests are sharp and well resolved, indicating that substitutional solutions are formed in all the cases. L5-1 In contrast, when the U.V photolysis (with the full emission of a high pressure mercury lamp Osram HBO 200 filtered by a Schott UG 11 filter) is carried out at 10 - 20 K, duryl radicals are always produced in durene doped with the above mentioned ortho-methyl substituted benzaldehydes as well as with 3,4-DMB or 3,4-DMA. These results show that the mechanism of duryl radical formation is not the same in the high (200-300 K) and in the low (10-20 K) temperature range. Duryl radi- cals are however not produced at 10 - 20 K during the photolysis of highly purified durene single crystals where no impurity emission was detected at these very low temperatures. This suggests that when guest molecules are present in tne crystal, they are probably acting as light traps to induce photochemical activity. incoi^scm- 1 lio.o^ascrrr 1 IKO.CH-UOcm- 1 1(0,0)- 430 cm" 1 II (0,0) f 898 A 1(0,0)4884, 5A The three first fluorescence bands of the duryl radicals di- rectly produced at 10 - 20 K in a sample of durene doped with 2,5-DMB are displayed in the figure on curve A. The 0,0 band and the two first vibronic bands present the similar complex fine structure. When this sample is subsequently annealed in the dark by raising its temperature up to 65 K and then cooled again at 10 - 30 K, the fine structure of L5-2 bands has changed. The new spectrum shown on curve B is identical to the fluorescence, measured at 10 - 30 K, of duryl radicals produced at room temperature either photochemical ly in durene doped with methyl substi- tuted benzaldehydes or radiochemical ly in purified durene (S. Leach and E.Migirdicyan, Chem. Phys. Letters l 9 21 (1967)). It is particularly noteworthy that the higher energy intense subspectrum I starting at 4884,5 A on curve A has nearly disappeared on curve B where the subspec- trum II starting at 4898 A is predominent. Similar spectroscopic obser- vations were made on samples of durene doped with the other methyl substituted benzaldehydes. The different subspectra present in the duryl radicals fluo- rescence are attributed to species produced photochemical ly in tran- sient defects created because of the stress experienced by durene single crystals during the cooling at 10 - 20 K. The first electronic transi- tion of radicals present in transient defects has a higher energy as compared to that of species present in annealed samples. This suggests that the duryl radicals first formed in transient defects are reorien- ted as the temperature increases, so as to achieve a better equilibrium with their environment. In order to elucidate the mechanism of duryl radical produc- tion in transient defects at 10 - 20 K, the fluorescence intensity Ip changes were monitored on the subspectrum 1(0,0) band at 4884,5 A during photolysis with constant light intensity. The plot of Ip versus time of irradiation presents an initial linear region starting at the coordinates origin. However, the Ip changes level off at high radiation doses because of secondary reactions and a possible inner filter effect of the radical produced. Experiments performed in order to get the light intensity dependence of duryl radicals production at 10 - 20 K are in progress. The question was raised as to whether duryl radicals are formed in de- fects through an excitonic mechanism involving direct light absorption L5-3 in the bulk crystal or through the photodissociation of the host molecu- les sensitized by the aldehydic guest. The absence of radical production in carefully purified durene single crystal is against the excitonic mechanism. However, in mixed crystals, guest molecules can act not only as light absorbers responsible for photochemistry in the host but also as traps for excitons migrating through the crystal, by inducing struc- tural imperfections in their surrounding matrix. In order to choose between these two possibilities, the rate of duryl radical production was measured when U.V. photolysis was performed in the absence and through a pyrex glass filter which eliminates the radiations directly absorbed in the durene crystal but not the radiations absorbed in the weak n-TT transition of the guest and partly in its strong tttt transition. A decrease by a factor of 5 to 8 was observed in the rate of radical production in the presence of the pyrex glass filter. This indicates that the dissociation reaction in the host is at least partly photo- sensitized by the guest at these yery low temperatures. L5-4 Photochemistry Of Electrons Trapped In Organic Glasses 1 G, C. Dismukes, S. L. Hager, D. P. Lin, G, H. Morine, and J. E. Willard Department of Chemistry, University of Wisconsin Madison, Wisconsin 53706 It has been known for some time that mobile electrons generated in organic glasses by y radiolysis or photoionization of solutes at "^77 K become physically trapped with significant yields. 2 These e give broad absorption spectra (>3eV) with maxima in the visible or near infrared. The spectra have bandwidths at half-height ranging from ^0.5eV to >leV, long tails on the higher energy side of X , shorter tails on the low max energy side, and extinction coefficients at A. of lO 3 -!©**. Trapped max electrons can be removed from their traps by photons and migrate to en- counter solutes with which they can react in competition with retrapping. Products may include molecular anions produced by reaction with mole- cules with a positive electron affinity (e.g., biphenyl) , carbanions (by reaction with trapped radicals) free radicals (by dissociative elec- tron capture) , and neutral molecules (by reaction with cations) . This paper reviews some of the photochemistry of such trapped electrons ob- served in recent and continuing work in our laboratory. Until relatively recently it was not known whether the broad spectra of e~, such as illustrated in Fig. 1, are representative of electrons bound in uniform traps or are envelopes of the spectra of electrons in trapping sites with a continuum of energies. Data such as those of Figs. 2 and 3 give strong support to the conclusion that the spectra represent continua of trapping energies. 2 Fig. 2 3 indicates that exposure to light in the 700 nm-1000 nm region of the spectrum of e in 3-methyl- L6-1 pentane (A = 1600 nm) increases the absorption above 1900 nm and de- max creases it below, the process being reversed in the dark. This can be explained if the detrapped electrons are retrapped in weaker traps which they cause to deepen by coulombic interaction with bond dipoles of the surrounding molecules. Fig. 3* illustrates energy-selective de- trapping of e~ in methyltetrahydrofuran (MTHF) at 25 K by 1338 nm light, and relatively uniform bleaching by 1064 light. Fig. 4 is a formalized model to suggest the type of continuum of e spectra which might be re- sponsible for the effects observed. Evidence that the detrapping of e is monophotonic, rather than occurring by a biphotonic mechanism in- volving promotion to an excited state of the trap followed by absorp- tion of a second photon, is given by the first order dependence of pho- toconductivity (Fig. 5) and photobleaching on the intensity of mono- chromatic light used for the detrapping. 5 When the trapped electrons are produced by y irradiation of the organic glass, the matrix also contains trapped radicals which compete with the trapped cations for capture of detrapped electrons. The photoionization thresholds of the carbanions formed lie in the region of 600 nm (2 eV) and the ir absorp- tion extends into the ultraviolet. When a y-irradiated hydrocarbon glass is exposed alternately to infrared and ultraviolet light, elec- trons may be repeatedly converted from the trapped state to the car- banion state and back with 5% still remaining after 7 cycles. 7 The heat released by electrons on reaction with cations plus radicals in y- irradiated glasses following photodetrapping has been measured by dif- ferential thermal analysis at 77 K. e It decreases with dose, consis- tent with an increasing ratio of radicals to cations (Fig. 6) . The L6-2 extrapolated value at zero dose sets a limit on the heat of solvation of the cation in 3MP. Current work in our laboratory is using bleach- ing with monochromatic light to further resolve the contributions of varying trap depths to the broad e spectra. References 1. This work has been supported in part by the U. S. Energy Research and Development Administration under Contract No. AT (11-1) -1715 and by the W. F. Vilas Trust of the University of Wisconsin. 2. For a review and references see J. E. Willard, J. Phys. Chem. , 79 , 2966 (1975). 3. S. L. Hager and J. E. Willard, Chem. Phys. Lett., 24, 102 (1974). 4. S. L. Hager and J. E. Willard, J. Chem. Phys., 61, 3244 (1974). 5. G. C. Dismukes, S. L. Hager, G. H. Morine, and J. E. Willard, J. Chem. Phys., 61, 426 (1974). 6. G. C. Dismukes, Ph.D. thesis, University of Wisconsin-Madison, 1975. 7. D. P. Lin and J. E. Willard, J. Phys. Chem., 78, 1135 (1974). 8. S. L. Hager and J. E. Willard, J. Chem. Phys., 63, 942 (1975). Fig. 1. Spectra of trapped electrons in 3MP-dn at 23 K: (A) electrons produced by photoionization of TMPD at 72 K and cooled to 23 K; (B) after partial photobleaching of A at 23 K with 7CO-1000 ran light. 04 A, - 03 - OD 02 v-**~\-^ - 01 i v^ 12 DO 1600 2000 2400 WAVELENGTH, nm L6-3 SOB 5 Ho. + ,.. If — 1 1 + c 1900 nm i i 2500 nm -T i i _ A 2150 nm b 0.9 1700 nm - 1 ^- 0.6 Jr i i d 1 2150 nm 03 L L 02 4>-7-J 01 B_ f 1 2500 nm 4> ! ? 1.3 h ' — 12 _/ 1700 nm ' i i 1 2 2 TIME, Fig. 2. Optical density changes resulting from illumination of trapped electron populations and of TMPD in 3MP glass. Onset of illumination indicated by (+) and termination by t- ). Ana- lyzing wavelength shown on each plot. (a-e). 700-1000 nm illumination at 72°K of electrons generated in 3MP-o'h by UV photoionization of TMPD at 72°K. (0, same as (a-e) except illumination at 23° K after TMPD ionization at 72° K. (g). 700- 1000 nm illumination of electrons generated by ^-irradiation, T- 72°K. (h), optical density at 1 700 nm during and after gen- eration of trapped electrons by UV photoionization of TMPD in 3MP-rf 14 ; (-) mark signifies end of 2 min UV illumination. epP of>° v rfP v>o° ^o° ose 2.6 x 10 19 eV g _1 ; maximum light intensity 1 . mW „™-2 HEAT OF REACTION OF TRAPPED ELECTRONS IN 3MP Fig. 6. Heat of reaction of photodetrapped electrons in y-irradiated 3MP as a function of Y dose. { ••'•'•.•; Figure 1 1 L9-5 "i 1 1 1 1 1 1 r CO J I I L i r 520 500 480 460 440 240 230 nm Ar:C 2 H 2 (60% D) - 200 + Discharged Ar. 14° K Figure 2 2700 2500 1300 1200 800 cm-1 700 (a) Ar:HCC1 3 = 200 ♦ Undischarged Ar (b) Ar:HCC1 3 = 200 ♦ Discharged Ar Figure 3 L9-6 2100 2000 1900 1100 1000 900 800 (a) Ar:DCC1 3 = 200 + Undischarged Ar (b) AnDCCl 3 = 200 + Discharged Ar Figure 4 700 cm~ 500 450 L9-7 Atmospheric Chemistry of Chlorofluorocarbons Mario J. Molina Department of Chemistry University of California Irvine, California 92717 The chlorofluorocarhon-ozone depletion theory has been the subject of close examination by several research groups. Many facets of the fluorocarbon problem are currently under investigation; we are concerned here with its chemical and photochemical aspects. The theory assumes a lack of reactivity for the chlorofluorocarbons in the lower atmosphere, and an appreciable reactivity in the upper stratosphere for their photo- decomposition products, which include chlorine atoms. This chemical reactivity — described in detail elsewhere 1 — consists mainly of the following reactions: OH 3 HC1 ^=^ CIO ^=i CI CH^ 0,N0 The various assumptions on tropospheric and stratospheric chemistry may be tested by two different procedures. First, additional chemical and photochemical reactions can be studied in the laboratory and evalu- ated for their possible significance. Second, measurements of atmos- pheric concentrations of the various species may be compared to the values predicted by theory. We have recently carried out some laboratory studies on the species C10N0 2 (chlorine nitrate) 2 . Our current results indicate that an appre- ciable fraction of the chlorine atoms in the mid-stratosphere may be Ml-1 present as CIONO2. The formation of this compound involves the following three-body process : CIO + N0 2 + M -* C10N0 2 + M The rate of this reaction has not yet been measured in the labora- tory, but it is likely to be of the same order of magnitude as the rate of reaction of CI atoms or OH radicals with N02- In the stratos- phere, chlorine nitrate may be destroyed by photolysis and by reaction with species such as atomic oxygen. In order to estimate lower limits to its atmospheric decomposition rates, we have measured the optical absorption cross sections in the UV of purified samples of gaseous CIONO2. These cross sections, when combined with values of solar intensities, provide estimates for the photodissociation rates at various altitudes: the photochemical lifetime is about 3 hrs at 30 km and about 1 hr. at U0 km (for overhead sun). If oxygen atom attack is nearly as fast as attack on the similar molecule C10C1, then, above 30 km the lifetime of CIONO2 will be at most a few minutes. The possible importance of CIONO2 for stratospheric chemistry can be assessed by calculating the distribution of chlorine among the species CIONO2, HC1, CIO, and CI, that is, by considering the formation and destruction of CIONO2 in addition to the reactions discussed earlier-*-. The amount of CIONO2 present at mid-stratospheric altitudes might be large enough for actual in situ detection to be feasible, per- haps by infrared spectoscopy. Furthermore, due to the nature of the photodissociation process, diurnal, as well as seasonal variability in CIONO2 concentrations, is to be expected. The seasonal effects Ml-2 observed recently by Lazrus and co-workers on stratospheric HC1 con- centrations may be related to the variability in CIONO2, since the formation of this later species results chiefly at the expense of HC1. On the other hand, even assuming a fast formation rate and no destruction processes for CIONO2 other than photolysis, the ozone depletion predictions are diminished only by about 20 to 30% as com- pared to earlier calculations, because ozone depletion occurs mainly at altitudes where CIONO2 is readily photolyzed, that is, above 35 km. As mentioned above, another aspect of the theory — namely, the stability of chlorofluorocarbons in the lower atmosphere — can also be tested by two separate procedures. First, we can evaluate the various proposed mechanisms or 'sinks' which lead to chlorofluorocarbon removal (e.g., reaction with ions, dissolution in the ocean, trapping in Antartic snow, etc.). To date, no such specific sinks have been found to be of importance when compared to stratospheric photolysis. Second, the amount of chlorofluoromethanes now found in the atmosphere can be compared with the cumulative amount already manufactured and released to the environment, taking into account that some stratospheric photolysis has already occurred. We have conducted a careful evaluation of this comparison for CCI3F ( f luorocarbon 11)3. w e conclude that a tropospheric residence time of less than 30 years is not compatible with experimental observations. The results are certainly consistent with the assumption that stratospheric photolysis is the only sink, but the existence of an additional tropospheric sink of similar magni- tude cannot be ruled out on the basis of these comparisons. One of the largest uncertainties in this evaluation results from Ml-3 difficulties in the absolute calibration of the analytic instrument used to measure CCloF levels in ambient air. We have carried out abso- lute calibrations in our laboratories by successive expansions and dilutions of a know initial amount of CCloF. Our results are in good agreement with those of several other research groups employing similar methods. References : 1. F.S. Rowland and M.J. Molina, Rev. Geophys. Space Phys. , 13, 1 (1975). 2. F.S. Rowland, John E. Spencer and Mario J. Molina, "The Stratos- pheric Chemistry of Chlorine Nitrate, CIONO2" , (to be published). 3. F.S. Rowland and Mario J. Molina, "An Experimental Estimate of the Importance of Tropospheric Sinks for CCloF (Fluorocarbon-ll)" , submitted to J. Chem. Phys. Ml-4 Ultraviolet Photoabsorption Cross-Sections of CF CI and CFC1« as a Function of Temperature Arnold M. Bass and Albert E. Ledford, Jr. Physical Chemistry Division Institute for Materials Research National Bureau of Standards Washington, DC 20234 Because of recent interest ' in the effect of commonly used halocarbons upon the stability of the earth's stratospheric ozone layer, we have undertaken a remeasurement of the photoabsorption cross- sections of f luorocarbons 11 and 12 (CFC1„ and CF Cl~) . Although some , . i . , , i . i ( lb , 2a , b ) measurements of this absorption have been reported previously , (3) the recent report of a temperature dependence of the absorption motivated us to measure this property of these substances in greater detail. Measurements were made over the wavelength region 185-230 nm, the wavelength region over which photodissociation occurs in the strato- sphere. In order to determine the effect of the reported temperature dependence upon the estimated dissociation rates of these substances in the stratosphere, absorption measurements were made at two temperatures, 296K and 223K. The measurements were made with a 0.75 m Fastie-Ebert monochroma- tor equipped with a 2400 groove mm grating. Absorption measurements were made at intervals of 0.2 nm with a spectral resolution of 0.05 nm. The absorption cell in which the gas samples were placed was made of stainless steel. It was 50 cm long, and by using the multiple pass design of White and Bernstein and Herzberg path lengths up to a maximum of 10 meters could be used. Temperature control of the gas M2-1 sample was obtained by circulation of a refrigerated fluid, usually methanol, through the outer jacket of the cell. The gas temperature was measured by means of calibrated chromel-constantan thermocouples inside the cell. At a cell temperature of 220 K the temperature varia- tion of the sample over the length of the cell was approximately 1°C. The absorption cell was placed in the exit beam of the monochroma- tor. Immediately in front of the cell a sapphire plate was used to split the light beam and a portion of the signal illuminated a photo- multiplier tube and served to monitor the variation of the light source. A broad band interference filter which passed radiation of wavelength between 175 and 250 nm was inserted between the hydrogen continuum source and the entrance slit. This filter effectively removed all radiation outside of this range, and eliminated all scattered radiation. Data acquisition was automated by photon counting in conjunction with a stepping motor control for the monochromator wavelength drive, as a -u a . -, (6) described previously Dichlorodif luoromethane (CF.C1 9 ) was purified by trapping with liquid nitrogen the exit stream from a gas chromatograph equipped with a squalane column and a thermal conductivity detector. The major im- purity, which was CF_C1H, could be removed entirely by this method. After purification the CF Cl~ was better than 99.99 mole % pure with respect to detectable organic impurities. The f luorotrichloromethane (CFC1 ) was better than 99.99 mole % pure with respect to detectable organic impurities, as received. Both components were used after low- temperature degassing. The results of these measurements are shown in Figure 1, and in M2-2 Table I (a complete table of measured cross-section is available upon request). The measurements at room temperature are essentially in agreement with those reported by Rowland and Molina , but are approximately 15-20% lower than the values reported by Huebner, et al . The only measurements reported at low temperature are those (3) by Rebbert and Ausloos at the wavelength of the zinc resonance line, 213.9 nm. The temperature effect on the absorption cross-section that we have observed between 296 K and 223 K is not as large as they re- ported, being approximately a factor of two reduction for CF Cl„, but only 30% in the case of CFC1 . References 1. a) M. J. Molina and F. S. Rowland, Nature _249, 810 (1974). b) F. S. Rowland and M. J. Molina, Rev. Geophys . Space Phys . 13 , 1 (1975). 2. a) J. Doucet, P. Sauvageau, and C. Sandorfy, J. Chem. Phys. 58 , 3708 (1973), b) R. H. Huebner, D. L. Bushnell, Jr., R. J. Celotta, S. R. Mielczarek, and C. E. Kuyatt, Nature 257 , 376 (1975). 3. R, E Rebbert and P. J. Ausloos, J. Photochem. ^, 419 (1975). 4. J. U. White, J. Opt. Soc. Am. 32., 285 (1942). 5. H. J. Bernstein and G. Herzberg, J. Chem. Phys. 16, 30 (1948). 6. A. M. Bass and A. H. Laufer, J. Photochem. 2, 465 (1974). Work supported in part by National Aeronautics Space Administration and Office of Air and Water Measurement, National Bureau of Standards M2-3 Table I, Absorption Cross-Sections for CF CI- and CFC1„ Waveleng th/ nm Cross -Section x i n+ 20 2 10 cm , -1 molec CF 2 C1 2 CFC1 3 296 K 223 K 296 K 223 K 230.0 — 0.34 0.15 ! 225.0 — 0.94 0.51 220.0 0.063 0.026 2.45 1.38 215.0 0.22 0.095 6.48 4.26 210.0 0.74 0.38 15.3 12.1 205.0 2.60 1.50 33.4 29.3 200.0 8.56 5.70 62.6 55.6 195.0 25.2 19.5 106.2 98.4 190.0 61.3 55.0 169.9 163.9 185.0 110.7 111.9 221.9 236.7 Cross-section (a) = n £ o 760 273 •£n £> IW II* lit WAVELENGTH! NH) m •M !«• Figure 1 M2-4 Photodissociation of CC1. 4 R. E. Rebbert and P. Ausloos Institute for Materials Research National Bureau of Standards Washington, DC 20234 1-3/ In several recent investigations of the photolysis of CC1, it was suggested that the elimination of a chlorine molecule from an ex- cited state of CC1, may occur under certain conditions. CC1. + hv -> CC1 + Cl (1) 4 2 2 i-2/to The CC1 formed in this process was thought — i — to insert readily into CC1. to form two CC1_ radicals 4 3 CC1 + CC1. -+ C C1, -> 2CC1 (2) 2 4 2 6 3 In order to assess the importance of process 1 relative to that of the Cl-atom producing fragmentation steps CC1. + hv -> CC1 + CI (3) 4 3 ■* CC1 + 2C1 (4) CC1, was photolyzed at 300°K in the presence of various selective re- agent gases such as Br_, HBr, HC1, C_H, and . Ethane served mainly 2. 2. D 2. as an efficient Cl-atom interceptor, while HC1 and HBr were found to scavenge CC1- species via the insertion reaction CC1 + HX -> HCC1 X (where X = CI or Br). No evidence could be found for the insertion of reaction 2. In the photolysis of CC1.-C-H, mixtures, CC1_ radicals 4 / b 2. * react by the recombination reaction CC1 + CH CH -»■ CH CH CC1 2 ■+ CH + CH_CC1«. A detailed analysis of the experimental findings leads to the following quantum yields M3-1 Table I Quantum Yields of primary Radicals Formed in the Photolysis of CC1. at 213.9 and 163.3 nm 4 213.9 nm 163.3 nm (CCl 3 ) 0.9+0.15 0.2+0.05 & 0. 4 \\ST*-— ^ \N p \ \^vgl ass beads 0. 3 xnoN 2 0. 2 \\ normal /^ > v ^2 0. 1 surface l j 'no N 2 ~i 10 20 30 Time, Min. Fig. 2: Effect of Surface M4-5 Stratospheric Reactions of Chlorofluoromethanes R.J. Donovan and H.M. Gillespie, Department of Chemistry, University of Edinburgh. and J. Wolf rum and K. Kaufmann, Max-Planck-Institut fur Stromungs f or s chung, 3U Gottingen. Introduction Upper atmosphere reactions involving chlorofluoromethanes ( freons ) 1-5 are of considerable current interest . It has been suggested that the diffusion of freons, from the troposphere into the stratosphere, where photochemical and reaction processes produce free chlorine atoms, could lead eventually to a depletion of ozone in the upper atmosphere 12 2 3 * , via the CIO cycle ' . The reaction of 0(2 D ) with freons 11-13 (CFC1 , CF 2 C1 2 and CF-.C1) is known to give rise to CIO radicals , 0("4>) + CF CI -> CIO + CF CI . (1) x y x y-1 and thus provides a means of direct entry into the CIO cycle. However, a number of other channels may contribute to the overall rate of removal of 0(D) , viz: 0( 1 D) + CF CI ■*■ 0( 3 P) + CF CI (2) x y x y CF CI n + CI (3) x y-1 CF CI + FC1 (or Clg) ... (U) Little is known about the importance of these other channels, or of the M5-1 subsequent radical reactions which can in some cases give rise to further release of CI atoms. The work to be presented was directed towards obtaining a more detailed understanding of the primary and secondary processes which take place in the photochemically initiated reaction between 0( D) and the chlorofluoromethanes. Experimental Two experimental techniques were employed: firstly, nozzle-beam mass spectrometric sampling from a fast flow system, coupled with flash photolysis (see figure l) and secondly, time resolved optical 5 1 spectroscopy . 