Geology and Resources of Titanium GEOLOGICAL SURVEY PROFESSIONAL PAPER 959 This volume was published as separate chapters A—H /4_ :"w UNITED STATES DEPARTMENT OF THE INTERIOR THOMAS S. KLEPPE, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Director CONTENTS [Letters designate the chapters] (A) Titanium contents and titanium partitioning in rocks, by E. R. Force. (B) Metamorphic source rocks of titanium placer deposits—a geochemical cycle, by E. R. Force. (C) Rutile and sphene in blueschist and related high-pressure facies rocks, by M. C. Blake, J r., and Benjamin A. Morgan. (D) Titanium deposits in anorthosite massifs, by Norman Herz. (E) Titanium deposits in alkalic igneous rocks, by Norman Herz. (F) Titanium minerals in deposits of other minerals, by E. R. Force. (G) Rutile in Precambrian sillimanite—quartz gneiss and related rocks, east-central Front Range, Colorado, by Sherman P. Marsh and Douglas M. Sheridan. (H) Alluvial ilmenite placer deposits, central Virginia, by J. P. Minard, E. R. Force, and G. W. Hayes. fiUS. GOVERNMENT PRINTING OFFICE: 1976 0—211-317/1 98 €3WV® GEOLOGY AND RESOURCES ~ OF TITANIUM ‘ "W‘mwmwwmwm mwm‘mgmmm ‘ a , m , , “ fiuFESSIONAL PAPER 9594A, B, C, D, E, . 4‘22 “mww SW 4 RN... - _ * v mmu x§$§u§n§mfix um», x.unmimmummnmmmn “W“ "I! mmnmxm smummmm wmuammmw W N Nam \ mnmwwwmmf m . , w, uu-mlHunk”:nuflfitbnluufiuwmm m COVER PHOTOGRAPH 5 I. Asbestos ore 8, Aluminum ore. bauxite, Georgia I 2 3 4 2, Lead ore~Balmat Mine. N, Y, 9 Native copper ore, Keweenawan 5 6 3. Chromite~chromium ore. Wash. Peninsula, Mich. 4. Zinc ore-Friedensville. Pa‘ 10 Porphyry molybdenum ore, Colo. 7 8 5. Banded iron formation.Palmer. 11 ZInC ore. Edwards. N. Y Michigan 12 Manganese nodules. ocean floor 9 10 6a Ribbon asbestos ore. Quebec. Canada 13' Botrymdal fluorite are 11 12 X3 14 7. Manganese ore. banded Poncha Springs. Colo. rhodochrosite 14. Tungsten ore. North Carolina Titanium Contents and Titanium Partitioning in Rocks By E. R. FORCE Metamorphic Source Rocks of Titanium Placer Deposits—— A Geochemical Cycle By E. R. FORCE Rutile and Sphene in Blueschist and Related High-Pressure- Facies Rocks By M. C. BLAKE, JR., and BENJAMIN A. MORGAN Titanium Deposits in Anorthosite Massifs By NORMAN HERZ Titanium Deposits in Alkalic Igneous Rocks By NORMAN HERZ Titanium Minerals in Deposits of Other Minerals By E. R. FORCE GEOLOGY AND RESOURCES OF TITANIUM GEOLOGICAL SURVEY PROFESSIONAL PAPER 959—A, B, C, D, E, F UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1976 UNITED STATES DEPARTMENT OF THE INTERIOR THOMAS S. KLEPPE, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Director Library of Congress Cataloging in Publication Data Main entry under title: Geology and resources of titanium. (Geological Survey Professional Paper ; 959) CONTENTS: Force, E. R. Titanium contents and titanium partitioning in rocks—Force, E. R. Meta- morphic source rocks of titanium placer deposits, a geochemical cycle—Blake, M. C., J r., and Morgan, Ben- jamin. A. Rutile and sphene in blueschist and related high-pressure—facies rocks—Hem, Norman. Titan— ium deposits in anorthosite massifs.—Herz, Norman. Titanium deposits in alkalic igneous rocks—Force, E. R. Titanium minerals in deposits of other minerals. Includes bibliographies. Supt. of Docs. no.: I 19.162959 A—F 1. Titanium ores—United States. 2. Petrology—United States. I. Series: United States. Geological Survey. Professional Paper ; 959. QE390.2.T5G46 553'462 75—619269 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock Number 024-001-02891-3 APPRAISAL OF MINERAL RESOURCES Continuing appraisal of the mineral resources of the United States is conducted by the US. Geological Survey in accordance with the provisions of the Mining and Minerals Policy Act of 1970 (Public Law 91—631, Dec. 31, 1970). Total resources for purposes of these appraisal estimates in- clude currently minable resources (reserves) as well as those resources not yet discovered or not presently profitable to mine. The mining of mineral deposits, once discovered, depends on geologic, economic, and technologic factors; however, identification of many de- posits yet to be discovered, owing to incomplete knowledge of their dis- tribution in the Earth’s crust, depends greatly on geologic availability and man’s ingenuity. Consequently, appraisal of mineral resources results in approximations, subject to constant change as known deposits are depleted, new deposits are found, new extractive technology and uses are developed, and new geologic knowledge and theories indicate new areas favorable for exploration. This professional paper discusses aspects of the geology of titanium as a framework for appraising resources of this commodity in the light of to- day’s technology, economics, and geologic knowledge. Other Geological Survey publications relating to the appraisal of re- sources of specific mineral commodities include the following: Professional Paper 820—“United States Mineral Resources” Professional Paper 907—“Geology and Resources of Copper” Professional Paper 926—“Geology and Resources of Vanadium Deposits” Professional Paper 933—“Geology and Resources of Fluorine in the United States” CONTENTS [Letters designate the chapters] Page (A) Titanium contents and titanium partitioning in rocks, by E. R. Force ___- A1 (B) Metamorphic source rocks of titanium placer deposits—A geochemical cycle, by E. R. Force __________________________________________________ Bl (C) Rutile and sphene in blueschist and related high-pressure-facies rocks, by M. C. Blake, Jr., and Benjamin A. Morgan _________________________ C1 (D) Titanium deposits in anorthosite massifs, by Norman Herz ______________ D1 (E) Titanium deposits in alkalic igneous rocks, by Norman Herz ______________ E1 (F) Titanium minerals in deposits of other minerals, by E. R. Force _________ F1 METRIC-ENGLISH EQUIVALENTS Metric unit English equivalent Metric unit English equivalent Length Specific combinations—Continued millimetre (mm) = 0.03937 inch (in) litre per second (l/s) : .0353 cubic foot per second metre (m) : 3.28 feet (ft) cubic metre per second kilometre (km) .62 mile (mi) Area 10.76 square feet (ft?) square metre (m9) .386 square mile (mi-'5) square kilometre (km?) hectare (ha) 247 acres Volume cubic centimetre (cm’l) 0.061 cubic inch (in3) 61.03 cubic inches 35.31 cubic feet (ft3) 1030081 acre-foot (acre—ft) litre (1) cubic metre (n13) cubic metre H II I! H H II II H H (7: cubic hectometre (hm3) . acre-feet litre 2.113 pints (pt) litre 1.06 quarts (qt) litre . 6 gallon (gal) cubic metre .00026 million gallons (Mgal or 106 gal) cubic metre : 6.290 barrels (bbl) (1 bh1:42 gal) Weight gram (g) = 0.035 ounce. avoirdupois (oz avdp) gram : .0022 pound, avoirdupois (lb avdp) tonne (t) = 1.1 tons, short (2,000 lb) tonne : .98 ton, long (2,240 lb) Specific combinations kilogram per square centimetre (kg/cm?) : 0.96 atmosphere (arm) kilogram per square centimetre = .98 bar (0.9869 atm) cubic metre per second (ma/s) : 35.3 cubic feet per second (ft3/s) per square kilometre [(m3/S)/km”l metre per day (III/d) metre per kilometre (m/km) kilometre per hour (km/h) metre per second (m/s) metre squared per day (mz/d) cubic metre per second (ms/S) cubic metre, per minute (ms/min) litre per second (l/s) litre per second per metre [(l/s)/m] kilometre per hour (km/h) metre per second (m/s) gram per cubic centimetre (g/cm“) gram per square centimetre (g/cui”) gram per square 4 cubic feet per second per square mile [(fts/s)/mii] : 3.28 feet per day (hydraulic conductivity) (ft/d) 2 5.28 feet per mile (ft/mi) : .9113 foot per second (ft/s) : 3.28 feet per second : 10.764 feet squared per day (t‘t’-‘/d) (transmissivity) : 22.826 million gallons per day (Mgal/d) =264.2 gallons per minute (gal/min) :: 15.85 gallons per minute : 4.513 gallons per minute per foot [(gal/min) /ft] : .62 mile per hour (mi/h) : 2.237 miles per hour : 62.43 pounds per cubic foot (lb/ft") : 2.048 pounds per square foot (lb/ft?) centimetre .0142 pound per square inch (lb/in?) Temperature degree Celsius (°C) : 1.8 degrees Fahrenheit ("1?) degrees Celsius (temperature) = [ (1.8 x °C) + 32] degrees Fahrenheit Titanium Contents and Titanium Partitioning in Rocks By E. R. FORCE GEOLOGY AND RESOURCES OF TITANIUM GEOLOGICAL SURVEY PROFESSIONAL PAPER 959-A «gr? arm Abstract __.._ CONTENTS Distribution of titanium among rock types ___________________________________ Earth’s crust _________________________________________________________ Igneous rocks ________________________________________________________ Sedimentary rocks __________________ ._ _________________________________ Metamorphic rocks ____________________________________________________ Distribution of titanium among minerals ____________________________________ Oxides Silicates Partitioning of titanium between oxides and silicates _____________________ References cited __________________________________________________________ FIGURE 1. TABLE 1. 2—4. 5—8. 9. ILLUSTRATION Variation diagram of Ti02 content and its partitioning in some metamorphic rocks ______________________________________ TABLES Average titanium abundance in the Earth’s crust ______________ Average Ti02 content of some: 2. Igneous rocks ________________________________________ 3. Sedimentary rocks ____________________________________ 4. Metamorphic rocks ___________________________________ Partitioning of titanium: 5. Among minerals of igneous rocks _____________________ 6. Among minerals of high-grade metamorphic rocks from northwestern Adirondacks, New York ______________ 7. Between rutile and biotite in some gneisses fnom Mason Mountain, Macon County, NC _____________________ 8. Among minerals of some other metamorphic rocks _______ Economic cutoff grades for some titanium ores in the United States Page A1 {DWWNNNHHHH Page A8 Page A1 III NNN cacao: GEOLOGY AND RESOURCES OF TITANIUM TITANIUM CONTENTS AND TITANIUM PARTITIONING IN ROCKS By E. R. FORCE ABSTRACT Basic (but not ultrabasic) and alkalic rocks tend to have the highest TIOz contents of igneous rocks. Few igneous, sedi- mentary, or metamorphic rock types have an average Ti02 content as much as an order of magnitude different from the crustal average. In many rocks, common silicate minerals rather than oxides contain most of the Ti02 in the rock. Because only oxide minerals. are economically useful at present, the. economic geology of titanium is as much a function of mineralogy as it is of chemistry. DISTRIBUTION OF TITANIUM AMONG ROCK TYPES EARTH’S CRUST The elemental abundance of titanium in the Earth’s crust has been estimated to be about 0.45 percent, or 0.75 percent as TiO2 (Clarke and Wash- ington, 1924; Goldschmidt, 1954; Vinogradov, 1962; Wedepohl, 1967). Lee and Yao (1970) revised these estimates to 0.64 and 1.07 percent, respectively, par- titioned as shown in table 1. Titanium is reported in chemical analyses of rocks and minerals as TiOz; all subsequent discussion refers to TiO2 rather than elemental titanium. IGNEOUS ROCKS Vinogradov (1962) stated that the average TiO2 content of some igneous rock types is 0.05 percent TABLE 1.—Avemge titanium abundance in the Earth’s crust from Lee and Yao (.1970) Percfent T’ ' 1: ° (£211;th pggdjetrfz‘) (3:33:11 Crust type Ti TiOz volume Total crust _____________ 0.64 1.07 100 Oceanic crust ______ .81 1.35 37 Continental crust ___ .53 .88 63 Shields _________ .55 .92 44 Fold belts ______ .50 .83 19 for ultramafic rocks, 1.5 percent for mafic rocks, 1.3 percent for rocks of intermediate composition, and 0.38 percent for felsic rocks. Nockolds (1954) and Beus (1971) averaged TiO2 analyses of a large num- ber of samples of each type for a wide variety of igneous rocks; the results are given in table 2. Basic and intermediate igneous rocks generally contain more TiOZ than ultramafic or acidic igneous rocks; for most rock types, alkalic varieties are richer in TiO2 than the nonalkalic varieties. Varia- tions are subtle, however, and departures of an order of magnitude from average crustal abundance are remarkable (table 2). Economic concentrations of titanium occur in the igneous rocks anorthosite, norite, and nepheline sye— nite, although the average TiO2 content of all three rocks is lower than the average crustal abundance (table 2). Apparently, the average TiO2 content of the rock and the formation of local concentrations Of valuable titanium minerals are quite separate in igneous rocks. Because of the local prominence of titanium minerals in Precambrian gabbro and anor- thosite, however, Rose (1969, p. 41) thought that these rocks contain most of the titanium in the. crust; the foregoing analyses show this is not the case. SEDIMENTARY ROCKS According to Goldschmidt (1954) and Vinogradov (1962), sedimentary rocks contain an average of 0.75 percent TiOZ; this is slightly lower than the average abundance in continental crust (table 1). Considerable variation is found among the average percentages of TiO2 in different sedimentary rock-s (table 3). Shales have the highest average TiO2 con- tent and limestones the lowest. Of the sandstones, those of low textural and (or) mineralogic maturity have the highest average values; placer concentra- tions of heavy minerals, however, are normally ma- ture and have a high TiO2 content. A1 A2 TABLE 2,—Average TiOe content (in weight percent) of some igneous rocks GEOLOGY AND RESOURCES OF TITANIUM TABLE 4.—Average Ti02 content (in weight percent) of some metamorphic rocks , Beus’ (1971) kagglsuéms‘t) V8.13: if: 2 gm Num- Num- ber ber ana- 8.118.- Rock type TiOz ’lyzed TiOz lyzed Ultramafic rocks: Dunite ______________ 0.20 9 0.07 118 Peridotite ___________ .81 23 .53 196 Alkali peridotite _____ 1.30 12 ___ ___ Pyroxenite __________ .53 46 .83 294 Alkali pyroxenite ___- 3.31 21 ___ ___ Hornblendite ________ 2.86 15 ___ ___ Kimberlite __________ 1.43 14 2.17 4.21 Basic rocks: Tholeiitic basalt _____ 2.03 137 ___ ___ Olivine tholeiite _____ 1.65 28 ___ ___ Normal alkali basalt __ 2.63 96 ___ ___ Continental basalt ___ ___ ___ 1.50 445 Geosyncline basalt ___ ___ ___ 1.67 360 Oceanic basalt _______ ___ ___ 2.67 148 Gabbro _____________ 1.32 160 1.13 762 Norite ______________ .89 39 ___ ___ Alkali gabbro _______ 2.86 42 ___ ___ Anorthosite __________ .32 17 ___ ___ Intermediate rocks: Diorite _____________ 1.50 50 1.00 678 Andesite ____________ 1.31 49 .83 866 Tholeiitic andesite ___- 2.60 26 ___ ___ Alkali andesite ______ 2.84 37 ___ ___ Acidic rocks: Tonalite ____________ .62 58 .77 42.6 Dacite ______________ .64 50 .57 480 Granodiorite ________ .57 137 62 523 Rhyodacite __________ .66 115 ___ ___ Adamellite __________ .56 121 ___ ___ Dellenite ____________ .42 58 ___ ___ Granite _____________ .37 72 .33 1967 Rhyolite ____________ 22 22 33 138 Alkalic rocks: Syenite _____________ .83 18 .68 426 Trachyte ____________ .66 24 .67 292 Monzonite ___________ 1.12 46 _-_ ___ Latite ______________ 1.18 42 ___ ___ Nepheline syenite ___- .66 80 .50 584 Phonolite ___________ .59 47 .40 245 METAMORPHIC ROCKS The range of TiO2 content of metamorphic rocks appears to reflect the diversity of parent rocks. The data are much less complete for metamorphic rocks (table 4) than for parent materials (tables 2 and Num- ber ana- Rock type TiOz lyzed Reference Amphibolite _________ 1.3.7 370 Beus (1971). Gneiss ______________ .58 410 Do. Schist _______________ .60 538 Do. Slate ________________ .80 8 Billings and Wilson (1965). Greenschist __________ 1.64 13 Hutton (1940). Quartzite ____________ .23 7 Billings and Wilson (1965), Clarke (1924). Serpentinite _________ .015 91 Clarke (1924), Abdullayev and Gusey- nova (1970). Glaucophane schist __ .78 5 Clarke (1924). Eclogite _____________ 1.27 16 Beus (1971). 3) ; a better compilation would be useful When the possible mobility of titanium during metamorphism is considered. Engel and Engel (1958, 1962a) have demonstrated slight decreases in the titanium con- tent of paragneiss and amphibolite with increasing metamorphism. DISTRIBUTION OF TITANIUM AMONG MINERALS OXIDES The polymorphs of TiOZ—rutile, anatase, and brookite—ar-e all fairly common in nature. Rutile is the stable phase (Schuiling and Vink, 1967; Jamie- son and Olinger, 1969) , this fact helping to explain the presence of rutile in such diverse geologic en- vironments as alteration products (leucoxene) formed at surface temperatures and pressures, vari- ous metamorphic rocks, and igneous rocks. Ilmenite (FeTiO-g) is a common accessory mineral in a Wide variety of rocks. Analyses of natural ilmen- TABLE 3.