0( D) atoms were produced by photolysis of 0_ in the Hartley Continuum (200-300 nm) and by photolysis of NpO (A^200 nm) . In the fast flow system, reagents were mixed just prior to entry into the flow tube (flow rates of typically 10 m s , with total pressures in the range 300-^00 N m and 0_ : halo carbon: He = 1:10:^0, were used). Products of the primary and secondary radical reactions were monitored as a function of time with the mass spectrometer (Extranuclear quadrupole) set to sample a preselected mass number. The analogue output from the mass spectrometer was then digitised and stored in a 200 channel signal averager (Datalab, DL.102A); in most cases it was necessary to average the results of sixteen experiments before adequate signal/noise ratios were observed. Little electronic interference from the flash was observed, however some variation in the interference with mass number was found and was thus checked for all experimental runs. The experimental conditions used for the time resolved optical studies were similar to those given in ref. 5. M5r*.2 Nozzle-Beam. Mass Spectrometric Sampling Results Formation of CIO was observed (m/e=5l) following the photolysis of 0~ in the presence of freons 11-13 and CCl^, supporting previous observations using ultraviolet absorption spectroscopy . No signal at m/e=51 was observed when 0^ was excluded from the flow mixtures under otherwise identical conditions. The yield of CIO was found to increase with increasing partial pressure of 0- (over the accessible range, —2 1 5-13 N m ) , as expected from the increase in 0( D) production. Other products observed immediately following photolysis of mixtures containing 0-, were FC1, CFC10 (both from CFC1-) and CF 2 from CFpCl 2 . Small yields of CClp and CFC1 were also observed from direct photolysis of CFCl^ and CF^Clp, respectively. As these transient species have the same charge/mass ratio as CFC10 and CF 0, a small correction to yields of the latter species was made. The formation of FC1 provides direct evidence for the primary process, 0( 1 D) + CFC1 + CClpO + FC1 (5) (atomic recombination processes would be too slow to account for our observations on FC1). Optical Spectroscopy Results Previous work on reaction (l) has been extended using N p as the source of 0( D) . It was shown that the reaction of chlorine atoms with NpO, to yield CIO, is negligibly slow under the conditions employed and thus any CI atoms formed in reaction (3) do not contribute to CIO formation. Furthermore, the formation of NO via M5-3 Oi 1 !)) + N 2 -* 2N0 (6) provides a convenient means of establishing the initial concentration of 0( D) produced "by photolysis. As the extinction coefficient of CIO in the ultraviolet is known the branching ratio into channel (l) can be determined as k..>0.5 (k +k 2 +k_+k. ) . Secondary reactions involving CF CI _. radicals and related studies involving bromomethanes will also be discussed. References 1. M.J. Molina and F.S. Rowland, Nature, 2j*9, 8l0, (197*0. 2. F.S. Rowland and M.J. Molina, Rev. Geophys. Space Phys., 13 , 1 (1975). 3. R.S. Stolarski and R.J. Cicerone, Can. J. Chem. , 52., l6l0 (197*0. U. J.N. Pitts, H.L. Sandoval and R. Atkinson, Chem. Phys. Lett., 29, 31 (197*0 . 5. H.M. Gillespie and R.J. Donovan, Chem. Phys. Lett., 37, U68, (1976). 145-4 i r Hh Hh [ H ... It, p< a) .- bO B . ■p J. u> td >. ft w w M • a c9 •h co H ft C to O !h ^ a) ■P .G O V ft « ^1 » •H 4-> CO o ■« 43 U >> ui -p cd td O H H « ft <« ft (/) U] H O H O 1 II O -H a s -h OSrl ■H ^ ft td a 3 -P V c: -p a> r-J O HO*! ft •H W Eh ft g ft M5-5 Some Fundamental and Applied Aspects of the Atmospheric Reactivity of Selected Organic Molecules J. N. Pitts, Jr. , R. Atkinson, K. R. Darnall, A. C. Lloyd, R. A. Perry, and A. M. Winer Statewide Air Pollution Research Center Department of Chemistry University of California Riverside, California 92502 In the last several years, there has been a gratifying increase in the use of fundamental photochemical and kinetic information to obtain a better understanding of atmospheric chemistry; this has tended to break down somewhat arbitrary distinctions between "basic" and "applied" research. Thus, for example, it is now widely recognized that the hydroxyl radical is the major chain carrier in the chemistry of both the natural and polluted troposphere and stratosphere. A second example involves the current national debate over the degree of control of photochemically reactive emissions from a variety of industrial and domestic sources. The significance of this issue has grown with the improvement in the efficiency of control of emissions from motor vehicles, which, of course, increases the relative importance of stationary source emissions. And important approach to this problem, developed over the last decade, has been the utilization of photochemical hydrocarbon reactivity scales as the basis for controlling the formulations of commercial solvents. Such phenomena as rates of NO to N0_ conversion, Nl-1 maximum ozone levels, or loss of hydrocarbon have been employed to date to generate such reactivity scales. The vast majority of such data have been obtained by irradiation of NO -HC-Air mixtures in smog chambers of x a variety of types. However, this use of secondary smog manifestations, while of great utility, neverthless has led to some significant inconsist ancies when comparisons of various reactivity scales are made. In view of this, we have developed recently a new reactivity scale based on the validated premise that the loss rates of hydrocarbons and nonphotochemically reactive organics can — under carefully selected experimental conditions — be ascribed solely to attack by the hydroxyl radical. This reactivity scale, then, is essentially a categorization of the absolute rate constants for the reactions of hydroxyl radicals with these organic molecules. We have used three sources of information in developing our scale: literature values for absolute rate constants, values determined directly by flash photolysis - resonance fluorescence techniques in our laboratory, and hydrocarbon loss data determined in experiments conducted in the SAPRC 6400-liter all-glass environmental chamber. The flash photolysis - resonance fluorescence technique is applicable to a wide range of pressure (15 to 650 torr total pressure) and temperatures (297 to 450K) , whereas the environmental chamber technique produces rate constants under simulated trospheric conditions. By employing these two techniques, a large number of organics have been investigated and incorporated into a scale in which the reactivities of the hydrocarbons are ordered relative to methane (equal 1) . The relative importance of alkanes, alkenes, and aromatic hydrocarbons — as well as other organics — will be discussed in Nl-2 terms of this reactivity scale and also in terms of the effect of irradiation times on both real and simulated atmospheres. It will be shown, however, that the ultimate utility of this scale depends to a significant degree upon a better understanding of the mechanisms of the initial attack of the OH radical on the organic molecule, as well as a better understanding of the subsequent steps which produce ozone in these systems. In this regard, new data relating alkane structures with sub- sequent photo-oxidation processes, and recent studies of the reactions of the hydroxyl radicals with the series of aromatic hydrocarbons at elevated temperatures which have interesting mechanistic implications will be presented. Nl-3 IR Fourier-Transform Spectroscopic Studies of Atmospheric Reactions H. Niki , P. Maker, C. Savage and L. Breitenbach Research Staff Ford Motor Company, Dearborn, Mich. 48121 Current knowledge of atmospheric reactions involving polyatomic radicals is severely limited due primarily to the lack of adequate ana- lytical methods for monitoring reactants and product. Fourier-transform IR spectroscopy offers a highly sensitive in situ detection method. In order to maximize system performance in signal detection and in computer- aided data processing and analysis, we have assembled a rapid scan, high resolution IR Fourier-transform facility based on an EOCOM Model 7001 interferometer, a PDP 11/40 on-line computer, and in-house software. The interferometer is equipped with a liq. N~ cooled HgCdTe de- tector and is capable of monitoring IR signals in the 600^4000 cm range with Av >_ 1/16 cm and scanning speed <10 sec. Spectra of prod- ucts are derived from interf erograms, ratioed against background, con- verted to absorbance, and analyzed using linear desynthesis with refer- ence spectra. The two photochemical reactor-IR absorption cells used in this study are a l-l Pyrex cylinder (50-cm long, double pass) and a 70— Jl Pyrex cylinder (1-m long, 40 pass). The standard spectra of reactants and products were recorded typically at 0.1 Torr and/or 1 ppm (=2.5 x 13 -3 10 molecule cm ) in the respective cells. The noise level of the product spectra was comparable to 0.5% of the reference spectra. To illustrate this technique, results obtained on the photo- oxidation of formyl radical (HCO) will be presented. The HCO is a cru- cial radical species in atmospheric chemistry. It is commonly assumed that it undergoes abstraction reaction with 0„ . The present study shows N2-1 that over 50% of this reaction forms peroxyformyl radical. The HCO radical was produced in the presence of 1 atm. air by C£-atom sensi- tized reaction of HCHO at ppm level. In this system, CO- and HCOOH as well as CO were observed as products. In addition, the IR spectrum of hitherto undetected product has been characterized, and has been tenta- tively assigned to performic acid. To further verify the formation of peroxyformyl radical, N0 9 was added to this system to form an organic nitrate analogous to peroxyacetyl nitrate (PAN) , a well-known product in photo-oxidation of CH_CH0-N0 9 mixture in air. These results are shown in Figs. 1 and 2. Time-resolved product spectra of HCHO-NO -C&„ mixture at 10 ppm each (Fig. 1) show the formation of a "new product." The spectrum of this compound is shown in Fig. 2(B), which was obtained by desynthe- sizing the difference spectrum (Fig. 2(A)), i.e., difference between 4 and min. spectra in Fig. 1. This spectrum has a number of well- resolved P, Q, and R branches and resembles those of nitric acid and PAN. The material balance for carbon indicates that over 25% of the HCHO consumed is incorporated in this compound. Therefore, the spectrum has been tentatively assigned to peroxyformyl nitrate (PFN) . The prod- uct is unstable, having a lifetime of about 10 min. in the dark under the present experimental conditions. Attempts to improve spectral char- acterization of these compounds are underway. N2-2 t = sec ** J - ■ i i f .r 3000 2000 1000 (cm -1 ) Fig. 1: Photolysis of HCH0-N0 2 -C£ System, 10 ppm Each N2-3 (A) Difference Spectrum (At = 4 min) HC00H(.15) / HNO (1.0) / HCHO(5.1.ppm) H (B) Spectrum of PFN (HC00N0 ) ? Z50 2000 1500 1000 (cm "S Fig. 2: Product Spectra of HCHO-NO -C£ System, 10 ppm Each N2-4 PHOTOCHEMICALLY GENERATED OZONE FROM ISOLATED STRONG POINT SOURCES D. D. Davis & W. Keifer Chemistry Department/ U. of Md. , College Pk. , Md. In spite of increasingly tight controls on a wide range of emission sources/ cities up and dcwn the East Coast of the United States continue to experience very high levels of air pollution. Of particular concern has been the high level of ozone observed both in urban as well as non- urban areas of this part of the country. In the Nation's Capital, the summer of 1975 delivered yet another bumper crop of days having air pollution index readings of 100 and over. In probing the question as to where this pollution is coming from, various answers can be had from various authorities in the field. These answers range from the city automobile being the major source to that of the lush natural summer vegetation of the East Coast being the cause. In general, however, there have been far more predictions made based on desk calculator experiments than on actual field studies . To be reported in this talk will be the results of an aircraft field sampling program carried out in the vicinity of Washington, D.C. to determine if the air quality of the Nation's Capital is being controlled from inner city sources, from outlying non-urban sources, or from both. The aircraft employed in the study was a essna 205 equipped with an overhead air sampling duct and sampling instrumentation which included: a Monitor Labs NO/N0 2 /NO detector; a Monitor Labs Ozone detector; a Meloy S0 2 , H S, CS 2 (Total Sulfur) Analyzer; and Environ- mental labs Condensation Nuclei Counter; an Ecolyzer CO Analyzer, and a Beckman total Hydrocarbon Analyzer. N3-1 The principal geographical region of investigation in this study was an approximate 12,000 square mile corridor south- southwest of Washington, D.C. One of the major reasons for the selection of this region was an earlier observation that summer winds from the south-southwest have resulted in the highest air pollution index readings recorded in the Washington/ D.C. area. A total of 45 to 50 flights have been made in this corridor over the period of October 1973 to June 1976. Much of this flying has been concentrated on detailed studies of power plants in the area and has taken place during late spring, summer, and early fall, periods during which Washington, D.C, like other cities, has experienced its highest air pollution index readings. As indicated earlier in the text, one of the major questions which we hoped to explore in this study was that of why SW or southerly winds typically result in the highest air pollution index readings in D.C. . The correlation between the air quality index and wind direction is most definitely not a simple one since a very large number of meteorological factors can be involved (i.e. wind variability in speed and direction, surface winds vs. winds at one to two thousand feet, humidity, cloud cover, and temperature) . The results of the study to date have shown that the high index readings recorded with SW or Southerly winds do not, in general, reflect strong sources being present in the immediate southern part of the greater Metropolitan Washington, D.C. area. In fact, what was observed was that typically under conditions of high temperature, high humidity, and high solar flux, ozone levels 1.5 to 3 times that of ambient air N3-2 were found blowing toward or into the Washington, D.C. area from distances going out to anywhere from 50 to 100 miles. During one of the worst air pollution periods in D.C. , with SW winds, levels of ozone in the Washington, D.C. -Richmond corridor reached into the 200 to 300 ppb range. The above information is not to be interpreted, we believe, as a direct indication that all air pollution problems in D.C. originate from out-of-town air masses. The air quality of any large city is most definitely the result of multidimensional sources. In the case of the Greater Washington, D.C. area, the high density of automobiles is clearly a major contributing factor. Our new findings, however, would also indicate that the air mass moving into the city from far out into the countryside also appears to be a major contributor of photochemical oxidant. One of the important questions to be answered concerning the observed high ozone levels to the southwest and south of D.C. is that of the source. Thus far, our studies have shown that at least four sources are involved, (1) automobile traffic on Inters tate-9 5 between Washington and Richmond, (2) two large power plants located at Possum Point, Va. on the Potomac River and Morgantown, Md. on the Potomac River, (3) a cellophane factory located at Fredericksburg, Va. and, (4) the urban plume from Richmond, Va. which itself is a complex mixture of point sources. A final source which should be mentioned but cannot be well-defined is the diffuse source still upwind of the corridor which we generally refer to as ambient air. Significant variations in the level of ozone in this upwind ambient air obviously also reflect multi- N3-3 dimensional sources consisting of highly diluted urban plumes , diluted strong point source plumes , regional photochemistry involving natural constituents from forest, and to some extent stratospheric injected ozone. Details on each of the sources mentioned in conjunction with the Washington, D.C. -Richmond corridor will be discussed. Suffice it to say at this point that a quantitative assessment of the contribution of ozone from each of these sources will require much further investiga- tion. A final comment that should be made to maintain some perspective on the problem being discussed is the following. Although Washington, D.C. has been placed as the focal point of a site being affected by external pollution sources, this in no way implies that D.C. itself is not a major air pollution source for other cities downwind from it. Data, in fact, will also be presented which shows this phenomenon. N3-4 Abstract Evidence for Alkoxy Radical Isomerization in C.-C Alkanes in NO -Air Systems x William P. L. Carter, Karen R. Darnall, Alan C. Lloyd, Arthur M. Winer, and James N. Pitts, Jr. Statewide Air Pollution Research Center University of California Riverside, California 92502 Most photochemical smog models, whether detailed or lumped, predict that the following are the most important reactions in the photooxidation of alkanes in the presence of oxides of nitrogen under ambient conditions: RH + OH ■+ R + HO where R - CH , C 2 H , etc. (1) R + 2 + R0 2 (2) RO + NO ->• RO + NO (3) RO -+ R 1 + R 2 CH0 (4) R 3 CH + -»• HO + R 3 CH0 (5) To obtain insight concerning the adequacy of this simple mechanism, data concerning other possible reactions of the major radical interme- diates is required. In view of the paucity of data obtained under ambient conditions concerning the mechanism and products of alkane oxidations, we have examined the extensive gas phase studies of free radical reactions and hydrocarbon oxidation carried out at temperatures of -250 C and above. A type of reaction commonly encountered at these temperatures is the isomerization of alkyl, alkoxy, and alkylperoxy N4-1 radicals via internal hydrogen (H) abstractions. Similar intramolecular reactions have been postulated and observed in liquid phase studies at temperatures close to ambient. However, to the best of our knowledge, these processes have not been incorporated in mechanisms proposed for the hydrocarbon oxidations occurring in photochemical air pollution. We have estimated Arrhenius parameters in order to permit calcula- tion of approximate rate constants (at 300 K) for intramolecular hydrogen shift reactions for alkyl, alkoxy, and alkylperoxy radicals which are expected to be involved in the photooxidation of C.-C, alkanes. To evaluate the potential significance of the radical isomer- izations under ambient conditions, we have compared their rates as calculated to the expected rates for the competing processes with 0. or NO which are believed to be important in the atmosphere. While isomer- izations of alkyl, and probably most alkylperoxy radicals, do not appear to be important, we find that the possibility of 1,5 hydrogen shifts in alkoxy radicals with 6 hydrogens must be considered. Figure 1 shows the reaction sequence and products predicted in the photooxidation of n-pentane with and without the proposed alkoxy radical 1,5 H shift isomer ization. If the OH radical abstracts a hydrogen from the 3 position in n-pentane, no rapid isomerization is possible while rapid isomerization is predicted following abstraction from the 1 or 2 position. In order to test these predictions and to ascertain the importance of intramolecular alkoxy radical isomerization in photochemical air pollution, we have performed a series of experiments under simulated atmospheric conditions in two environmental chambers. Irradiations of NZp2 alkane (n-butane, n-pentane or n-hexane)-N0 systems in air were carried out in a 5800-liter evacuable, Teflon-coated environmental chamber using a 25,000 watt solar simulator. Analogous studies were carried out in a 6400-liter all-glass environmental chamber equipped with fluorescent black lights. Detailed results from the experiments in both chambers will be presented. Computer kinetic model calculations with and without the 1,5 H shift reaction of alkoxy radicals were carried out simulating the experimental systems studied. Results of calculations of the yields of the pentanone isomers for a selected n-pentane experiment (Run 164) are shown in Table 1. The pentanone yields predicted when alkoxy radical isomerization is not included reflect the relative rates of OH abstrac- tion shown in Figure 1. On the other hand, the calculated yields of 1- and 2-pentanone are significantly lower when isomerization of their precursors is assumed. The experimental results for four n-pentane-NO -air irradiations are also shown in Table 1. It is clear from the results of Run 164 (glass chamber) that both the relative and absolute yields of the pentanone isomers are consistent with the predicted importance of 1,5 hydrogen shift isomerization of the pentoxy radicals. Results from the evacuable (Teflon-lined) chamber for the same initial n-pentane concen- tration are in excellent agreement. Data for both higher and lower n-pentane concentrations give the same relative yields of the pentanone isomers, although the absolute amounts vary as expected. Additional support for the occurrence of intramolecular alkoxy radical isomeriza- tion is found in the yields of acetaldehyde and propionaldehyde, both N4-3 of which are expected decomposition products of the 2- and 3-pentoxy radicals under the experimental conditions employed. Similar results will be presented for studies of the n-butane and n-hexane photooxidation systems. On the basis of our work, it appears that alkoxy radical isomer- izations, which are well known from elevated temperature free radical and oxidation studies as well as in certain liquid phase systems around room temperature do, in fact, occur under photochemical smog conditions at ambient temperatures. The implications of these results for alkane oxidation under ambient conditions, including the formulation of photo- chemical smog models which incorporate higher alkanes (>_C ) , will be discussed. Acknowledgements This work was supported in part by the National Science Founda- tion-RANN (Grant No. AEN73-02904-A02) . The contents do not necessarily reflect the views and policies of the National Science Foundation-RANN, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. The authors also gratefully acknowledge support of the California Air Resources Board (Grant No. 5-067-1 and Contract No. 2-377) for the construction of the chamber facility employed in this study. N4-4 Figure 1. Suggested sequence of major reactions following abstraction from n-pentane by hydroxy 1 radicals in b-C.H _-N0 -air mixtures. The relative percentages for the OH reactions are derived from the data of Greiner [1970], C-C-C-C-C n-C 5 H 12 + QH 00- N0 c-c-c-c-c — { — > C-C-C-C-C N0 2 > C-C-C-C-C 00* C-C-C-C-C C-C-C-C-C 00- ? C-C-C-C-C NO -> C-C-C-C-C NO 2 NO » — I — > C-C-C-C-C N0 2 HO, -> C-C-C-C-C ^FAST ISOM. OH ■> C-C-C-C-C -> C-C-C-C-C HO, .FAST ISOM, OH > C-C-C-C-C ■*■ + HO, -> C-C-C-C-C Table 1. Calculated and observed absolute yields and percentages of pentanone isomers from the n-pentane-NO x -air photooxidation under a variety of conditions. Product Yields in ppb (% of total pentanones) 1-Pentanone 2-Pentanone 3-Pentanone MODEL CALCULATIONS a) No Isomerization With Isomerization 31 ppb (20%) 0.2 (1%) 83 ppb (54%) 9 (23%) 40 ppb (26%) 30 (77%) Initial Run Chamber HC (ppm) NO y 164 Glass 135 Evacuable 4 (Teflon-lined) 165 Glass 166 Class 35 0.4 ofl 0.1 o.i EXPERIMENTAL DATA b) b) b) b) 8 (21%) 30 (79%) 8 (16%) ^2 (84%) 32 (17%) 160 (83%) <1 (0-17%) 6 (100-83%) a) Using initial conditions of Run 164. Similar % yields were obtained using initial conditions of other runs and are omitted for clarity. b) None observed; detection limit 0.5 ppb. N4-5 Collisional Destruction of Rovibronic Levels in S.. Glyoxal: Electronic, Vibrational and Rotational State Changes L.G. Anderson, A.E.W. Knight, and C.S. Parmenter Department of Chemistry, Indiana University Bloomington, IN 47401 USA At pressures above 100 mtorr, the decay of S.. glyoxal (CHOCHO) is dominated by collisions. In addition to collision-induced vibrational and rotational relaxation within the S 1 state, every collision partner so far tried is effective in destroying the S.. electronic state itself. The competition between these channels and some fine details of collisional state changes can be worked out with unusual clarity be- cause of the open structure in both S -S_ fluorescence and absorption spectra. Early studies identified the collisional destruction of the elec- tronic state as collision-induced intersystem crossing . Subsequent investigation of fluorescence lifetimes and quenching from low S vibrational levels (e ., < 2000 cm ) characterized this crossing as vib ~ ° 2-4 "small molecule" S..