—-Average Ti02 content (in weight percent) of some sedimentary rocks Pettijohn’s (1963) value Turekian and Wedepohl’s (1961) Beus’ (1971) value recalculated value ROCR type TiOz Number analyzed TiOz Ti02 Number analyzed Sandstone: 0.25 253 0.25 0.52 211 Orthoquartzite ___ .2 26 ___ ___ ___ Lithic arenite ___- .3 20 ___ ___ ___ Graywacke ______ .6 61 ___ ___ ___ Arkose __________ .3 32 ___ __- ___ Siltstone ____________ .59 235 ___ ___ ___ Shale _______________ .65 78 .77 63 252 Limestone ___________ ___ ___ .07 20 364 TITANIUM CONTENTS AND TITANIUM PARTITIONING IN ROCKS ites fall on both sides of the theoretical TiOz content of 52.5 percent (see Mackey, 1972) ; variations are due, among other things, to the presence of inter- growths with hematite and magnetite and to solid solution with hematite (low values) or leaching of iron during alteration (high values). Material called ilmenite in the literature on beach-sand mining can contain more than 60 percent TiO2 as a result of en- richment during weathering and transport. Much of this material actually consists of alteration products pseudomorphous after ilmenite (see Markewicz, 1969; Garnar, 1972). Leucoxene is a loose term for the end products of this alteration of ilmenite. The titanium is contained in fine-grained aggregates of rutile, brookite, ana- tase, hematite, and (or) sphene (Bailey and others, 1956; Baker, 1962; Temple, 1966). Leucoxene can form by hydrothermal alteration as well as by weathering (Ross, 1941). Another oxide of titanium, perovskite (CaTiog), is a rock-forming mineral in some alkalic rocks and has economic importance. Normally, it has less TiO2 than its theoretical content of 59 percent because of the presence of rare earths and niobium. SILICATES Although many rock-forming silicate minerals contain Ti02, those most important to the geochem- istry of titanium are sphene, biotite, hornblende, and titanaugite. Sphene (CaTiSiOs) ideally has 41 percent TiO2 but may contain less (Leonova and Klassova, 1964). It is a common accessory mineral in a large number of igneous and metamorphic rocks. Biotite and hornblende contain relatively small amounts of titanium, but, because they are usually much more abundant than titanium-rich accessory minerals, they are important carriers of titanium. In common igneous rocks of calc-alkalic compositions, biotite has as much as 5.9 percent TiOZ, and horn- blende as much as 2.7 percent TiO2 (Deer and others, 1962b, 1963a; Al’mukhamedov, 1967 ; Lyakhovich, 1970). In common metamorphic rocks, biotite con- tains as much as 5 percent TiO'g, and hornblende as much as 3 percent; in both minerals, TiO2 increases with metamorphic grade (Force, 1976). The con- centration of TiO2 in hornblende and biotite is large- ly independent of the TiO2 content of the rock in both igneous and metamorphic rocks. In many volcanic and (or) alkalic rocks, titanaug- ite is a characteristic phase and may contain as much as 9 percent TiO2 (Deer and others, 1963a). A3 In alkalic igneous rocks, two more unusual min- erals may also contain titanium. Melanitic andradite garnet contains as much as 17.1 percent TiO2 (Deer and others, 1962a), and kaersutitic amphiboles as much as 10.3 percent TiO2 (Deer and others, 1963a). The common rock-forming silicates that do not contain much titanium also require comment prepar- atory to a discussion of the geochemical partitioning of titanium. The work of Deer and others (1962a, b, c, 1963a, b) was used as a source of this information, as were the following references cited. In general, minerals that lack large amounts of titanium include silica minerals, feldspars, and other framework sili- cates (Rankama and Sahama, 1950), muscovite, chlorite, and serpentine (Kwak, 1968; Ernst and others, 1970; Lyakhovich, 1970; Abdullayev and Guseynova, 1970), common garnets (Howie, 1955; Engel and Engel, 1960), kyanite-group minerals, olivine, and epidote~group minerals. Orthopyroxene and metamorphic clinopyroxene contain little ti- tanium, although metamorphic clinopyroxene con- sistently contains more than coexisting orthopyrox- ene (Eskola, 1952; Howie, 1955) . PARTITIONING OF TITANIUM BETWEEN OXIDES AND SILICATES For any element that occurs in valuable nonsilicate minerals, the amount of the element available to form these valuable minerals depends on the extent to which host rocks used the element in silicates (Sullivan, 1948). This is applicable to titanium, since all presently mined titanium minerals are oxides. A study of the partitioning of titanium is thus of potential significance to economic geologists. Following Ramdohr (1940), authors have tradi- tionally said that oxide minerals contain most of the titanium in rocks (cf. Rankama and Sahama, 1950; Rose, 1969). It will be shown in this section that such a situation is a special, though interesting, case. Several treatises cover the partitioning of titani- um between oxides and silicates in a general way. Ramberg (1948, 1952, p. 72—75, 156—161) ap- proaches the subject from the viewpoint of thermal stability of titanium-bearing silicates (as does Force (1976)). Verhoogen (1962) discussed the effect of oxidation on the partitioning. Shcherbina (1971) mentioned the effect of high Mg/ Fe ratio in limiting the stability of ilmenite and of high alumina (A1203) in limiting the stability of sphene and clinopyroxene relative to plagioclase and orthopyroxene. Others have noted that magmatic rocks having a high ratio of orthopyroxene to clinopyroxene are richer in il- A4 GEOLOGY AND RESOURCES OF TITANIUM TABLE 5.—Pa’rtitioning of titanium among minerals of igneous rocks TiOz in Min- min- . enal l’erceni‘. Rock ifype (weight percent (5331.: (3:331; 3021131102 in parentheses) Reference Mineral percent) percent) mineral 1- Olivine PYroxenite Abdullayev and Diopside 0.134 47.4 74.0 (0-085) - Guseynova (1970) . Enstatite .073 22.2 20.0 Olivme .00 16.3 .0 Serpentine .017 11.0 2.4 Magnetite .080 3.1 2.4 2. Olivine pyraxenite ____do ______________ DiOpside .105 42.8 61.6 (0-073) - Enstatite .070 32.3 30.1 Olivine .00 12.5 .0 Serpentine .027 10.9 4.1 Magnetite .087 1.5 1.4 3. Peridotite (0.048) _____ ___-do ______________ Diopside .102 22.0 47-9 Enstatibe .080 13.3 22.9 Olivine .00 44.4 .0 Serpentine .060 18.3 22.9 Magnetite .108 2.3 4.2 4. Olivine nori’oe (0.79) ___ Wage-r and Mitchell Hypersthene .93 28.5 25 (1951, p. 142). Olivine _________ 21.5 _____ Plagioclase _________ 48.0 _____ 5. Picritic diabase (0.59) _ _ Al’mukh amedov Pyroxene .50 23.38 20.34 (1967). Olivine .07 41.37 5.09 Biotite 5.86 3.12 30.53 Plagioclase .23 19.83 8.47 Magnetite 3.72 3.86 23.72 Ilmenite 52.6 .13 11.85 6. Diabase (1.32) _______ ____do ______________ Clinopyroxene .79 36.36 21.96 Olivine .19 7.85 .76 Plagioclase .13 48.71 ‘ 4.54 Magnetite 18.88 1.43 20.54 Ilmenite 49.56 1.26 47.03 7. Biotite pyroxene- Samoylov and Pyroxene .64—1.29 _ _ _ _ 1 3—58 calcite carbonatite. Razvozzhayeva Biotite 1.89—3.49 _ _ _ _ 39—72 ( 1972) . 8. Amphibole-pyroxene- ____do ______________ Pyroxene 64—129 ___ _ 90 calcite carbonatite. Amphibole 62—334 _ _ _ _ 5—8 9. Ijolitic carbonatite _____ ____do ______________ Sphene _________ ___ _ 54—90 10. Nepheline syenite ______ Popolitov, Kovalenko, Pymxene and _________ _ _ _ _ 1 1—19 and Znamensky amphibole (1966). Biotite _________ ___- 32—42 Titanomagnetite _________ _ _ _ _ 29—41 and ilmenite 11. Syenite _______________ ___..do ______________ Pyroxene and _________ ____ 16—25 amphibole Biotite _________ ____ 15—19 Titanomagnetite _________ _ _ _ _ 53—57 and ilmenite 12. Syenite (0.57) ________ Leonova and Klassova Biotite 3.45 3.0 18.2 (1964). Pyriboles .78 13.0 17.8 Magnetite 4.64 4.5 36.6 13. Granite _______________ Znamensky (1957 ) . “Silicates” _________ _ _ _ _ 69—7 7 14. Granite (0.20) ________ Leonova and Klass‘ova Biotite 2.06 3.9 40.2 (1964). Sphene 22.42 .3 33.6 Magnetite .8 .7 3.0 15. Granite (0.14) ________ ____do ______________ Biotite 4.2 2.7 80.7 Sphene 17.8 .14 17.8 Magnetite 1.2 .3 2.5 TITANIUM CONTENTS AND TITANIUM PARTITIONING IN ROCKS menite (Goldschmidt, 1954, p. 413; Deer and others, 1962c, p. 31). Relatively few sources, however, give precise figures for the partitioning of titanium because pre- cise figures require chemical and modal analyses of a rock as well as chemical analyses of all titanium- bearing phases in the rock. Such data have seldom been acquired. Table 5 shows some published information for the partitioning of titanium in igneous rocks. The type of data varies considerably and consists of unsup- ported statements (No. 13), summaries of partition- ing studies of rock suites (Nos. 7—11), and partition- ing analyses of individual rocks (the others). Not all rocks listed are common. Table 6 shows the partitioning of titanium among minerals of some high-grade quartzofeldspathic par- agneisses of simple mineralogy. Information is se- A5 lected from Engel and Engel (1958, 1960); the calculations are mine. Titanium oxide not present in the silicate phases probably is in ilmenite (rutile is not present). Table 6 also shows TiOZ present in hornblende of associated amphibolites (Engel and Engel, 1962a, b) ; again, the remainder must be in ilmenite. Table 7 shows the partitioning of titanium in a suite of high-grade: gneisses of simple mineralogy from the North Carolina Blue Ridge. Titanium is present virtually only in rutile and biotite in these rocks; this situation is unusual but allows an easy calculation of partitioning without mineral analyses. Partitioning for some other metamorphic rocks is shown in table 8. For those rock types that are un- usual even in the terranes in which they occur (tec- tonic inclusions and eclogites) , representative analy- ses were selected from those available. The figures TABLE 6.—Partitioning of titanium among minerals of high-grade metamorphic rocks from northwestern Adirondacks, New York, calculated from Engel and Engel (1.958, 1960, 1962a, b) QUARTZOFELDS PATH IC PARAGNEISS TiOz Percent of rock TiOz present in mineral (remainder presumably in ilmenite) Rock number (weight percent) Biotite Garnet Sphene Emeryville and West Balmat groups (amphibolite facies) Qb 228 ___________ 0.87 79 0 Absent Qb 235 M _________ .67 81 Absent Do. Qb 230 ____________ .65 83 0 Do. Qb 231 ____________ .66 82 Absent Do. Qb 236 ____________ .81 86 ______ do ____ Do. Qb 3 ______________ .62 95 ______ do ____ Do. Bgn 27 ____________ .70 94 ______ do ____ Do. Russell and Pierrepont groups (intermediate) Bgn 14 ____________ 1.02 89 0 Absent Bg'n 4 _____________ .56 82 0 o. Bgn 9 _____________ .51 77 0 D0. Bgn 6 _____________ .68 88 0 Do. Bgn 25 ____________ .76 98 0 3 Colton group (granulite facies) Bgn 20 ____________ 0.67 74 0 Absent Bgn 19 ____________ .57 621 1 Do. Bgn 21 ____________ .52 74 1 Do». AMPHIBOLITE Rock b Ti02 T102 clfintednt of Homblerfide in Pfircfint gecdeifh num er (weight percent) (weigolil;l peerrcznt) (weighltogercent) mainderl preghtmalirlyeinrialmenite) Emeryville group (amphibolite facies) A3 ________________ 1.20 1.37 73 85 AE317 ____________ 1.17 1.32 66 75 AE415 ____________ 2.48 1.32 77 41 AE334 ____________ 2.11 1.19 76 43 AE337 ____________ 2.07 1.45 75 53 AE326 ____________ 2.13 1.27 71 42 AE338 ____________ 2.04 1.30 68 43 Colton group (granulite facies) A10 ______________ 1.40 2.49 31 56 A67 ______________ .99 2.26 12 28 A104 _____________ 1.60 2.24 23 32 A105 _____________ 1.56 2.26 21 31 A0342 ____________ 1.71 2.61 41 64 A0358 ____________ 1.74 2.49 58 83 A0348 ____________ 1.78 2.30 26 34 A0341 ____________ 1.55 2.71 44 72 A0362 ____________ 1.74 2.42 24 34 A6 TABLE 7.—Pa7'titioning of titanium between mtile and biotite in some quartz-gamet-feldspar-biotite-sillimnite-cumming— tonite—pyroxene—gmphite gneisses from Mason Mountain, Macon County, N.C. [Mineral percentage established by 2,000-point counts corrected for spe- cific gravity; rapid-rock analyses by U.S. Geological Survey, samples pro- vided by F. G. Lesure] Percent rock TiOz Biotite Rutile in in in rutile TiOz rock rock ( remainder (weight (weight (weight mostly Number percent) percent) percent) biotite) MM 1 ______ 0.84 3.3 0.36 43 MM 4 ______ 1.6 18.5 .72 45 MM 8 ______ .96 3.6 .94 98 MM 11 _____ 1.0 absent 1.05 100 MM 14 _____ .78 do. .72 92(?) MM 17 _.-___ .73 30.1 .43 59 GEOLOGY AND RESOURCES OF TITANIUM are inexact; the analyses are not sufficiently accurate for this purpose, and I attempted to correct the modal analyses for specific gravity. As is the situa- tion for igneous rocks, information on most common types of metamorphic rocks cannot be found. Avail- able analyses nec-essarily are heavily biased toward higher grade assemblages having grains coarse enough to be easily counted. On the basis of tables 5—8, it can be said that, in general, silicates contain much of the titanium in most rocks. Silicates contain more than 90 percent of the TiO2 in the ultrabasic rocks listed, about half the TiO2 in basic volcanics, almost all the TiO2 in carbonatites, and more than 60 percent of the TiO2 TABLE 8.—Pai'titioning of titanium among minerals of some other metamorphic rocks Rock type .and number (weight percent T102 in parentheses) Mineral Mineral in rock (weight percent) T102 in mineral Percent rock TiOz (weight percent) ' in each mlneral Tectonic blocks from California serpentinites (Ernst and others, 1969, 1970) Glaucophane schist 5OCZ60 (2.17) ________ Glaucophane 0.22 17 2 White mica .27 15 2 Chlorite .09 7 0 Actinolite .11 21 1 Garnet .25 13 1 ISEplildote ___ 23 __ p ene ___ . Rutile ___ 0'1 Remalnder? Amphibolite 1893 (1.26) _________________ Hornblende 1.07 99 Sphene ___ 1 Remainder? Garnet amphibolite CAT 4 (2.41) ________ Homblende .90 40 15 Garnet .34 50 Rutile _ _ _ 1 Remainder ? Msfic quartzo-feldspathic schist, oligoclnse zone (Mason, 1962) Callery-Waiho R. (1.72) _________________ Quartz ___ 14 _. Plagioclase ___ 34 __ Homblende 1.03 2 1 Biotite 1.70 22 22 Almandine .57 25 Opaque ___ 5 Remainder? Madras charnockites (Howie, 1955; Howie and Subramanism, 1957) Garnet enderbite Ch113 (0.99) ___________ Quartz ___ 20 __ K-feldspar ___ 14 __ Plagioclase ___ 32 __ Garnet 03 14 0 Hypersthene 30 13 4 Biotite ___ 1 0 (probably) “Ores” ___ 2 Remainder? Chamockite 4639 (0.56) _________________ Quartz ___ 41.8 __ K-feldsvar .02 28.5 1 Plagioclase ___ 19.5 __ vaersthene .94 7.9 13 “Ores” _ _ _ 2.2 Remainder ? Chamockite 6436 (0.40) _________________ Quartz ___ 34.6 __ K-feldsuar 02 26.5 1 Plag'ioclase .16 31.8 13 vaersthene 1.78 4.8 2 “Ores” ___ 2.1 Remainder? Garnetiferous leptynite 3708 (0.31) ______ Quartz ___ 50.5 __ K-feldspar 0 38.7 0 Garnet 07 4.5 “Ores” _ _ _ 2.1 Remainder? TITANIUM CONTENTS AND TITANIUM PARTITIONING IN ROCKS TABLE 8.——Partitioning of titanium among minerals of some other metamorphic rocks—Continued A7 Rock type and number (weight percent TiO-g in parentheses) Mineral TiOz in mineral (weight percent) Mineral in rock (weight percent) Madras charnockites (Howie, 1955; Howie and Subramaniam, l957)—Continued Percent rock TiOz in each mineral Intermediate rock 137 (0.77) _____________ Quartz ___ 20.7 __ K-feldspar ___ 33.5 __ Plagioclase ___ 36.6 __ Homblende 2.06 1 .7 4 Hypersth-ene .15 3.4 1 “Ores” ___ 3.6 Remainder? Intermediate rock 2270 (1.61) ____________ Quartz __- 6.1 __ K-felespar .0 3.8 0 Plagioclase .0 52.9 0 Hornblende 1.85 1.4 2 Biotite 4.61 7.9 22 Hypersthene .15 5.5 1 Augite .24 13.7 2 Magnetite 1.53 4.8 4 Ilmenlite 43.55 2.8 76 ? Dioritic charnockite 4642a (1.12) ________ Plagioclase .0 45.0 0 Hornblende 1.72 0.9 0 Hypersthene .11 25.2 2 Augite .30 24.8 7 Ilmenite 48.90 2 85 Magnetite 6.38 2 10 “Norite” 2941 (1.18) ____________________ Plag'ioclase .05 37.8 2 Hornblernde .50 9.3 4 Hypensthene .65 30.0 16 Aug'ite .70 19.4 1 2 “Ores” ___ 3.2 Remainder? “Pyroxenite” 3709 (1.32) ________________ Homblende 1.48 8.7 10 Hypersthene .10 56.4 4 Augite .68 25.4 13 Hercynite 1.42 4.2 5 “Ores” ___ 5.1 Remainder? Other charnockites (Howie, 1958, 1965) Mafic charnockite, Sudan 7286 (1.04) _____ Hornblende 2.10 20 40 Aug'ite .38 1 7 6 Ferrohypersthene .18 15 3 Hypersthene diorite, Uganda S347 (1.52) __ Hypersthene ___ ___ 1 Augite _ __ _-_ 3 Hornblende ___ ___ 15 Almandine _ _ _ _ _ _ 1 Plagioclase ___ ___ 0 Orthoclase __- ___. 0 Magnetite _ _ _ - _ _ 1 Ilmenite _ _ _ _ _ _ 68 Biotite _ _ _ _ _ _ 14 Quartz _ _ _ _ _ _ 0 Eclogite (Coleman and others, 1965 ; Ernst and others, 1969, 1970) B102A (0.98) __________________________ Garnet 0.26 61 16 Clinonvroxene .79 13 10 Homblende .87 18 15 Rutile _ _ _ 1 Remainder? 36NC62 (2.1) __________________________ Omnhacite .18 34 3 Garnet .54 29 7 Glaucophane .31 18 3 Efilgielrée " ' $3 Remainder? 102RGC58 (1.5) ________________________ Omnhacite .15 67 7 Garnet 38 1 6 4 Chlorite _ _ _ 10 _ _ Istgltlieige __ _ 2'; Remainder? A8 I I I I I I I I I I A 2.56 A A AA 2.41 ) 273 percent)/ 3.06 2.39 percent 1-8 _ (percent percent percent ‘ 1.6 — [j x A w _ . + O _I 1.4 * fl x o 0 III 0: _ _ 2 I A n. 1.2 — - a: D 0 O E El _ < _ E E El 1.0 XE] a O E x o“ — — i: ' O X ’— 0 Z 0.8 — X _ 8 ‘3 o a: X td.‘ _ O O O F . 0% A I o 6 - 0 c o . E D D O 3 _ OO _ 0.4 - El _ 0.2 — _ 0 i I I I I I I | 1 l O 20 40 60 80 100 WEIGHT PERCENT TiO2 IN SILICATES EXPLANATION O High-grade paragneiss from the Adirondack Moun- tains, New York (Engei and Engel, 1958. 1960) X High-grade gneiss from Mason Mountain, Macon County, North Carolina [I Charnockites of Madras, India (Howie, 1955; Howie and Subramaniam, 1957) A Amphibolites from California coast ranges (Ernst and others. 1969,1970) O Eclogites from Japan (Ernst and others. 1969, 1970) + Eclogites from California (Coleman and others. 1965) FIGURE 1.—-Variation diagram of TiOa content and its par- titioning in some metamorphic rocks. GEOLOGY AND RESOURCES OF TITANIUM in felsic intrusive rocks. In metamorphic rocks, the partitioning varies rather widely; this variation will be discussed by Force (1976) . Figure 1 shows that, for some metamorphic rock suites, the partitioning is quite independent of the TiO2 content of the rocks involved (some suites part- ly listed in table 8 are fully represented in fig. 1). The diagram as a whole, and the data for each suite, shows a scattering of points. For two rocks having a given TiO2 content (and, in fact, most rock types do have TiOZ contents Within the same order of mag- nitude) , the mineralogy of the titanium can be strik- ingly different. The processes that control the min— eralogy are most worthy of attention. Ore deposits of titanium are all examples of an extreme of partitioning in which the titanium occurs in oxide form. Although some titanium ores are en- riched in titanium far beyond the average crustal abundance, others are not, as is shown in table 9. In fact, the rutile sands of eastern Australia now being mined have TiO2 values well under the average crust- al abundance! The value of these low TiOZ ores is in their mineralogy and ease of beneficiation. It is misleading to assume that the TiO2 content of potential titanium resources is an indicator of their value. For example, Thomas and Berryhill (1962) reported high TiO2 values in Alaskan beach sands; probably very little titanium of economic value is present because the high TiO2 values occur in non- magnetic fractions of the sand in which sphene, au- gite, and hornblende are reported as prominent com- ponents and titanium-bearing oxides are sparse or absent. Therefore, it would be wasted effort to pros- pect here by chemical analysis for Ti02. TABLE 9.—-Economic cutoff grades for some titanium ores in the United States TiOz cutoff grade Location Host rock Ore mineral (percent) Sanford Lake Ilmenite nook IlmenIite and 13.5 (Tahawus), in anorthosite ilmenite- N .Y.1 and norite magnetite intergrowths Lakehurst, Fossil river (?) Altered ~1.5 N. J .2 sandI placer ilmeni-te Trail Ridge, Fossil beach Altered ~1.5 Fla.“ sand placer ilmenite ‘ Gross (1968). 3 Markewicz (1969). 3 Pirkle and Yoho (1970). TITANIUM CONTENTS AND TITANIUM PARTITIONING IN ROCKS REFERENCES CITED Abdullayev, Z. B., and Guseynova, S. F., 1970, Trends in the distribution of titanium and vanadium in the ultraba- sites of the low Caucasus in Azerbaijan: Geochemistry Internat., v. 7, p. 1012—1016. Al’mukhamedov, A. L., 1967, Behavior of titanium during dif- ferentiation of basaltic magma: Geochemistry Internat, v. 4, p. 47—56. Bailey, S. W., Cameron, E. N., Spedden, H. R., and Weege, R. J., 1956, The alteration of ilmenite in beach sands: Econ. Geology, v. 51, p. 263—279. Baker, George, 1962, Detrital heavy minerals in natural ac- cumulates: Australasian Inst. Mining and Metallurgy Mon. Ser. 1, 146 p. Beus, A. A., 1971, Titanium distribution in the lithosphere: Chem. Geology, v. 8, p. 247—275. Billings, M. P., and Wilson, J. R., 1965, Chemical analyses of rocks and rock minerals from New Hampshire: New Hampshire Div. Econ. Devel. Mineral Resources Survey, pt. 19, 104 p. Clarke, F. W., 1924, The data of geochemistry (5th ed.): U.S. Geol. Survey Bull. 770, 841 p. Clarke, F. W., and Washington, H. S., 1924, The composition of the Earth’s crust: U.S. Geol. Survey Prof. Paper 127, 117 p. Coleman, R. G., Lee, D. E., Beatty, L. B., and Brannock, W. W., 1965, Eclogites and eclogites—Their differences and similarities: Geol. Soc. America Bull., v. 76, p. 483~ 508. Deer, W. A., Howie, R. A., and Zussman, Jack, 1962a, Rock- forming minerals, v. 1, Ortho- and ring silicates: New York, John Wiley and Sons, 333 p. 1962b, Rock-forming minerals, v. 3, Sheet silicates: New York, John Wiley and Sons, 270 p. 1962c, Rock—forming minerals, v. 5, Non-silicates: New York, John Wiley and Sons, 371 p. 1963a, Rock—forming minerals, v. 2, Chain silicates: New York, John Wiley and Sons, 379 p. 1963b, Rock-forming minerals, v. 4, Framework sili- cates: New York, John Wiley and Sons, 435 p. Engel, A. E. J., and Engel, C. G., 1958, Progressive meta- morphism and granitization of the major paragneiss, northwest Adirondack Mountains, New York. pt. 1, Total rock: Geol. Soc. America Bull., v. 69, p. 1369—1414. 1960, Progressive metamorphism and granitization of the major paragneiss, northwest Adirondack Mountains, New York, pt. 2, Mineralogy: Geol. Soc. America Bull., v. 71, p. 1—58. 1962a, Progressive metamorphism of amnhibolite, northwest Adirondack Mountains, New York, in Petro- logic studies, a volume in honor of A. F. Buddington: New York, Geol. Soc. America, p. 37—82. 1962b, Hornblendes formed during progressive meta- morphism of amphibolites, northwest Adirondack Moun- tains, New York: Geol. Soc. America Bull., v. 73, p. 1499—1514. Ernst, W. G., Seki, Yotari, Onuki, Hitoshi, and Gilbert, M. C., 1969, Phase proportions and physical properties of minerals and rocks from the Franciscan and Sanbagawa. metamorphic terranes: Tokyo, Japanese Soc. Promotion Science, 85 p. A9 1970, Comparative study of low-grade metamorphism in the California coast ranges and the outer metamorphic belt of Japan: Geol. Soc. America Mem. 124, 276 p. Eskola, P. E., 1952, On the granulites of Lapland: Am. Jour. Sci., Bowen Volume, p. 133—171. Force, E. R., 1976, Metamorphic source rocks of titanium placer deposits—A geochemical cycle, in Geology and re sources of titanium: U.S. Geol. Survey Prof. Paper 959— B, 16 p. Garnar, T. E., Jr., 1972, Economic geology of Florida heavy mineral deposits, in Puri, H. S., ed., Proc. 7th Forum on Geology of Industrial Minerals: Florida Bur. Geology, Spec. Pub. 17, p. 17—21. Goldschmidt, V. M., 1954, Geochemistry: Oxford, Clarendon Press, 730 p. Gross, S. 0., 1968, Titaniferous ores of the Sanford Lake District, New York, in Ridge, J. D., ed., Ore deposits in the United States 1933—1967 (the Graton-Sales volume), v. 1: New York, Am. Inst. Mining Metall. Petroleum Engineers, p. 140—153. Howie, R. A., 1955, The geochemistry of the charnockite series of Madras, India: Royal Soc. Edinburgh Proc., v. 62, p. 725—768. 1958, African charnockites and related rocks: Congo Belge Service Geol. Bull., v. 8, pt. 2, 14 p. 1965, The geochemistry and mineralogy of some char— nockites from Baffin Island, Brazil, British Guiana, Congo, Madras, and Uganda: Indian Minerals, v. 6, p. 67—76. Howie, R. A., and Subramaniam, A. P., 1957, The para- genesis of garnet in charnockite, enderbite, and related granulites: Mineralog. Mag, v. 31, p. 565—587. Hutton, C. 0., 1940, Metamorphism in the Lake Wakatipu region, western Otago, New Zealand: New Zealand Dept. Sci. and Indus. Res. Geol. Mem. 5, 90 p. Jamieson, J. C., and Olinger, Bart, 1969, Pressure-tempera- ture studies of anatase, brookite, rutile, and TiOg, pt. 2: Am. Mineralogist, v. 54, p. 1477—1481. Kwak, T. A. P., 1968, Titanium in biotite and muscovite as an indication of metamorphic grade in almandineam- phibolite facies rocks from Sudbury, Ontario: Geochim. et Cosmochim. Acta, v. 32, p. 1222—1229. Lee, Tan, and Yao, Chi-lung, 1970, Abundance of chemical elements in the Earth’s crust and its major tectonic units: Internat. Geology Rev., v. 12, p. 778-786. Leonova, L. L., and Klassova, N. S., 1964, On the geochemistry of titanium during the intrusive process: Geochemistry Internat., v. 1, no. 2, p. 290—295. Lyakhovich, V. V., 1970, The mean contents of Ta, Nb, and Ti in granitoid biotite: Geochemistry Internat, v. 7, p. 825—834. Mackey, T. S., 1972, Alteration and recovery of ilmenite and rutile, pt. 1, World. occurrence, production, reserves, de- mand: Australian Mining, v. 64, no. 11, p. 18—26. Markewicz, F. E., 1969, Ilmenite deposits of the New Jersey coastal plain, in Geology of selected areas in New Jersey and eastern Pennsylvania and guidebook of excursion (field trip 6): New Brunswick, N.J., Rutgers Univ. Press, p. 363—382. A10 Mason, Brian, 1962, Metamorphism in the Southern Alps of New Zealand: Am. Mus. Nat. History Bull., v. 123, p. 211—248. Nockolds, S. R., 1954, Average chemical compositions of some igneous rocks: Geol. Soc. America Bull., v. 65, p. 1007— 1032. Pettijohn, F. J ., 1963, Chemical composition of sandstones— excluding carbonate and volcanic sands: U.S. Geol. Sur- vey Prof. Paper 440—8, 21 p. Pirkle, E. C., and Yoho, W. H., 1970, The heavy mineral ore body of Trail Ridge, Florida: Econ. Geology, v. 65, p. 17—30. Popolitov, E. 1., Kovalenko, V. I., and Znamensky, E. B., 1966, Geochemistry of titanium in the middle Paleozoic intrusives of northeastern Tuva: Geochemistry Internat., v. 3, p. 253—260. Ramberg, Hans, 1948, Titanic iron ore formed by dissocia- tion of silicates in granulite facies [Greenland]: Econ. Geology, v. 43, p. 553—569. 1952, The origin of metamorphic and metasomatic rocks: Chicago, 111., Univ. of Chicago Press, 317 p. Ramdohr, Paul, 1940, Die erzmineralien im gewohnlichen magmat‘ischen gesteinen: Akad. Wiss. Gettingen Abh., Math-Phys. Kl., no. 2, 43 p. Rankama, Kalervo, and Sahama, Th. G., 1950, Geochemistry: Chicago, I11., Univ. of Chicago Press, 912 p. Rosie, E. R., 1969, Geology of titanium and titaniferous de- posits of Canada: Canada Geol. Survey Econ. Geology Rept. 25, 177 p. Ross, C. S., 1941, Occurrence and origin of the titanium de- posits of Nelson and Amherst Counties, Virginia: U.S. Geol. Survey Prof. Paper 198, 59 p. GEOLOGY AND RESOURCES 0F TITANIUM Samoylov, V. S., and Razvozzhayeva, E. A., 1972‘, Behavior of rare elements in formation of carbonatites of rare metals: Internat. Geology Rev., v. 14, p. 440—449. Schuiling, R. D., and Vink, B. W., 1967, Stability relations of some titanium minerals (sphene, perovskite, rutile, and anatase): Geochim. et Cosmochim. Acta, v. 31, p. 2399— 2411. Shcherbina, V. V., 1971, Geochemistry of titanium: Geok— himiya, no. 3, p. 302—309. Sullivan, C. J., 1948, Ore and granitization: Econ. Geology, v. 43, p. 471-498. Temple, A. K., 1966, Alteration of ilmenite: Econ. Geology, v. 61, p. 695—714. Thomas, B. I., and Berryhill, R. V., 1962, Reconnaissance studies of Alaska beach sands, eastern Gulf of Alaska: U.S. Bur. Mines Rept. Inv. 5986, 40 p. Turekian, K. K., and Wedepohl, K. H., 1961, Distribution of the elements in some major units of the Earth’s crust: Geol. Soc. America Bull., v. 72, p. 175—192. Verhoogen, John, 1962, Distribution of titanium between sili- cates and oxides in igneous rocks: Am. Jour. Sci., v. 260, p. 211—220. Vinogradov, A. P., 1962, Average contents of chemical ele- ments in the principal types of igneous rocks of the Earth’s crust: Geokhimiya, no. 7, p. 641—664. Wager, L. R., and Mitchell, R. L., 1951, The distribution of trace elements during strong fractionation of basic magma—A further study of the Skaergaard intrusion, East Greenland: Geochim. et Cosmochim. Acta, v. 1, p. 129—208. Wedepohl, K. H., 1967, Geochemie: Berlin, Walter de Gruyter and Co., 220 p. Znamensky, E. B., 1957, Distribution of titanium in granites: Geokhimiya, no. 2, p. 112. Metamorphic Source Rocks of Titanium Placer Deposits— A Geochemical Cycle GEOLOGY AND RESOURCES OF TITANIUM GEOLOGICAL SURVEY PROFESSIONAL PAPER 959-B CONTENTS Page Page Abstract _________________________________________ B1 Formation of placer deposits ______________________ B6 Introduction ______________________________________ 1 Stream placers _______________________________ 6 Low-grade metamorphism _________________________ 1 Beach placers ________________________________ ’7 The lowest grades ____________________________ 1 Prelithification diagenesis of placers ____________ 7 Greenschist facies ____________________________ 1 Influence of source rock on placer deposits _______ 7 Lower amphibolite facies ————————————-_ ————————— 2 Application of the source-rock model to titanium Rutile in lower grade metamorphic rocks _________ 2 mineral provinces ___________________________ 8 Higher grade metamorphism ——————————————————————— 2 Western coast of Australia ____________________ 8 Titanium oxides in the sillimanite zone _________ 2 Eastern coast of Australia _____________________ 8 Titanium oxides in the granulite facies _________ 3 Southern India and Ceylon ____________________ 10 Weathem‘? "““i “““ ,‘““““"_‘ “““““““ 5 Western Africa _______________________________ 10 Chemlcal behavmr during weathering __________ 5 _ Weathering of titaniferous silicates ________ 5 Eastern Africa “"""'""“""""""'7“ 10 Weathering of titanium oxides _____________ 5 Eastern United States ———————————————————————— 10 Economic significance of weathering ____________ 6 References cited __________________________________ 13 ILLUSTRATIONS Page (FIGURE 1. Rutile in the Blue Ridge—Great Smoky part of the Knoxville 2- degree quadrangle _______________________________________ B3 2. Photomicrographs of rutile in high-grade metamorphic rocks ____ 4 3. Relationship of placers, drainage, and metamorphism along the western coast of Australia _______________________________ 9 4. Relationship of placers, drainage, and metamorphism in southern India and Ceylon ________________________________________ 11 5. Relationship of fossil and modern placers and metamorphism in the southeastern United States ___________________________ 12 TABLE Page TABLE 1. Provenance and types of deposits in some Ti02 provinces _______ B8 GEOLOGY AND RESOURCES OF TITANIUM METAMORPHIC SOURCE ROCKS OF TITANIUM PLACER DEPOSITS— A GEOCHEMICAL CYCLE By E. R. FORCE ABSTRACT Progressive normal (Barrovian) metamorphism shifts ti- tanium from sphene, biotite, and hornblende in lower meta- morphic grade rocks to ilmenite and rutile in rocks of silli- manite and higher metamorphic grade. Ilmenite and rutile, if present in the parent rock, are 'residually concentrated in its weathered product; when freed, they can form placer con— centrates in situations where contamination by other mineral suites is low and when hydraulic energy is great but hy- draulic equivalence is not attained. Many valuable placers are the result of an optimal geo- chemical cycle of high-grade metamorphism, repeated intense weathering, and short dispersal path. Complicating factors in this model are rock composition and oxygen pressure during metamorphism and intermediate sedimentary hosts during dis- persal. Titanium mineral provinces consist of different weathering and erosion products of the same source terrane. Saprolite and fluvial and beach placer deposits (of any age, if unmeta- morphosed) may each form resources in the millions of tons of contained Ti02. IN TRODUCTION In this paper, one possible geochemical cycle of titanium is followed from Barrovian metamorphism through weathering into the realm of sedimentation. The purpose is to put the formation of titanium placer deposits into a geologic context that may have predictive value. The cycle starts With the beginning of metamorphism because, although the parent-rock types may vary widely, metamorphism of even the lowest grades usually repartitions the titanium in the rock. Even during slight metamorphism of placer deposits, the mineralogy of titanium is usually reset, an indication that the “cycle” is complete. LOW-GRADE METAMORPHISM This section includes discussion of metamorphic grades as high as the kyanite zone, which is substan- tially higher than common usage of the term “low- grade metamorphism.” The most important carrier of titanium in the lower grade metamorphic rocks is fine-grained sphene; intergrown anatase is also a significant car- rier. As prograde reactions produce biotite and horn- blende, they become important carriers. At low metamorphic grades, ilmenite and rutile from precursor rocks are unstable with respect to sphene and anatase, even in rocks rich in TiO2 such as slightly metamorphosed placer deposits or mag- matic segregations of titanium minerals (Harker, 1932, p. 6; Ross, 1941; Goldschmidt, 1954, p. 415; Carroll and others, 1957; Carpenter and others, 1966). In heavy-mineral laminations of sandstone from the Great Smoky Group in the biotite zone of metamorphism, i'ntergrowths of sphene and an un- known opaque mineral form pseudomorphs of titani- um oxides. THE LOWEST GRADES For those rocks that show zeolite facies and prehn- ite-pumpellyite facies metamorphism, sphene is present as a metamorphic mineral in many rock types (Coombs and others, 1959, p. 60; Seki, 1961; Turner, 1968, p. 265—266). In some literature (see Hutton, 1940, p. 62; Seki and others, 1971), sphene in these lowest grade rocks is referred to as “leu- coxene” because of its occurrence in very fine grained aggregates and dusty coatings. GREENSCHIST FACIES Sphene continues as a stable metamorphic phase through the greenschist facies and is present in many rock types (Turner, 1968, p. 271—284). Hut- ton (1940) described a transition within the chlorite zone; in the low-grade textural subzone, fine-grained “leucoxene” is in skeletal aggregates as pseudo- morphs after ilmenite and, together with chlorite, replaces titanaugite, biotite, and matrix. In the higher grade textural subzone, sphene occurs as rounded or spindle-shaped coarse and discrete grains. B1 B2 Although partitioning of titanium among minerals of low-grade rocks is difficult to study because of the fine grain sizes involved, most of the titanium in rocks of the chlorite zone of the greenschist facies must be in sphene because other minerals that con- tain titanium are absent, unless relict minerals per- sist. Anatase may be present as intergrowths with sphene. Magnetite may be present but contains very little titanium in these low-grade rocks (Abdullah and Atherton, 1964). If metamorphism reaches the biotite zone of the greenschist facies, some of the titanium can enter biotite. Brown (1967) showed that chlorite from the chlorite zone averaged 0.1 percent TiOZ, whereas biotite in adjacent rocks of the biotite zone averaged 1.2 percent Ti02. LOWER AMPHIBOLITE FACIES Partitioning of titanium in biotite-bearing rocks of the lower amphibolite facies probably remains un- changed from that of the biotite zone of the green- schist facies. Force (1976a, table 8) shows a parti- tioning analysis for such a rock; the analysis was calculated from data gathered by Mason (1962). In mafic rocks, hornblende is a carrier of significant amounts of titanium. RUTILE IN LOW’ER GRADE METAMORPHIC ROCKS Rutile can occur in the lower grade rocks—it is common, for example, in slate as an alteration prod- uct of elastic biotite (Harker, 1932, p. 46), but it appears to be restricted to pelitic or other aluminous rocks (Hutton, 1940; Zen, 1960; Espenshade and Potter, 1960; Mason, 1962; Albee and others, 1965). In mineral assemblages from which rutile has been described, the CaO content is low, which inhibits the formation of sphene, and (or) the A1203 content is high (Shcherbina, 1971) . In the rutile-bearing rocks investigated by Zen and by Espenshade and Potter, the ratio A1203:Ca0 is over 200. In other rocks of lower metamorphic grades, rutile forms by introduc- tion of fluids that cause the breakdown of ilmenite or biotite. The fluids may contain magnesium (Chi- dester, 1962, p. 69; Southwick, 1968), sulfur (Force, 1976b) , or C02 (Schuliling and Vink, 1967) . Occurrences of rutile in lower grade terranes may be numerous, but, because they are controlled by local alteration or special rock chemistry, they are not comparable to regional occurrences of rutile in higher grade terranes. Detrital rutile is generally de- stroyed in the lower grades of metamorphism, except in special circumstances where rutile is stable rela- tive to sphene at a low grade, as it is in an extremely GEOLOGY AND RESOURCES OF TITANIUM aluminous sediment. Rutile in higher grade rocks is thus generally metamorphic rather than detrital. HIGHER GRADE METAMORPHISM Several texts (Rankama and Sahama, 1950, p. 563; Deer and others, 1962b, p. 31; Turner, 1968, p. 325, 329) discuss in a general way the formation at the onset of granulite-facies metamorphism of ti- tanium oxides at the. expense of titaniferous silicates. Ramberg (1948, 1952, p. 72—75, 156—161) has cov- ered the topic in more detail than others. Except for Ramberg’s (1948) work, discussions of the process in a context of economic geology are few and obscure. The mechanism for the segregation of titanium into oxides seems clear on a superficial level. The titanium-rich silicates—biotite, hornblende, and sphene—have progressively restricted stability fields at the higher grades of metamorphism and may not be present at all in pyroxene granulites (Howie, 1955; Turner, 1968, p. 329). Silicates that form in their place (hypersthene, diopside, pyrope-almandine garnet, plagioclase, and potash feldspar) contain much smaller amounts of titanium that their pre— cursors; the excess Ti02 in the rock then forms 0x- ides. Which oxides form is partly a function of oxygen pressure. TITANIUM OXIDES IN THE SILLIMANITE ZONE Some authors have said that, in rocks of silliman- ite grade, biotite breaks down to sillimanite and oth- er minerals including ilmenite and rutile (Harker, 1932, p. 58; Abdullah and Atherton, 1964; Over- street and others, 1968, p. 15). This point remains moot, however, because other authors have inter- preted the same textural evidence as growth of silli- manite without destruction of biotite (Chinner, 1961 ; Evans and Guidotti, 1966). Reports of quartzofeldspathic gneisses and schists containing both sillimanite and rutile are too numer- ous to list. In many such rocks, sphene becomes un- stable with respect to rutile at a slightly higher grade than that at which sillimanite appears but, appar- ently, before the onset of granulite facies. Figure 1 shows the distribution of rutile in metamorphic rocks in parts of the Blue Ridge and Great Smoky Mountains of North Carolina. Rutile is primarily restricted to the sillimanite zone and is most common in or near sillimani-te-orthoclase— or hypersthene- zone rocks. Rutile in the kyanite zone was found only in pelitic rocks; none was found in staurolite and lower zones. The fact that isograds and rutile dis- METAMORPHIC SOURCE ROCKS OF TITANIUM PLACER DEPOSITS B3 360§3‘:OO' ‘ 3’0' 83 °OO" ‘ ‘ . ‘ \ 30' ("82 00’ ‘ EXPLANATION i\ ”‘FFFRSON ’ Newport 1 \('RI “F I grin/9% .. . / n max D: 3 x \r/ (,m KE 3’\\_, °_/ Burnsulg Rutile present ‘ o / S H " ‘ N\ :3 MADISON :9; Y AN( m; RZLJ o r' «in: Rutile absent ‘ l" / “3W:\ld/. " ; M H‘ ] (sph t ' t 1 3 ~ »~ / ene presen 1n mos samp 85)] p // \/}“L\ /// //‘;\PX\ /« /_}‘> /°; . ——'_" . l' ii mr x / ° 4/ \ x‘ Metamorphic isog'rad, i Ht M g f I w ° 0/“), //°/(/ dashed where approximate Rock-unit boundary (shown only near isograds) 30’ NRA HAM / / /// ,m. a} HYIJD’ERSTHE /"'/ i .