-T mixing . This transforms to intermediate mix- ing as increasingly higher levels are pumped by the exciting light, and finally, statistical decay is observed from very high S, levels . Quenching cross sections a for destruction of the S 0° level by collision- induced triplet formation are known for over twenty collision 2-4 partners . Several correlations of these cross sections have been f\ 7 given ' . We find another and particulary simple relationship, namely In a aje , where e is the potential well depth between pairs of quenching molecules (He-He, CO-CO, etc.). This is effective not only for correlation of cross sections for S glyoxal quenching, but also 01-1 for correlation of cross sections in numerous other studies spanning vibrational and rotational relaxation as well as collision-induced pre- dissociation. When considering cross sections of a series of quenchers Q acting on a single excited species S, the parameter "V^Zl is pro- portional to the well depth £ Qn . Thus the correlation is indicative of interactions using primarily the attractive part of the potential, and consistent with this view, we observe that the correlation is a poor representation of relative cross sections in cases where collisions are inefficient in bringing about state changes. We have deduced the absolute cross sections for rotational, vi- brational and electronic state changes after pumping narrow rotational distributions in the 0° or the 7 1 levels of the S state (V7 is the lowest frequency fundamental, 235 cm ). We observe that the cross sections are all large and competitive with each other. They are given specifically in Table I for the collision partners S n glyoxal and Ar. They are in the form of the second order rate constants in units f\ _ 1 _i of 10 torr s . A hard sphere collision has in these units a rate constant of about 10. Table 1 Rate constants (in 10 torr"' 5 / for destruction of Au levels in qlyoxal Hord sphere ss 10 state change "from G level 0° Ar from G level 7! Ar el I 0.7 I 0.7 vib 3 0.7 13 3 rot 21 12 10 II 01-2 The analysis of rotational structure in 12 S--S n absorption bands by Ramsay and co-workers has allowed us to detect a selection rule for rotational relaxation in experiments which observe fluorescence at ca. 0.5 cm resolution after excitation of small groups of rotational levels with an Av = 0.33 cm Ar laser line. We observe in glyoxal- glyoxal collisions that cross sections for AK = ±2 collisions far ex- ceed those of other rotational state changes. A curious selection rule appears in vibrational relaxation effected 2 by glyoxal-glyoxal collisions after pumping the level 7 . Fluorescence structure shows that the cross section for AU7 = -2 is large whereas that for the vibrational change AU7 = -1 is relatively small. The fluorescence spectra display this in two ways. First, we observe at pressures where collisional effects are first introduced that the 0q band is substantially brighter in fluorescence than the 7} band (these relative intensities represent the respective 0° and 7 1 level popu- lations). Second, we observe that the "spiky" rotational structure in the 7£ fluorescence band at these low pressures is mirrored by spiky structure in the developing 0q band, whereas the rotational structure in the 7 \ band is substantially without these "non-Boltzmann" features. Apparently the cross section for 0° population after 7 2 excitation is competitive with that for rotational relaxation within the 7 2 level it- self. As a result, the rotational disequilibrium produced in 7 2 upon excitation is largely retained during the initial stages of 0° pop- ulation. On the other hand, the cross section for 7 1 population is sufficiently small so that appreciable progress towards rotational equilibration in the 7 2 level occurs before collisions populate 7 1 . 01-3 Thus a near Boltzmarm rotational distribution is transferee! to the level 7 during its initial stages of growth. 1. L.G. Anderson, C.S. Parmenter, H.M. Poland, J.D. Rau, Chem. Phys. Lett. 8, 232 (1971). 2. J.T. Yardley, G.W. Holleman, and J.L. Steinfeld, Chem. Phys. Lett. 10, 266 (1971). 3. L.G. Anderson, C.S. Parmenter, and H.M. Poland, Chem. Phys. 1_, 401 (1973). 4. R.A. Beyer, P.F. Zittel, and W.C. Lineberger, J. Chem. Phys. 62^, 4016, 4024 (1975). 5. R. van der Werf, E. Schutten, J. Kommandeur, Chem. Phys. 11 , 281 (1975). 6. J.E. Selwin and J.I. Steinfeld, Chem. Phys. Lett. 4_, 217 (1969). 7. C.A. Thayer and J.T. Yardley, J. Chem. Phys. 57, 3992 (1972). 01-4 The Vibronic Dependence of Glyoxal Photodissociation' George H. Atkinson and C. G. Venkatesh Department of Chemistry Syracuse University Syracuse, New York 13210 The selective excitation of single vibronic levels (SV1) in poly- atomic molecules has been established as a yery useful method for examining state-to-state relaxation. The relaxation mechanism is usually monitored through the detection of radiative decay channels (mostly fluorescence). As a consequence, only the net effect of all nonradiative channels is usually measured even though several processes may be simultaneously contributing to the mechanism (e.g., internal conversion, intersystem crossing, vibrational relaxation and photo- chemistry). A significantly more detailed view of the relaxation mechanism can be obtained by monitoring only one of these nonradiative channels in conjunction with SVL excitation. We report in this paper a study of the photodissociation mechanism of glyoxal based on the direct observation of a dissociation product following the selective excitation of single vibronic levels in the parent molecule. The study was carried out in a low pressure regime where the collisional perturbations on excited-state processes can be controlled. It is intended to establish the relationship between the vibronic character of the initially populated *A level and the net efficiency for dissociation into a specific product, CO. Selective excitation was obtained from a tuned dye laser and the concentrations of CO are measured by resonance fluorescence. To our knowledge, this is the first reported study combining these two experimental techniques 02-1 for measurements of SVL photodissociation. Single vibronic levels in glyoxal vapor were selectively populated by a tuned (0.1 a bandwidth) nitrogen laser pumped dye laser. The absolute wavelength of the dye laser radiation was set by means of a one meter spectrometer working in the second order of an 1800 g/mm grating (resolution > 0.08 ft). The pulsed laser radiation was detected by a photomultiplier and the signal analyzed with a box car integrator. Carbon monoxide was detected by resonance emission from the fourth 2 positive system of CO in the vacuum ultraviolet. A xenon lamp, powered by a microwave discharge (2450 MHz), initiated the resonance emission from CO which was detected by a solar blind photomultiplier (10 6 gain) with a Csl photocathode. The photomultiplier signal was amplified and the total signal offset in order to suppress contri- butions from dark current and scattered resonance lamp emission. The resulting signal was measured on a voltmeter. The signal was cali- brated by using known pressures of high purity CO measured by a capacitance manometer. The error in the linearity of the entire de- tection technique was less than 1%. The relationships between the C q increased with pressure below approximately 1 torr and (2) decreases with pressure above approximately 1 torr, (3) the value of p at a given pressure depends on the vibronic level initially populated, and (4) the rate of 02-2 00 CD -do 00 3> Q13IA lAiniNvno d CVJ d 5- 3 en 02-3 decrease of CQ with pressure above 1 torr is a function of the SVL initially populated. The change in pressure dependence near 1 torr reflects the presence of two competing decay channels. Below 1 torr, collisionally- induced intersystem crossing from *A to 3 A levels appears to control the net efficiency for CO production by virtue of observation 1. Above 1 torr, control of the mechanism shifts to the competive decay of the intermediate designated as H. The intermediate H was first proposed by 3 4 Yardley in an analysis of data obtained by Parmenter. It is formed subsequent to the population of 3 A glyoxal and undergoes first order decay to yield CO. Col 1 isional quenching of H occurs in competition with CO formation and accounts for the inverse pressure dependence of p versus the pressure of glyoxal for this pressure region are linear for all SVL studied. The dependence of CQ on the initial SVL populated obtains from the presence of a non-Boltzmann distribution of energy among the vi- brational levels in the } A state in this pressure region. Fluorescence and coll isional ly-induced vibrational relaxation (VR) and intersystem crossing (ISC) to 3 A all compete to relax the initially pumped SVL. ISC leads to CO production via the decay of 3 A levels and the specie H. VR redistributes energy within the *A state among a variety of vibrational levels. This distribution remains non-Boltzmann up to pressure of approximately 10 torr. SVL fluorescence spectra have been obtained which establish the existence of these unique vibronic distri- butions in this pressure regime. At a given pressure and for a specific SVL excitation, this unique non-Boltzmann distribution determines the 02-4 total among of ISC occuring from *A to 3 A and thereby controls the (J)p . Thus, it is from these distributions that the vibronic dependence of cf> c0 at a given pressure arises. In a completely analogous way, the rate of decrease of ty-r, with glyoxal pressure for excitation of specific SVL also derives from these non-equilibrated vibrational distributions within *A . u 1. Supported by the U.S. Army Office of Research - Durham under Grant DAHC04-75-G-0104. 2. T. G. Slanger and G. Black, J. Ctiem. Phys., 55, 2164 (1971) and references therein. 3. J. T. Yardley, J. Chem. Phys., 55, 6192 (1972). 4. C. S. Parmenter, J. Chem. Phys., 41, 658' (1964). 02-5 Behaviour of Benzene in Low Vibrational Levels S.A. Lee, J.M. White and W.A. Noyes, Jr . Department of Chemistry University of Texas Austin, Texas 78712 Fluorescence, triplet state, and i somen' zati on yields from benzene vapor have been determined at several wavelengths but principally at 266.8 nm where B ? molecules are formed in the zeroth vibrational level. Fluorescent yields were determined using 0.18 for benzene vapor at 20 torr and 253 nm as a standard. Triplet yields were determined by two methods: 1) The sensitized triplet emission of biacetyl and 2) The sensitized isomerization of the 2-butenes. The former has the advantage of using much smaller amounts of foreign gas and hence collisional vibrational equilibration is also smaller. When used under comparable conditions the two methods give comparable results. It is well known that there are many isomers of benzene, but benzvalene seems to be the only one which should concern us here. It absorbs strongly from 210 to 230 nm. The production and destruction of benzvalene was studied using a conventional photolysis system. In these experiments cis-2-butene was mixed with the benzene prior to irradiation as a means of removing triplet benzene. Differential absorption spectra were measured after the photolysis with a Cary-14 spectrometer using a benzene blank. The relative amount of benzvalene produced was estimated semiquantitatively by measuring the area under the absorption curve in the region 210-230 nm. In some experiments, a benzvalene-benzene mixture, produced by photolysis at one wavelength were rephotolyzed at a second wavelength. From the resulting data we are able to draw some conclusions about the wavelength dependence of 03-1 the probability of benzvalene formation and destruction. The B ? molecules in their zeroth vibrational level show no variation of fluorescent yield Qf with pressure until pressures of 20 torr are reached. At high pressure Q f decreases to a value near 0.18, the same value reached at 253 nm and at 20 torr. This might be inter- preted to be the yield from B« molecules with vibrational energy equilibrated with the surroundings. Thus at low vibrational levels Q^ appears not to increase with increases in vibrational energy. Below about 20 torr and at 266.8 nm, Q f is constant at 0.22. This result implies that the variation in Q f with pressure at 253 nm or 259 nm originates with the occupation of different vibronic levels. Presum- ably equilibration of vibrational energy in the B ? state does not give the yield of 0.22 because increasing the pressure of inert gases causes the yield to be lowered. The acquisition of vibrational energy at the zero-point energy level is apparently much slower than the loss of vibrational energy from high levels. To explain the difference between these two yields, it is suggested that the thermal equilibration of vibrational energy following absorption at 253 or 259 nm leads to appreciable population of the 237 cm" vibrational level (this is the same mode as the 398 cm" vibrational level in the ground state and is an out-of-plane motion). Emission from this level to the ground state would not be symmetry forbidden but should have a very low transition probability due to the Frank-Conden principle. Population of 237 cm vibrational level as the result of collisions would make this level important after excitation at shorter wavelength but it would not be so important in the processes occuring after excitation at 266.8 nm. 03-2 In the isomerization study, it was found there was no benzvalene formation observed at 266.8 nm, but at shorter wavelengths benzvalene is formed and the steady state concentration is higher at shorter wave- lengths. Illumination at 253 nm up to the steady state followed by irradiation at 266.8 nm caused benzvalene to disappear. Even though benzvalene is not formed at 266.8 nm, it is destroyed by a benzene sensitized reaction, since benzvalene quenches B ? (v=0) molecules presumably energy transfer can occur from benzvalene to B« molecules and results in such electronic relaxation. Illumination at 253 nm followed by irradiation at 247.2 nm caused the benzvalene to increase. This suggests that the photostationary state depends on the excitation wavelength. It has been assumed that photons absorbed by benzene caused fluorescence, triplet state formation, and isomerization. A fourth possibility, a radiationless transition from the B ? state (or possibly 3 the B, state) to the ground state, has been neglected. At 266.8 nm the isomerization yield seems to be zero at low pressures and the sum of Q.p and triplet yield is about 0.97 and thus nearly all the input photons can be accounted for. The energy balance is not found at any shorter wavelength, and is partly due to benzvalene formation, but since the primary isomerization yields are not known, it is difficult to have complete accountability of the absorbed photons. Benzvalene appears to quench the B ? state of benzene so that fluorescent yields tend to decrease with increase in exposure. 03-3 Dual Lifetime Fluorescence From Pyrimidine Kenneth G. Spears and Mahmoud El-Manguch, Department of Chemistry, Northwestern University, Evanston, Illinois 60201. The 1,3 diazabenzene (pyrimidine) 1 B 1 (n-»rr*) electronic state has a nonradiative relaxation characterized by a dual lifetime in low pres- sure gas phase emission. The molecule is especially interesting as a system with non-bonded electrons, substantial spin-orbit coupling, and a small S-T energy gap (1800 cm" 1 ). A primary goal is to understand the nature of the S-T coupling with respect to the low triplet state density and non-bonding electrons. Early work on the molecule has been substantial and a partial list of references include a critical review of the electronic state spectro- scopy of all azabenzenes (1), the detailed spectroscopy of pyrimidine X B! (2) and condensed phase analysis of fluorescence and phosphore- scence yields (3). The recent work of Parmenter and associates (k) has analyzed the fluorescence emission spectroscopy (SVL) and re-assigned some bands in the absorption spectrum. Recent work by Parmenter and Knight has shown that the molecule is very sensitive to collisions with self quenching rates higher than expectations based on hard sphere collision diameters. We have reported our early measurements on pyri- midine excited to several higher vibrational levels above the 0-0 band (^2000 cm -1 ) (5). We found a dual emission lifetime for these states with lifetimes near 1 nsec and 100-200 nsec at 0. 1 Torr. Very recent work of Uchida, Yamazaki and Baba (6) has also reported dual emission but no precise measurements of longer lifetimes were possible. We have extended our results to many absorption bands, including the 0-0 band. We find that the lower bands fit the model of Lahmani, 04-1 Tramer and Trie (7) for a reversible intersystem crossing. The model treats the case of a forward crossing rate k., a reverse crossing rate k , and a singlet fluorescence decay rate k_. For the case of observed dual lifetime emission (k is the fast decay) in the form l(t) = -k t -k t C e s + Ce L we can derive the individual parameters k,, k , k„ , s L r i' r' f ' the number of effective states, N, coupling to the singlet state, the average coupling matrix element V q , and the average triplet density of states p . The following equations are used to derive these para- meters. C /C. = k./k (2) s' L i! r v k s (k. + k ) 2 k. _ V 1 X' + _JL L f r r (3) k -k k r (k) L * k. + k 1 r k s =2nV s 2 T P T ( 5 ) N=TT 2 V ST P^ (6) Our experiments measure C , k , C , k as well as relative quan- S S J_i Li turn yields for various bands. For lower states we obtain zero pressure lifetimes as long as 2. 5 usee that are quenched with very large cross sections (10 x hard collision). The absolute amount of short lived emission is very small for all bands, but the amount of long lived emission ranges from > 100 times larger (near 0-0 band) to S.. conversion must be extremely rapid, at least as compared to S emission, since it is usually observed that the major emission originates from S, even when there is much larger S "** S~ oscillator strength. In the absence of high concentrations of fluorescence quenchers, the S 1 " > S„ fluorescence spectrum in solution is invarient to excitation wavelength. On the other hand, under the same conditions, the S. -> S n fluorescence quantum yield may exhibit im- 2 portant variations. These are conveniently expressed by the ratio of the quantum yield when excitation is at some wavelength A to the yield when excitation is into the first absorption system. This ratio, after 2 suitable correction , is usually interpreted to be a measure of the overall efficiency, g (A ), with which the A excitation to some state S converts to the emitting vibronic levels of the state S. , and, as n 1 expected from this interpretation, is always observed to be less than or equal to unity. For dilute solutions of benzene or toluene in some suitable trans- parent saturated hydrocarbon solvent (e.g., isooctane) , g generally declines as the excitation wavelength decreases and then at a wave- length close to the maximum of the third absorption system (^ 185-190nm) 06-1 reverses direction and begins to rise. 3-6 For p-xylene and generally more complex molecules, 3 remains about equal to unity over this entire spectral range. The origin of these variations has never really been satisfactor- ily explained. Clearly there exists for benzene and toluene an inef- ficiency in the S ■*■ S, conversion process but its source is unknown. More importantly still, the overall mechanism of the conversion process has never been elucidated. It has generally been conjectured that S "*" S, conversion proceeds via a cascade involving successive S "*" S n 1 n n-1 internal conversions but there is, in fact, no experimental proof of this for any polyatomic molecule. Also to be considered is that ioniza- tion thresholds in condensed phases may be lowered sufficiently from their gas phase values that for many aromatic molecules, at relatively low excitation energies, the non-radiative conversion to S, could im- portantly involve electron- ion pairs as intermediates in the process. Indeed, Laor and Weinreb, have reported for excitation wavelengths within the third absorption systems of benzene and toluene, a decrease in 3 in the presence of electron-scavenging additives. Also Fuchs, 9 Heisel and Voltz have conjectured that the upswing in 3 of benzene and toluene at ^ 180 nm is somehow due to the onset there of an autoioniza- tion. In two recent publications we reported observation of a very weak S~ ■* S-. and S„ "*" S„ fluorescence from a number of aromatic molecules when excited into their third absorption systems at 184.9 nm. ' With the capability now of monitoring the emission from the three elec- tronic states S_, S and S. rather than just the terminal state S , we 06-2 have begun a series of studies directed towards the elucidation of the mechanism of S -+ S n radiationless conversion. Our first results for n 1 p-xylene are presented below. The fluorescence quantum yields of p-xylene in isooctane excited at 184.9 nm have been determined for the transitions S -> S ($.-.), S -> S n ($») and S..-> S ($- ) as functions of the concentration of p-xylene (c ) and of the concentration of added quenchers (CC1, , CHC1~, c_ - C ? F ,) and compared with the S 1 -> S n fluorescence quantum yield (to 1 ) for excitation at 253.7 nm in the same solutions. Also vapor fluores- cence quantum yields have been re-determined for the S„-*- S„ and S„-* S n transitions. It has been found that the total non-radiative decay rate constants of S„ (2. X 10 sec ) and of S ? (l« n x 10 sec ) are essentially the same for dilute solution and vapor phases. Also it is found that in dilute solution, $_ is independent of concentration of quencher whereas $_ is strongly quenched and to about the same extent as is the ratio 6 - $, /^p,. In more concentrated solutions (c > 3 M), $ rapidly declines to unobservable levels whereas and B appear to increase-but only slightly. On the basis of these results the follow- ing is concluded for a dilute solution of p-xylene in isooctane excited 9 into S_ (at 184.9 nm) ; i) The S„ state either radiates (k = 3.0 X 10 3 3 r -1 14 -1 sec ) or makes internal conversion to S ? (k = 2.,. X 10 sec ). No autoionization occurs nor do any other decay channels such as S_-> S or S_->- S have appreciable probability; ii) The S_ state either radia- 8 —1 tes (k = 1.3 X 10 sec ), makes internal conversion to S, (k = l.