- w / (‘ilEROKyi/g ///?/ // \ Cashiers \ o ' ) liayesvil c/"' / 2? 35°00 1% \«f/ gov {mum mom-ms , . Marmn \ < M(‘ DOW ELI. (3%“ ,,,,, M «>33 " KUl“HERPORD . Fietclzcr Rock hinim y f /\ \ \ H ltN [JERSON o’ ,2 o / . r n Hendersonvillv / ,/ 0‘7 )r/e rd / /~o’ OTRA/SYI VANIA POLK a \0&Qc:// e/’ L //Ye ‘V 31‘3“ SI‘ARTANBUKU a GREEN \‘l LLB ’1 20 MlLES O 10 20 KILOMETRES FIGURE 1.—Rutile in the Blue Ridge—Great Smoky part of the Knoxville 2-degree quadrangle (Hadley and Nelson, 1971). All rock-unit boundaries and zoned isogvrads except hypersthene after Hadley and Nelson. Rutile distribution and hyper- sthene isograd (boundary of granulite facies) based on study of about 500 thin sections obtained from Richard Gold- smith, F. G. Lesure, J. B. Hadley, and A. E. Nelson (U. S Geological Survey). Hypersthene is roughly coextensive with sillimanite-orthoclase pairs. Information from Franklin (F) and Dellwood (D) 71/2-min quadrangles (shaded areas) is summarized. tribution together cross numerous lithologic bound- aries in the sillimanite zone implies limits to the in- fluence of stratigraphic or compositional control of rutile appearance. In the area shown, many rocks have either sphene or rutile, but rocks having both are rare. A few rocks in zones lower in grade than sillimanite-orthoclase have rutile rimmed by sphene or ilmenite, but sphene probably is not in equilibri- um with rutile even in these rocks. Biotite is present in both sphene- and rutile-bearing rocks. Overstreet and others (1968, pl. 2) also observed that rutile distribution directly correlates with silli- manite distribution from the Piedmont province in North Carolina and South Carolina. Sphene is pres- ent in rocks of all grades locally present, but rutile is common only within the sillimanite zone (Over- street and Griflitts, 1955, p. 561). TITANIUM OXIDES IN THE GRANULITE FACIES Rutile and ilmenite are apparently stable phases in metamorphic facies lower than granulite, but they are much more common in granulite—facies rocks. In that facies, the stability fields of the titanium-rich silicates—sphene, hornblende, and biotite—are greatly restricted. Magnetite may also decrease. The disappearance of each mineral will be discussed in turn. Sphene—Many authors have mentioned the disap- pearance of sphene in the granulite facies (cf. Howie, 1955). Cooray (1962) found that rutile takes the place of sphene in leucocratic granulites. The most thorough treatment is given by Ramberg (1952, p. 72—75), who discussed reactions in which rutile and ilmenite form by the breakdown of sphene. In some lime-rich rocks, however, sphene remains a B4 stable phase in the granulite facies (Ramberg, 1952, p. 160; Deer and others, 1962a, p. 75) . Turner (1968, p. 325) has stated that the disap- pearance of sphene in amphibolites is a useful indi- cation of granulite facies. Engel and Engel (1962a, p. 53—57 ) found that the disappearance of sphene in amphibolite coincides roughly with the appearance of metamorphic clinopyroxene and precedes the ap- pearance of orthopyroxene. The titanium thus lib- erated from sphene apparently is incorporated in ilmenite, which shows an increase in average abun- dance from 2 percent in the lower grade rocks to 3.5 percent in the granulite facies. The difference is at- tributable to reactions involving hornblende and magnetite as well as sphene. Hornblende.—By definition, hornblende is stable with orthopyroxene in the hornblende granulite sub- facies (lower grade than pyroxene granulite subfa- cies (Turner, 1968) ). Hornblende contains as much as 3.9 percent TiOZ in granulite~facies rocks (Ram- berg, 1948, table 1; Eskola, 1952, p. 133; Howie, 1955, table 6), whereas in lower grade rocks horn- blende contains less TiO2 (Engel and Engel, 1962b; Bushlyakov, 1970). The increase of TiO2 in horn- blende is more than compensated for by the decrease in the amount of hornblende (Force, 1976a, table 6) , and the TiO2 present as hornblende decreases as metamorphic grade increases. In the pyroxene gran— ulite subfacies, the silicates that replace hornblende as stable phases contain little Ti02, and the TiO2 from hornblende forms oxides (Force, 1976a, table GEOLOGY AND RESOURCES 0F TITANIUM 8). In mafic rocks, ilmenite is the most common min- eral to form, but rutile is present in mafic granulites near Franklin in Macon County, N.C., and presum- ably elsewhere. In more silicic hypersthene gneisses, rutile is a more common metamorphic mineral (fig. 2A). Biotite.——Biotite is also stable in the lower sub- facies of the granulite facies; in quartzofeldspathic gneisases it may be the major mafic mineral (cf. Engel and Engel, 1958). Biotite in granulite~facies rocks has a TiO2 content as great as 6 percent (Ram- berg, 1948, table 2; Kretz, 1959; Engel and Engel, 1960, table 13). Although the titanium content of biotite increases with metamorphic grade (Engel and Engel, 1960; Kwak, 1968; Zakrutkin and Gri- gorenko, 1968; Bushlyakov, 1970), the total amount of TiO2 present in biotite (as in hornblende) de- creases as the amount of biotite decreases (Force, 1976a, table 6). Garnet, sillimanite, pyroxene, quartz, and feldspar that take the place of biotite are low in Ti02, and the excess TiO2 forms oxides (rutile in fig. 28). MagnMite—Metamorphic grade is one of the two factors that control the amount of magnetite in a metamorphic rock; the other is oxygen pressure. As metamorphic grade increases, the amount of mag- netite decreases and the amount of TiO2 present in the magnetite increases (Overstreet and Griflitts, 1955, p. 561; Buddington and others, 1963; Abdul- lah and Atherton, 1964) . The overall effect is to de- crease the TiO2 present in magnetite in granulite- FIGURE 2.——Photomicrographs of rutile in high-grade metamorphic rocks from the Franklin area of Macon County, N.C. A, Rutile (R) with pyroxene (P), pale amphibole (A), quartz, and two feldspars in gneiss of intermediate composition. Plane-polarized light, 250x. B, Rutile (R) with sillimanite (S), garnet (G), biotite (B), and perthitic potash feldspar (K) in sillimanite gneiss. Plane-polarized light, 250x. METAMORPHIC SOURCE ROCKS OF TITANIUM PLACER DEPOSITS facies rocks (calculations from Buddington and oth- ers, 1963, table 2). Under ideal conditions, the prog- ress of reactions reducing magnetite (and increas- ing ilmenite) might be mapped by using textures on aeromagnetic maps. High oxygen pressure also has the effect of de- creasing the amount of magnetite and increasing the amounts of hematite, ilmenite, rutile, or inter- growths of these minerals (Buddington and others, 1963; Lindsley, 1962) . Sphene, hornblende, and biotite may disappear a1- together, and the abundance of magnetite may de- crease in rocks of pyroxene granulite subfacies, in which no silicates contain appreciable amounts of Ti02, and the rock’s titanium is mostly in ilmenite and (or) rutile. Where eclogites are associated with granulites (Coleman and others, 1965), the charac- teristic high rutile content of eclogite makes it an ’ end member along with pyroxene granulite (Gold- schmidt, 1954; Turner, 1968; cf. Force, 1976a, table 8). WEATHERING The behavior of titanium and titanium-bearing minerals during weathering is not understood in detail. It seems clear, however, that titania as a chemical entity is residually enriched and that grains of rutile and “ilmenite” are residually enriched. Il- menite begins its alteration to “leucoxene” during weathering. CHEMICAL BEHAVIOR DURING WEATHERING The ratio of ionic charge to ionic radius of titani- um is such that during weathering it forms insoluble compounds (Mason, 1966, p. 163) that remain in the weathered rock. A classical study by Goldich (1938) showed TiO2 enriched over granitic gneiss parent rock by a factor of 4 in a residual clay (constant weight calculation). Studies in temperate and tropi- cal zones show residual enrichment of TiO2 in Weathered felsic and mafic rock (Goldich, 1938; Jackson and Sherman, 1953; Wells, 1960; Short, 1961; Dennen and Anderson, 1962; Loughnan, 1969, p. 90, 112; Parker, 1970). In Hawaii, Sherman (1952) observed soils over basalt that contain 25 percent Ti02, present mostly as comcretions; only under extraordinarily reducing conditions is TiO2 leached from these soils. In bauxitos enrichment of TiO2 over parent rocks by factors of 2 to 4 is charac- teristic (Patterson, 1967, table 4; Valeton, 1972, table 32) . B5 The chemical enrichment of titanium during weathering is reflected mineralogically by two proc- esses: (1) Titanium—bearing silicates and titanifer- ous magnetite alter to clays, iron oxides, and fine- grained dispersed titanium oxides and (2) titanium oxides resist weathering. WEATHERING OF TITANIFEROUS SILICATES Titanium-bearing silicates, with the exception of some occurrences of sphene (cf. Overstreet and oth- ers, 1963), break down during weathering. Gold- schmidt (1954, p. 419—420) recorded that sagenitic rutile is a product of weathering of biotite and that other fine rutile needles are a product of weathering of other silicate minerals. In bauxite, TiO2 occurs mostly in grains finer than 325 mesh (40p), which form by weathering of titaniferous silicates (Hart- man, 1959, p. 1380). Anatase is the most common al- teration product containing titanium in both tropi- cal (Valeton, 1972, p. 32,186) and temperate soils (McLaughlin, 1955). WEATHERING OF TITANIUM OXIDES Ilmenite is a relatively stable mineral in the weath- ering process (Jackson and Sherman, 1953, p. 232— 235; Loughnan, 1969, p. 25). In Liberia, West Africa, rocks that contain both ilmenite and magne- tite weather to saprolite that contains ilmenite but not magnetite (cf. White, 1972, p. 86) . Gullies cut in saprolite are floored with ilmenite and quartz grains. In Cleveland County, N.C., Overstreet and others (1963) found ilmenite to be a common constituent and magnetite a rare constituent of saprolite over many types of metamorphic rock; Goldich (1938) recorded an enrichment of ilmenite and impoverish- ment of magnetite in soil over a Minnesota gneiss. Ilmenite is also a residual mineral in high-alumina saprolite over basalt (Roedder, 1956; Hoste-rman, 1969, p. 35; Patterson, 1971, p. 26). In other baux- . ites over a variety of parent rocks, ilmenite grains are present, though commonly altered (Hartman, 1959) . “Ilmenite” as mined in placer deposits commonly contains much more TiO2 than the chemical formula for ilmenite allows and actually consists partly of fine-grained Ti02 minerals, Which form as alteration products (Force, 1976a). Some studies show that in situ weathering of parent rock can produce “ilmen- ite” that consists partly of fine-grained anatase and rutile alteration products (Hartman, 1959 (bauxite) : Carroll and others, 1957, p. 180; Houston and Murphy, 1962, p. 31 (weathered sandstone); BG Rumble, 1973 (weathered schist)). Jackson and Sherman (1953, p. 289—290) show that “leucoxene,” the TiOZ-rich end product of this alteration process, can form if weathering is intense. Cannon (1950, p. 209) noted that some alteration of ilmenite grains to material richer in TiO2 takes place during alluvial transport. Leucoxene can also be the result of post- depositional weathering of an unconsolidated placer deposit above the water table (Pirkle and Yoho, 1970). Rutile is considered by some (summarized by Pet- tijohn, 1957, p. 502—508) to be among the minerals most resistant to weathering. It is a constituent of saprolite in Cleveland County, NC (Overstreet and others, 1963), and Virginia (Fish, 1962). In Macon County, NC, gullies cut in gneiss saprolite are floored with grains of quartz, garnet, and rutile. Un- like ilmenite grains, rutile grains normally do not form other alteration products; microcrystalline rutile itself is an alteration product formed in the weathering zone, so metamorphic (or igneous) rutile is generally already stable there. Where weathering is intense, rutile high in niobium and iron may be unstable and recry‘stallize to aggregates of fine anatase, ECONOMIC SIGNIFICANCE OF WEATHERING Probably TiO2 is enriched by all normal weather- ing processes over all rocks, and the amount of en- richment is proportional to the intensity of weather- ing. Where the parent rocks contain TiO2 in poten- tially useful form—that is, oxides coarser than silt size, as in the high-grade metamorphic rocks previ- ously described—the friable weathering product is enriched in these minerals, which may even be leached of undesirable elements. Saprolite deposits of titanium are potential ores in some areas (Fish, 1962; Herz, 1976), and, because so little attention has been directed toward saprolite, the discovery of more titanium resources in saprolite is likely. FORMATION OF PLACER DEPOSITS The formation of placer deposits is a complicated topic approached here only in the context of the geo- chemical cycle of titanium. Udden (1914) observed that heavy minerals are finer than the light minerals with which they occur in elastic deposits; Rubey (1933) noted that small heavy minerals and large light minerals have equal settling velocities in water. Since then, most discus- sions of placer deposits have been approached from the viewpoint of Rittenhouse’s (1943) concept of GEOLOGY AND RESOURCES OF TITANIUM hydraulic equivalence (cf. Baker, 1962; Briggs, 1965, p. 940—941; Tourtelot, 1968). Simply stated, hy- draulic equivalence prevails when minerals of the same settling velocity are found together. More detailed sedimentologic studies, discussed herein, show that, in general, hydraulic equivalence does not prevail in natural deposits and that hy- draulic factors other than settling velocity enter in. Certainly, complications are involved in the forma- tion of rich placer concentrates; any single hydraulic mechanism can, at best, explain only the observed size relations of heavy minerals to light minerals but not the concentration of heavy minerals relative to light minerals that characterizes a rich placer. The complications may be hydraulic, such as local erosion or wind ablation, or partially hydraulic, such as na— tural heavy-media separations or lack of constant relative availability of mineral sizes. STREAM PLACERS The first possible site of placer concentration is in upstream deposits near the mineral source. Stream deposits of this type are most important for minerals having specific gravity greater than 7, such as gold and cassiterite (Emery and Noakes, 1968). Locally, they offer possibilities for significant concentrations of the titanium minerals rutile and ilmenite (Herz and others, 1970) and other heavy minerals having specific gravity around 5, such as monazite (Over- street and others, 1968). Upstream placer deposits are closely tied to particular source rocks, and the small drainage basin limits dilution by barren debris to a minimum. In the small drainage basins, how- ever, hydraulic sorting is generally poor. The volume of placer deposits may be small because of either bed- rock stream beds or narrow valleys or both. In Sierra Leone, an exceptional rutile deposit, probably a weathered upstream placer, is of both large volume and high grade (Spencer and Williams, 1967 ) . Farther downstream, dilution by sediment poor in valuable minerals is more severe (see summary of this and related processes by Pettijohn (1957, p. 556—57 3) ). Hydraulic sorting, however, is common- ly more effective downstream than in upstream areas and may locally have produced placer concentra- tions (cf. McGowen and Groat, 1971). Van Andel (1950) showed that deposition patterns of heavy minerals in rivers are not always in accord with predictions from hydraulic equivalency; sorting by size but not density is prevalent, and concentration of a given mineral occurs because it is supplied to the stream in a size that is locally concentrated METAMORPHIC SOURCE ROCKS OF TITANIUM PLACER DEPOSITS (Briggs, 1965; White and Williams, 1967). Experi- mental and theoretical work (Grigg and Rathbun, 1969; Brady and J obson, 1973) has shown that ero- sional (“reentrainment”) processes, which produce residual enrichments of fine-grained particles inde- pendent of density, are a factor in producing alluvi- al concentrations (cf. Hjulstrom, 1939). BEACH PLACERS Most important placer concentrations of ilmenite and rutile are modern or fossil beach deposits. Dunes associated with beach placers may also contain con- centrations of titanium minerals. In the beach de- posits, titanium minerals are concentrated along with other heavy minerals in lenses of black sand. Concentrations of heavy minerals can be observed to form in the upper swash zone during storms; dif- ferential erosion of larger light-mineral grains leaves behind a residue of heavy minerals (Cannon, 1950, p. 206; Whitworth, 1956; Gillson, 1959, p. 425). Overstreet (1972) stated that a natural jig- ging action of the sand on the beach by normal wave energy improves the concentrate by subjecting the rising light minerals to erosion. Differential remov- al of large light grains by wind may also improve the concentrate. Storm-generated beach concentrates may be dis- persed between storms, unless they are deposited be- yond normal high tide (Whitworth, 1956) . Deposits may be preserved from erosion by