„ X r 10 13 -1 ^ 14 -1 10 ' sec ) or autoionizes (k > 10 sec ). All other decay channels are of negligible importance; iii) In the absence of quencher, the 06-3 ejected electron returns to the parent positive ion to regenerate S„ and not a lower state. In the presence of quencher the electron is scavenged and no longer can contribute to any emission. Possible mech- anisms for the disappearance of $„ at high c are considered. Prelim- inary results for benzene and toluene will also be presented. *This work was supported by the U.S. Energy Research and Development Administration, Document No. COO-913-61. 1. J.B. Birks, Photophysics of Aromatic Molecules (Wiley-Interscience, London, 1970) and references cited therein. 2. C.W. Lawson, F. Hirayama and S. Lipsky, J. Chem. Phys. _5_1 1590 (1969). 3. C.L. Braun, S. Kato and S. Lipsky, J. Chem. Phys. 39 1645 (1963). 4. S. Feinleib, Ph.D. Dissertation, University of Minnesota, 1965. 5. U. Laor and A. Weinreb, J. Chem. Phys. 43 1565 (1965). 6. J.B. Birks, J.C. Conte, and G. Walker, J. Phys. B (Proc. Phys. Soc . ) Ser. 2, 1 934 (1968). 7. R.C. Jarnagin, Accts. Chem. Res. 4- 420 (1971) and references cited therein. 8. U. Laor and A. Weinreb, J. Chem. Phys. 50 94 (1969). 9. C. Fuchs, F. Heisel and R. Voltz, J. Phys. Chem. 75 3867 (1972). 10. F. Hirayama, T.A. Gregory and S. Lipsky, J. Chem. Phys. 58 4696 (1973). 11. T.A. Gregory, F. Hirayama and S. Lipsky, J. Chem. Phys. 58 4697 (1973). 06-4 Photosensi tization Of The 2-Butenes By Benzaldehyde In The Gas Phase A. J. Yarwood, Department of Chemistry, McMaster University, Hamilton, Ontario, Canada; G. R. De Mare and M. Termonia, Laboratoire de Chimie Physique Moleculaire, Faculte des Sciences, Universite Libre de Bruxelles, 50 av . F.-D. Roosevelt, B-1050 Brussels, Belgium. The sum of the quantum yields of the benzaldehyde pho- tosensitized (X = 366 nm) cis->trans and trans + cis isomeri- zations for both the 1 , 3-pentadienes (independent of pres- sure over the range studied: 10 - 400 Torr) and the 1,2- dichlor oe thy lenes (for substrate pressures > 100 Torr) is near unity, indicating that the quantum yield for intersys- tem crossing in benzaldehyde irradiated at 366 nm is also near unity (1,2). Supporting evidence for this high inter- system crossing yield for benzaldehyde in the gas phase was given by Berger , Goldblatt and Steel (3) who found $(C ) - 0.85 for the benzaldehyde sensitized decomposition of 2,3- diazabi eye lo (2 . 2 . 1 J hep t-2-ene. The fall-off in the quantum yields of i somer i z ati on of 1 ,2-dich lor oe thy lene with decreasing pressure below 100 Torr was interpreted as being the consequence of the competition of unimolecular deactivation steps of triplet benzaldehyde with the bimolecular energy transfer (1). That benzaldehy- de phosphorescence actually competes with the energy trans- fer to 1 , 2-dichloroe thy lene has now been confirmed with 07-1 laser and flash excitation sources. Preliminary competition experiments have shown that energy transfer from triplet benzaldehyde to cis-l,2-di- . . 4 chloroethy lene , requiring an average of 10 collisions (1), proceeds at about 1/40 the rate of energy transfer to trans- 1 , 3-pentadiene . This was not unexpected because the energy transfer to 1 ,2-dichloroe thy lene is either slightly endo- thermic or thermoneutral while the transfer to 1, 3-pentadi- ene is exothermic by over 20 Kcal/mole. For the 2-butenes the energy transfer from triplet benzaldehyde should be en- dothermic by about 4 to 6 Kcal/mole. Surprisingly, competi- tion experiments showed that the energy transfer to both cis- and trans-2-butene proceeds at about 1/35 the rate of energy transfer to trans- 1 , 3-pentadiene (it is thus slight- ly faster than the energy transfer to 1 ,2-dichloroethy lene) . We have therefore undertaken a detailed study of the benzal- dehyde sensitized isomerization of the 2-butenes in the gas phase . Experimental The experiments were performed using a cylindrical Pyrex cell (11.9 cm long; 3.8 cm diameter) which was enclo- sed in a box fitted with glass windows. The temperature was regulated at 60.00 ± 0.05 *C. The light source, a Philips Philora HPK 125 watt "Wood" lamp, supplied an inci- -9 . -2 -1 dent intensity of 0.46 ± 0.02 x 10 Einstein cm s at X = 366 nm (the mercury lines, 365.0 - 366.3 nm) . The 07-2 absorption coefficient of benzaldehyde vapour for this light is e(decimal) = 4.4 cm 1 mole (4). Results and discussion Benzaldehyde sensitization of the 2-butenes at 366 nm and 60.0 °C leads only to geometric cis £ trans isomeriza- tion. No extraneous peaks were observed when the reaction mixture was analyzed by gas chromatography with a flame ion- ization detector. Also, prolonged irradiation causes no ob- servable (-0.1 Torr) pressure change in the reaction cell. The quantum yields of the cis -> trans, ( $ (c+t)), and trans -* cis, ($ ( t->-c) ) , isomerizations were determined under initial conditions (less than 2 % i somer izat ion) for 2-but- ene pressures ranging from 9.9 to 250 Torr and benzaldehyde pressures ranging from 3.3 to 8.2 Torr. $(c->-t) = 0.533 ± 0.009 and $(t-*c) = 0.258 ± 0.010, independent of the 2-butene pressure (see Figure). Thus $ ( c+t) / $ ( t+c) = 2.07 ± 0.11. 0.5 a 0.4 _j llj >- 0.3 "*> Z) i— — ?■ < 0.2 z> o 0. ♦-*-»- . (0.533^0.009) CIS— TRANS <0.258±0.010) TRANS— CIS 100 200 2-BUTENEJORR 07-3 300 The ratios of the quantum yields or of the initial rates for 2-butene isomer izat ion are given for a few sensi- tizers. The increase in the ratio c-*t/t->c indicates that the branching ratio for triplet 2-butene is different for sensitizers with triplet energy below 80 Kcal/mole or that the sensitization mechanism has changed. Sensitizer E, Kcal/mole Ratio c-^-t/t^-c Ref . Benzaldehyde 71.9 Sulfur dioxide 73.4 Pyrazine (313nm) 76 Benzene 84 . 4 Cadmium 87.7 Mercury 112.7 2.07 2 1 .60 1 .02 1 .00 1 .00 This work 5 6 7 8» 10 9, 10 Financial aid from the N.R.C. (Canada), the F.N.R.S. (Belgium) and N.A.T.O. is gratefully acknowledged. References e, Fontaine, Huybrechts and Termonia, J. Photochem 2/73) 289. ia, D.Sc. Thesis, U.L.B. 1976, in pi , Goldblatt and Steel, J.A.C.S., 9_5 e, Fontaine and Huybrechts, Bull. So 72) 171 . rn and Gusten, Z .Naturf or s ch . , 27a 1 . De Mar 1 (197 2. Termon 3. Berger 4. De Mar 5. £1 < 19 Penzho Demer j 6. Jones Lahman 199. 7. Lee , D 8. Hunzik 9. Cundal and Pa and De 10 . Tsuna »r epar at l on . ( 1973) 1717. c . Chim. Be lges , i an and Calvert, In t . J . Chem. Kinetic and Brewer, J.A.C.S., 9_4 (1972) 631 i, Delouis and Le Gouill, J. Photoc (1972) 1401; s, VII (1975)45 ; Ivanof f , hem. , 2 (1973) enschlag and Haninger, J . Chem. Phy s . , 4_8_ (1968)4547. er, J. Chem.Phys. , 50 (1969) 1294. 1, Prog. Reaction Kinetics, 2_ (1964) 165; Cundall lmer, Tr ans . F ar aday Soc, 5_6 (1960) 1211; Termonia Mare, Chem. Phys . Let ters , 25_ (1974) 402. shima and Sato, Bui 1 . Chem. Soc . Japan , 41 (1968) 284. 07-4 Near Infrared Detection of Peroxyl Radicals in Mercury Photosensitized Reactions Heinrich E. Hunziker IBM Research Laboratory 5600 Cottle Road San Jose, California 95193 While peroxyl radicals play a pervasive role in low temperature oxidation and photooxidation processes, our knowledge about these transient species has been relatively limited due to difficulties in applying most of the powerful free radical monitoring techniques to them. This situation can be improved by making use of the unique property of specific electronic absorption bands in the 1 to 2 ym region for detection of peroxyl radicals. We have found that these bands can be observed by phase sensitive detection of absorption when a modulated concentration of the radicals is generated in a mercury photosensitized reaction. This method has made it possible to observe the lowest excited electronic state of HO with sufficient details of vibrational and rotational structure to estimate its geometry and illuminate its chemical nature. More recently we have also observed analogous electronic band systems of CH 0_, C H , etc., and CH .CO.O , in the same frequency region but with characteristic shifts of the electronic term energies relative to HO and interesting differences of vibronic structure. Pl-1 Just as for HO these spectra are dominated by a progression of the 0-0 vibration, but additional structure due to internal rotation around C-0 bond is introduced by changes of the internal rotational barrier between the ground and first excited electronic states. Since these band systems occur in a generally empty spectral region and are specific for particular peroxy radicals, they are very useful for determining radical formation, reaction pathways and kinetics in photochemical and other types of processes. Applications of this kind will be illustrated with several examples. Plr-2 Reactions Of Hydrogen Atoms With Fluorinated Ketones D.W. Grattan and K.O. Kutschke, Division of Chemistry, National Research Council of Canada, Ottawa, Canada, K1A 0R6 In an unpublished study of the reactions of CF_ radicals, produced by the near uv photolysis of (CF ) CO (HFA), with H , the product ratio (CF H + C F^)/2C0 rose beyond the expected limit of one for some condi- tions. A radical exchange reaction between H atoms and the ketone, which formed CF radicals, was thought to be responsible for this anomaly . In order to study such a possible reaction, H atoms were generated in the presence of HFA by the Hg-sensitized decomposition of H or of C Hr,. Attempts were made to evaluate relative rate constants by com- petitive methods. Each competition studied yielded unexpected kinetic results; possible mechanistic explanations are suggested. The constants for the quenching of Hg( P ) by C H, , C H n-C,H and HFA were determined with the nitrous oxide method. Experimental . A conventional apparatus was employed which contained a quartz reaction cell (^200 cm , optical path 10 cm) illuminated by a low pressure Hg lamp. The cell was part of a loop consisting of a glass cir- culating piston pump, U-tube cold trap and a second U-tube thermostatted at 26°C and containing a drop of Hg. The loop was connected to apparatus for volume measurement, and for chromatographic and/or mass analysis. Results , (l) Quenching Constants . Those for C„H, , C_H-, and n-C, H., ,. " ^ 4' 3 o — 4 10 were determined by the NgO method in the usual way 1 ' 2 . The method meas- ures the total quenching constant, k q x = k a X + k b X' for compound X, Hg( 3 Pi) + X ka » x > Hg( 1 S ) + X* (or other product) [a] k b,X Hg( 3 P ) + X [b] P2-1 and is based on the diminution in the rate of formation of Lfr.„ ") when d. \ a 2 ' 1 2 an addend is present during the photodecomposition of NO ' . For HFA, the determination was made in the presence of C H, , at three ratios of [c H, ]/[N 0], In order to scavenge the atoms formed. The kinetic analysis yields [c]; results are collected in Table 1. I a k q,H20 ^CalU [^j k q?HFA [HFA] r N 2 ~ k a,N 2 + k a,N 2 Py)] + k ajN20 ^ N 2 0] (2_) H + HFA; method I . A method which had been devised by Cvetanovic for the determination of the rates of addition of H atoms to olefins relative to that for H + C H, was employed. H atoms were gener- ated by the Hg(3p ) decomposition of 600 torr of H and their addition to a constant amount of C H, was in competition with that to other ole- fins (here to HFA). The yield of n-C.H , produced by combination of the CLH,. formed in the addition H + C^H, , was measured as a function of 2 5 2 4' [HFA]/[C H. ]. The standard mechanism and treatment is shown below. HgC^) + hv 253>? ■* Be(\) Kg( 3 -P ± ) + H 2 + 2H + HsC^q) H + C 2 H U ■* C 2 H 5 [1] H + HFA ■+ CF CHO + CF [2] 2C 2 H 5 * n-C 1+ H 10 (C 2 H 6 + C^) [3a(b)] 2CF 3 + C 2 F 6 [4] 3 + C^H + CF 3 C 2 H 5 (CF 3 H + C^) L5a(b)] CF Assuming that all radical-radical interaction rates are equal, and that there is no direct quenching of Hg( P ) by either C H, or HFA, then, representing the rate of formation of n-C,H in the presence and in the absence of HFA by (rBu) and (rBu) respectively, P2-2 ((rBu) l/2 - (rBu) l/2 )/(rBu) 1/2 = k^HFAD/k^C^]. The results (Fig.l), obtained at four pressures of CI, show that this function depends on the absolute pressures of C H, and/or HFA and not solely on their ratio. Presumably this arises from the neglect of [6]. C H + HFA -► C H COCF + CF [6] Resolving the kinetics including [6] and deleting [2] yields: (rBu^UrBu^-CrBu^/CrBuK 2 = ( (l/2)k 6 k l/2 /k) [HFA] where 2k E 2(k„ + k„ ) = k,_ + k^ = kj,. This olot also is shown in 3a 3b 5a 5b 4 Fig.l which shows that the points now all coalesce about a single line. If log k is taken as 13. 5(& mol s~ units), k^ = 1.0 x 10 I mol" s~ . During prolonged photolyses more C H, and HFA were consumed than is predicted by the mechanism; this is evidence of some additional reaction(s) between HFA and C H, . (_3) H + HFA; method II . In an attempt to overcome the disadvantages of the previous method, H atoms were generated from the quenching of Hg( P ) by C Ho ' . The rate of H formation was studied as a function of [HFA]/[C Hp~J; the following mechanism was expected to apply. Hg( 3 P x ) + C 3 Hg + H + C 3 H T + Hg( 1 S Q ) [7] Hg( 3 P n ) + HFA -»■ HFA* + Hg^S ) [8] 1 o H + C 3 H 8 + H 2 + C 3 H ? [9] H + HFA -»■ CF CHO + CF [2] 2C 3 H ? + C 6 H ll+ (C 3 H 8 + C 3 H 6 ) ClOa(b)] 2CF 3 ■> C 2 F 6 [4] CF 3 + C 3 H T + CF 3 C 3 H T (CF 3 H + C^) [lla(b)] ^tt /r H 2 /r H 2 = (1 + k 2 [HFA]/k 9 [C 3 H 8 ])(l + k [HFAJ/kgE^Hg] ) 9 From (l), k /kg is 10.2 and Baldwin gives k = 1.27 x 10 5 I mol" 1 s 1 , P2-3 A t low [HFAl/fc 3 H^lthe measured ^/^ were reasonably constant and give k = 8.5 x 10 5 I mol~ 1 s"" 1 . Above [HFA]/[C H g ] - 0.04, the k /k 2 so deduced increase very rapidly; presumably additional processes become important at higher [HFA]/[C_Hq] ratios. (k) Product Analysis . Heavier products of the overall reaction (600 torr H , 10 torr HFA) were studied. Many products were detected, most due to secondary processes; CF CHO was identified. Conclusions . The values of the quenching constants agree well with those of other workers. This gives some confidence to the k „_. = q,iirA 13 x 10 10 i mol" 1 s ; cf the much larger value of 39 x 10 10 I mol" 1 s" 1 for acetone measured by Cvetanovic . The result that addition of C'H to HFA proceeds at some 1000 times 7 the rate of that of H atoms also is consistent with other work . 1. R.J. Cvetanovic, J. Chem. Phys. 2^, 1208 (1955) 2. R.J. Cvetanovic in Progress in Reaction Kinetics , ed. G. Porter, vol. £, p 39. MacMillan Co., New York, 1964 3. R.J. Cvetanovic and L.C. Doyle, J. Chem. Phys. ££, 4705 (1969) 4. R.A. Back, Can. J. Chem. 3j£, 1834 (1959) 5. R.J. Cvetanovic, W.E. Falconer and K.R. Jennings, J. Chem. Phys. 3^, 1225 (1961); K.R. Jennings and R.J. Cvetanovic, J. Chem. Phys. 35, 1233 (1961) ^ 6. R.R. Baldwin, Trans. Farad. Soc. $j), 527 (1964) 7. A.S. Gordon, W. P. Norris, R.H. Knipe and J.H. Johnston, Int. J. Chem. Kinetics £, 15 (1975) 8. Y. Rousseau and H. E. Gunning, Can. J. Chem. i^L, 465 (1963). 9. G. London, A.C. Vikis and D.J. Le Roy, Can. J. Chem. j^8, 1420 (1970) OO.J.V. Michael and G.N. Suess, J. Phys. Chem. Jjj, 5 (1974) P2^4 Cmpd. HFA Table 1. Quenching Constants, k (a) C A C 3 H 8 S- C U H 10 »-10 This work 13.0 31.0 1.27 3.35 (h) Others Others (c) 30.0 1.13 (b) (a) 1.28 (a) I mol s 3.65 (b) Based, on k (b) . „_(e) , A ,(f) 3.75 » ^.o3 (a) k in 10 xw I mol M s _J ". (b) Based on k = lU.3, ref.(2). (c) Based on k _ „ = 30.0, ref. (2). (d) Ref. (8). (e) Ref. (9). q,L 2 M lt (f) Ref. (10). HFA /ETHYLENE Fig. 1. Points refer to [C-H.] = 20 (+), 15 (•), 10 (p) and 5 (A) torr, lower abscissa. Open circles refer to upper abscissa. P2-5 THE REACTION OF NH 2 WITH OLEFINS STUDIED BY FLASH PHOTOLYSIS R. LESCLAUX and PHAM VAN KHE Laboratoire de Chimie Physique A, Universite de Bordeaux I, 33405 TALENCE (France) . In spite of the large amount of studies on the photochemistry or radiochemistry of ammonia, little attention has been given to the reac- tivity of NH 2 radical, even though this radical is generally an inter- mediate in the dissociation processes of ammonia. However, the few available data concerning the NH2 reactivity, indicate that this radical is weakly reactive compared to the isoelectronic OH radical. In our first experiments, we have studied the reactions NH2 + N0(1) and NH2 + NH2 + M (2) which are fast processes, by using a conventional flash photolysis apparatus. It is particularly important to know the kinetics of the recombination process since if often competes with other NH2 slower reactions. The rate constants of this process were determined in the fall off region (0-1000 Torr of N2) , by using different third bodies (NH3, N2 and Ar) and in the temperature range 300-500 K (2). The reaction of NH2 with olefins was studied by flash photolysis, the NH2 absorptions being measured by using a CW dye laser operated in single mode. The reaction being fairly slow, it was never possible to completely eliminate the radical-radical recombination processes by using the conventional apparatus (white analysing light source + monochromator) because the NH2 concentration was too high. The CW dye laser was then used in order to increase the sensitivity of radical detection by renonance absorption (laser spectral width narrower than the absorption line) . The flash photolysis set up is essentially the same as the one previously described (1). NH2 radicals are produced by photolysis of ammonia, using two flash lamps delivering 40-60 joules in 20 ys. An absorption optical path of about 30 m is obtained by placing the reac- tion cell in a multipass system. The temperature can be varied from room temperature up to 240°C. The single mode CW dye laser, Spectra Physics model 580 has a spectral width lower than 50 MHz. Absorption measurements were performed at 597.73 run which corresponds to one of the strongest NHo absorption line. The signals are detected by using two photodiodes and a differen- tial amplifier on the oscilloscope. One of the photodiodes measures the absorption signal, the other, the reference laser intensity. This system reduced considerably the perturbations due to the laser inten- sity variations and allows good measurements of absorptions less than 10 %. An important increase in the sensitivity of NH2 radical detection was obtained by comparison to the experiments using an analysing light source. For an equivalent absorption and signal to noise ratio the NH 2 concentration was about 30 times smaller with the laser. Moreover, the measured optical density is now a linear function of the radical con- centration. P3-1 The initial concentration was evaluated by measuring the second order kinetics of NH2 recombination which rate constant was determined in various conditions (2) . The NH2 concentration was found to be from 2.10~9 to 10~8 M for 10 % of light absorption, according to the pressure broadening of the absorption line, for pressures up to 700 Torr. This important increase in the sensitivity of radical detection is essentially due to the fact that absorption measurements can be made at the maximum of the absorption line and also to the large number of pass in the cell possible with the laser beam. A new increase in the sensi- tivity should be obtained by averaging the signals from a large number of flashes. The sensitivity of detection obtained in our system is in the same range as that of most other methods of radical detection with the exception of the laser excited fluorescence which is much more sensi- tive at low pressure. However, the very efficient collisional quenching of the NH2 (^Aj) excited state reduces considerably the fluorescence intensity at high pressure (3) . The advantage of resonance absorption is that it is not too sensitive to experimental conditions and can then be applied to various systems (atmospheric conditions, flammes etc..) The reaction kinetics of NH£ with ethylene, propylene and 1-butene were determined in the temperature range 300-500 K. The pressure of olefins were varied from 10 to 100 Torr. In all conditions the disappea- rance kinetics of NH2 were first order and the reciprocal lifetimes were a linear function of the olefin pressure (fig. 1). Thus no radical recombination contributed to the NH2 disappearance. The temperature dependance of the rate constants (fig. 2) is accounted for by the follo- wing Arrhenius expressions : k. m ; , = 1.2.1 8 exp - 3.95 (± 0.2) / RT M^s" 1 NH 2 + L 2 H 4 k^ + c H = 2.9. 10 8 exp - 4.3 (± 0.2) / RT m'V 1 2 3 6 Sou* ,_C,H =2.8.10 8 exp-4.1(±0.2) / RT mV Z 4 8 which correspond to,.the following room temperature rate constants. 1.6 x 10 , 2.2 x 10 , 3.0 x 10 5 M^s -1 for C H,, C^, and 1-C^Hg respectively. Activation energies are given in kcal/mole and the errors on the rate constants are estimated to be smaller than 20 %. The reaction is thus fairly slow at room temperature compared to the equivalent reaction of the isoelectronic OH radical. The difference is essentially due to the significant activation energy found in the case of NH2 radicals whereas there is no activation energy and even a slight negative temperature coefficient in the case of OH radical (4). No information are given in this work about the products of the reaction. However, the addition reaction is the most likely to occur since hydrogen abstraction reactions from hydrocarbons by NH2 radicals seem to be very slow : < 10-* M~'s~^ (5). P3-2 Experiments are still in progress in order to extend this study to a series of unsaturated hydrocarbons. REFERENCES 1 - R. LESCLAUX, P.V. KHE, P. DEZAUZIER, and J.C. SOULIGNAC, Chem. Phys . Letters 35, 493 (1975) 2 - P.V. KHE, J.C. SOULIGNAC, and R. LESCLAUX - to be published. 3 - J.B. HALPERN, G. HANCOCK, M. LENZI, and K.H. WELGE, J. Chem. Phys. 63, 4808, 1975. 4 - R. ATKINSON and J.N. PITTS, Jr., J. Chem. Phys. 63, 3591 (1975). 5 - R. LESCLAUX and P.V. KHE, J. Chem. Phys. 70, 119 (1973). ^o-^-iOf 1 ) P_ (Torr) 80 L 36 Fig. 1 :Reciprocal NH„ lifetime as a fonction of propylene pressure, Ln k N V C 3 H 6 Fig. 2 : Arrhenius plot for the reaction of NH with propylene P3-3 Detection And Reactions Of NH(X^2 )Radicals In The Vacuum Uv Flash Photolysis Of NH^ Using Resonance Fluorescence I.Hansen, K.Hoinghaus, G.Zetzsch and ff.Stuhl; Physikali- sche Ghemie I, Ruhr-Universitat , 46$0 Bochuro, W. -Germany In the course of an investigation of the kinetics of imino radicals we have recently studied the quenching of 1 + 1 ^ metastable NH and ND radicals in their (b £ ) states J . In the present paper we would like to describe a method to study the kinetics, of ground state NH(X^£~) radicals. Kinetic data on reactions of NH(X<£~) radicals is sparce. There are only a few estimates of rate constants in the literature which are based on experimental work such as those for the reactions with NH^, N-H^, NH, NO, HN,, 0~, CpH^ and CA-* None of these rate constants have been ob- tained from more than one experiment. In the most direct studies, previously NH(X^.£ - ) was generated either by elec- tron bombardment or flash photolysis of NH^ or HN^ and was detected by time resolved absorption spectroscopy. With these techniques concentrations of NH in the approxi- 13 -5 15 -5 mate range of 3x10 ^ cm ^ to 2x10 y cm v were produced and lifetimes "CL40.2 ms were observed. In the presence of NHv 2) it was concluded that NH is removed efficiently by NH^ ' A) and even more efficiently by NH ^ . In the present study the apparatus consisted of a pulsed light source, a reaction chamber, an NH emission P4-1 lamp and a resonance fluorescence detection system. A similar photolysis apparatus for the study of the kine- tics of OH has been described previously ' . Vacuum uv light pulses (Ty 2 = 1.5/105 nm) . The approximate range of the initial 11 -6 concentration of NH was estimated to be from 2x10 cm ^ 12 -3 to 2x10 XcL cm ° . The NH resonance emission was generated by passing a small flow of a mixture of NH^ at 0.35 Torr and Ar at 1.6 Torr through a microwave discharge. The NH(A IT-^X^ )- emission of the lamp is shown in Fig. 1. The dominant feature of this spectrum is the Q-branch of the (0;0)- band. Furthermore the R-and P-branches of this band and the Q-branch of the (1;l)-band are clearly observed in this spectrum. Care had to be taken to avoid generation of Np(G^7T ;v'=0 -*B^7r; v"=0) emission at 337 nm in the discharge, It should be noted also that under the present conditions 1 1 emission from NH(c TT-*a A ) at 325 nm appears to be absent. The resonance fluorescence from NH radicals was moni- tored at right angles to the incident flash light and to the NH resonance light. It was detected by a photomulti- plier through an interference filter, the transmission of which is included in Fig. 1 (dashed line). The signal of P4-2 the photomultiplier was fed to a multichannel scaler where up to 64 single runs were accumulated and averaged. Several experiments were performed to show that the ob- served fluorescence signal indeed originates from the pre- sence of NH(X^£~) radicals. In the present system NH radicals can be observed up to 1 s after the light pulse. In the presence of NH^ the removal of NH(X"!£~) appears to be rather complex. Fig. 2 shows two decay curves of the NH resonance fluorescence under the same experimental condi- tions but with and without inert gas present. The semi- logarithmic plot shows that the decay of NH cannot be accounted for by a single first order reaction. In Fig. 3 the inverse of the intensity of the fluorescence is plotted vs. time for these two runs. Similarly this figure shows that no single second order process causes the removal of NH(X^£~). From decay curves like these, limits of rate constants can be obtained for the reactions of NH with NH and with NHv. Furthermore, it seems to be likely that NH can survive collisions with the reactor walls during its long lifetime. Upon addition of NO to this system the NH lifetime de- creases resulting in a rate constant for the reaction NH + NO in agreement with a value previously reported ^ . These and other results will be discussed. References: 1) G.Zetzsch, F.Stuhl, Ghem.Phys. Letts. 3£, 375 (1975) P4-3 2) K.A. Mantei, E.J. Bair, J.Chem.Phys. 49, 3248 (1968) 3) G.M. Meaburn, S.Gordon, J.Phys.Chem. 72, 1592 (1968) 4) F.Stuhl, H.Niki, J.Chem.Phys. _5£, 3671 (1972) 5) S.Gordon, W.Mulac, P.Nangia, J.Phys.Chem. £5, 2087 (1971) 3i0 339 338 337 336 335 33A 333 332 X/nm Viz- 1 Emission spectrum of the NH lamp. The transmission of the interference filter used is included. Fig;. 2: Semilogarithmic plot Fig;. 3: Plots of the inverse of the resonance fluores- of the resonance fluores- cence intensity (arb. units) cence intensity (arb. units) vs. time vs. time Experimental conditions: o0.15 Torr NH,; •0.15 Torr NH, + 700 Torr He D P4-4 Interpretation Of The Arrhenius Plots For Reactions Of Oxygen Atoms With Olefins R.J. Cvetanovic Division of Chemistry, National Research Council of Canada, Ottawa, Canada Relative rates of the ground state oxygen atom, 3 0( P), reactions with a number of representative olefins and their variation with temperature were determined in this laboratory almost two decades ago . Since that time, and in particularly in the last few years, several techniques have been developed and used to measure the absolute values of the rate constants of these reactions and of their tem- 2-5) perature dependences . As a result of these develop- ments , the absolute values of the rate constants are now known with great accuracy. However, the reactions are very fast and are therefore generally not affected strongly by temperature. As a consequence, determinations of the Arrhenius parameters are experimentally very demanding. Nevertheless, temperature dependence studies reported in the literature indicate for some of these reactions two somewhat unexpected features: l) negative values of the 2a, 2b ' Arrhenius energy parameters 3c, 5) and 2) curved logarithmic Arrhenius plots"' "' ' . A definitive confirmation of these two observations, especially of the curvatures of the Arrhenius plots, requires precise determinations of the P5-1 reaction rates over relatively wide temperature intervals. The present paper will report the final results of the P d ) recent study in this laboratory by D.L. Singleton and R.J. Cvetanovic of the temperature dependence of the rates of oxygen atom reactions with several olefins. The major part of the oral presentation will be devoted to a discus- sion of the observed negative values of the Arrhenius energy parameters and the curvatures of the Arrhenius plots and of the broader implications of these observations. The background experimental information will be briefly summa- rized in the following. The determinations of the rate constants have been 2) made by the phase-shift technique . Study of the tempe- rature dependence of the rates was made possible by a consi- derable further improvement ' in the precision of the measurements, with the standard deviation reduced to about 3%. The reactions of the following olefins were studied: ethylene, propylene, 1-butene, 3-methyl-l-but ene , isobutene, cis-2-butene and t et ramethy lethylene . (The results for 2c) t et ramethy lethylene have been published and those for the other olefins have been submitted for publication ). When (least squares) linear logarithmic Arrhenius plots are imposed on the experimental points, negative values of the Arrhenius activation energy parameter are obtained for tetramethylethylene , cis-2-butene and isobutene (-0.77s -0.27 and -0.10 kcal/mole, respectively). For ethylene, P5-2 propylene, 1-butene and 3-methyl-l-but ene the Arrhenius activation energies are positive (1.68, 0.T2, 0.66 and 0.53 kcal/mole, respectively). For the olefins studied so far by the phase-shift technique the preexponent ial factors are all very similar (1.2x10 for t etramethylethylene and close to 7x10 l.mole .sec for the other olefins studied). These results are in very good agreement with the relative data from competitive experiments and, with minor excep- 3) tions, with the resonance fluorescence results Although for the temperature interval (298-^80K) of these experiments ' the observed temperature depen- dences can be approximately expressed empirically by the Arrhenius equation, the data seem to be correlated better by slightly curved logarithmic Arrhenius plots. This is particularly noticeable for the reactions of 3-methyl-l- butene, 1-butene, propylene and isobutene. A theoretical treatment based on the transition state theory shows that the curvatures can be explained by the temperature depen- dence of the entropies of activation and do not therefore 3c) require the postulate that abstraction of the allylic hydrogen from the olefins competes efficiently with the oxygen atom addition to the olefinic double bond. Great similarity of the overall behavior and of the apparent Arrhenius parameters for propylene, 1-butene and 3-methyl- 1-butene in spite of the vastly different lability of the allylic CH bonds rules out an important contribution from P5-3 the hydrogen abstraction reaction. Product analysis 2e ) study for the reactions of oxygen atoms with 1-butene and 3-methyl-l-but ene supports this conclusion. Several explanations have been suggested in the literature for the observed negative values of the Arrhe- nius activation energy parameter. There is at the moment no conclusive evidence in support of one of these sugges- tions to the exclusion of the alternative possibilities. One of the potential explanations, suggested originally in connection with the oxygen atom studies in this labo- 12a) ratory, ' postulates formation of intermediate molecular complexes ("charge transfer" complexes) between oxygen and olefins. It seems to be consistent with all the currently available experimental observations and will be discussed in some detail. 1) R.J. Cvetanovic, J. Chem. Phys . 30., x 9 (1959); i"bid 33., 1063 (i960); Can. J. Chem. 38 » l6 T8 (i960); Adv. Photochem. 1, 115 (1963). 2)a)R. Atkinson and R.J. Cvetanovic, J. Chem. Phys. 5_5_, 659 (1971); ibid 56., U32 (1972); b) S. Furuyama, R. Atkin- son, A. Colussi and R.J. Cvetanovic, Int. J. Chem. Kinet, 6, 7U1 (197*0. c) D.L. Singleton, S. Furuyama, R.J. Cvetanovic, and R.S. Irwin, J. Chem. Phys. 6_3, 1003 (1975); d) D.L. Singleton and R.J. Cvetanovic (sub- P5-4 mitted for publication); e) R.J. Cvetanovic and L.C. Doyle (to be published). 3)a)D.D. Davis, R.E. Huie, J.T. Herron, M.J. Kurylo, and W. Braun, J. Chem. Phys . £6, U868 (1972); b) D.D. Davis, R.E. Huie, and J.T. Herron, J. Chem. Phys. 59., 628 (1973); c) R.E. Huie, and J.T. Herron, J. Chem. Phys. 59., 628 (1973); c) R.E. Huie, J.T. Herron, and D.D. Davis, J. Phys. Chem., 76, 3311 (1972); d) M.J. Kurylo, Chem. Phys. Lett. ~lk , 117 (1972). k) F. Stuhl and N. Niki , J. Chem. Phys. 5_5_, 395^ (1971). 5) A. A. Westenberg and N. de Haas, Twelfth (International) Symp. Combust., Poitiers, France, 289 (1969). P5-5 Absolute Rate Of The Reaction Of 0( 3 P) With Hydrogen Sulphide P. A. Whytock and R. B. Timmons, Department of Chemistry, The Catholic University of America, Washington, D.C. 20017 and s 3 J. H. Lee, J. V. Michael, W. A. Payne and L. J. Stief, Astrochemistry Branch, NASA/Goddard Space Flight Center, Greenbelt, Maryland 20771 3 The rate constant for the reaction of 0( P) with H.S has been the subject of considerable interest (K. Schofield, J. Phys. Chem. Ref. Data 2, 25 (1973): I. R. Slagle, F. Biaocchi and D. Gutman, Astracts of the 12th Int. Symp. on Free Radicals, Laguna Beach, California (1976)), arising in part from the possible importance of this reaction in the chemistry of combustion processes, in polluted terrestrial atmospheres and in the atmosphere of Venus. Previous investigations, all employing the discharge-flow technique give room temperature rate constants which vary by more than a factor of 3. The results from the two variable temperature studies also do not agree. Gutman and co-workers obtained k = 3.3x10 exp(-4000/ 3 -1 -1 1.987 T) cm molecule ' sec " between 250 and 500 K, whereas Hollinden, Kurylo and Timmons (J. Phys. Chem. 74, 988 (1970)) reported k = 2.9 x 10" 13 exp(-1500/1.987 T) cm 3 molecule" 1 sec" 1 from 205 to 300 K. The disagreements noted above, which probably result from inac- curate assessment of the overall reaction stoichiometry, have motivated the present study using the flash photolysis-resonance fluorescence technique for the temporal measurement of [o]. This technique allows 11 -3 very low atom concentrations to be used (<10 cm ) and affords experi- mental conditions under which the first step in a reaction scheme may P6_i be isolated. Thus uncertain stoichiometric corrections do not compli- cate the analysis even in systems like the present where fast secondary reactions can contribute to atom removal. The apparatus and techniques have been described in detail pre- viously (R. B. Klemm and L. J. Stief, J. Chem. Phys. 61, 4900 (1974)). In the present study, 3-component mixtures of H-S, 0. and argon diluent 3 11 -3 were flash photolyzed at X > 136 nm to produce 0( P) (~ £ 10 cm ) in an excess of both H«S and the source compound, 0_ H 2 + PH 2 . (1) Their neglect of reaction (1) is based on the work of Norrish and Oldershaw who studied the flash photolysis of PH- and concluded that reaction (1) is important only at high temperatures. Their arguments are not compelling. Furthermore, there are qualitative results from three separate studies which are not consistent with reaction (1) being negligibly slow at room temperature. These include product analysis in 8 9 a flow system, product analysis in gamma radiolysis and an ESR-fast flow study. To our knowledge there has not been a single determina- tion of the rate constant k and, although the primary processes PH_ + hu -* PH_ + H is plausible and consistent with experimental re- sults, ' the formation of H is not established by direct detection. We have obtained information on both of these questions by monitoring Lyman-a resonance fluorescence in the vacuum UV flash photolysis of phosphine. 12 The apparatus and techniques have been described previously. Dilute mixtures of PH_ in He were flash photolyzed at \ > 105 nm to P8-1 produce H-atom concentrations of the order 10H cm~3 or less while the 13 -3 PIL reactant concentration was ~ 10 cm or larger. In order to min- imize photodecomposition of the reactant and to avoid accumulation of reaction products, reaction mixtures were flowed through the reaction cell at a rate just sufficient to replenish the sample between sub- sequent pulses of vacuum UV radiation. Under the conditions employed here, with [PH„] » [h], H-atoms detected by the pulse counting reson- ance fluorescence technique decayed exponentially with time due to re- action with PH_ and to diffusion out of the reaction viewing zone: in[H] ■ " "observed * + tn[H] o (2) where the observed pseudo-first order rate constant is given by k , . = k-OHj + k,. (3) observed 13d The diffusion term k, was determined independently using mixtures of CH, and He at the appropriate pressures and temperatures employed in this study. Typically k, was of the order of 20% of k , J Jr J d observed It was observed under all conditions employed that equation (2) was obeyed and the decay of H was first order over at least two decay lifetimes. In addition, plots of k , vs [PH„] exhibited good line- obs. 3 arity as required by equation (3) and the intercepts yielded diffusion rate constants k, which agreed with those obtained independently. Rate data were obtained at seven temperatures in the range 209 to 495 K over a wide range of experimental conditions, i.e. [PH„J, total pressure and flash energy (atom and radical concentration). An example of the range of experimental conditions employed is shown in the table for data at 3 -1 -1 298 K. Mean values of k- (units cm molecule sec ) at the other temperatures were: (10.3±0.6)xl0" 12 at 495 K, (7. 64±0.95)xl0 _12 at 420K P8-2 (5.93±0.41)xl0~ 12 at 353 K, (2.21±0.25)xlO"12 at 2 50 K, (1.77±0.19) xlO" 12 at 228 K and (1.37±0. ll)xlO" 12 at 209 K. A least squares treat- ment of the data leads to the Arrhenius expression k. = (4.52±0.39) xlO" 11 exp(-1470±50/1.987 T) cm 3 molecule" 1 sec~ 1 . Rate Data for H + PH at 298 K PH 3 Flash == j No. of k l a He torr Energy J Expt. l mTorr 10" •12 3 - -1 cm molec sec 0.8 20 20-144 5 2.99 + 0.61 1.0 50 183 1 4.37 2.0 50 20-144 9 3.43 + 0.47 4.0 100 20-183 12 3.38 ± 0.32 6.0 10,20 110-183 3 4.14 + 0.20 8.0 200,400 36-144 7 3.41 ± 0.25 9.0 15,30 81-183 4 3.59 ± 0.09 (av.) 3.45 ± 0.46 a. uncertainty in k. is the standard deviation 1. NAS/NRC Resident Research Associate 2. NAS/NRC Senior Resident Research Associate 3. Department of Chemistry, University of Essex, Colchester; on leave at Catholic University of America, Washington, D. C. 4. S. T. Ridgway, Bull. Am. Astron. Soc, 6, 376 (1974). P8-3 5. J. D. Bregman, D.F. Lester, and D.M. Rank, Astrophys. J., 202 , L55 (1975). 6. R.G. Prinn and J.S. Lewis, Bull. Am. Astron. Soc. , 7, 381 (1975); R.G. Prinn and J.S. Lewis, Science, 190 , 274 (1975). 7. R.G.W. Norrish and G.A. Oldershaw, Proc. Roy. Soc., A262, 1 (1961). 8. D.M. Wiles and C.A. Winkler, J. Phys. Chem. , 61, 620 (1957). 9. J.W. Buchanan and R.J. Hanrahan, Radiation Research, 42_, 244 (1970) ibid, 44, 305 (1970). 10. R. B. Timmons, Private communication. 11. D. Kley and K. H. Welge, Z. Naturforschg. 20a , 124 (1965). 12. R. B. Klemm and L.J. Stief, J. Chem. Phys., 61^, 4900 (1974); W. A. Payne, Jr. and L.J. Stief, J. Chem. Phys., 64, 1150 (1976). P8-4 2 2 * Energy-Dependent Cross Sections for Quenching of Li( P) and Na( P) , t John R. Barker, Shen-Maw Lin, and Ralph E. Weston, Jr. , Chemistry Department, Brookhaven National Laboratory, Upton, New York 11973 2 2 Li( P) and Na( P) atoms were produced with varying amounts of translational energy by photodissociation of Lil with light of wave- lengths from 200-230 nm or Nal with wavelengths of 215-240 nm. In the presence of a quenching gas of concentration [Q], the lifetime t of an excited metal atom is related to the bimolecular quenching rate constant ^ V 1 = T 0" 1 + k Q [( " > <» where T- is the natural lifetime of the excited atom. The single-photon time-correlation apparatus used to determine lifetimes is illustrated in Fig. 1. The light source has a pulse width of ^ 4 nsec (FWHM) and a pulse repetition rate of ^ 10-20 kHz. Each light pulse generates a start pulse for the TAC, and a stop pulse is generated if a fluorescence photon is detected by photomultiplier P2 before the next start pulse. A voltage proportional to the time elapsed between start and stop pulses is stored in the multichannel analyzer, and after many pulses have accu- mulated, the number of counts in each channel is proportional to the population of excited atoms at the corresponding time. Thus, a fluores- cence decay curve is obtained directly. Due to the finite width of the exciting light pulse relative to t , numerical deconvolution methods were used to calculate lifetimes. This research sponsored by the U. S. Energy Research and Development Administration. f Present address: Stanford Research Institute, Menlo Park, CA 94025. Ql-1 AMP | — TdTscTV - | T^ I — | DISC 2 DL STOP _ START B A 1 MCPHA TTY REC Fig. 1 - Block diagram of single-photon time-correlation apparatus. HV, high voltage supplies; L, lamp; PI and P2, photomultipliers; M, mono chroma tor; FURN, furnace; C, sample cell; LP, light pipe; F, interference filter; AMP, amplifier; DISCI and DISC2, discrimi- nators; T-S, timer-sealer; DL, delay line; TAC, time-to-amplitude converter; BA, biased amplifier; MCPHA, multichannel pulse-height analyzer; TTY, teletype printer and paper tape punch; REC, recorder. From the measured values of k n , cross sections for quenching (S ) can be calculated from the relation 00 = | S Q (g)P(g)gdg (2) where g is the relative collision speed and P(g) is the speed distribu- tion function. The latter quantity is calculated from Franck-Condon factors for transitions from populated vibrational levels of the ground state to the upper dissociative state of the alkali iodide. The Franck- Condon factors are derived from observed fluorescence spectra. Ql-2 Because Eq. (2) cannot be uniquely deconvo luted, a "phenomeno- loglcal" cross section, < S n >, is defined in terms of the velocity g corresponding to the maximum in P(g) by the expression < Y " V"W • (3) 2 2 For both Li( P) and Na( P) , the cross sections were found to be inverse- ly dependent on relative collision energy; this is particularly obvious in the case of Na when our results are combined with those obtained at lower temperatures in experiments with Na vapor. Using Eqs. (2) and (3) , we compared experimental cross sections with those calculated from two models for two-body collisions: (1) The orbiting collision model for particles subject to a long-range attractive potential V(R)=-C,/R , which leads to a cross section S (E) = w(3tt/2)(2C 6 /E) 1/3 , (4) where w is a quenching efficiency. (2) The absorbing-sphere model, in which it is assumed that every incoming trajectory reaching a capture radius R leads to quenching, with a cross section given by S_(E) = irR, 2 [l - V(R )/E] . (5) Q C C Results of this analysis are given in Table I. Ql-3 Table I. Parameters for quenching cross section expressions Orbiting collision model Absorbing-sphere model Li (2 2 P) Na(3 2 P) Li (2 2 P) Na(3 2 P) Gas C, w C, w R V(R ) R V(R ) 6 6 c c c c H 2 0.86 0.55 1.29 0.20 2.33 - 40 1.13 -125 D 2 0.86 0.52 1.29 0.15 2.22 - 43 1.26 - 50 N 2 2.00 0.52 2.95 0.28 2.18 -133 1.54 -173 CO 2.20 0.73 3.26 0.50 2.98 - 50 2.46 - 90 co 2 3.00 0.67 4.46 0.63 2.58 -200 2.52 -200 C 2 H 4 4.70 0.74 7.05 0.52 3.66 - 24 3.45° - 40 c h 10.00 1.80 15.43 2.2° 6.56 19 6.6 C - 98 c _ CO r S given by Eq. (4); C, in units of 10 erg cm . S given by Eq. (5); R in A, V(R ) in meV. B. L. Earl, R. R. Herm, S. M. Lin, and C. A. Mims, J. Chem. Phys . 56, 867 (1972). The orbiting-trajectory model gives reasonable agreement with our results for Li and Na, but in the latter case it fails to predict the large cross sections observed by other workers at low energies. The cross sections calculated from the absorbing sphere formula are in better agreement over a large energy range. The physical basis for this is the "harpooning" or curve-crossing model, in which the potential 2 2 curves for the covalent states M( P) + Q and M( S) + Q intersect a strongly attractive curve for the ionic state M Q . The quantities R and V(R ) of Eq. (5) can then be identified with the distance at which Ql-4 'the upper covalent and ionic curves cross, and the potential energy at this distance. A more detailed calculation, based on the Landau-Zener probabilities of non-adiabatic transitions, is in agreement with this. Our results fit qualitatively into the pattern of quenching and reactive cross sections obtained for other alkali metal atoms with a variety of quenching gases. qi-5 Electronic-To-Vibrational Energy Transfer Reactions: Na(3 2 P) + CCKX 1 ^ , v=0) D.S.Y. Hsu# and M.C. Lin Physical Chemistry Branch, Chemistry Division Naval Research Laboratory, Washington, D.C. 20375 Extensive experimental data exist on the quenching of excited Na atoms by various gases. Most of these studies deal with overall quench- ing cross sections. 1 To understand the mechanism by which the electron- ic energy is released during collisions, it would be extremely helpful to have detailed microscopic data. Hassler and Polanyi 2 have previously indicated that CO vibrational excitation up to v=3 was observed in the Na - * + CO reaction, employing the infrared chemiluminescence method. However, no detailed vibrational distribution was given. Using an infrared laser resonance absorption method, we have obtained the initial vibrational population distribution of CO produced from the E-"V energy transfer reaction, Na(3 2 P) + COCX 1 ^, v=0) - Na(3 2 S) + C0(X a i: + , v*8) AE = - 48.5 kcal/mole The results are found to be consistent with the curve- crossing model 3 by Bauer, Fisher and Gilmore and the impulsive model 4 by Levine and Bernstein. In order to measure the initial vibrational population distribu- tion of the CO, a stablized cw CO laser (preset at the various vibra- tional-rotational CO lines) and a Chromatix CMX-4 dye laser (tuned to one of the Na doublets) were directed colinearly along the axis of a Q2-1 low-carbon stainless steel reaction tube, which contained Na vapor and mixtures of CO in Ar. An aluminum block oven is used to maintain a constant temperature. The time-resolved absorption curves, signal- averaged over 1024 shots of the dye laser, were used to obtain the initial vibrational population distribution. The population distribu- tions obtained from using different pumped Na states and CO concentra- tions (0.5%, 1%, 2%) agree closely with one another. The fact that the distributions differ very little regardless of whether the 3 2 P! / 2 or 3P 3 / 2 state of Na is pumped indicates that both states are effectively in equilibrium, due to the large cross sections for doublet mixing,5j6 before appreciable energy transfer to CO has taken place. The lack of dependence on CO concentration is evidence that the excited CO is indeed produced in the primary step, and not by secondary reactions. Figure 1 presents the average of these four sets of experimental vibra- tional population distributions at 528°K. Our results show unequivocally that in the Na* + CO reaction, the E~»V transfer process is nonresonant; a resonant transfer would result in a narrow distribution peaking at v=8. That CO vibrational excita- tion was observed up to v=8 clearly indicates that CO is vibrationally excited up to the limit of the available electronic energy. The non- statistical distribution, as revealed by the existence of a maximum at v=2, suggest strongly that the E-'V energy transfer process occurs through an impulsive mechanism. In fact, our experimental distribution can be correlated satisfactorily with the "near-colinear" impulsive model by Levine and Bernstein, 4 using (AE>/hV = 2.83 as the average vibrational quanta transfered in their Poisson distribution. Our experimental distribution also agrees surprisingly well with the Q2-2 prediction of the curve-crossing model by Fisher and Smith. 