wind transport into the dune system beyond the beach (Cannon, 1950, p. 207). Neiheisel (1958) and Gillson (1959, p. 425) noticed that the heavy-mineral fraction of dune sands is poor in the heaviest minerals and that the highest dunes contain the least amount of heavy minerals. Apparently, wind cannot concentrate heavy minerals as efficiently as waves, but it can car- ry most of a concentrate formed on the beach well above high tide. Where modern beach placers are mined, the deposit is renewed to some extent by storm waves (Whitworth, 1956; Gillson, 1959, p. 428; Overstreet, 1972). The grain size of a mineral in beach deposits is generally in inverse proportion to its density. Rutile, however, is commonly of finer grain size than its density would predict, probably because rutile is fine grained in original source rocks (Gillson, 1959, p. 423). Some detailed studies have shown that light and heavy minerals are generally not hydraulically equivaler' in beach deposits, and the difference has been ascribed to the lack of constant relative avail- ability of minerals of all sizes (McMaster, 1954, p. B7 197, 211) or to the effects of preferential wave ero- sion of coarser light grains that stick up out of the bed (McIntyre, 1959, p. 294—296; Hand, 1967, p. 516). Hand (1967 ) found that the heaviest minerals de- parted the most from hydraulic equivalency with the light minerals. Martens (1935) found that the heavy-mineral suites of lean natural concentrates are poor in the heaviest minerals, whereas those of the richest concentrates consist almost entirely of the heaviest minerals. The combination of both lines of evidence suggests that rich concentrates are the result of significant departures from hydraulic equivalence. PRELITHIFICATION DIAGENESIS OF PLACERS During early diagenesis of placer deposits, un- stable heavy minerals may disappear or become skeletal in shape (Hutton, 1959; Neiheisel, 1962) and thus enrich a placer concentrate in the more stable titanium minerals. Ilmenite may be leached of iron and other elements (Dryden and Dryden, 1946; Temple, 1966) or even complete its alteration to “leucoxene” if it is above the water table (Pirkle and Yoho, 1970), but ilmenite and rutile are re- sistant to complete destruction by intrastratal solu- tion (Pettijohn, 1957, p. 502—520). INFLUENCE OF SOURCE ROCK ON PLACER DEPOSITS The mineralogy of placer deposits is inherited from source rocks, with modifications by the weath- ering process. Formation of a heavy-mineral concen- trate is apparently the result of several hydraulic and some nonhydraulic factors, one of which, the- relative abundance of different sizes of each mineral, is largely inherited from source rocks. Although the formation of placer deposits can be studied from a hydrodynamic viewpoint, the influence of source rocks is certainly strong enough to warrant a search for optimal conditions where mineralogy and grain size of the source rock are favorable and dilution by other debris is low. Formation of placer concentrates of ilmenite and rutile raises the TiO2 content of the depdsit far be- yond its average value in sandstone. The converse, however, is not true; anomalous TiO2 values in heavy-mineral concentrates need not reflect the presence of ilmenite and rutile but may be caused by fine magnetite-ilmenite intergrowths (Talati, 1971), titaniferous magnetite (Wright, 1964), or sphene (Thomas and Berryhill, 1962), all of which locally B8 form heavy-mineral concentrations. In unusually immature deposits, a high TiO2 content may even reflect the abundance of titanaugite and hornblende (Thomas and Berryhill, 1962) . APPLICATION OF THE SOURCE-ROCK MODEL TO TITANIUM MINERAL PROVINCES Where processes of metamorphism, weathering, and sedimentation are optimal, provinces of titani- um deposits can occur. In these provinces, titanium deposits are likely to be of several types, including saprolites, modern or fossil fluvial placers,‘a.nd mod,- ern or fossil marine placers (table 1). Any type may contain millions of tons of Ti02, as in saprolite near Roseland, Va.; river deposits at Sherbro, Sierra Leone; modern beach deposits of India; or fossil beach deposits at Trail Ridge, Fla. (values from Klemic and others, 1973). The most valuable de- posits may consist of titanium minerals that have passed through many of these stages without sub- stantial dilution, such as, for example, a weathered fossil beach deposit, material for which was derived from weathered older sedimentary rocks, which were, in turn, derived from a deeply weathered acidic granulite-facies terrane. Relations between source rocks and placer deposits can be studied in some detail where intermediate host rocks are not important or where ancient dispersal patterns are known. WESTERN COAST OF AUSTRALIA For ilmenite and rutile beach placer deposits on the western coast of Australia, an almost direct con- nection can be made between placers and source rocks. The rivers drain limited areas along the coast; the interior is poorly drained. Placer deposits are thus near their possible source rocks, and dilution by other debris is low. GEOLOGY AND RESOURCES OF TITANIUM Beach placer deposits of ilmenite at Cape] and Bunbury (fig. 3) occur in a suite, which also con- tains zircon, garnet, sillimanite, and pyroxene simi- lar to detritus from the Cape Naturaliste area and gneisses to the east (Carroll, 1939). Granulite—facies metamorphic rocks are common in both areas, and laterite cover is widespread (Lowry and others, 1967; Wilson, 1964, 1969). Nearby Tertiary and Permian sedimentary rocks also have probably con- tributed debris to the placer deposits but have a rela- tively impoverished heavy-mineral suite (Carroll, 1939); they could not have been the predominant source, A recently discovered rutile-ilmenite placer de- posit at Eneabba, western Australia (fig. 3), is an immature Tertiary beach deposit formed by the re- working of a fossil alluvial fan (Baxter, 1972). The immediate source rocks for the fan and beach de- posit are continental Mesozoic sandstones. The Meso- zoic rocks and the placer deposit are lateritized. The ultimate source rocks are gneisses that are locally of granulite facies and predominantly of acidic com- position (Wilson, 1964, 1969). Whitworth (1956) mentioned a beach placer ru- tile deposit n-ear Albany (fig. 3) that apparently is not being worked. Nearby streams drain small areas in which granulite-facies gneisses are prominent (Wilson, 1964, 1969). EASTERN COAST OF AUSTRALIA The beach deposits of rutile on the eastern coast of Australia, at present the world’s most important rutile reserve, only very weakly support a case con- necting placer to source rock; despite some careful work, the full transportation history of this rutile is not known. Streams draining Permian and Mesozoic sandstones are the immediate source of the rutile (Whitworth, 1956; Connah, 1961), but the mineral assemblage in the placer deposits is unlike that in TABLE 1.—Prove'mmce and types of deposits in some TIOz provinces [X, titanium resource; '!, not documented but likely; (2), two sedimentary cycles; P, present; __, not known] ' - . er Modern Modem migzgigldiigc $322232; Oldhzzlzimggifiary ‘Zgilmerfgarf’;d stream marine bedrock bedrock host rock placer placer Cape] and Bunbury, western coast of P P P ? ‘I X Australia. Eneabba, western coast of Australia.-- P ? (2) , P X __ __ Eastern coast of Australia _________ ? ? P ? ? X Southern India and Ceylon _________ P P X X ? X Sherbro, Sierra. Leone ______________ P X, 'Z X X __ __ Man, Ivory Coast __________________ P P __ __ X __, Roseland, Va. _____________________ X X X __ X __ Franklin, N.C. ____________________ P P X P X _._ Lakehurst, N.J. ___________________ P __ (2) , P, X X __ __ Trail Ridge, Florida-Georgia _______ P (2), P, X X __ __ METAMORPHIC SOURCE ROCKS OF TITANIUM PLACER DEPOSITS B9 1 10° 1 15° 120° I | | INDIAN OCEAN Port Hedland 2 °—7 . ~— 0 f, 92 °“" Dean) 5 a ' “'3. g EXPLANATION ‘ X Ilmenite (I) or rutile (R) deposit being mined “‘46 20(0)? ,, / 5a , . . . . Riyflr ~ ”‘4‘; Occurrence of Ilmemte or rutile, economlc \_ significance unknown \ \ Camarvon 5 Gasw ”9 Re 255 _ ner _ it '\ Granulite facies metamorphic terrane ~ Other high-grade metamorphic terrane Sandstone . Q Geraldton Possible intermediate sedimentary host EN EABBA (1R 0 30° — — .Kalgoorlie CAPEL -BUNBURY(I) CAPE NATURALISTE I D ° ° :0 o H D D .. 35°“ 6% Albany —< 0 100 200 300 400 MILES I l I . I I I 4 I I I I I I I I 0 100 200 300 400 KILOMETRES I I I I I FIGURE 3.—Relationship of beach placer titanium deposits, modern drainage, possible intermediate sedimentary hosts, and granulite-facies metamorphic terrane along the western coast of Australia. Metamorphism is from the Geo- logic Map of Western Australia (Horwitz, 1966) and Wilson (1969). Titanium deposits are from Regional Divisions and Reported Mineral Occurrences in Western Australia (Western Australia Department of Mines, 1968) and Baxter (1972). any local primary rocks (Whitworth, 1956). Over- grade than any locally present must have formed street (1967, p. 84—91) found, on the basis of mona- the source. Mesozoic drainage patterns (Brown and zite composition, that rocks of higher metamorphic others, 1968) were such that the assemblage may B10 have been derived from far in the interior where several large areas of granulite-facies rocks are present (Wilson, 1964, 1969). Most of these hypo- thetical source areas are at present poorly drained, but nearby desert sand contains rutile (Layton, 1966). A better picture of possible source rocks could be given if the distribution of sillimanite-bear- ing rocks were known, since reactions at sillimanite grade are known in other areas to produce rutile, as reactions at granulite grade do. SOUTHERN INDIA AND CEYLON India has long been an important producer of high-TiO2 ilmenite. Figure 4 shows the locations of the mining operations, all of which are on modern beaches. Rutile and monazite are coproducts; garnet, sillimanite, and hypersthene are also common; mag- netite is present but in small amounts (Jacob, 1956). The deposits are localized by headlands and, in most places, can be linked to specific streams (Jacob, 1956). The immediate sources of the ilmenite are young but weathered coastal plain sands (Jacob, 1956; Gillson, 1959; Overstreet, 1972, p. 48—57) and thick laterite deposits (Gillson, 1959); these have the same heavy-mineral assemblage as the beach placers (Menon, 1966; Rajamanickam, 1968). The ultimate source is deeply weathered crystalline rock of the interior, of which granulite-charnockite ter- rane forms a prominent part. The most important sources of ilmenite, rutile, and monazite are appar- ently leptynites (quartz-garnet-feldspar gneisses) and charnockites (Jacob, 1956; Menon, 1966). The distribution of placer deposits matches fairly well the small drainages having headwaters in granulite terrane (fig. 4) ; the fit in sillimanite-bearing rocks might be better. Relatively modern geomorphic evolution of the drainage has been important in controlling precise deposit locations in India (Gillson, 1959, fig. 2, p. 429). Unknown geomorphic changes are likely to have had great effect on the dispersal of titanium minerals into older intermediate sedimentary host rocks. Fine tuning of relations between source rocks and placers is probably not worthwhile. It might be thought that ilmenitic placers could be derived from the Deccan traps, but heavy-mineral concentrates on the coast near the traps consist part- ly of titaniferous magnetite and titanaugite (Talati, 1971). The previously discussed placers mined in southern India are deficient in these two minerals (Jacob, 1956). GEOLOGY AND RESOURCES OF TITANIUM Placer deposits of ilmenite, rutile, and monazite are also important in Ceylon. The major deposit is at Pulmoddai, on the eastern coast. Metamorphic rocks of the pyroxene granulite facies are so preva- lent in Ceylon that relating them to the location of placer deposits becomes trivial. Cooray (1962, table 1, p. 263) found that leptynites and khondalites com- mon within the granulite facies contain rutile, whereas other gneisses at lower grades contain sphene. Overstreet (1972) stated that the location of modern placers is controlled to a significant degree by an intermediate Pleistocene host, which is itself an important ilmenite resource. WESTERN AFRICA The literature on western African metamorphic belts and titanium deposits is insufficient to make a definitive association between them. Some informa- tion from Sierra Leone and the Ivory Coast is suggestive. Sherbro, in Sierra Leone, is a large rutile deposit, probably formed as an alluvial placer in a small drainage area. Some saprolite may be present in the deposit also. The deposit itself has been lateritized. Besides rutile, the sands contain zircon, ilmenite, monazite, sillimanite, and kyanite (Spencer and Wil- liams, 1967). Local source rocks are the Kasila belt of gneisses in the granulite facies (Andrews-J ones, 1966, p. 22—23). Rutile in alluvial concentrates is closely associated with granulite terrane throughout the belt (Junner, 1929) . In the Ivory Coast, alluvial placer deposits of ru- tile near Danane are found in a restricted drainage and are reportedly derived from rutile-bearing lep- tynites (Bagarre, 1964). In the Whole district, the occurrence of alluvial rutile is closely associated with the granulite-charnockite province of Man. EASTERN AFRICA Bloomfield (1958) noted that, in the granulite- facies terrane (Morel, 1961) of Malawi, rutile oc- curs in biotite-free varieties of acidic garnet-pyrox- ene gneiss and in associated migmatite and garnet pegmatite. Colluvium of the adjacent Shire River plain contains rutile, ilmenite, and garnet; resource information has apparently not been published. EASTERN UNITED STATES Deposits of ilmenite in the coastal plains of New Jersey and Florida-Georgia are the most important placer deposits of titanium minerals in the United States. Alluvial deposits, probably of less impor- METAMORPHIC SOURCE ROCKS 0F TITANIUM PLACER DEPOSITS Bl]. 74° 76° 78° 80° 82° N/ | l V l ‘ A /v \\ T , 9““ z 1; _ 14°—— ’ ‘ g «v I 3 ,9» to f‘ ‘P I w / I N D A 31 ' Madras Z .Bangalore 0‘ <9) _ 12° —— V EXPLANATION BAY OF 5% Ilmenite deposit being mined 5% 141‘ BENGAL 5% Concentration of ilmenite, economic 10., _ significance unknown Granulite-charnockite terrane Possible intermediate sedimentary hosts 8°— MANAVALAKURICHI X CEYLON 0 100 200 MILES l l l 1 l | [ I l l | l 1 0 100 200 KILOMETRES 6"— INDIAN OCEAN l | l I l FIGURE 4,—Relationeship of beach placer titanium deposits, modern drainage, possible intermediate sedimentary hosts, and granulite~facies metamorphic terrane in southern India and Ceylon. Metamorphism and sedimentary rocks are from the India Geological Survey (1963), Narayanaswami and Mahadevan (1964), and Cooray (1968). Titanium deposits are from Jacob (1956) and the U.N. Economic Commission for Asia and the Far East (1963). tance, in the Blue Ridge province will be discussed In the Blue Ridge province near Roseland, Va., a first, however, since the association between deposits Precambrian alkalic anorthosite body is intruded and source rocks is simpler. into gneisses that locally are in the granulite facies B12 (Herz, 1968). Anorthosite and adjacent gneiss con- tain rutile; anorthosite, dike rocks, and gneiss con- tain ilmenite. This area was formerly an important titanium mining district (Ross, 1941). Much of the titanium resource of the area is in saprolite over anorthosite, dike rocks, and gneiss (Fish, 1962). A Cambrian sandstone nearby contains a fossil placer of ilmenite and subordinate rutile (Bloomer and de Witt, 1941). Modern placer deposits of ilmenite and rutile are forming in streams draining anorthosite and gneiss (Herz and others, 1970). Ilmenite placers are also forming in streams draining gneiss alone. Near Franklin, in Macon County, N.C.,.also in the Blue Ridge province, some gneisses of upper amphib- olite facies (sillimanite grade) and most gneisses of granulite facies contain rutile (fig. 1). Saprolite over these gneisses is residually enriched in rutile and i1- menite, and alluvium from natural washing of sapro- lite in small gullies consists primarily of quartz, gar- net, rutile, and ilmenite. Larger streams draining 85" GEOLOGY AND RESOURCES OF TITANIUM the area carry the heavy minerals garnet, ilmenite, kyanite-sillimanite, zircon, and monazite in average order of abundance. Five short channel samples of sandy alluvium from the Little Tennessee River and six samples from the river’s larger tributaries in this area average about 0.25 percent rutile and about 1.0 percent ilmenite. Large-volume deposits of this material have not been discovered; placer and re- sidual accumulations of rutile have been mined at nearby Shooting Creek. Precambrian(?) sandstones to the north and east locally contain detrital concen- trations of titanium minerals (Carpenter and others, 1966), pseudomorphous after rutile and ilmenite. “Ilmenite” deposits in young coastal plain forma- tions near Lakehurst, N.J., were discovered by Markewicz (1969) , who used a source-rock and dis- persal-path model. A major hypothetical source rock is gneiss, much of it in granulite facies, of the New Jersey highlands, and the hypothetical dispersal path is an ancestral Delaware River. The deposition- l i E Knoxvillefl/ i ./' . /. I; W. J - ' \yB: ham§\ / 2\{‘, a 30° 7» MEXICO O 50 100 150 MILES i—T—Lrfii__l O 50 100 150 KILOMETRES TE‘ )0, . .{r H \t ‘ ~M‘ \, 1\ Raleigh) /, 1 a is A ‘ .