3 It can be concluded that the Na -35 " + CO E-V energy transfer reaction occurs impulsively, with 18.2 kcal/mole of its 48.5 kcal/mole of electronic energy channelling into vibrational and the remaining into translation and/or rotation. O - <2Xpf • -- curve- crossing model A— impulsive? model #NRC/NRL Resident Research Associate (Sept 1975-present) Q2-3 References 1. C. Th. J. Alkemade and P.J. Th. Zeeger, in "Spectro- Chemical Meth- ods of Analysis," ed. by J.D. Winefordner, Wiley Interscience, New York, 1971. This review summarizes many earlier work on the total quenching cross sections of excited alkali atoms. 2. J.C. Hassler and J.C. Polanyi, Disc. Faraday Soc. 44 (1967) 182. 3. (a) E. Bauer, E.R. Fisher and F.R. Gilmore, J. Chem. Phys. 51 (1969) 4173; (b) E.R. Fisher and G.K. Smith, Appl. Opt. 10 (1971) 1803. 4. R.D. Levine and R.B. Bernstein, Chem. Phys. Letters 15 (1972) 1. 5. (a) J. Pitre, F.S.C. and L. Krause, Can. J. Phys. 46(1968) 125; (b) 45 (1967) 2671. 6. H.L. Chen and-S. Fried, IEEE J. Quantum Electron. QE-11 (1975) 669, Q2-4 Energy Transfer In The Collision Of Metastable Excited Ar 3 P 2 Atoms With Ground State H 2 S Atoms . By Penelope B Monkhouse , Kyle D Bayes* and Michael A A Clyne , Department of Chemistry, Queen Mary College, London El 4NS f U.K. The interaction of metastable excited argon ( 3 P 2 o) atoms with ground state H 2 S atoms presents a very fundamental problem in energy transfer. Conservation of energy and momentum show that H should rec- eive almost the full energy defect (1.34 eV) . From a study of the shape and width of the H atom Lyman-a (121.6 nm) line emission profile, info- rmation was derived regarding the nature of the collision process, what kinetic energy is actually present in the H atom, and whether any thermalization occurs. In addition, we have compared our results for the Ar 3 P 2 + H system with the line profiles of Lyman-a lines obtained from a microwave discharge plasma. Metastable argon atoms were produced by flowing purified argon through a low-power (220 V, 5 mA) hollow cathode discharge which has been described . Total pressures were near 1.3 Torr. H 2 was diluted to the required strength (^ 10%) in argon and passed with argon carrier gas through a microwave discharge. The reagent flow was mixed coaxially with the metastable Ar flow downstream of the hollow-cathode discharge, in front of a LiF observation window. The Lyman-a 121.6 nm line, H 2p 2 P - 2 S, was examined using a 1.5 m Fastie-Ebert vacuum monochromator (Jarrell-Ash) in the fourth order of a 1200 line mm -1 grating blazed at 500 nm and using a slit width of 30 ym. The Lyman-a line from the Ar 3 P 2 o + H interaction was found to * On sabbatical leave from University of California at Los Angeles. Q3-1 be broadened (HBW = 0.013 nm) , and to possess a profile close to a box- car (rectangular) shape. The edges of this line are steep, as expected from the instrumental profile (Fig.l) . The CK 1 ^ - 2 P 3/ ) emission '2 11 line at 138.0 nm from the reaction Ar 3 Pg 2 + Cl 2 was chosen to deter- mine the slit function; it was unreversed, near Gaussian shape, and thermal (HBW = 0.003 nm) . The top of the Lyman-a profile was effect- ively flat, with only limited self reversal at relatively high atom particle densities in the order of 10 13 cm -3 . The Lyman-a and chlorine line profiles were then digitised manual- ly and processed computationally through a deconvolution program which 2 utilized a Fourier transform routine . The resulting spectrum confirm- ed our preliminary estimations of an essentially rectangular profile. The results will be discussed in terms of the model proposed by 3 Biondi and Connor in studies of non-thermal line shapes of Ne prod- uced by electron -ion recombination. Deconvolution of the observed line shape has led to the conclusion that all (^ 90%) of the excess energy (1.34 eV for Ar 3 P2) in the Ar* + H interaction is found as kinetic energy of the H atom, allowing for the very small fraction that is transferred to the Ar *S atom. This indicates that negligible thermalization had occurred under our conditions, as expected. The overall rate constant for the process Ar 3 Po a + H 2 S -> Ar 1 S + H(2s,2p) o 4 has been measured using the Ar 3 P2 + Kr reaction as a standard . The value obtained was (2.8 ± 1.5) x lO -10 cm 3 molecule -1 s _1 , i.e. close to unit collisional efficiency at 298 K. Q3-2 Lyman-a profiles from microwave plasmas operating in pure argon and helium have also been investigated. In the case of argon, a comp- osite profile consisting of a 0.013 nm wide rectangle surmounted by a 'slightly warm' (^ 500 K) Doppler spike. These results suggest that a major excitation process for Lyman-a in the microwave discharge in argon is by collision of H with metastable argon atoms. The helium plasmas showed much narrower H atom line profiles tending towards the limit of the instrumental line width at low [hJ . Thus it would seem likely that either the predominant excitation of H(2p) in helium plasmas did not involve metastables or that H(2s) was formed and then partly thermalized by collisions with He, before quenching to the 2p state could occur. REFERENCES 1. W.C. Richardson and D.W. Setser, J.Chem.Phys. , (1973) ,58, 1809. 2. Unpublished data of M. Heaven, Queen Mary College. 3. R.T. Connor and M.A. Biondi, Phys. Rev. (1965) , 140A , 778. 4. L.A. Gundel, D.W. Setser, W.S. Nip, M.A. A. Clyne and J. A. Coxon , J.Chem.Phys., (1976), in course of publication. Q3-3 HBW=0.003 nm — _ -J HBW=0.013 nm— Fig.lA. Emission profiles of (a) the Cl ( 4 P and (b) H (2p 2 P- 2 S) at 121.6 nm. fc - 2 P: ) at 138.0 nm The H atom line is a composite of 4 experimental scans. I ', O O ,' 1 t o," 1 o Dt' \> p> ■. •Q" ;o» ■Of, •' o •&', | l i MX \ 1 8 4 AACprtO 4 8 Fig. IB. Idealized profiles of the Lyman-a line (----) and the Cl line (. . . .) , with preliminary manual esti- mation of the slope of the true (unfolded) Lyman-a profile (M). Q3-4 3 3 Quenching Rate Constants for Ar( P_) , Kr( P^) and 3 Xe( P_) by Halogen-containing Molecules, and * Branching Ratios for XeF and KrF Formation. J. E. Velazco , J. H. Kolts and D. W. Setser Department of Chemistry Kansas State University Manhattan, Kansas 66506 1 2 The discovery ' of rare gas-halogen excimer formation from the reaction of rare gas metastable atoms with halogen or halogen-containing molecules, has led to the development of a new class of ultraviolet 3 lasers. However, scaling-up of these prototype devices faces some practical as well as theoretical problems: halogen donors must be found that give excimer formation with high efficiency, do not absorb at the laser frequency and are reasonably easy to handle. Modeling of these systems requires the knowledge of the rate constants (and branching ratios) for the relevant steps. In this work total 3 3 3 (thermal) quenching rate constants for Ar( P~), Kr( P ? ), and Xe( P ? ) with a variety of fluorine donor reagents are reported; also, the branching ratios for the formation of KrF* and XeF* were measured. Emission from a second excimer state of XeF, XeCl, KrCl and KrF, which 2 correlated with the ( P , ) state of the rare gas ion will be reported, as will the observation of vibrational relaxation of the excimers in the 1-10 torr pressure range. Total quenching rate constants (k n ) were measured using a flowing 4 -1 after-glow apparatus with a linear flow velocity of ^80 msec . The Q4-1 Ar metastable atoms were produced by passing a flow of purified Ar through a hollow cathode discharge. Total metastable concentrations ^5 x 10 cc , were measured at the first observation point, with 3 3 3 3 Ar( P 2 ) and Ar( P ) in a ^6:1 ratio. The Kr( P ) or Xe( P 2 ) metastables 3 were produced by energy transfer reactions from Ar( P~ 9 ) by adding small quantities (^0.3%) of Kr or Xe to the Ar flow before the discharge. The decay of the metastables along the tubular reactor as a function of quencher concentration was followed by atomic absorption using Pen Ray source lamps: ^ the Ar 811.5 nm, Kr 811.2 nm, Xe 881.9 nm lines are used. Fluorine donors were diluted in Ar and stored in a passivated stainless vessel connected to the inlet part of the flow reactor by 1/4" ss tubing. Less reactive reagents were stored and metered through glass lines. The quenching rate constants, k_, from these experiments are listed in Table I. The reproducibility of k_ for the highly corrosive F-donors is + 20% due mainly to difficulties in purification of samples. For chlorine and fluorides that do not attack glass, the reproducibility is + 10%. A comparison technique was developed to measure the branching ratio for excimer formation, k^/k , which eliminates the need of a direct and absolute measurement of \,y In the absence of quenching or non-radiative processes, the excimer emission intensity is given by (1). ^ W M * ][RX] (1) Hence, if k^— is assigned for a reference reaction, and if experiments » are done for the same [M*] , we can obtain other k^ from (2) , Q4-2 C /[R ' X1 "" = "« g/ro ' (2) by simply comparing the relative emission intensities for known RX and RX' concentrations. In order to perform these intensity comparisons, a system was built that consists of two gas handling sections (one grease-free but of glass and the other stainless steel) and the flow reactor section. Pressures and flow rates were monitored 3 with a pressure transducer. Emission from the reaction of Kr( P„) 3 or Xe( P_) (produced a described above) with added halogen donor, was detected by a computer controlled 0.75 m Jarrell-Ash monochromator equipped with a cooled photomultiplier tube and SSR photon counter. For low pressure spectra, the intensity (counts/sec) was recorded o at 1 A intervals; whereas, for high pressure spectra shorter intervals were used. The data were stored in magnetic tapes and subsequently corrected for spectral response. Integrated emission intensities were obtained by simple addition of the counts in each interval. Molecular chlorine was selected as the reference donor (RX) because the absolute values of k Tr __ and k _, could be estimated, and also KrCl XeCl because Cl„ can be handled in a relatively simple and reliable way. Columns 4 and 5 of Table 1 list the rate constants for excimer formation relative to CI- (see eq. (2)). The absolute value of k^ from the Kr* + CI reaction was calculated by using the absolute rate constant for the formation of the CI * 257 nm continum from Ar* + CI -11 -1 -1 (1.7 x 10 ' ' cc molec sec ). Comparing the emission intensities of the CI * 257 nm continuum and the KrCl excimer emission gives (3). Q4-3 ^ = ^ [Cl^ [Ar( 3 P Qt2 )] (3) X KC1* ScrCl [Cl 2 ] [Kr( 3 P 2 )] The ratio of Ar and Kr metastable concentrations was determined and the result for equal [Cl ] was k = 52 x 10 ml molec sec 2. KrCl This value gives a branching ratio of 0.7 for excimer formation from Kr* + Cl_. This value however, may have an uncertainty of as much as + 50% because the intensity comparisons required to determine k * are ci 2 at the extreme ends of our calibrated spectral response curves. Furthermore, it is likely that some KrCl* emission extends into the v.u.v. and, therefore, is not included in I T , _,*. Careful recalibra- KrCl tion experiments now under way will decrease the uncertainty of this branching ratio. The lack of reliable oscillator strengths for transi- tions in the 6s-6p manifold precludes a determination of Ic. r1 by the method outlined above. However, evidence will be presented that supports the assignment of unity branching ratio for excimer formation from Xe* + CI . Extensive vibrational relaxation of the excimers occur in the 1-40 Torr pressure range as shown in Figure 1 for XeF from Xe* + ONF. Similar results were obtained for XeCl*, KrCl* and KrF*. Since significant relaxation is evident even at < 5 Torr, simple considera- tion of the gas-kinetic collision frequency at these pressure (y 4.3 x 10 at 5 Torr) suggest a lifetime > 50 ns for the rare gas-halide excimers. Figure 1 also shows the XeF emission at ^ 265 nm, 2 originating from an upper excimer excited state ( tt ,_) that correlated 2 5 with the ( P. .J) state of the rate gas ion. The state responsible for Q4-4 2 the main excimer emission correlates with the P.,„ ion state. Emission from the upper state was also identified for XeCl*, KrF* and KrCl*. The separation between the maximum of the primary emission and that due to this new state is 9422 cm" 1 (1.17 eV) for XeF, 9867 cm" 1 (1.22 eV) for XeCl, 5050 cm" 1 (0.63 eVO for KrF and 5380 cm" 1 (0.67 eV) for KrCl. The dynamics of these reactions as deduced from the emission spectra of the excimers will be briefly discussed in terms of analogy to alkali metal atom reactions. References : * This work was supported by U.S. Army Research Office (DAHC04-75-60018) and by U.S. Energy Research and Development Administration (E (11-1) -2807) 1. a. J. E. Velazco and D. W. Setser 4th Conference on Chemical and Molecular Lasers Oct. 21-23, 1974; IEEE J. Quant. Electron. QE-11 , 708 (1975). b. J. E. Velazco and D. W. Setser, J. Chem. Phys. 62, 1990 (1975). 2. M. F. Golde and B. A. Thrush, Chem. Phys. Lett. 29, 486 (1974). 3. a. K. Searles and G. A. Hart, Appl. Phys. Lett. 27^, 243 (1975). b. J. J. Ewing and C. A. Brau, Appl. Phys. Lett. 27^, 350 (1975). c. E. R. Ault, R. S. Bradfor and M. L. Bhaumik, Appl. Phys. Lett. 27, 413 (1975). d. G. C. Tisone, A. K. Hays, and J. M. Hoffman, Opt. Commen. 15 , 188 (1975). 4. L. G. Piper, J. E. Velazco and D. W. Setser, J. Chem. Phys. 59_, 3323 (1973). 5. C. A. Brau and J. J. Ewing, J. Chem. Phys. 63, 4640 (1975). Q4-5 TABLE I. Rate Constant Summary Reagent . , in ll 3 . -1 -1, k (10 cm molec sec ) Ar( J P 2 ) Kr( J P 2 ) Xe( J P 2 ) Relative "k rCl Sot "x eCl CI, 71 73 72 100 c 100 c F ? 75 72 Br 2 65 C1F 3 IC1 61 0F 2 57 53 CF OF 43 42 SF 6 16 - SeF 6 71 TeF, 6 58 S0 2 F 2 42 NF 3 14 16 N 2 F 4 31 33 NOF 36 47 N0C1 48 HC1 37 HBr 57 HI 75 CF 3 I 75 100 100 60 62 b 33 b 50 57 82 80 47 35 25 28 %o ^0 65 63 9 7 14 22 27 30 <5 C 40; (43) 51 56 61 52 BrF, 0;(33)' 0;(23)' a. Cl_ was used as the reference reaction for assigning \ ify . b. Both chloride and fluoride excimer emission was observed. c. NO (y) and (3) emissions were the main product. This value is based on an estimate of the KrF emission intensity. d. NO (y) bands were also observed. This value includes the contribu- tion of the Y~bands. e. MBr rather than MF was the product. Q4-6 Xe + FNO v'-o-° NO(Y) 2 3 4 5 6 La-*"** mjw 1 Torr 12 Torr 36 Torr 200 250 300 350 400 nm Q4-7 V-V Energy Transfer in H p - Additive Gas Mixtures using a Stimulated Raman Excitation Technique Richard G. Miller* and J. K. Hancock Chemistry Division Naval Research Laboratory, Washington, D. C. 20375 Introduction We have measured gas phase vibrational energy transfer rates for the following inelastic collision process, H 2 (v=l) +A(v=0) -» H p (v=0) + A (v > 0) + Ae (cm -1 ) where A was various heteronuclear diatomics and triatomics. A measurable population in the H p (v=l) state was produced using a stimulated Raman excitation technique. Ducuing, Joffrin, and Coffinet demonstrated the utility of this technique in measuring V-T 2 deactivation rates in H p gas. Matsui, Resler, and Bauer have more recently measured V-V energy transfer rates in H p -C0 and D p -C0 gas mixtures using a stimulated Raman - IR fluorescence technique. We have extended this excitation scheme to our system utilizing two cells which has enabled us to observe V-V energy transfer in H p -A gas mixtures at pressures much less than an atmosphere. Experimental Our experimental apparatus is depicted in the accompanying diagram. The excitation source was a doubled Nd : YAG laser (international Laser Systems, Inc. ) with a TEM^ mode output of ~ 85 mjoules/pulse and a 15 nsec pulse duration. The 532. Onm beam was Q5-1 bent and focused into the center of Cell 1 which contained 50 psi of H p . Stimulated Stokes-shifted beams (at 683. 2 nm) and anti-Stokes- shifted beams (at Vj5« 6 nm) subsequently propagated from Cell 1 in both directions coaxially with the laser beam. The Stokes-shifted beam and the laser beam are then re- focused into the center of the second cell resulting in the excitation of the H p (v=l) state. Energy transfer rates from H p (v=l) to the additive molecules A were obtained by monitoring the . IR fluorescence from the IR-active vibrational modes of the A species. The A species which we have 12 15 studied include HC1, DC1, HBr, DBr, COp, NpO, CO, CO, and NO. The IR-fluorescence from these species was monitored with an InSb photoconductive detector. The temporal profile of the fluorescence was processed with a Biomation 8100 transient 'digitizer interfaced with a Nicolet IO72 signal averager. The respective V-V and V-T rates were such that the desired V-V rate was contained in the fall time of the fluorescence from the additive molecule. Results The Hp-HCl gas mixtures were studied extensively. Pressures of 85 to 692 torr and 1. 06 to 73 torr of H and HC1, respectively, were -1 -1 -1 irradiated. The (P T ) values ranged from 20 to 310 sec " torr . l o W + VTn The composite V-V and V-T energy transfer rate constant (k ) for the Hp to HC1 transfer was evaluated to be 1510 ± 210 sec" . -1 _ ,. 3,4 . ..< . TT . „n . W + VT torr . Two other groups •" have obtained H p to HC1 k values by exciting a mixture of H p and HC1 containing a trace of 5 -1 -1 HF with an HF laser. Bott's^ value was 1250 ± 150 sec " torr " and Q5-2 k -1 -1 Pirkle and Cool reported a value of 1730 ± 250 sec torr . We have also made some measurements on the HF-H p -HCl system using HF laser excitation and obtained (P t )" values that were in good agreement with our results obtained with the stimulated Raman pumping scheme. The complete results shown in the accompanying table will be compared further with available results of others. The importance of V-V transfer to the overtone and combination bands of the A species studied will be discussed. Reasonable correlations of our H p -H(D)X results with V-V results on various homonuclear diatomic - H(D)X gas mixtures can be shown. Q5-3 References * NRC Postdoctoral Fellow. 1. J. Ducuing, C. Joffrin, and J. P. Coffinet, Opt. Coram., 2, 6 (1970). 2. H. Matsui, E. L. Resler,Jr. , and S. H. Bauer, J. Chem. Phys. , 63, 4171(1975). 3. J. F. Bott, preprint. k. R. J. Pirkle and T. A. Cool, preprint. Q5-4 Vibrational Energy Transfer Rates in H 2 (v=l) + Additive M -> H 2 (v=0) + M(v=l,2) + AE Mixtures at 296°K vv + vt X-m Additive Molecule M Energy_ 2 Defect (cm ) t -1 (sec -1 N torr ) Probability HC1 1274 -1508(v=2) 1510 + 210 8.62 x 10~ 5 DC1 2069 34(v=2) 689 + 30 3.95 10" 5 HBr 1604 -865 (v=2) 224 + 18 1.26 10" 5 DBr 2321 527(v=2) 212 + 36 1.19 x 10" 5 co 2 1811(001) 551(021) 444(101) 497 + 30 2.01 x 10' 5 N 2 1936(001) 795(021) 679(101) 462 + 14 1.82 x 10" 5 12 co 2017 -100 (v=2) 12.3 + 9.5 6.02 x 10" 7 13 co 2069 -6(v=2) 9.7 + 1.9 NO( 2 n 1/2 ) 2284 436 (v=2) 42 + 17 2.47 x 10~ 6 NO( 2 n 3/2 ) 2159 311 (v=2) Q5-5 ScvJ CJ a/ DC O ^n J— sx V CO CJ o J r— LU r- mmm H~ r»- Q Q5-6 Experiments Concerning the Laser Enhanced Reaction Between t and NO Kin-Kwok Hui and Terri II A. Cool , School of Applied and Engineering Physics, Cornell University, Ithaca, New York 14853 Several recent studies of the laser-enhanced reaction of nitric oxide with vibrationa I ly excited ozone NO + 3 + + N0*( 2 B ) + 2 (la) and NO + * -+ N0 2 t ( 2 A ] ) + 2 (lb) (1-5) have been reported. In these studies the visible chemi I uminescence 2 from NO *( B.) has been observed to be significantly enhanced in re- sponse to the laser-induced vibrational excitation of ozone, i.e., 3 + hv -v 3 + (2) The initial experiments of Gordon and Lin revealed a lag in the 2 rise of fluorescence from NO *( B.) produced in reaction (la). It was suggested that this lag was associated with the v, •*■ v_ coupling of vibrational modes in 0,. However, later experiments showed that the v, ■+ v„ coup I ing time was an order of magnitude longer than that neces- sary to explain the fluorescence lag reported by Gordon and Lin. Other puzzling aspects of the experimental measurements on reac- (4 7) (4) tions (la) and (lb) have been reported. ' Kurylo e_t aj_. , measured the rate sum for reactions (la) and (lb) as a function of temperature and observed a complicated temperature dependence which could not be ex- (4) pressed with a single choice of Arrhenius parameters. Kurylo et a I . have suggested that an understanding of this temperature dependence requires the resolution of the influences of intermode vibrational transfer in , of vibrational Rl-1 deactivation of , and the relative temperature dependences of the two reaction channels (la) and (lb). We have performed experimental measurements with an apparatus de- signed to yield new information concerning reactions (la) and (lb). A Q-switched CO laser operated on the P(30) transition of the 9.5 micron band is employed for laser excitation of . The laser pro- duces I mJ pulses of I ysec (FWHM) pulse width and 170 Hz repetition rate. A continuous flow reactor has been employed which was carefully designed for more rapid and uniform gas mixing than had been accomplish- ed in previous work. The apparatus was designed for operation over the range 150-400 K, and permits observations of vibrational fluorescences from and N0 ? in addition to the electronic fluorescence previously observed. Measurements have been performed of the temporal variations in fluorescence from the NO *( B ) and NO „•( A.) states as a function of gas composition over a wide range of parameters. 2 The fluorescence from the NO *( B ) molecules does not show a de- layed rise, of the type reported by Gordon and Lin, under any experi- mental conditions we have explored to date. These include both high and low partial pressures of with and without large amounts of various diluents. Conditions for which all three modes of 0, should be equilibrated and conditions for which only the v. and v modes partici- pate were examined. We believe the delayed fluorescence rise reported by Gordon and Lin was not of kinetic origin, but rather was caused by 2 incomplete gas mixing. The dependence of the decay of NO *( B ) on NO pressure gives a value of (7.5 I 0.8) x 10 cm mole sec for the sum of rate constants for reactions (la) and (lb) at a temperature of 308 t 3 K in good agreement with recent measurements. ' Observations of vibrational fluorescence from the NO '( A.) state at 3.7 microns show a rise time determined by collisional quenching of + 2 NO T ( A.) and a characteristic decay rate determined by the combined rate of reactions (la) and (lb). A rate constant for these combined reactions of (8.6 t 0.8) x 10 cm mole sec at 308 i 3 K is obtained from the NO "N A.) fluorescence decay, in agreement with the value 2 found from the relaxation of NO *( B ). The rise time data for N0 9 ^C"A.) fluorescence give values of (I.I t 0.2) x 10 cm mole sec for the quenching of NO + ( 2 A ) by NO and (3.1 + 0.8) x I0 M cm 3 mole~' -I +2 sec for the quenching of NO ' ( A.) by ? . These values are in good ( R ) agreement with the results of previous measurements. 