- .Charlott L._ _. E (A b 009 Charleston EXPLANATION 1 X \ Ilmenite deposit being mined 5% Concentration of ilmenite, economic significance unknown E Sillimanite zone FIGURE 5,—Relationship of fossil and modern placer titanium deposits, possible intermediate sedimentary hosts, and meta- morphic terrane in the southeastern United States. Metamorphism is from Morgan (1972). METAMORPHIC SOURCE ROCKS OF TITANIUM PLACER DEPOSITS a1 environment of the main formation in which “il- menite” was finally deposited is in dispute; it may have been a beach. “Ilmenite” here was formed by weathering and is actually a black, slightly magnetic aggregate of fine rutile and iron oxides much higher in TiO2 than is true ilmenite (Markewicz, 1969). “Ilmenite” deposits of Trail Ridge, in Florida and Georgia, are in a Pleistocene beach deposit probably formed as a spit connected to an eroding headland of older coastal plain deposits (Pirkle and Yoho, 1970). “Ilmenite” is a weathered aggregate contain- ing fine-grained rutile (Garnar, 1972) and occurs with rutile and zircon. There were probably two main sources of titanium minerals (fig. 5); one is the previously discussed high-grade metamorphic terrane in the Blue Ridge of North Carolina, and the other is a high—grade metamorphic terrane in the Piedmont province of North and South Carolina (Overstreet and others, 1968). Their relative con- tribution as sources is uncertain. A history of the Trail Ridge deposits would probably record trans- portation and deposition in intermediate host rocks. REFERENCES CITED Abdullah, M. I., and Atherton, M. P., 1964, The thermometric significance of magnetite in low-grade metamorphic rocks: Am. Jour. Sci., v. 262, p. 904—917. Albee, A. L., Bingham, Elizabeth, Chodos, A. A., and Maynes, A. D., 1965, Phase equilibria in three assemblages of kyanite-zone pelitic schists, Lincoln Mountain quadrangle, central Vermont: Jour. Petrology, v. 6, p. 246—301. Andrews-Jones, D. A., 1966, Geology and mineral resources of the northern Kambui schist belt and adjacent granu- lites: Sierra Leone Geol. Survey Bull. 6, 100 p. Bagarre, E., 1964, Prospection pour rutile au sud-ouest de Man, region Ganleu-Singouine: Soc. pour le Devel. Minier de la cote d’Ivoire Rap. 70, 11 p. Baker, George, 1962, Detrital heavy minerals in natural ac- cumulates: Australasian Inst. Mining and Metallurgy, Mon. Ser. 1, 146 p. Baxter, J., 1972, The geology of the Eneabba area, western Australia, in Report of the Department of Mines, western Australia, for the year 1971: Perth, Western Australia Dept. Mines, p. 105—106. Bloomer, R. 0., and de Witt, Wallace, Jr., 1941, Titaniferous sandstone near Buena Vista, Virginia: Econ. Geology, v. 36, p. 745—747. Bloomfield, Keith, 1958, The geology of the Port Herald area [Nyasaland]: [Nyasaland] Geol. Survey Dept. Bull. 9, 76 p. Brady, L. L., and Jobson, H. E., 1973, An experimental study of heavy-mineral segregation under alluvial-flow condi- tions: U.S. Geol. Survey Prof. Paper 562—K, 30 p. Briggs, L. I., 1965, Heavy mineral correlations and proven- ances: Jour. Sed. Petrology, v. 35, p. 939—955. Brown, D. A., Campbell, K. S. W., and Crook, K. A. W., 1968, The geological evolution of Australia and New Zealand: London, Pergamon Press, 409 p. B13 Brown, E. H., 1967, The greenschist facies in parts of east- ern Otago, New Zealand: Contr. Mineralogy and Petrol- ogy, v. 14, p. 259—292. Buddington, A. F., Fahey, Joseph, and Vlisidis, Angelina, 1963, Degree of oxidation of Adirondack iron oxide and iron-titanium oxide minerals in relation to» petrogeny: Jour. Petrology, v. 4, p. 138—169. Bushlyakov, I. N., 1970, Titanium content of amphibole and biotite from granitoids as an indicator of the conditions under which they formed: Acad. Sci. USSR Doklady, Earth Sci. Sec., v. 186, p. 171—174. Cannon, H. B., 1950, Economic minerals in the beach sands of the southeastern United States, in Snyder, F. G., ed., Proceedings of the symposium on mineral resources of the southeastern United States: Knoxville, Univ. of Tennessee Press, p. 202—210. Carpenter, R. H., Wedow, Helmuth, Jr., and Maher, S. W., 1966, “Fossil” placers in the Precambrian 0coee Series, Cocke County, Tennessee [abs.]: Geol. Soc. America, Southeastern Sec., Ann. Mtg, Athens, Ga., April 13—16, 1966, Abs. with program, p. 17—18. Carroll, Dorothy, 1939, Beach sands from Bunbury, western Australia: Jour. Sed. Petrology, v. 9, p. 95—104. Carroll, Dorothy, Neuman, R. B., and Jafl‘e, H. W., 1957, Heavy minerals in arenaceous beds in parts of the 0coee Series, Great Smoky Mountains, Tennessee: Am. Jour. Sci., v. 255, p. 175—193. Chidester, A. H., 1962, Petrology and geochemistry of selected talc-bearing ultramafic rocks and adjacent country rocks in north-central Vermont: U.S. Geol. Survey Prof. Paper 345, 207 p. Chinner, G. A., 1961, The origin of sillimanite in Glen Clova, Angus: Jour. Petrology, v. 2, p. 312—323. Coleman, R. G., Lee, D. E., Beatty, L. B., and Brannock, W. W., 1965, Eclogites and eclcgites—their differences and similarities: Geol Soc. America Bull., v. 76, p. 483—508. Connah, T. H., 1961, Beach sand heavy mineral deposits of Queensland: Queensland Geol. Survey Pub. 302, 31 p. Coo-mbs, D. 8., Ellis, A. J., Fyfe, W. S., and Taylor, A. M., 1959, The zeoliite facies, with comments on the interpreta- tion of hydrothermal syntheses: Geochim. et Cosmochim. Acta, v. 17, p. 53—107. Cooray, P. G., 1962, Charnockites and their associated gneisses in the Pre-Cambrian of Ceylon: Geol. Soc. London Quart. Jour., v. 118, p. 239—273. 1968, Provisional geologic map of Ceylon: 1 :2,000,000. Deer, W. A., Howie, R. A., and Zussman, Jack, 1962a, Rock- forming minerals, v. 1, Ortho- and ring silicates: New York, John Wiley and Sons, 333 p. 1962b, Rock-forming minerals, v. 5, Non-silicates: New York, John Wiley and Sons, 371 p. Dennen, W. H., and Anderson, P. J., 1962, Chemical changes in incipient rock weathering: Geol. Soc. America Bull., v. 73, p. 375—384. Dryden, A. L., Jr., and Dryden, Clarissa, 1946, Comparative rates of weathering of some common heavy minerals: Jour. Sed. Petrology, v. 16, p. 91—96. Emery, K. 0., and Noakes, L. C., 1968, Economic placer de- posits of the continental shelf: U.N. ECAFE CCOP Tech. Bull., v. 1, p. 95—111; reprinted in Offshore, v. 29, no. 3, 1969. scale B14 Engel, A. E. J., and Engel, C. G., 1958, Progressive meta- morphism and granitization of the major paragneiss, northwest Adirondack Mountains, New York, pt. 1, Total rock: Geol. Soc. America Bull., v. 69, p. 1369—1414. 1960, Progressive metamorphism and granitization of the major paragneiss, northwest Adirondack Mountains, New York, pt. 2, Mineralogy: Geol. Soc. America Bull., v. 71, p. 1—58. 1962a, Progressive metamorphism of amphibolite, northwest Adirondack Mountains, New York, in Petrologic studies, a volume in honor of A. F. Buddington: New York, Geol. Soc. America, p. 37—82. 1962b, Hornblendes formed during progressive meta- morphism of amphibolites, northwest Adirondack Moun- tains, New York: Geol. Soc. America Bull., v. 73, p. 1499— 1514. Eskola, P. E., 1952, On the granulites of Lapland: Am. Jour. Sci., Bowen Volume, p. 133—171. Espenshade, G. H., and Potter, D. B., 1960, Kyanite, silli- manite, and andalusite deposits of the southeastern States: U.S. Geol. Survey Prof. Paper 336, 121 p. Evans, B. W., and Guidotti, C. V., 1966, The sillimanite- potash feldspar isograd in western Maine, U.S.A.: Contr. Mineralogy and Petrology, v. 12, p. 25—62. Fish, G. E., Jr., 1962, Titanium resources of Nelson and Am- herst Counties, Virginia, pt. 1, Saprolite ores: U.S. Bur. Mines Rept. Inv. 6094, 44 p. Farce, E. R., 1976a, Titanium contents and titanium parti- tioning in rocks, in Geology and resources of titanium: U.S. Geol. Survey Prof. Paper 959—A, 10 p. 1976b, Titanium minerals in deposits of other minerals, in Geology and resources of titanium: U.S. Geol. Survey Prof. Paper 959—F, 5 p. Garnar, T. E., Jr., 1972, Economic geology of Florida heavy mineral deposits, in Puri, H. 8., Proc. 7th Forum on Geology of Industrial Minerals: Florida Bur. Geology, Spec. Pub. 17, p. 17—21. Gillson, J. L., 1959, Sand deposits of titanium minerals: Min- ing Eng., v. 11, p. 421—429. Goldich, S. S., 1938, A study in rock weathering: Jour. Geology, v. 46, p. 17-58. Goldschmidt, V. M., 1954, Geochemistry: Oxford, Clarendon Press, 730 p. Grigg, N. S., and Rathbun, R. E., 1969, Hydraulic equival- ence of minerals with a consideration of the reentrain- ment process: U.S. Geol. Survey Prof. Paper 650—B, p. B77—B80. Hadley, J. B., and Nelson, A. E., 1971, Geologic map of the Knoxville quadrangle, North Carolina, Tennessee, and South Carolina: U.S. Geol. Survey Misc. Geol. Inv. Map I—654, scale 1:250,000. Hand, B. M., 1967, Differentiation of beach and dune sands, usirng settling velocities of , light and heavy minerals: Jour. Sed. Petrology, v. 37, p. 514—520. Harker, Alfred, 1932, Metamorphism—A study of the trans- formations of rock masses: London, Metheun, 360 p. Hartman, J. A., 1959, The titanium mineralogy of certain bauxites and their parent materials: Econ. Geology, v. 54, p. 1380—1405. Herz, Norman, 1968, The Roseland alkalic anorthosite massif, Virginia, in Isachsen, Y. W., ed., Origin of anorthosite and related rocks: New York State Mus. and Sci. Serv- ice Mem. 18, p. 357—36 [1969]. GEOLOGY AND RESOURCES 0F TITANIUM 1976, Titanium deposits in alkalic igneous rocks, in Geology and resources of titanium: U.S. Geol. Survey Prof. Paper 959—E, 6 p. Herz, Norman, Valentine, L. E., and Iberall, E. R., 1970, Rutile and ilmenite placer deposits, Roseland district, Nelson and Amherst Counties, Virginia: U.S. Geol. Sur- vey Bull. 1312—F, 19 p. Hjulstriim, Flip, 1939, Transportation of detritus by moving water, in Trask, P. D., ed., Recent marine sediments: Tulsa, Okla., Am. Assoc. Petroleum Geologists, p. 5-31. Horwitz, R. C., 1966, compiler, Geologic map of western Australia: East Perth, Western Australia Geol. Survey, scale 1:2,500,000. Hosterman, J. W., 1969, Clay deposits of Spokane County, Washington: U.S. Geol. Survey Bull. 1270, 96 p. Houston, R. S., and Murphy, J. F., 1962., Titaniferous black sandstone deposits of Wyoming: Geol. Soc. Wyoming Bull. 49, 120 p. Howie, R. A., 1955, The geochemistry of the charnockite series of Madras, India: Royal Soc. Edinburgh Proc., v. 62, p. 725—768. Hutton, C. 0., 1940, Metamorphism in the Lake Wakatipu region, western Otago, New Zealand: New Zealand Dept. Sci. and Indus. Res. Geol. Mem. 5, 90 p. 1959, Mineralogy of beach sands between Halfmoon and Monterey Bays, California: California Div. Mines and Geology Spec. Rept. 59, 32 p. India Geological Survey, 1963, Geologic map of India: Delhi, scale 122,000,000. Jackson, M. L., and Sherman, G. D., 1953, Chemical weather- ing of minerals in soils, in Norman, A. G., ed., Advances in agronomy, v. 5: New York, Academic Press, p. 217— 319. Jacob, K., 1956, Ilmenite and garnet sands of the Chowghat, Tinnevelly, Ramnad, and Tanjore coasts: India Geol. Survey Rec., v. 82, p. 567—602. Junner, N. R., 1929, Report of the Geological Department for part of the year 1927 and for the year 1928: Sierra Leone Geol. Survey Dept. Rept., p. 1—17. Klemic, Harry, Marsh, S. P., and Cooper, Margaret, 1973, Titanium, in Brobst, D. A., and Pratt, W. P., eds., United States mineral resources: U.S. Geol. Survey Prof. Paper 820, p. 653—665. Kretz, R. A., 1959, Chemical study of garnet, biotite, and hornblende for gneisses of southwestern Quebec with emphasis on distribution of elements in coexisting min- erals: Jour. Geology, v. 67, p. 371—402. Kwak, T. A. P., 1968, Titanium in biotite and muscovite as an indication of metamorphic grade in almandine-am- phibolite facies rocks from Sudbury, Ontario: Geochim. et Cosmochim. Acta, v. 32, p. 1222—1229. Layton, W., 1966, Prospects of off—shore mineral deposits on the eastern seaboard off Australia: Mining Mag., v. 115, p. 344—351. Lindsley, D. H., 1962, Investigations of the system FeO- Fezou-TiOg: Carnegie Inst. Washington Year Book, v. 61, p. 100—106. Loughman, F. C., 1969, Chemical weathering of the silicate minerals: New York, American Elsevier Publshing Co., 154 p. METAMORPHIC SOURCE ROCKS 0F TITANIUM PLACER DEPOSITS Lowry, D. C., Horwitz, R. C., and Gemuts, 1., 1967, Busselton and Augusta, western Australia: Australia Bur. Mineral Resources, Geology and Geophysics Bull., 1:250,000 geo— logical series, Australian national grid, sheets SI/50 and SI/50—9. Markewicz, F. E., 1969, Ilmenite deposits of the New Jersey coastal plain, in Geology of selected areas in New Jersey and eastern Pennsylvania: New Brunswick, N.J., Rutgers Univ. Press, p. 363—382. Martens, J. H. C., 1935, Beach sands between Charleston, South Carolina, and Miami, Florida: Geol. Soc. America Bull., v. 46, p. 1563—1596. Mason, Brian, 1962, Metamorphism in the Southern Alps of New Zealand: Am. Mus. Nat. History Bull., v. 123, p. 211—248. 1966, Principles of geochemistry (3d ed.) : New York, John Wiley and Sons, 320 p. McGowen, J. H., and Groat, C. G., 1971, Van Horn Sand- stone, west Texas—An alluvial fan model for mineral exploration: Texas Univ. Bur. Econ. Geology Rept. Inv. 72, 57 p. McIntyre, D. D., 1959, The hydraulic equivalence and size distributions of some mineral grains from a beach [On- tario]: Jowr. Geology, v. 67, p. 278—301. McLaughlin, R. J. W., 1955, Geochemical changes due to weathering under varying climatic conditions: Geochim. et Cosmochim. Acta, v. 8, p. 109—130. McMaster, R. L., 1954, Petrography and genesis of the New Jersey beach sands: New Jersey Dept. Conserv., Ser. Bull. 63, 239 p. Menon, K. K., 1966, Sequence and source of the Tertiaries neair Karrachal, Trivandrum District: Geol. Soc. India Bull., v. 3, p. 75—78. Morel, S. W., 1961, Amphibolite-granulite facies assemblages in southern Nyasaland: Mineralog. Mag., v. 32, p. 725— 730. Morgan, B. A., 19721, Metamorphic map of the Appalachians: U.S. Geol. Survey Misc. Geol. Inv. Map I—724, scale 1 :2,500,000. Narayanaswami, S., and Mahadevan, T. M., 1964, Geology of Trivandrum and Madras areas: Internat. Geol. Cong, 22d, New Delhi, India, 1964, Guide to excursion C—23, 13 p. Neiheisel, James, 1958, Origin of the dune system on the Isle of Palms, S.C.: South Carolina Div. Geology Mineral Industries Lab. Bull., v. 2, p. 46—51. 1962, Heavy-mineral investigations of Recent and Pleistocene sands of the lower coastal plain of Georgia: Geol. Soc. America Bull., v. 73, p. 365—374. Overstreet, W. C., 1967, The geologic occurrence of mona- zite: U.S. Geol. Survey Prof. Paper 530, 327 p. 1972, Ilmenite beach placer, Pulmoddai, Ceylon, and ilmenite resources of the Pleistocene and Holocene for- mations of Ceylon: U.S. Geol. Survey open-file rept., 70 p. Overstreet, W. C., and Griflitts, W. R., 1955, Inner Pied- mont belt, in Russel. R. J., ed., Guides to southeastern geology: Boulder, Colo., Geol. Soc. America, p. 549—577. Overstreet, W. C., Yates, R. G., and Griflitts, W. R., 1963, Heavy minerals in the saprolite of the crystalline rocks in the Shelby quadrangle, North Carolina: U.S. Geol. Survey Bull. 1162—F, 31 p. Bl5 Overstreet, W. C., White, A. M., Whitlow, J. W., Theobald, P. K., Jr., Caldwell, D. W., and Cuppels, N. P., 1968, Fluvial monazite deposits in the southeastern United States: U.S. Geol. Survey Prof. Paper 568, 85 p. Parker, Andrew, 1970, Chemical and mineralogical analyses of some basic and ultrabasic rocks and their initial weathering products: Reading Univ. Dept. Geology Geol. Rept. 4, 44 p. Patterson, S. H., 1967, Bauxite reserves and potential alumi- num resources of the world: U.S. Geol. Survey Bull. 1228, 176 p. 1971, Investigations of ferruginous bauxite and other mineral resources on Kauai and a reconnaissance of ferruginous bauxite deposits on Maui, Hawaii: U.S. Geol. Survey Prof. Paper 656, 74 p. Pettijohn, F. J ., 1957, Sedimentary rocks (2d ed.) : New York, Harper and Row, 718 p. Pirkle, E. G., and Yoho, W. H., 1970, The heavy mineral ore body of Trail Ridge, Florida: Econ. Geology, v. 65, p. 17—30. Rajamanickam, G. V., 1968, Heavy mineral studies of the Cretaceous-Tertiary formations of Pondicherry, South India, in Cretaceous-Tertiary formations of South India: Geol. Soc. India Mem. 2, p. 234—242. Ramberg, Hans, 1948, Titanic iron ore formed by dissociation of silicates in granulite facies [Greenland]: Econ. Geol- ogy, v. 43, p. 553—569. 1952, The origin of metamorphic and metasomatic rocks: Chicago, 111., Univ. of Chicago Press, 317 p. Rankama, K. K., and Sahama, Th. G., 1950, Geochemistry: Chicago, 111., Univ. of Chicago Press, 912 p. Rittenhouse, Gordon, 1943, Transportation and deposition of heavy minerals: Geol. Soc. America Bull., v. 54, p. 1725— 1780. Roedder, Edwin, 1956, Report on the mineralogy of the ti- tanium in bauxite from Salem Hills, Oregon, in Cor- coran, R. E., and Libbey, F. W., eds, Ferruginous bauxite deposits in the Salem Hills. Marion County, Oregon: Oregon Dept. Geology and Mineral Industries Bull. 46, 53 p. Ross, C. S., 1941, Occurrence and origin of the titanium de- posits of Nelson and Amherst Counties, Virginia: U.S. Geol. Survey Prof. Paper 198, 59 p. Rubey, W. W., 1933, The size distribution of heavy minerals within a waterlaid sandstone: Jour. Sed. Petrology, v. 3, p. 3—29. Rumble, Douglas, III, 1973, Fe—Ti oxide minerals from region- ally metamorphosed quartzites of western New Hamp- shire: Contr. Mineralogy and Petrology, v. 42, p. 181— 195. Schuiling, R. D., and Vink, B. W., 1967, Stability relations of some titanium minerals (sphene, perovskite, rutile, and anatase): Geochim. et Cosmochim. Acta, v. 31, p. 2399—2411. Seki. Yotaro, 1961, Pumpellyite in low-grade metamorphism: J our. Petrology, v. 2, p. 407—423. Seki. Yotaro, Onuki, Hitoshi, Oba, Tadamichi, and Mori, R., 1971, Sanbagawa metamorphism in the central Kii Penin— sula, Japan: Japanese Jour. Geology and Geography, v. 41, p. 65—78. Shcherbina, V. V., 1971, Geochemistry of titanium: Geok- himiya, no. 3, p. 302—309. B16 Sherman, G. D., 1952, The titanium content of Hawaiian soils and its significance: Soil Sci. Soc. America Proc., v. 16, p. 15—18. Short, N. M., 1961, Geochemical variations in four residual soils: Jour. Geology, v. 69, p. 534—571. Southwick, D. L., 1968, Mineralogy of rutile- and apatite- bearing ultramafic chlorite rock, Harford County, Mary— land: U.S. Geol. Survey Prof. Paper 600—C, p. 038—044. Spencer, R. V., and Williams, F. R., 1967, The development of a rutile mining industry in Sierra Leone: Common- wealth Mining and Metallurgical Cong., 8th, Australia and New Zealand, 1965, Proc.—General: v. 6, p. 1337— 1341. Talati, D. J., 1971, Ilmenite sands of Gujarat coast: Mineral Wealth (Gujarat Div. Geol. Min.), v. 7, p. 9—13. Temple, A. K., 1966, Alteration of ilmenite: Econ. Geology, v. 61, p. 695-714. Thomas, B. I., and Berryhill, R. V., 1962, Reconnaissance studies of Alaska beach sands, eastern Gulf of Alaska: U.S. Bur. Mines Rept. Inv. 5986, 40 p. Tourtelot, H. A., 1968, Hydraulic equivalence of grains of quartz and heavier minerals, and implications for the study of placers: U.S. Geol. Survey Prof. Paper 594—F, p. F1—F13. Turner, F. J., 1968, Metamorphic petrology—Mineralogical and field aspects: New York, McGraw-Hill Book Co., 403 p. Udden, J. A., 1914, Mechanical composition of elastic sedi- ments: Geol. Soc. America Bull., V. 25, p. 655—744. U.N. Economic Commission for Asia and the Far East, 1963, Minerals map: sheet 3, scale 1:500,000. Valeton, Ida, 1972, Bauxites: New York, American Elsevier Publishing Co., 226 p. Van Andel, Tj. H., 1950, Provenance, transport, and deposi- tion of Rhine sediments—A heavy mineral study on river sands from the drainage area of the Rhine: Wagen- ingen, Germany, Veenman and Zonen, 129 p. GEOLOGY AND RESOURCES OF TITANIUM Wells, N., 1960, Total elements in topsoils from igneous rocks—An extension of geochemistry: Jour. Soil Sci., v. 11, p. 409—424. Western Australia Department of Mines, 1968, Regional di- visions and reported mineral occurrences in western Australia: East Perth, Western Australia Dept. Mines, scale 1:2,534,400. White, J. R., and Williams, E. G., 1967, The nature of a fluvial process as defined by settling velocities of heavy and light minerals: Jour. Sed. Petrology, v. 37, p. 530— 539. White, R. W., 1972, Progressive metamorphism of iron- formation and associated rocks in the Wologizi Range, Liberia: U.S. Geol. Survey open-file rept., 90 p. Whitworth, H. F., 1956, The zircon-rutile deposits on the beaches of the east coast of Australia with special refer- ence to their mode of occurrence and the origin of the minerals: New South Wales Dept. Mines Tech. Rept., v. 4, p. 7—61. Wilson, A. F., 1964, The petrological features and structural setting of Australian granulites and charnockites: Inter- nat. Geol. Cong, 22d, New Delhi, 1964, Rept., pt. 13, p. 21— 44. 1969, Granulitic terrains and their tectonic setting and relationship to associated metamorphic rocks in Australia: Geol. Soc. Australia Spec. Pub. 2, p. 243- 258. Wright, J. B., 1964, Iron-titanium oxides in some New Zea- land iron sands: New Zealand Jour. Geology and Geo- physics, v. 7, p. 424—444. Zakrutkin, V. V., and Grigorenko, M. V., 1968, Titanium and alkalies in biotite in metamorphic facies: Acad. Sci. USSR Doklady, Earth Sci. Sec., v. 178, p. 124—127. Zen, E—an, 1960, Metamorphism of lower Paleozoic rocks in‘ the vicinity of the Taconic range in west-central Ver- mont: Am. Mineralogist, v. 45, p. 129—175. Rutile and Sphane in Blueschist and Ralated High-Pressure-Facies Rocks By M. c. BLAKE, 111., and BENJAMIN A. MORGAN GEOLOGY AND RESOURCES OF TITANIUM GEOLOGICAL SURVEY PROFESSIONAL PAPER 959-0 CONTENTS Page Abstract ________________________________________________________________ Cl Introduction ______________________________________________________________ Description of individual localities __________________________________________ Western California ____________________________________________________ New Caledonia _______________________________________________________ Sanbagawa belt, Japan ________________________________________________ Central Alps _________________________________________________________ Soviet Union _________________________________._____I _________________ Venezuela ____________________________________________________________ Discussion _______________________________________________________________ Resource potential _______________________________________________________ References cited __________________________________________________________ mmwwWWNNF-IHH ILLUSTRATION Page FIGURE 1. Photomicrographs of rutile-sphene relations in high-pressure meta- morphic rocks _____________________________________________ 03 III GEOLOGY AND RESOURCES OF TITANIUM RUTILE AND SPHEN E IN BLUESCHIST AND RELATED HIGH-PRESSURE-FACIES ROCKS By M. C. BLAKE, JR., and BENJAMIN A. MORGAN ABSTRACT Sphene and rutile are characteristic accessory minerals of blueschist-facies metamorphic rocks. Only sphene, how- ever, is present in lower grade blueschist assemblages. In many areas of the world, these blueschists pass into, or are in fault contact with, higher grade glaucophane schists, rutile-bearing amphibolites, and eclogites. The origin of the rutile may be related to a prograde metamorphic reaction that alters sphene to rutile, but few data are available to sup- port this hypothesis. It seems more probable that the rutile crystallized directly from ilmenite or titanium-rich clinopyrox- ene and magnetite in basalts and gabbros during the con- version of these rocks to amphibolites and eclogites contain- ing omphacite, amphibole, and garnet. Regardless of its origin, rutile in the higher grade blueschists is a potential titanium resource and is currently being mined from eclogite and glaucophane-ecologite rocks in the Soviet Union, where re- sources of contained Ti02 could range as high as millions of tons. INTRODUCTION Low-grade metamorphic rocks of the blueschist facies, characterized by minerals such as glauco- phane, jadeitic pyroxene, aragonite, and sphene, have been described from many parts of the world (Irwin and Coleman, 1972). According to most workers, blueschist-facies rocks formed under con- ditions of relatively high pressure and low tempera- ture (5 to 7 kbar and 150° to 300°C) and are most likely the products of subduction in island arcs and active continental margins (Coleman, 1972). De- tailed petrologic studies in a number of areas, in- cluding California (Coleman and Lee, 1963; Blake and others, 1967; Ernst and others, 1970), Japan (Hashimoto and others, 1970; Ernst and others, 1970), New Caledonia (Brothers, 1970; Brothers and Blake, 1973), Venezuela (Morgan, 1969, 1970), and the Alps (van der Plas, 1959; Bearth, 1959, 1966; Ernst, 1971, 1973), have demonstrated that, with increasing metamorphism, these rocks locally pass into amphibolites, eclogites, or glaucophane eclogites containing garnet, blue-green hornblende, epidote, omphacite, and rutile in addition to glauco- phane. Oxygen isotope studies suggest that the high- er grade rocks formed at temperatures on the order of 400° to 550°C and at pressures of 9 to 10 kbar (Taylor and Coleman, 1968). Belts containing these higher grade amphibolites and eclogites usually lie closer to the inferred subduction zones than the low- er grade blueschists do, and the sense of the meta- morphic progression is believed to mark the direc- tion of lithospheric underthrusting (Ernst, 1971). Although this division into two high-pressure min- eral facies appears to be reasonably well established, differentiation in specific geographic areas is difl‘i- cult. Problems include: (1) Differences in mineral assemblages related to host-rock chemistry (for ex- ample, epidote is known to occur in low-grade blue- schists containing abundant ferric iron as well as in higher grade rocks), (2) faulting of the higher grade rocks such that they occur as either isolated fault blocks or tectonic inclusions in lower grade blueschists, and (3) retrogression of many higher grade blueschists to low-grade mineral assemblages. No documentation of a prograde reaction convert- ing sphene to rutile was found in this review of detailed studies on the petrography of blueschists. In fact, only the retrograde sequence has been de- scribed from many areas; rutile in high-grade blue- schist rocks is rimmed and partially replaced by sphene. If the sphene-rutile transition is to represent a potentially valuable isograd, the reaction must be defined. The present data strongly suggest that rutile does not form at the expense of sphene in blueschist- facies metamorphic terranes. DESCRIPTION OF 'INDIVIDUAL LOCALITIES WESTERN CALIFORNIA The presence of rutile and sphene in metamorphic rocks of the Franciscan assemblage has long been C1 C2 known (Taliaferro, 1943, p. 170; Bailey and others, 1964, p. 97). Borg (1956) pointed out the associa- tion of sphene with glaucophane and lawsonite schists and of rutile with eclogites in part of north- ern California. Coleman and Lee (1963) subdivided similar rocks into three metamorphic types in part on the basis of the presence of sphene versus rutile. They showed that rutile-bearing high-grade schists and eclogites are metabasalts occurring as tectonic inclusions (“knockers”) in less metamorphosed Franciscan rocks; subsequent studies have verified this conclusion over large areas of the Coast Ranges (Bailey and others, 1964; Ernst and others, 1970) . In a very detailed study, Lee and others (1966) described one of these “knockers,” a glaucophane schist, from the Cazadero area and showed that the primary metamorphic mineral assemblage, which consists of glaucophane, actinolite, epidote, garnet, and rutile, was more or less replaced by retrograde pumpellyite, chlorite, and sphene. Modal analyses of four parts of this rock showed sphene ranging from 6.1 to 8.1 volume percent, whereas rutile made up only 0.1 to 0.3 percent (fig. 1A). Coleman and Lee (1963, p. 286) suggested the following retro- grade reactions: (1) Pumpellyite + lawsonite + sphene :epidote+rutile+water+quartz and (2) lawsonite + sphene:epidote + rutile + water + quartz. Both re- actions suggest that the rutile—bearing assemblages would be favored by higher pressures and (or) high- er temperatures; this inference is supported by the previously mentioned isotopic studies. Chemical analyses of the rutile- and sphene-bearing meta- basalts, as well as of unmetamorphosed basalts from the Franciscan assemblage, indicate that these rocks are all very similar and suggest that the metamor- phism was essentially isochemical. Undoubtedly, the most extensive area of rutile— bearing “high-grade” metamorphic rocks in the Cali- fornia Coast Ranges is found on Catalina Island, where Bailey (1941) mapped an area of about 10 km2 of gneisses containing albite, hornblende, diop- side, rutile, and rare kyanite and structurally over- lying lower grade lawsonite— and sphene-bearing schists along a low-angle fault. Detailed mineralogi- cal and chemical studies of coexisting phases from a few of these rocks were described by Ernst and others (1970); the titanium-bearing phases, how- ever, were not included in their study. Rutile-bearing metamorphic rocks also occur in the Sierra Nevada foothills in central California. In this area, tectonic blocks of garnet amphibolite and rare eclogite are found along serpentinite-marked fault zones. Neither lawsonite nor glaucophane has GEOLOGY AND RESOURCES 0F TITANIUM been reported from these rocks, but recent work by Morgan (1973) suggested that, in many respects, these rocks are similar both chemically and miner- alogically to high-grade amphibolite “knockers” in the Franciscan assemblage to the west. The higher grade assemblage studied by Morgan consists of gar- net-amphibole-albite-rutile that is being retrograded to epidote, chlorite, ilmenite, and sphene. Rutile in these rocks occurs as scattered grains as large as 0.5 mm and is either retrograded to sphene or to ilmen- ite and minor sphene (fig. 13) . NEW CALEDONIA A remarkably continuous metamorphic terrane is present in the northwestern part of New Caledonia (Espirat, 1963; Coleman, 1967; Lillie and Brothers, 1970). Here one can walk from unmetamorphosed Eocene cherts and siliceous shales to Cretaceous phyllites with lawsonite-glaucophane-sphene and on to schists and gneisses containing several mappable metamorphic isograds including garnet, epidote, and blue-green hornblende within a radius of 15 km. The rocks strike northwest and dip to the southwest; their structural thickness is on the order of 20,000 m (Lillie, 1970). The highest grade part of this se- quence crops out along the northwestern coast be- tween Balade and Touho, a distance of about 100 km. Locally, these rocks consist of extremely coarse grained schists and gneisses with individual crystals of glaucophane, blue-green hornblende, epidote, om- phacite, albite, phengite, and garnet locally as large as 2.5 cm. Rutile is found as 1- to 2-mm grains in glaucophane-rich metabasalt and amphibolite (fig. 10) and in veins cutting these rocks. At least some of these coarse-grained rocks are found in serpentin- ite-rich mélange zones and may not be conformable with the enclosing schists. The detailed mineralogy and chemistry of these rocks and included phases were being studied in 1973 by Phillipa Black at Auckland University in New Zealand. SANBAGAWA BELT, JAPAN Numerous detailed studies have been carried out on the metamorphic rocks of the Sanbagawa belt of Japan, which extends along the Pacific Ocean side of the Japanese islands for more than 1,000 km. The width of the belt is 20 to 50 km. The highest grade rocks are to the north and west, and metamorphic grade decreases south and east into upper Paleozoic and Mesozoic sedimentary rocks (Miyashiro, 1972; Seki and others, 1964; Ernst and others, 1970). In Sikoku, where numerous detailed studies have been RUTILE AND SPHENE IN BLUESCHIST AND RELATED ROCKS C3 FIGURE 1.—Photomicrographs of rutile-sphene relations in high-pressure metamorphic rocks. A, Garnet-glaucophane amphib- from Cazadero, Calif.; plane-polarized light showing cluster of rutile grains mantled by sphene. olite (sample 500Z60) Other minerals include garnet, phengite, and glaucophane. Field of View is 1.4 mm. B, Garnet amphibolite (sample 2126) from Chinese Camp 71/2-min quadrangle, western Sierra Nevada, California; plane-polarized light showing two grains of rutile mantled by ilmenite, which in turn is mantled by minor sphene. Large gray grains are amphibole. Field of View is 1.4 mm. C, Glaucophane amphibolite (sample 9677) from New Caledonia; plane-polarized light showing a group of dark rutile crystals with no mantling by sphene. Other minerals are phengite and glaucophane. Field of view is 1.4 mm. D, Eclogite-amphibolite (sample 1115) from Puerto Cabello, Venezuela (Morgan, 1970); plane-polarized light showing grain of rutile completely mantled by sphene. Other minerals include paragonite and amphibole. Field of view is 0.35 mm. made (Hide, 1961; Iwasaki, 1963; Suzuki, 1964; 1 hornblende), sphene, white mica, garnet, and sodic Ernst and others, 1970) , three metamorphic mineral zones have been mapped. In zone I, mafic schists con- tain epidote, chlorite, actinolite, albite, sphene, and minor glaucophane. Farther north in zone 11, the mafic rocks are characterized by coarser grain size and the assemblage of epidote, chlorite, calcic amphi- bole (actinolite grading northward into blue-green amphibole. Within zone III, the mafic rocks contain abundant porphyroblasts of albite plus blue—green hornblende, with epidote, chlorite, garnet, rutile, and sphene; rutile also has been reported from siliceous rocks of this zone (Ernst and others, 1970). Else- where in Sikoku, rutile—bearing eclogites represent an even higher grade of metamorphism (Banno, C4 1964, 1966), but these rocks seem to be tectonic slices similar to those described from California. CENTRAL ALPS The metamorphic zonation in the central Alps is not developed in a simple, roughly continuous suc- cession but represents a series of complex nappes, metamorphosed and emplaced over a long period of time (Ernst, 1971, 1973). In the north, the Pennine, Sesia—Lanzo, and Helvetic nappes collectively form a relatively high pressure metamorphic terrane, which is structurally overlain to the south and east by the Austroalpine nappes (and southern Alps) representing a relatively high temperature meta- morphic terrane. The tectonic contact between these two contrasting terranes has been named the Alpine Suture (Ernst, 1973). Within the high-pressure terrane, three metamorphic events have been recog- nized, which began during the Cretaceous and con- tinued into the Oligocene: (1) Early Alpine blue- schist-type facies, (2) a later greenschist-type facies, and (3) a late Alpine higher temperature event ranging from greenschist to amphibolite facies. Dur- ing the early Alpine event, a progressive metamor- phic sequence was formed with zeolite- and prehn- ite—pumpellyite-facies rocks grading southward through a zone composed of both greenschists and blueschists into an albite-amphibolite facies contain- ing eclogite. The southernmost rocks contain rutile and grade northward into sphene-bearing lawsonite blueschists and laterally into greenschists (Bearth, 1962, 1966; Chatterjee, 1971). The details of this transition, however, are not given. SOVIET UNION Rutile-bearing eclogite and garnet-glaucophane schists are found in the southern Ural Mountains of the Soviet Union (Chesnokov, 1960). The rocks oc- cur as scattered blocks or slabs as long as 1 km and as thick as 150 to 200 m surrounded by quartz- muscovite schists, graphite-mica schists, and garnet- glaucophane-muscovite-quartz schists. According to Chesnokov, the rocks probably represent isolated metamorphosed gabbro bodies in a metasedimentary terrane. The amount of rutile in the eclogite and gamet- glaucophane schists ordinarily does not exceed 1 or 2 percent, but locally higher values exist, and Ti02 content is as high as 5.3 percent, presumably all as rutile. Size analyses show that the rutile crystals generally fall between 0.1 and 0.2 mm. Four poten- tial orebodies are mentioned and described by Ches- GEOLOGY AND RESOURCES OF TITANIUM nokov, and, according to N. L. Dobretsov (oral commun., 1973) , at least one of these deposits is cur- rently being mined. VENEZUELA Eclogite, eclogite-amphibolite, garnet-amphibolite, and garnet-glaucophane schist are found in north- ern Venezuela in a broad belt extending along the coast west from Caracas for nearly 200 km (Dengo, 1950; Morgan, 1969, 1970). The eclogites and re- lated rocks occur as irregular lenses in rocks of the epidote-amphibolite facies. This terrane is structur- ally overlain to the south by thrust sheets of green- schist- and lower grade blueschist-facies rocks (Shagam, 1960; Piburn, 1968). Although field rela- tions are complicated by extensive thrust faulting, the regional metamorphism can be attributed to former subduction zones that presumably lay north of the present coastline (Bell, 1972) . Within the lower grade blueschist rocks, rutile is absent and sphene is rare. Rutile as small (less than 0.5 mm) round grains, however, is a common acces- sory mineral in the eclogites and related high-grade rocks. Rutile grains are rimmed by sphene in those eclogites that are boudins in marble. This alteration may be related to the reaction ruti1e+quartz+cal- citeasphene+CO2 (fig. 1D). According to Morgan (1970) , the conditions of metamorphism for the eclogites and associated rocks include temperatures of about 525°C and pressures of about 7 kbar. DISCUSSION Rutile-bearing eclogites, amphibolites, and garnet- glaucophane schists occur in the areas just described as (1) tectonic inclusions in lower grade sphene- bearing blueschists (California) and (2) thrust sheets overlying lower grade rocks (central Alps) and in (3) relatively coherent metamorphic terranes (New Caledonia, Venezuela, and Japan), where