2 Results for the AC/DC fluorescence ratios for both the NO *( B ) and NO + ( A ) states indicate a surprisingly low yield of NO ^( A ) by reaction (lb) which suggests that products of the laser enhanced reac- tion are predominantly excited in the trans lat iona I and rotational degrees of freedom rather than in vibration. Measurements currently in progress devoted toward an understanding of reactions (la) and (lb) will be discussed. References 1. R. J. Gordon and M. C. Lin, Chem. Phys. Lett. 22_, 262 (1973). 2. M. J. Kurylo, W. Braun, A. Kaldor, S. M. Freund, and R. P. Wayne, J. Photochem. 3_, 71 (1974/1975). 3. W. Braun, M. J. Kurylo, A. Kaldor and R. P. Wayne, J. Chem. Phys. 6J_, 461 (1974). 4. M. J. Kurylo, W. Braun, C. N. Xuan, and A. Kaldor, J. Chem. Phys. 62, 2065 (1975). 5. R. J. Gordon and M. C. Lin, J. Chem. Phys. 64, 1058 (1976). 6. K. K. Hui, D. I. Rosen, and T. A. Cool, Chem. Phys. Lett. 32_, 141 (1975). 7. A. E. Redpath and M. Menzinger, J. Chem. Phys. 62, 1987 (1975). 8. M. F. Golde and F. Kaufman, Chem. Phys. Lett. 29_, 480 (1974). JU-4 ■k Infrared Laser Enhanced Reactions: Chemistry of NO(v=l) with O3 J. C. Stephenson and S. M. Freund National Bureau of Standards Washington, D.C. 20234 A variety of recent experiments demonstrates that bimolecular chemical reaction rates may be enhanced by optically exciting the vibrational levels of the reactant molecules with an infrared laser. The molecular system studied in greatest detail is the reaction of ozone with nitric oxide to give nitrogen dioxide and molecular oxygen, a process of importance in the atmosphere. The thermal reaction rate constant has been thoroughly investigated. The enhancement of the reaction rate following excitation of the Oo (001) vibrational level by a COo laser was first reported by Gordon and Lin, and then studied in 9—/ greater detail by Kurylo et al, and Freund and Stephenson. The present work continues the investigation of the laser-enhanced reaction between NO and 0-j. Because it is the ozone O2-O bond which breaks during the reaction, it was expected and then proven that vibrational excitation of the 0„ asymmetric stretching motion increases the rate. For the reactions k B 2 3 + NO 5 N0 2 ( B 1 ) + 2 k' + B 2 3 + NO + N0 2 ( B 1 ) + 2 3 + NO + N0 2 ( 2 A 1 ) + 2 k' 3 + + NO + A N0 2 ( 2 A 1 ) + 2 4. where Oo represents vibrationally excited ozone, the room temperature R2-1 rate constant enhancement factors k' B /k B = 4.1+2 and k'./k. =17.1 2 +4.3 were determined. Since the NO bond does not break in the reaction, it was not obvious that there would be any increase in rate for the analogous reactions k' N0 + + 3 + 1 N0 2 ( 2 B 1 ) + 2 (!') NO + 3 i N0 2 ( 2 B 1 ) + 2 (1) k' N0 + + 3 + 1 N0 2 ( 2 A 1 ) + 2 (2') NO + 3 i N0 2 ( 2 A 1 ) + 2 (2) where NO denotes NO(v=l) . However, in the experiment described below, we excited the NO(v=l) state with a CO laser, and found k' /k- =4.7 2.6 +1.4 and (k' 2 + k M _ )/k? — 18 * A necessary part of the experi- ment was the determination of the rate at which NO is vibrationally deactivated by 0~ . k N0-0 NO + 2 + NO + 2 for which k N0 _ Q = 920 + 80 sec -1 Torr" 1 . All of the experimental measurements were performed on mixtures of gases flowing in a pyrex cell. The reactants were monitored by means of calibrated flow meters and were admitted to the cell through con- centric glass tubing which terminated in a "shower head" configuration. The laser excitation beam traversed the center of the cell perpendicular to the gas flow and visible and infrared fluorescence was viewed in a direction perpendicular to both the flow direction and the laser beam. The distance between the fluorescence viewing window and the mixing jet R2-2 could be easily adjusted by sliding the glass tubing in or out through an O-ring connector to the desired position. Polished BaF 2 windows were fastened to the cell with epoxy at the Brewster angle laser entrance and exit positions and at the central fluorescence viewing port. The pressure drop AP along the cell was small (less than 5 percent of the total pressure for conditions under which data was taken) . The commercially obtained reactant gases were used directly from the cylinders without further purification. The ozone was pro- duced in a commercial generator which gave a maximum mole fraction of 0- in 2 of 0.045. The N molecules were pumped by radiation from a liquid- nitrogen-cooled cw CO laser which produced about 3 watts of power on the fundamental (TEM OQ ) mode of the 1884 cm -1 line (P(13) transition of the 9+8 12 C 16 band) . The beam was focused by a 10 cm focal length BaF 2 lens and was chopped at the focal point by a servo-controlled chopping wheel which produced a square-wave modulated beam (100 percent modula- tion) at a frequency of about 100 Hz. It was then collimated by a second BaF„ lens and directed into the cell by three front-surface- reflecting mirrors. The beam diameter at the entrance to the reaction cell was approximately the same size as the tubing through which it subsequently passed. Since the center of this laser line is 732 MHz from the transition in nitric oxide, N0(v=0, J=3/2) 2 n 3/2 -»■ N0(v=l, J=5/2) 2 ^ 3 / 2 ' the molecular transition must be tuned into coincidence by applying a magnetic field. The polarization of the laser was such that E (laser) was parallel to B (magnet) . The magnetic field required to maximize R2-3 the absorption was 760 gauss; this field was measured to be homogeneous to + 10 gauss over the region of the cell near the fluorescence port. Measurements of the laser power absorbed in a short cell containing NO placed at the center of the magnetic coils gave an absorption coefficient at room temperature of 0.0209 cm Torr (standard devia- tion of 0.0003 cm Torr -1 ) for absorption of the 1884 cm laser line in the pressure region P < 4 Torr. The laser frequency was not stabilized by locking to the top of the gain profile or to an external cavity. The reproducibility of the laser absorption coefficient measurements showed that any drift in laser frequency was small com- pared to the Doppler width of the NO transition. Visible fluorescence from the N0 2 ( B-,) product, hereafter desig- ■k nated NO2 , was detected with a photomultiplier tube cooled to -30°C. The N0(v=l) fluorescence was monitored with a large area photovoltaic InSb detector. The analog signals which developed across appropriate load resistors were amplified by low noise amplifiers (DC to 1 MHz band pass) , processed by a transient recorder which digitized the —8 signal in real time (minimum 10 sec/channel) , and stored in a multi- channel analyzer. The electronics were triggered by the signal from an InSb detector which monitored the laser light transmitted through the cell. In a typical AI/Iq experiment, 40,000 cycles of the square wave signal were summed. R2-4 REFERENCES 1. R. J. Gordon and M. C. Lin, Chem. Phys. Lett. 22.. 262 (1973). 2. M. J. Kurylo, W. Braun, A. Kaldor, S. M. Freund, and R. P. Wayne, J. Photochem. 3, 71 (1974). 3. M. J. Kurylo, W. Braun, C. N. Xuan, and A. Kaldor, J. Chem. Phys. 62, 2065 (1975). 4. S. M. Freund and J. C. Stephenson, "Laser Enhanced Chemical Reaction between 0^(001) and NO," to be published, Chem. Phys. Lett R2-5 Kinetic Energy - And Internal State Dependence Of The NO + O3 + NO2* + 2 Reaction. Anthony E. Redpath, Michael Menzinger Department of Chemistry, University of Toronto, Ontario . The numerous degrees of freedom of nuclear and electronic motion of both reactants and products make the title reaction rich in dynamical detail. Associated with the problem of Energy Consumption are the questions of influence of (a) reactant kinetic energy E, (b) of the NO( 2 II, ,_, . ) spin-orbit states and (c) of 03(^1^2^3) vibrational excitation, on the dynamics and rates of the different product channels. (d) The effects of NO vib-rotational and O3 rotational energy are believed to be small and we have not investigated them. The problem of Energy Disposal has its spicy sides since (e) at least two electronic N0 2 states ( Aj groundstate and an excited state, presumably B ^) are populated, but primary vibrational state distributions are unavailable due to the complexity of the * electronic N0 2 chemiluminescence spectrum and to extensive vib-rotational quenching in the A^ groundstate. (f) * 1 1 + Furthermore, the 9 ( A , I ) states are energetically z. g g accessible, but Gauthier and Snelling have shown' ' them to be produced with negligible probability in the thermal reaction. (g) Given the long radiative lifetimes of N0 2 *(t r * ~ 10~ 4 sec) and N0 2 +(t r + ~ 10~ 2 sec?) it has to be R3-1 kept in mind that bulk experiments are usually conducted at pressures where secondary collisions dominate the kinetics. In our "single collision" beam/gas experiments, a He or H 2 seeded supersonic NO beam issues from a heatable nozzle (300-650 K) , is skimmed, chopped and shot through an O., filled scattering chamber (variable temperature 150-300 K to populate 0~ vibration) and monitored. Using lock-in techniques chemiluminescence is detected by (1) a bare photomultiplier (2) a monochromator/photomultiplier combination (3) a PbS infrared detector plus filter for electronic N0 2 * emission (1-3 um) and (4) PbS detector and another filter for vibrational NCU* (3-4 um) emission. The beam energy is varied via (a) the seeding ratio and (b) the nozzle temperature. Beams are velocity analyzed by time- of-f light. It is found that NO beams from a heated (600 K) nozzle yield ca. 4x more visible CL than beams of the same nominal velocity issuing from a room temperature nozzle. It is shown that the former have a higher population of the NO ( n_,_) spin-orbit component than the latter. We conclude that NO ( II . ) + 0^ yields primarily electronically excited * 9 2 N0 2 / while NO(^H, /2 ) + 3 leads to groundstate N0 2 ( A 1 ) . The only (NO'O^ collision complex which correlates with N0 2 *( 2 B 1 ) + 2 ( 3 O and N0 2 = M 2 A 1 ) + 2 ( 3 Z~) rather than with 2 (-'-A -, -*-£ ) d) has Cr symmetry: NO approaches along the o ^ \„ bisecting plane and attacks the central rather than R3-2 an end-standing O-atom. This "bisecting C5" reaction pathway is expected to be greatly facilitated by the 0^ (2 ) bending mode v 2 , which furthers bond formation between the end-standing oxygen: A cyclic 0_ isomer of very similar energy as the "classical" bent 0^ groundstate, that has been predicted on theoretical grounds^ supports this view. However, competing channels in which end-standing may be transferred cannot be ruled out on the basis of the circumstantial evidence presented here. We observe enhancement of the CL crossection through thermally populating 3 internal modes. The excitation functions for production of visible and for IR (1-3 ym) electronic emission are almost identical: till ~ 3 kcal above threshold they are well represented by CL (E) = C(E - E Q )n E > E Q = E < E with C = 7.1 x 10~ 5 A 2 , E Q = 3.0 ± .3 kcal/mole and n = 2.3 ± .3. This yields an activation energy in good (5) agreement with measured values. The CI spectra, recorded at E - 15 ± 1.5 and 12 ± 1.2 kcal/mole and at thermal energies show a strong blue shift with increasing collision energy: the short wavelength R3-3 cutoff invariably agrees with the total available energy, demonstrating that in a small (unknown) fraction of * collisions the total energy is funneled into internal N0 2 and subsequently emitted as light. The energy dependence of the vibrational IR emission is markedly distinct from electronic emission: no threshold is discernible even at the lowest collision energy E . = 1.1 kcal/mole, and the signal rises less rapidly as E increases. In the spirit of information theory a c _(E) or rather k CL (E) = va(E) = p(E')a)(E) is decomposed ^ ' into a calculable (a priori) product phosespace density p(E*) and the dynamical "averaged state to state transition probability" 00(E). It comes somewhat as a surprise that the time -honoured hard sphere/line-of- (8) center model for reaction crossections gives an excellent representation of w(E) « (1 - E /E) . For the penta-atomic title reaction, the phosespace factor dominates a(E). We express our thanks to Dr. Akimichi Yokozeki for the information theoretical analysis, and to Professor Tucker Carrington for the correlation diagrams and for numerous chemil luminating discussions in the course of this work. Financial support stems from the National Research Council of Canada and is gratefully acknowledged. R3-4 References (1) Gauthier, M. , and Snelling, D.R., Chem. Phys. Lett., 20, 178 (1973). (2) (a) Gordon, R.J., and Lin, M.C., Chem. Phys. Lett., 22, 262 (1973). (b) Gordon, R.J., and Lin, M.C., J. Chem. Phys., 64 , 1058 (1976). (3) Wright, J.S., Can. J. Chem., 5_1, 139 (1973). (4) Shih, S., Buenker, R.J., and Peyerimhoff, S.D., Chem. Phys. Lett., 2_8, 463 (1974). (5) Clough, P.N., and Thrush, B.A., Trans. Far. Soc, 63_, 915 (1967). (6) Bernstein, R.B., Levine, R.D.,: "Role of Energy in Reactive Molecular Scattering: An Information Theoretic Approach" in Advances in Atomic and Molecular Physics, Vol. II. Ed. Bates, D.R., Academic Press, 1975 (7) Pruett, J.G., Grabiner, F.R., Brooks, P.R., J. Chem. Phys. , 63, (75) 1173. (8) Levine, R.D., Bernstein, R.B., Molecular Reaction Dynamics, Oxford U.P., 1974. R3-5 The Effect of Infrared Laser Excitation on Reaction Dynamics: + C^H. 1 " and + OCS + 2 4 Ronald G. Manning , Walter Braun, and Michael J. Kurylo Physical Chemistry Division National Bureau of Standards Washington, DC 20234 Introduction A variety of experiments have been reported in the literature recently regarding the energetics and dynamics of atom exchange processes (A + BC ■> AB + C) . These studies have utilized the infrared arrested fluorescence technique to analyze the energy distribution in the diatomic product and thermal , microwave , chemical , and infrared laser excitation methods to investigate the effect of internal energy in the diatomic reactant on the reaction rate. Calculations for these systems show that, depending on the nature of the reaction hypersurface, reactant vibrational energy can be very effective or virtually ineffective in influencing the rate of chemical reaction . A number of more complex systems have also been investigated using either IR laser excitation of 7 8 9 a reactant molecule ' ' or chemical or flash photolytic production of 10 11 12 an excited transient (i.e. 0H(u>0) ' and CN(u>0)) . Most of these systems exhibited rate constant enhancements of factors of three to four for reaction of the vibrationally excited species when compared with the thermal process. In the molecular system studied in greatest detail (0 + NO) , it was found that the temperature coefficient remained essentially identical to that for the Boltzman reaction . These results R4-1 indicate that while there appears to be little coupling between the reactant vibrational energy and the reaction coordinate a small but significant effect can be observed on the overall reaction rate. In this study we have investigated the effect of vibrational 3 energy on the reactions of 0( P) atoms with OCS and C^H, : k l + OCS * CO + SO (1) ,kt o + ocs i co + so (i ) k 2 + C H. 4 CH n + CHO (2) 2 4 3 + C 2 H, i CH + CHO (2 + ) Experimental The experimental apparatus for the kinetic measurements (depicted) schematically in Fig. 1) consists of a flash photolysis resonance fluorescence (FPRF) apparatus modified to allow passage of IR laser radiation. The basic FPRF assembly consists of a reaction cell of arbitrary geometry to which are attached a flash lamp, resonance lamp, and fluorescence detecting photomultiplier on perpendicular axes. The sensitivity and precision of the FPRF method enable one to discern changes in experimental conditions which cause perturbation of the measured decay rate by several percent . The CO laser shown in the diagram can be operated cw over numerous lines in the 9-10y region with a power level of 3-6 watts. Consequently, under conditions which minimize vibrational deactivation, steady state concentrations of vibrationally excited species can be achieved such that ([S]u>0/[S]u = 0) _> 0.05. By comparing the atom decay rate for laser on vs. laser off conditions, any significant change in the rate constant due to reactant vibrational R4-2 excitation can be detected. The limiting value of the laser enhancement that can be discerned depends ultimately on two factors: 1) the equilibrium concentration of vibrationally excited reactant that can be experimentally realized (i.e. absorption of laser flux vs. deactiva- tion losses) and 2) the magnitude of the activation energy for the thermal reaction. This determines the extent of a purely thermal (heating) effect. Results t The excited substrate molecules (S ) can be formed by two methods: direct IR absorption and V-V transfer from another molecule which is pumped by the laser. Such V-V transfer is generally much faster than V-T deactivation in the system and this latter scheme is f often successful in achieving high enough concentrations of S to t 13 facilitate measuring the smallest possible change in k /k . In f these experiments CH F has been used as a transfer agent because of its strong absorption of the P20 line, 9.6y band CO laser radiation. Under these conditions we observed little or no effect of vibra- tional energy in C~H, on the rate of reaction with atoms. Because of a somewhat higher activation energy, it was considerably more difficult to measure an effect in the + OCS reaction. Within the limits of detectability, all observations in this system could be attributed to heating effects. A factor of 1.5 increase in the rate constant for + C~H. over that for + C H. and a factor of 2 4 2 4 f 3 for + OCS over + OCS would have been detectable in these experiments. The results seem to indicate that there is little or no coupling of vibrational energy to the reaction coordinate leading to R4-3 activated complexes in these two reaction systems. Experiments are in progress to determine the effect of vibrational energy on other reaction systems. References I. K. G. Anlauf, P. J. Kuntz, D. H. Maylotte, P. D. Pacey, and J. C. Polanyi, Disc. Faraday Soc. 44, 183 (1967). 2a. L. J. Kirsch and J. C. Polanyi, J. Chem. Phys. 5_7, 4498 (1972). b. J. H. Birely, J. V. V. Kasper, F. Hai, and L. A. Darnton, Chem. Phys. Letters 31, 220 (1975). 3a. R. F. Heidner and J. V. V. Kasper, Chem. Phys. Letters 15 , 179 (1972). b. L. B. Sims, L. R. Dosser, and P. S. Wilson, Chem. Phys. Letters 32_, 150 (1975). 4. D. J. Douglas, J; C. Polanyi, and J. J. Sloan, J. Chem. Phys. 59., 6679 (1973). 5a. T. J. Odiorne, P. R. Brooks, and J. V. V. Kasper, J. Chem. Phys. 55, 1980 (1971). b. D. Arnoldi, K. Kaufman, and J. Wolf rum, Phys. Rev. Letters 34, 1597 (1975). c. Z. Karny, B. Katz, and A. Szoke, Chem. Phys. Letters _35, 100 (1975). d. S. R. Leone, R. G. Macdonald and C. B. Moore, J. Chem. Phys. 63, 4735 (1975). 6. J. C. Polanyi, Accounts of Chemical Research 5_, 161 (1972). 7. M. J. Kurylo, W. Braun, C. Nguyen Xuan, and A. Kaldor, J. Chem. Phys. 62, 2065 (1975); 63, 1042 (1975). 8. R. J. Gordon and M. C. Lin, J. Chem. Phys. 64, 1058 (1976). 9. E. N. Chesnokov, V. P. Strunin, N. K. Serdyuk, and V. N. Panfilov, Reaction Kinetics and Catalysis Letters 3^, 131 (1975). 10. S. D. Worley, R. N. Coltharp, and A. E. Potter, J. Phys. Chem. 76, 1511 (1972). II. G. E. Streit and H. S. Johnston, J. Chem. Phys. 64, 95 (1976). R4-4 12. V. H. Schacke, K. J. Schmatjko, and J. Wolf rum, Ber. Buns. Phys Chem. 77, 248 (1973). 13a. M. J. Kurylo, W. Braun, and A. Kaldor, Chem. Phys. Letters 27_, 249 (1974). b. W. Braun, M. J. Kurylo, and A. Kaldor, Chem. Phys. Letters 28, 440 (1974). c. D. I. Rosen and T. A. Cool, J. Chem. Phys. 62, 466 (1975). R4-5 a: O O k CO LjJ LjZ < i/) K (J =j z l/) u 1- 3 2 w Q_ > UJ _J UJ •H I O co e CO CO H 4J O rt •u o 43 cO a • a a. (3 co o CO /-n -u CO P< CO H OJ 4J Mh Ph -H Pn CJ CU v-* X 43 CU 4-1 CU CJ 4J M-t c c O CU CO O 4-1 g co a CO CU CO u u a> M O ^ CO 3 •H r-H ^1 T3 M-i CU CO CJ CU CO •H O rH 4-" C cO CO PS 6 C H cu o 43 CO !-i CJ CU O en u m CU M tot) •rl R4-6 Reaction of Flash Photo lytically Produced CN(X Xtv) Radicals With 0( 3 P) Atoms K. J. Schmatjko and J. Wolf rum MPI fur Stromungsforschung, D-3400 Gottingen, West Germany The reaction (1) CN(X 2 21 + , v) + 0( 3 P) — D> CO (X*2 + , v) + N( 4 S, 2 D) was studied using a flash-photolysis arrangement combined with a discharge -flow reactor. CN-radicals were produced by flash-photo- lysis of C N diluted in He and their concentrations monitored by visible absorption spectroscopy. The flash discharge (17 kV, 0. 2 uF) 3 is switched by a thyratron with a repetition rate up to 5 Hz. 0( P) atoms present in excess with respect to the CN radicals are continu- ously generated in a microwave discharge. The concentrations of the CO(X "X , v) molecules formed in reaction (1) were obtained from time-resolved infrared laser absorption spectroscopy (s. Fig. 1). PZT k* a IN, P'JMP ,f Jt CRATING -J VI v \j- lr* -He, N,.C0 in MICROWAVE DISCMARGE (S CO -LASER fT M *'° 2 r— CjN 2 PUMP MULTIPASS MIRROR — SYSTEM Ij^^zjl, y - ryB—^ Fig. 1 Schematic of the experimental apparatus for infrared laser resonance absorption -gain measurements The CO -Laser could be tuned by a grating on single rotational lines of the vibrational transitions of CO(v) between (1,0) and (26,25). For R5-1 transitions below (3,2) the laser output had to be etabilized continuously between measurements. Tomaintain isothermal conditions and to avoid the rapid exchange of vibrational energy during the formation of the CO(v) molecules very low concentrations have to be applied. This was achieved by using an internal multi-pass mirror system. The adjustable mirrors allowed up to 60 traversals of the analysis laser beam along the flow tube. From the absorption or gain signals the absolute concentration of CO(v) formed in the reaction was evaluated starting with the highest vibrational transition of CO(v) where [CO(v+l)] = was found. No absorption or gain was observed for lines in the (26,25) to the (15,14) bands. From the measured concentration -time profiles (s. Fig. 2a) the distribution of CO(v) molecules formed in reaction (1) could be obtained directly. The shape of the distribution measured at early times (s. Fig. 2b) shows only small changes as the reaction time proceeds. The rate constant k = (1 , 1 + 0, 3) 10 cm mol s at 295 K could be obtained from the CO(v, t) profiles. No systematic variation was found using different CO(v) levels. "CO -3 10 ' 2 : 9 • 5-:o" - \ i 10" 5-10'° — \/\ i \ o I.I.I REACTION TIME A 8 12 VIBRATIONAL LEVEL OF CO Fig. 2a Formation of CO(v) in Reaction (1) Fig. 2b Distribution of CO(v) formed in Reaction (1) R5-2 The observed energy distribution (s. Fig. 2b) shows two different parts which can be understand by considering the correlation diagram of reaction (1). As can be seen in Fig. 3 several pathways exist: one re- 1 + 4 action path direct to CO(X 2. ) and N( S) atoms and two reaction paths 2 1 + to N( D) atoms and CO(X ^ ) either directly or via the stable inter- 2 mediate NCO(X T|~). To simulate the dynamics of reaction (1) three- dimensional classical trajectory calculations were carried out using two different empirical adiabatic potential energy surfaces. The tra- 4 jectory calculations show that on the path to N( S) atoms the reaction energy is predominantly channeled into vibrational excitation of the CC molecule. o On the path to N( D) via NCO(X 2 TT) the formation of CO(v=0) predominates and the reaction ener- gy is mainly converted into electro- nic excitation of the metastable 2 N( D) atom. If one summes up the measured absolute population of the CO vibrational levels according to the shapes of the distribution pre- kJ/mol 200 n* n i's) --200 400 Fig. 3 Correlation diagramm for Reaction (1) dieted by the trajectory calculations one obtains 0( P) + CN -^-t>CO(v) + N( D) (0. 85 + 0. 05) -k— C>CO(v) + N( 4 S) (0. 15 + 0. 05) Observation of the electron state of the nitrogen atoms produced in re- action (3) provides an independent way to obtain information on the im- portance of the two pathways (la) and (lb). Observation of the nitrogen atoms produced in the reaction was made by time-resolved resonance absorption in the vacuum UV (s. Fig 4). R3V3 ,', ' 1 1 flash tueYI i MICROWAVE DISCHARGE LAMP VUV MONOCHROMATOH Fig. 5 shows the measured time- resolved absorption signals from 2 the N( D) atoms produced in (1). 4 While the concentration of N( S) remains nearly constant during the observation time a decay of 2 the N( D) concentration due to 2 Fig. 4 Detection of nitrogen atoms quenching of the metastable N( D) formed in Reaction (1) by atomg by the componen ts of the resonance absorption reaction mixture is observed. g 0.1 ■D I 0.2 N( 2 D) 200 jjs/div Fig. 5 Oscillogram of the vacuum ultraviolett absorption of N( D) formed in Reaction (1) 2 From the measured absorption signals a ratio of 5: 1 for N( D) versus 4 N( S) formation in reaction (1) in good agreement with the result from the CO(v) distribution is obtained. R5M Vibrational Photochemistry: The Relaxation of HCl(v=l) and DCl(v=l) by Bromine Atoms R.D.H. Brown, I.W.M. Smith and S.W.J. Van der Merwe, Department of Physical Chemistry, University Chemical Laboratories, Cambridge CB2 1EP, England. The atom-transfer reaction Br + HCl(v) *- HBr + CI (l) is establishing itself as the prototype of endoergic reactions whose rates are selectively enhanced by vibrational excitation of the bond that is broken in the reaction. Recently the relative rates of (l) for 1 < v < k have been determined -using a 'chemiluminescence depletion' method and the total rates for reactive plus inelastic processes, i.e. Br + HCl(v) — -+~ Br + HCl(v'^v), (2) have been measured for v=l,2 in laser induced vibrational fluorescence 2 experiments at room temperature. This paper reports measurements by the second technique of (k + k ) for HCl(v=l) and DCl(v =1). The experimental results are interpreted in the light of the results of Monte Carlo trajectory calculations. Experimental Method and Results A description of our general experimental method has been given 3 elsewhere. Gas samples were prepared in a flow-discharge system and some of the HC1 contained in this mixture was excited to v=l with the output from a pulsed HC1 chemical laser. The decay of the vibrational fluorescence was measured with different concentrations R6-1 of Br atoms present. These were prepared by means of the reactions + Br — *- BrO + Br (3) + BrO — •► Op + Br (h) that convert atoms quantitatively to Br. The atomic oxygen was produced by partial dissociation of in a microwave discharge and its concentration was determined, prior to addition of Br , by titration with NO . By measuring the energy absorbed from the laser it was shown that 1 - 1.5$ of the HC1 was excited to v=l. The degree of excitation was purposely kept thj.s low to avoid problems that might arise through fast V-V exchange of vibrational energy followed by rapid removal of HC1 from levels with v > 1. Table 1 summarises the results obtained in our experiments. Table 1. Summary of Observed Rate Constants Excited molecule T K 5a 5b k 5b 10" 13 3 . . -1 -1 cm molecule s in -13 3 10 cm molecule s HCl(v=l) 210 2.8 2.6 | 2.8 2a HCl(v=l) 295 5.6 k.Q 2.7 2b HCl(v=l) 371 9.5 1-1 I 2.6 2c DCl(v=l) 295 9.k 9.U Chemical reaction between Br and HCl(v=l) is so endoergic that it cannot remove HCl(v=l) at low temperatures. There remain, however, two energy transfer processes that may contribute to the overall rate: R6-2 -rBr*( 2 P , ) + HCl(v=0) (5a) Br( P , ) + HCl(v=l)^ ' J/ ^ Br(^P 3/2 ) + HCl(v=0) (5b) Process (5a) is 2.3 kcal/mole endoergic but it can provide a route for loss of HCl(v=l) in our experiments as Br* is rapidly quenched by o . The rate constant k._ can be calculated from k _ for the 2 5a -5a k reverse process. Assuming k to be independent of temperature, pa the values of k._, listed in the last column of Table 1 are obtained. 5b Our value of k,_, at 295 K is almost twice that determined in 5b other recent investigations. In these Br atoms were formed by partial dissociation of Br in a discharge. Because Br deactivates 2a HCl(v=l) efficiently, the acceleration brought about by the atoms was only **H0% as against a threefold increase in our experiments. Nevertheless it appears that an as yet unidentified systematic error may cause the difference between these two sets of results. Theoretical Results and Discussion To assist with the interpretation of the experimental data a three-dimensional quasiclassical trajectory study is now being made of the dynamics of Br + HCl(v) and Br + DCl(v) collisions. These allow one to examine the electronically adiabatic routes for removal of HCl(v). The computations are being carried out using procedures that have been described previously with a single parameter LEPS potential. Calculations have been carried out on collisions between Br and HCl(l < v < k) with translational and rotation energy distributions corresponding to temperatures of 333.3 and 1000 K. The results are summarised in Table 2. R6-3 Table 2. Calculated Rate Constants 1000 k 3 2 1 T k l + k 2 k l 100k 7 V -m" 11 3 ~ , "I "I nn -ll 3 " . "I "I . " K 10 cm molecule s 10 cm molecule s k.. +k 8.6 T.l 2.7 0.1 5.6 U.5 1.5 0.05 66. T 63.U 56.5 50 333.3 U 3 2 1 U.6 2.3 0.3, 2.1 1.2 0.09 U6.T 51.5 28 The relative values of the rate constants for different v are in reasonable agreement with the experimental results of Douglas et al . For v > 2 processes (l) and (2) have roughly equal rates with relaxation becoming more effective than reaction as v and T are lowered. Energy transfer occurs in collisions where at some stage in the trajectory ^ r HCl^e HCl^ > ^ r HBr^ r e HBr^' i,e * reaction 'nearly' takes place. The calculations provide strong evidence that process (5b) does not occur in electronically adiabatic collisions. The absolute magni- tude of the observed rate constant, and the fact that k,_, for DC1 is 5b greater than for HC1, indicate that relaxation proceeds via the kind 6 of curve- crossing process that has been proposed by Nikitin. The 7 rate of this process should be approximately proportional to v. Consequently three detailed mechanisms may make roughly equal con- R6-4 tributions to the removal of HCl(v=2) by Br atoms, although this proposal is at variance with the conclusions reached by Douglas, Polanyi and Sloan. References (a) D.J. Douglas, J.C. Polanyi and J.J. Sloan, J. Chem. Phys., 22, 6679 (1973) and Chem. Phys., 13, 15 (1976). 2 (a) S.R. Leone, R.G. Macdonald and C.B. Moore, J. Chem. Phys., 63 , U735 (1975); (b) D. Arnoldi, K. Kaufman and J. Wolfrum, Phys. Rev. Letters, 3k, 1597 (1975); (c) Z. Karny and B. Katz, Chem. Phys. Letters, 38, 382 (1976). 3 (a) R.D.H. Brown, G.P. Glass and I.W.M. Smith, Chem. Phys. Letters, 32, 517 (1975) and (b) J.C.S. Faraday II, _J_1, 1963 (1975). S.R. Leone and F.J. Wodarczyk, J. Chem. Phys., 60, 31 ^ (197*0. 5 (a) I.W.M. Smibh and P.M. Wood, Mol. Phys., 2£, kkl (1973; 6 (b) I.W.M. Smith, J.C.S. Faraday II, Jl, 1970 (1975). E.E. Nikitin and S. Ya. Umanski, Faraday Disc. Chem. Soc, 53 , 7 (1972). ' I.W.M. Smith, Acct. Chem. Res., £, in press (1976). R6-5 Vibrational Relaxation of HF(v= 1, 2, 3) In The Presence Of H 2 , N 2 , And C0 2 J. F. Bott The Aerospace Corporation P. O. Box 92957 Los Angeles, California 90009 The study of vibrational relaxation has greatly increased in the last few years because of the interest in chemical lasers. In addition, these chemical lasers have provided the means for performing the ex- perimental studies. Until recently, experiments have been largely con- fined to relaxation rate measurements for molecules in the first vibra- tional levels because of the difficulties involved in the controlled production of the higher levels so that specific upper level processes could be isolated. The techniques of direct laser pumping of the second vibrational level, low-pressure combustion with spectroscopic diagnos- tics, reactive flows, in medium-pressure flow tubes, and laser-induced fluorescence by sequential absorption have been used to study upper vibrational level deactivation. Laser-induced fluorescence was used in this study to measure the relaxation rates of HF(v= 1, 2, 3) in H ? , N-,, 2 , HC1, and C0 2 . A laser-induced fluorescence experiment in which the sequential absorption of photons produces the higher vibrational levels is relatively easy to perform, and the interpretation of the data is straightforward. The experiments can be performed with a large ratio of the collision partner to HF, thereby reducing the HF-HF V-V deactivation processes to negligible contributions. A knowledge of the upper level deactivation rates is important for R7-1 understanding and modeling the performance of the HF chemical laser since the F + H ? pumping reaction directly populates levels v = 1,2, and 3. The values for HF(v)-CO, deactivation obtained in this experiment can be compared with the theoretical calculations of Dillon and Stephen- son and the measured values obtained in more complicated experiments. The apparatus for this experiment included a HF TEA laser, a fluorescence cell, and several detectors. The TEA laser operated on SF/ and H-, and produced laser pulses of a few millijoules with time durations of approximately 100 to 300 nsec. The laser was operated at a repetition rate of ~ 3 Hz without substantial degradation of the fluores- cence signal. A concave mirror and a partially transmitting flat formed the laser cavity. No spectral line selection line selection was necessary, and the laser output contained various lines of the 3—" 2, 2— 1, and 1— "0 vibrational bands. By appropriate adjustments in the pressures of SF, and H ? , the laser was made to pump the HF to vibrational levels v= 1,2, and 3 or just v= 1 and 2. The pressure tuning determined whether the 1~*0, 2— -1, and 3— *2 laser transitions occurred in the proper sequence or overlapped sufficiently to produce upper level pumping of the HF in the fluorescence cell. The fluorescence from HF(v= 3) and HF(v= 2) were monitored with a RCA C-31034 GaAs photomultiplier and a RCA 7102 (S-l) photomulti- plier, respectively. An InSb detector was used to monitor the fluores- cence from HF(v= 1). Spectral filters were used to pass only the desired 3 — 0, 2 — 0, or 1—0 fluorescence from HF. The signals from the detector or photomultiplier were recorded with a Biomation 805 transient recorder and transferred to a Nicolet Model 1072 signal R7-2 averager in which 32 to 512 experiments could be stored and averaged before being displayed on an X-Y recorder. The relaxation rates of HF(v= 3) were measured in gas mixtures containing 0. 02% HF, to 99. 98% H 2 , and a balance of Ar at a total pressure of 20 Torr. The slope of the rates plotted versus H ? concen- tration was interpreted as the sum of the V-V transfer rates from HF(v = 3) to H 2 and the V-R, T deactivation rate of HF(v = 3) by U^. The measured relaxation rates extrapolated to zero H ? concentration in- dicated a value that is compatible with the HF(v= 3)-HF deactivation rate obtained by Osgood and coworkers. Data were also obtained for HF(v= 1) and HF(v= 2). Similar data were obtained for the deactivation of HF(v) by N 2 , C> 2 , HC1, and CO r Although this technique is not applicable to the study of all mole- cules, it has many advantages for those diatomic molecules capable of laser action on several vibrational, rotational transitions. This tech- nique can most likely be extended to HF(v >3) by the use of such fuels as HBr or HI in the pumping laser. The rates of vibrational energy transfer from the upper vibrational levels of HF to N ? , ? , HC1, and CO-, increase faster with v than the predicted rates for harmonic oscillators having no energy defect. Part of this increase can be attributed to the decrease in the energy defect, AE(v), with v so that less energy must be absorbed by the rotational and translational degrees of freedom for the transfer of a quantum of energy. However, the dependence of the rates on v can be explained only qualitatively in terms of the anharmonicity of HF and the resulting energy defect dependence on v. R7-3 2 2 Quenching of NO(B n ) , Produced by the Reaction of N( D) with NO G. Black , R. L. Sharpless, and T. G. Slanger Stanford Research Institute, Menlo Park, California 94025 2 Very little information is available on the quenching of N0(B II ) , largely because the state has not been observed in NO fluorescence. * However, NO(B il ) can be made and its quenching studied in many other ways. We have used the reaction 2 2 N( D) + NO - N + N0(B n ) (1) 2 2 as the source of N0(B II ), generating the N( D) by photodissociation of r 2 NO (Ref. 2). This reaction produces most of the N0(B II ) in the v'=0 level . 2 The apparatus for these measurements has been fully described. 2 Previously, we had shown that over the region 1100-1500 A, the largest 2 o o quantum yield of N( D) from NO occurs between 1150 A and 1200 A. Most o of the measurements were therefore made at 1175 A. A Wratten No. 18A o filter was used to isolate the 3000-4000 A region in which the strongest (0, v ) NO 3 bands occur. For these measurements, a cooled EMI 9635AM was used (dark count at -20 C, ~ 20 counts/sec). The experiments were 2 performed in a very similar way to those described previously. Figure 1 shows plots obtained for N = and 17.2 torr. In all cases, single exponential decays were observed covering at least one decade change in intensity. Figure 1 shows two effects of adding N --the decay rate increases and the intercept decreases. The increasing decay rate with quenching gas addition has previous- . 1 -14 4-6 2 ly been used to determine rate coefficients for quenching N( D) . The decay rates shown in Figure 1 give a rate coefficient of 1.96 x 10 3,-1 -1 * , . c ,2 x t cm molec sec for the quenching of N( D) by N , in good agreement R8-1 10,000 with the two most recently 6,7 reported values of (1.5 + 0.1) x 10~ 14 and (1.85 + 0.15) x lo" 14 3,-1 -1 cm molec sec Our main purpose here is to determine rate coefficients for 2 quenching N0(B H ) from the decrease in the intercept with quenching gas addition (Figure 1) . [Recently, we used similar raeasur- ments to study the quenching of 3 _ Se (B S )]. Absorption by the added gas was always negligible over the ~ 10 cm from the LiF window to the center of the region viewed by the photomultiplier . Therefore, for plots like those shown in Figure 1, we can write 100 . u l—S-1 1 1 L_ 0.1 0.2 0.3 0.4 0!"» 0.6 0.7 TIME (msec) Fig. 1. Quenching of N( D) and N0(B 2 n r ) v i =0 . Ar = 3 torr; N 2 = 28 mtorr. 20 min experiments at 1175 A. o I Tk [Q] 1 + Tk Ar + Tk rN o 0] Ar L J N 2 2 (2) where I is the intercept at time zero (the time the lamp is pulsed) o with no quenching gas, I is the intercept with quenching gas addition [Q] , T is the radiative lifetime of N0(BTi ) (3.0 x 10 sec), and 2 k , k , and k N Q are the rate coefficients for quenching N0(B II ) , by Q, Ar, and ISLO, respectively. Figure 1 gives I /I = 1.91 + 0.09 for N = 17.2 torr. This point o — 2 and values of I /I at several other nitrogen pressures are shown plotted o as suggested by Equation (2) in Figure 2. The straight line predicted by Equation (2) is indeed observed. The slope S of this Stern-Volmer plot can be written R8-2 Tk 2.0 I /I 1.5 1.0 _ r _ r i i r i/..i__i 2 4 li 3 10 12 14 Hi IE [l\! 2 | PRESSURE (to,r) Fig. 2. Stem-Volmer plot for N, 2 quenching of N0(B II ) , . Ar = 3 torr; N 2 = 28 mtorr. S = N. 1 + Tk [Ar] + Tk [NO 2 (3) To determine k N , we must know 2 k and k M n . The rate coeffi- Ar N 2 cient k was determined with Ar sufficient accuracy for substitu- tion in Equation (3) by making measurements similar to those described above in argon-N mixtures containing NO = 15 mtorr and neglecting quenching by NO. Below, we describe how k. T _ was ' N 2 determined. The quenching of the intercept by CO was studied at several NO pressures from 40 to 760 mtorr. For these experiments, Equation (3) was rearranged: -1 Tk k [Ar] V N 2 01 Ar L J 2 + "~ : + — : (4) CO. CO. CO. where S , the reciprocal slope of the Stern-Volmer plots, is the half- quenching pressure. The data for the CO experiments are shown plotted as suggested by Equation (4) in Figure 3. The slope of this graph gives kjj o/^CO = **•"* — 0»5» an< * tne intercept, after correcting for the small argon contribution, gives a "true" half-quenching pressure (the pressure at which the rate of quenching by CO equals the rate of radiation) for -12 3 -1 CO of 1.47 + 0.2 torr. Hence, k CQ = (7.0 + 1.0) x 10 cm molec -1 -6 2 -11 3 sec (since T = 3.0 x 10 sec) and k N = (3.2 + 0.8) x 10 cm -1 -1 2 molec sec . Table 1 shows true half-quenching pressures and corres- ponding rate coefficients for several gases, together with the values R8-3 10 of Campbell and Thrush for CO , H , and NO, and the value of Melton and Klemperer for the 2 quenching of N0(B II ) by NO. 1 0.2 0.3 0.4 5 0.6 0.7 0.8 0.9 [N 2 0] PRESSURE (torr) Fig. 3. C0 2 quenching data. The reciprocal of the slope of Stern- Volmer graphs vs [NO], Ar = 3 torr. References 1. A. B. Callear and I.W.M. Smith, Trans. Faraday Soc. 59, 1720 (1963). 2. G. Black, R. L. Sharpless, T. G. Slanger, and D. C. Lorents, J. Chem. Phys. 62, 4266 (1975). 3. K. H. Becker and K. H. Welge, Z. Naturforsch. , 20a, 442 (1965). 4. G. Black, ■ T. G. Slanger, G. A. St. John, and R. A. Young, J. Chem. Phys. 51, 116 (1969). 5. T. G. Slanger, B. J. Wood, and G. Black, J. Geophys . Res. 76, 8430 (1971). 6. T. G. Slanger and G. Black, J. Chem. Phys. 64, xxxx (1976). 7. D. Husain, S. K. Mitra, and A. N. Young, J. Chem. Soc. Farad. II 70, 1721 (1974). 8. G. Black, R. L. Sharpless, and T. G. Slanger, J. Chem. Phys. 64, xxxx (1976). 9. M. Jeunehomme and A.B.F. Duncan, J. Chem. Phys. 41, 1692 (1964). 10. I. M. Campbell and B. A. Thrush, J. Quant. Spectrosc. Radiat. Transfer 8, 1571 (1968). 11. L. A. Melton and W. Klemperer, J. Chem. Phys. 5_9, 1099 (1973). R8-4 Table 1 Quenching of NO(B II ) , „ at 298 K H 6 v r v'=0 Half-Quenching Pressures and Rate Coefficients Rate Coefficient Other values [10,11] 3 -1-1 3 -1 -1 (cm molec sec ) (cm molec sec ) (2.9 + 0.5) x 10" 13 (2.9 + 0.4) x 10~ 13 (6.1 + 1.1) x 10" 13 (7.0 + 1.0) x 10" 12 1.0 x 10" 11 (1.03 + 0.26) x 10" 2.0 x lo" (2.5 + 0.6) x lo" 11 (2.9 + 0.3) x lo" 11 Ha If -Quenching Pressure Gas (torr) He 36 + 6 Ar 36 + 5 N 2 17 + 3 co 2 1.47 + 0.2 H 2 1.0 + 0.25 CH, 4 0.41 + 0.09 CO 0.36 + 0.04 NO 2 NO °2 0.064 + 0.013 NH 3 0.038 + 0.009 C 2 H 4 0.032 + 0.008 (3.2 + 0.8) x lo" 11 4.5 x lo" 11 (1.4 + 0.1) x 10 -10 (1.6 + 0.3) x lo" 10 (2.7 + 0.6) x 10" 10 (3.2 + 0.8) x 10" 10 R8-5 AUTHOR INDEX (Underscoring indicates author presentation) Abramowitz, S. B2 Adamson, A. W. Gl Addington, J. W. G4 Akimoto, H. H8, K8 Albery, W. J. II Albrecht, A. C. E8 Allen, E. R. K7 Anastasi, C. P7 Anderson, J. G. Fl Anderson, L. G. 01 Archer, M. D. II Asprey, L. B. L2 Atkinson, G. H. 02_ Atkinson, R. Nl Ausloos, P. M3 Austin, E. R. H7 Back, M. H. C3 Back, R. A. C3, H6 Baldwin, G. C. C8 Balzani, V. G6 Barker, J. R. Ql Baronavski, A. E5 Bass, A. M. M2 Bauer, S. H. C4 Bayes, K. D. Q3 Bergmark, W. R. 12 Bergstrbm, S. J3 Bersohn, R. H5 Black, G. F2, R8 Bogan, D. J. B4 Bolletta, F. G6 Bondybey, V. E. LI Bott, J. F. R7 Bottenheim, J. Kl, K3, K4 Bowman, W. D. A4 Bowmen, W. R. II Braun, W. B2 , R4 Breitenbach, L. N2 Bremer, N. L3 Brown, B. J. L3 Brown, R. D. H. R6 Brune, D. C. J4 Brus, L. E. LI Bufalini, J. J. K5 Calvert, J. G. Kl, K3 Campion, A. L7 Carter, W. P. L. N4 Chameides, W. L. F4 Chien, K. R. C4 Cicerone, R. J. F4 Clark, H. M. C8 Clyne, M. A. A. E3, Q4 Cody, R. J. E4 Coleman, W. F. L2 Cool, T. A. Rl Corkum, R. C3 Crutzen, P. J. F6 Cu, A. D3 Curran, A. H. E3 Gandini, A. H6 Cvetanovic, R. J. F8, P5 Gay, B. W. K5 Damon , E . K3 Gillespie, H. M. M5 Darnall, K. R. Nl, N4 Gratzel, M. 15 Davidson, J. A. F5 Graham, R. E. H4 Davis, D. D. F3, N3 Grattan, D. W. P2 De Armond, M. K. G3 Gregory, T. A. 06 De Mare, G. R. 07 Groth, W. Dl Demas, J. N. A4, G4 Gutierrez, A. R. Gl DeMore, W. B. M4 Gutman , D . H4 Despres, A. L5 Hackett, P. A. H6 Dismukes, G. C. L6 Hager, S. L. L6 Dodge, M. C. K9 Hakala, D. C8 Donnelly, V. M. E2 Hancock, J. K. Q5 Donovan, R. J. M5_ Hansen, I. P4 Dwyer, P. D5 Hanst, P. L. K5 El-Manguch, M. 04 Harris, E. W. A4, G4 El-Sayed, M. A. L7 Harvey , R . B . F4 Eyler, J. R. E6 Heicklen, J. K6 Fajer, J. ii Henry, M. S. D4 Felder, W. 11 Hoffman, M. Z. D4, 13 Ferreira, M. I. C. 11 Hoggard, P. E. £5 Finkenberg, E. G7 Hogo, H. K9 Fisher, P. G7 HSinghaus, K. P4 Fontijn, A. Bl Holms trbm, B. J3 Fournier, J. H9 Hoshino, M. K8 Frazier, G. F. A5 Howard , C . J . F5 Freund, S. M. C6, R2 Hsu, D.S.Y. 91 Gafney, H. D. 91 Hui, K-K Rl Hunziker, H. E. PI Lampe, F. W. H7 Inoue, G. K8 Larson, H. B. J3 Ishiwata, T. H8 Ledford, A. E. M2 Jacox, M. E. L9 Lee, E. K. C. C2, El Jaffe, S. MA Lee, J. H. P6, P8 Janda, K. D5 Lee, S. A. 03 Janson, T. R. li Lee , S . -J . H5 Jones, G. 12 Le j eune , V . L5 Jones, L. H. L2 Lesclaux, R. P3_ Juris, A. G6 Levine , S. Z. Kl Kaduk, B. B3 Lewis , R . B . L2 Katz, J. J. IA Lewis, R. S. C2 Kaufman , F . E2 Lichtin, N. N. 13 Kaufman, K. M5 Lin, D. P. L6 Kawasaki, M. H5 Lin, M. C. H2, Q2 Keifer, W. N3 Lin, S-M Ql Khe, P. V. P3 Lindsay, J. M. A5 Killus, J. P. K9 Lip sky, S. 06 Kirk, A. D. G5 Litvak, H. H5 Klapthor, R. D5 Lloyd, A. C. Nl, NA Klose, J. Z. C7 Lyman , J . L . C5 Klosterman, N. E. A2 Maestri, M. G6 Knight, A. E. W. 01 Maker, P. N2 Koeppe, M. D5 Manning, R. G. RA Kolts, J. H. QA Martin, T. W. G2 Krauss, M. B2 Masanet , J. H9 Kurylo, M. J. B2, RA Mataga, N. J2 Kutschke, K. 0. P2 McAfee, J. F6 Lahmani, F. D2 McBride, R. P. AA, GA McDermid, I. S. E3 Peterson, S. G4 McDonald, J. R. E5 Philen, D. L. 12 McDowell, R. S. L2 Phillips, D. 05 Menzinger, M. R3 Pitts, J. N. Nl, N4 Merrill, J. G3 Plummer, B. F. D8 Michael, J. V. 11> P6, P8 Porter, G. B. G5 Migirdicyan, E. L5 Rebbert, R. E. M3 Miller, R. G. C2, PI Redpath, A. E. R3 Molina, M. J. Ml Reeves, R. R. C8 Monkhouse, P. B. £3 Ritter, J. J. C6 Mori, T. J2 Rockley, M. G. G5 Morine, G. H. L3, L6 Rockwood, S. D. C5 Niki, H. N2 Sabety-Dzvonik, M. E4 Noonan, R. C. K5 Santhanam, M. 12 Noyes, W. A. 03 Savage , C . N2 O'Connor, D. V. Jl Schiff, H. I. F5 Okada, T. J2 Schmatjko, K. J. R5 Okuda, M. K8 Schmeltekopf , A. L. F5 Ors, J. A. D7 Schwerzel, R. E. A2 Ott, W. R. A3 Scott, L. J. D8 Overend, R. K2 Serpone, N. G6 Paine, R. T. L2 Setser, D. W. QA Paraskevopoulos, G. K2 Sharp less, R. L. R8 Parmenter, C. S. 01 Shedd, A. D5 Parsons, J. M. H6 Shortridge, R. G. H2 Pasch, N. F. D6 Simons, J. P. HI Paulson, J. F. E7 Singleton, D. L. F8 Payne, W. A. F7, P6, P8 Slanger, T. G. F2, R8 Perry, R. Nl Smith, I. W. M. ZZ> R6 Spears, K. G. 04 Srinivasan, R. D2 Stedman, D. H. F4 Stephenson, J. C. R2 Stevens, B. D7 Stevens, M. V. G2 Stief, L. J. P6, P8 Stratton, K. D5 Streit, G. E. F5 Stuhl, F. P4 Su, F. K3 Swords, M. D. 05 Tanaka, I. H8 Tang, K. Y. El Taylor, D. G. A4 Tchernev, D. I. 16 Tedder, S. H. L4 Termonia, M. 07 Testa, A. C. D3 Thomas, T. F. E7_ Thorsell, D. L. K3 Timmons, R. B. P6 Toby, F. S. B3 Toby, S. B3 Trappen, N. L8 Umstead, M. E. H2 Van, S. P. J4 Van der Merwe,S.W.J.R6 Velazco, J. E. Q4 Venkatesh, C. G. Verber, C. M. Vermeil, C. Walters, R. T. Wampler, F. B. Ware, W. R. Washida, N. Watson, R. T. Webber, S. E. Welge, K. H. Weston, R. E. Wettack, F. S. Whitbeck, M. R. White, J. M. Whitten, G. Z. Whytock, D. A. Wijnen, M. H. J. Wildes, P. D. Wilkerson, T. D. Willard, J. E. Willis, C. Windsor, M. W. Winer, A. M. Wolf rum, J. Wright, R. E. Yardwood, A. J. Zalewski, E. F. Zellner, R. Zetzsch, C. 02 A2 H9 Gl K4 Jl K8 F3 D6, L4 CI Kl 03 K9 F7, P6_, P8 H3 11 A5_ L3, L6 C3 G5 Nl, N4 M5, R5 Gl 07 Al P7 P4 *U.S. GOVERNMENT PRINTING OFFICE: 1976 624-270/623 1-3 PENN STATE UNIVERSITY LIBRARIES ADOQDTQIDEflEE '•^6-l9l &