rutile is presumably replaced by sphene at a lower grade, although the details of this transition are not known. In nearly every area, there is petrographic evidence for retrograde metamorphism of rutile to sphene, but nowhere has the progression from sphene to rutile been documented. Because sphene is stable in Barrovian metamorphic terranes well into the sillimanite zone, it might appear that the formation of rutile is related to unusual chemistry of the host rocks. As Force (1976) pointed out, however, rutile is found in low-grade metamorphic rocks low in CaO; in the blueschists, nearly all of the rocks are metamorphosed basalts high in CaO. A more con- RUTILE AND SPHENE IN BLUESCHIST AND RELATED ROCKS vincing hypothesis is that the rutile is derived from titanium-rich minerals such as ilmenite, magnetite, and clinopyroxene in the original basalt. During metamorphism, titaniferous augite and plagioclase are converted to omphacitic pyroxene, amphibole, and garnet, the titanium being left to form rutile. Examples of the conversion of titanium-bearing clinopyroxene to omphacite and rutile in the central Alps have been documented by Miller (1970) and by Chinner and Dixon (1973). During subsequent retro- grade metamorphism, the breakdown of omphacite and other minerals releases calcium, which reacts with the rutile to form rims of sphene. This hypothe- sis certainly fits the observed field and petrographic data for tectonic slices from California, but it re- mains to be seen whether rutile is derived directly from primary minerals in more coherent terranes where gradational cases might be observed. RESOURCE POTENTIAL The large grain size and abundance of rutile in many eclogites and related high-pressure rocks war- rant its classification as a titanium resource. As was mentioned earlier, rutile is currently being mined from these kinds of rocks in the Urals, and similar potential orebodies may exist in New Caledonia, on Catalina Island (California), and in northern Vene- zuela. In all of these areas, the mineral potential is greatly enhanced by the possibility of shoreline plac- er deposits of rutile. The magnitude of rutile resources in deposits of high-pressure rocks has not been published, but the figures cited above for grade and size of low-grade resources in eclogitic tectonic blocks of the Urals imply millions of tons of contained rutile. REFERENCES CITED Bailey, E. H., 1941, Mineralogy, petrology and geology of Santa Catalina Island, California: Unpub. Ph.D. dissert., Stanford Univ., Palo Alto, Calif. Bailey, E. H., Irwin, W. P., and Jones, D. L., 1964, Franciscan and related rocks, and their significance in the geology of western California: California Div. Mines and Geology Bull. 183, 177 p. Banno, Shohei, 1964, Petrologic studies on Sanbagawa crystal- line schists in the Bessi-Ino district, central Sikoku, Japan: Toyko Univ. Fac. Sci. Jour., sec. 2, v. 15, p. 203—319. 1966, Eclogite and eclogite facies: Japanese Jour. Geology and Geography, v. 37, p. 105—122. Bearth, Peter, 1959, Uber Eklogite Glaukophanschiefer und metamorphe Pillowlaven: Schweizer. Mineralog. u. Petrog. Mitt., v. 39, p. 267—286. C5 1962, Versuch einer Gliederung alpinmetamorpher Serien den Westalpen: Schweizer. Mineralog. u. Petrog. Mitt., v. 42, p. 127—137. 1966, Zur mineralfaziellen Stellung der Glaukophange- steine der Westalpen: Schweizer. Mineralog. u. Petrog. Mitt., v. 46, p. 13—23. Bell, J. S., 1972, Geotectonic evolution of the southern Carib- bean area: Geol. Soc. America Mem. 132, p. 369—386. Blake, M. 0., Jr., Irwin, W. P., and Coleman, R. G., 1967, Upside-down metamorphic zonation, blueschist facies, along a regional thrust in California and Oregon: U.S. Geol. Survey Prof. Paper 575—0, p. 01-09. Borg, I. Y., “1956, Glaucophane‘ schists and eclogites near Healdsburg, California: Geol. Soc. America Bull., v. 67, p. 1563—1584. Brothers, R. N., 1970, Lawsonite-albite schists from northern- most New Caledonia: Contr. Mineralogy and Petrology, v. 25, p. 185—202. Brothers, R. N., and Blake, M. 0., Jr., 1973, Tertiary plate tectonics and high-pressure metamorphism in New Cale- donia: Tectonophysics, v. 17, p. 337—358. Chatterjee, N. D., 1971, Phase equilibria in the Alpine meta- morphic rocks of the environs of the Dora-Maira Massif, western Italian Alps, pts. 1 and 2: Neues Jahrb. Min- eralogie Abh., v. 114, p. 181—245. Chesnokov, B. V., 1960, Rutile bearing eclogites from the Shubino village deposit in the southern Urals: Internat. Geology Rev., v. 2, p. 936—945. Chinner, G. A., and Dixon, J. E., 1973, Some high-pressure paragenesis of the Allalin Gabbro, Valais, Switzerland: Jour. Petrology, v. 14, p. 185—202. Coleman, R. G., 1967, Glaucophane schists from California and New Caledonia: Tectonophysics, v. 4, p. 479—498. 1972, Blueschist metamorphism and plate tectonics, in Internat. Geol. Congress, 24th, Montreal 1972, Proc.: sec. 2, p. 19—26. Coleman, R. G., and Lee, D. E., 1963, Glaucophane—bearing metamorphic rock types of the Cazadero area, California: Jo-ur. Petrology, v. 4, p. 260—301. Dengo, Gabriel, 1950, Eclogitic and glaucophane amphibolites in Venezuela: Am. Geophys. Union Trans, v. 31, p. 873—878. Ernst, W. G., 1971, Metamorphic zonations on presumably subducted lithospheric plates from Japan, California and the Alps: Contr. Mineralogy and Petrology, v. 34, p. 43~59. 1973, Interpretative synthesis of metamorphism in the Alps: Geol. Soc. America Bull., v. 84, p. 2053—2078. Ernst, W. G., Seki,Yotaro, Onuki, Hitoshi, and Gilbert, M. C., 1970, Comparative study of low-grade metamorphism in the California Coast Ranges and the outer metamorphic belt of Japan: Geol. Soc. America Mem. 124, 276 p. Espirat, J. J ., 1963, Etude géologique de régions de Nouvelle- Calédonie septentrionale: Unpub. thesis, Univ. of Cler- mont, France, 210 p. Hashimoto, Mitsuo, Igi, Sachio, Seki, Yotaro, Banno, Shohei, and Kojima, 1970, Metamorphic facies map of Japan: Kawasaki City, Japan Geol. Survey, map ser. no. 13, scale 1:2,000,000. Hide, Kei, 1961, Geologic structure and metamorphism of the Sanbagawa crystalline schists of the Besshi-Shirataki mining district in Sikoku, southwest Japan: Hiroshima CB GEOLOGY AND RESOURCES 0F TITANIUM Univ. Geol. Rept. 9, p. 1—87 [in Japanese with English abs.]. Irwin, W. P., and Coleman, R. G., 1972, Preliminary map showing global distribution of alpine-type ultramafic rocks and blueschists: U.S. Geol. Survey Misc. Field Studies Map MF—340, scale 1:40,000,000. Iwasaki, Masao, 1963, Metamorphic rocks of the Kotu-Bizan area, eastern Sikoku: Tokyo Univ. Fac. Sci. Jour., sec. 2, v. 15, p. 1—90. Force, E. R., 1976, Metamorphic source rocks of titanium placer deposits—A geochemical cycle, in Geology and resources of titanium: U.S. Geol. Survey Prof. Paper 959—B, 16 p. Lee, D. E., Coleman, R. G., Bastron, Harry, and Smith, V. C., 1966, A two-amphibole glaucophane schist in the Fran- ciscan Formation, Cazadero area, Sonoma County, Cali- fornia: U.S. Geol. Survey Prof. Paper 550—0, p. 0148— 0157. Lillie, A. R., 1970, The structural geology of lawsonite and glaucophane schists of the Ouégoa district, New Cale- donia: New Zealand Jour. Geology and Geophysics, v. 13, p. 72—116. Lillie. A. R., and Brothers, R. N., 1970, The geology of New Caledonia: New Zealand Jour. Geology and Geophysics, v. 13, p. 145—183. Miller, Christine, 1970, Petrology of some eclogites and meta- gabbros of the Oetztal Alps, Tirrol, Austria: Contr. Min— eralogy and Petrology, V. 28, p. 42—56. Miyashiro, Akiho, 1972, Metamorphism and related magma- tism in plate tectonics: Am. Jour. Sci., v. 272, p. 629— 656. Morgan, B. A., 1969, Geologia de la region de Valencia, Cara- bobo, Venezuela: Venezuela Dir-eccion Geologia, Bo]. Geologia, v. 10, no. 20, p. 3—136. 1970, Petrology and mineralogy of eclogite and garnet amphibolite from Puerto Cabello, Venezuela: Jour. Petrology, v. 11, p. 101—145. 1973, Tuolumne River ophiolite, western Sierra Ne- vada, California [abs.]z Geol. Soc. America Abs. with program, v. 5, p. 83. Piburn, M. D., 1968, Metamorfismo y estructura del grupo Villa de Cura, Venezuela septentrional: Venezuela Di- reccion Geologia, Bol. Geologia, v. 9, no. 18, p. 183—289. Seki, Yotaro, Oba, Tadamichi, Mori, Ryuji, and Kuriyagawa, S., 1964, Sanbagawa metamorphism in the central part of the Kii Peninsula: Japa ese Assoc. Mineralogists, Petro- logists, and Econ. Geologi ts Jour., v. 52, p. 73—89. Shagam, Reginald, 1960, Geology of central Aragua, Vene- zuela: Geol. Soc. America Bull., v. 71, p. 249—302. Suzuki, T., 1964, The Mikabu greenrocks in Sikoku (1): Petrology and geology of the Mikabu greenrocks in the district of Osugi, Kochi Prefecture: Kochi Univ. Res. Rept., V. 13, p. 93—102 [in Japanese with English abs.]. Taliaferro, N. L., 1943, Franciscan-Knoxville problem: Am. Assoc. Petroleum Geologists Bull., v. 27, p. 109—219. Taylor, H. P., Jr., and Coleman, R. G., 1968, 018/01“ ratios of coexisting minerals in glaucophane-bearing metamorphic rocks: Geol. Soc. America Bull., v. 79, p. 1727—1756. van der Plas, L., 1959, Petrology of the Northern Adula region, Switzerland: Leidse Geol. Meded., v. 24, pt. 2, p. 418—598. Titanium Deposits in Anorthosite Massifs By NORMAN HERZ GEOLOGY AND RESOURCES OF TITANIUM GEOLOGICAL SURVEY PROFESSIONAL PAPER 959-D CONTENTS Page Abstract _________________________________________________________________ D1 Introduction ______________________________________________________________ 1 Types of anorthosite massifs ______________________________________________ 2 Labradorite anorthosite massifs ________________________________________ 2 Andesine anorthosite massifs __________________________________________ 3 Adirondack Mountains, New York ________________________________ 3 Allard Lake, Quebec, Canada ______________________________________ 3 Roseland, Virginia ________________________________________________ 4 Prospecting for titanium deposits in anorthosite ______________________________ 5 References cited __________________________________________________________ 6 TABLES Page TABLE 1. Analysis of Lac Tio ilmenite ore __________________________________ D4 2. Analyses of selected rutile samples ________________________________ 5 III GEOLOGY AND RESOURCES OF TITANIUM TITANIUM DEPOSITS IN ANORTHOSITE MASSIF S By NORMAN HERZ ABSTRACT Anorthosite massifs are economically important sources of titanium minerals. The most valuable titanium deposits oc- cur in andesine anorthosite massifs rather than in llabradorite anorthosite massifs or stratiform anorthosites and are of two types: (1) Ilmenite bodies associated with gabbro in the border zones of nonalkalic andesine anorthosite massifs and (2) rutile and ilmenite disseminated along the borders of alkalic andesine anorthosite massifs. Deposits of titanium min- erals in anorthosite commonly contain millions of tons con- tained Ti02. INTRODUCTION The term anorthosite has been applied to all ig- neous rocks that are composed almost entirely of plagioclase of calcic to intermediate composition, though different processes can form a rock that con- forms to this definition. Anorthosites occur both as massifs and in stratiform complexes, the latter de— veloping by gravity stratification in ultramafic and mafic complexes. These stratiform complexes, gen- erally Precambrian but occasionally as young as Tertiary, are characterized by rhythmic layering, show many cumulate features, do not develop giant megacrysts, and have a plagioclase composition in the bytownite-anorthite range. Associated rocks are olivine- and pyroxene-rich ultramafics, generally layered and gravity stratified and occurring with chromitite and sulfides. Nearly all the important igneous deposits of ti- tanium minerals, however, are associated with mas- sif-type anorthosites and their related rocks. The largest ilmenite deposit in the United States occurs in the Adirondack anorthosite of New York, and the largest rutile deposit in the Roseland anorthosite of Virginia. The largest ilmenite deposit in Canada is the Lac Tio deposit in the Allard Lake anorthosite of Quebec. Large rutile deposits are associated with the Pluma Hidalgo anorthosite of Oaxaca, Mexico, and the St. Urbain anorthosite of Quebec, Canada. Large ilmenite deposits occur in anorthosite at Egersund- Sogndal, Norway, and in the Ukraine, U.S.S.R. No- ritic gabbro, an end member of common anorthositic assemblages, contains ilmenite deposit , such as those at Otanm‘aki, Finland. Deposits in granulite- facies wall-rock gneisses may be closely related to those in adjacent anorthosite, as they are at Roseland. Some massif anorthosites, such as those in Nain and Michigamau, Labrador, show gravity layering, but, in most complexes, gravity layering is generally poorly developed or absent. Ages determined to date on all massif anorthosites are Precambrian, about 1,300:200 m.y. (Herz, 1969). These rocks crystal- lized under slow cooling conditions in the upper man- tle or lower crust; the resulting megacrysts are com- monly tens of centimeters in length. The feldspars are andesine, in places antiperthitic, or labradorite; pyroxenes are the most common accessory minerals, but olivine may be present, as it is in the Laramie anorthosite of Wyoming. Some of the minerals show kink banding, and plagioclase twin lamellae may be bent or broken. The name applied to a rock of the anorthosite suite depends on the amounts of mafic minerals present. The name anorthosite is reserved for rocks contain- ing 90 percent plagioclase or more; gabbroic, noritic, or troctolitic anorthosites have 10 to 22.5 percent mafic minerals; anorthositic gabbro, norite, or troc- tolite have 22.5 to 35 percent mafic minerals (Isach- sen, 1968, p. 435). According to the terminology of Buddington and others (1963), titanium-bearing oxides in anortho- sites can include ilmenite, rutile, ilmeno-magnetite (ilmenite as fine intergrowths in magnetite), ilmeno- hematite (ilmenite as fine intergrowths in hematite), hemo-ilmenite (hematite as fine intergrowths in ilmenite) , titano-magnetite (titanium present in magnetite structure), and ulvospinel (FezTiOQ. Ti- D1 D2 taniferous magnetite is a generic term that does not specify presence or type of phases other than mag- netite. The titanium-rich minerals crystallize late in the magmatic history of anorthositic complexes. The be- havior of titanium during the crystallization of any magma is controlled by several factors, including the initial abundance of titanium in the magma, the chemical activities of silicon, aluminum, and iron, the partial pressure of oxygen, the formation of ti- tanium-rich immiscible melts, and the temperature of crystallization (Verhoogen, 1962). In a normally crystallizing tholeiitic magma having sufficiently high oxygen fugacity, titanium and iron will form oxides early in the magmatic history. These oxide minerals may segregate out and accumulate by grav- ity settling because of their high specific gravity. Another possible factor in the formation of titanium oxide minerals is the formation of an immiscible liquid that has the composition of iron and titanium oxides plus phosphate within a predominantly sili- cate liquid magma chamber (Philpotts, 1967). The formation of some ilmenite- and titaniferous magne- tite-apatite orebodies has been ascribed to this phenomenon. The early crystallization history of anorthositic plutons is a much debated topic (Yoder, 1968; Green, 1969) , but it is generally agreed that the rocks rich in iron-titanium oxide and mafic minerals are em- placed later than the anorthosites themselves (Isach- sen, 1968). If th: later rocks have crystallized from a magma, it must be one enriched in iron and titani- um either by processes of differentiation or by for- mation of an immiscible liquid. TYPES OF ANORTHOSITE MASSIFS Anorthosite massifs can be divided into two dis- tinct types based on their plagioclase and oxide com- positions (Anderson and Morin, 1968). At St. Ur- bain, Quebec, one type can be seen intruding the other, but this relationship is exceptional; any one massif generally consists of only one type. The two types recognized are (1) labradorite anorthosite massifs characterized by plagioclase of AneHs com- position and either titano-magnetite or its oxidized equivalent, magnetite+ilmenite, and (2) andesine anorthosite massifs containing plagioclase An48_25, hemo-ilmenite, and EnzAn ratios (En content of orthopyroxene and An content of plagioclase) great- er than one. The latter is generally called “Adiron- dack type” (Buddington, 1931) and is associated with the world’s principal ilmenite deposits. An unusual type of andesine anorthosite massif, found GEOLOGY AND RESOURCES OF TITANIUM in Roseland, Pluma Hidalgo, and St. Urbain, con- tains feldspar megacrysts that are strongly antiper- thitic and has a K20 content of 3 to 4 percent, which is about double the normal amount. This type has been called “alkalic anorthosite” (Herz, 1968) and has associated deposits of rutile as Well as ilmenite. Some typical labradorite anorthosite massifs in- clude the Michigamau of Labrador (Emslie, 1970) and the Duluth Gabbro Complex of Minnesota (Tay- lor, 1964) ; andesine anorthosites include massifs of the Adirondack Mountains in New York (Budding- ton, 1931), Allard Lake in Quebec (Hargraves, 1962), and Roseland, Va. (Herz, 1968). These mas- sifs will be discussed below as typical examples of the two types of anorthosite massifs. LABRADORITE ANORTHOSITE MASSIFS The Michigamau anorthosite pluton occurs in west-central Labrador and covers an area of 65x40 km (Emslie, 197 O). The main rock units, in order of crystallization, are a border zone of olivine gabbro, which is fine grained at the contact and probably chilled, a thick series of layered leucotroctolite, a considerable thickness of anorthosite that grades up- ward into leucogabbro and gabbro, and ferrous iron- rich rocks. Structures and textures indicative of crystal accumulation are developed in leucotrocto- lite, anorthosite, and leucogabbro. The ferrous iron-rich rocks range from mafic diorite through granodiorite to syenite and locally transgress the anorthosite. The compositions of the mafic minerals indicate a fractionation trend toward enrichment in iron. Concentrations of ilmenite and vanadium-bear- ing magnetite are in the gabbroic border zone in pods and layers forming a subzone 300 m Wide and several kilometers long. They are clearly a product of the trend toward iron enrichment that became pro- nounced after the crystallization of the main mass of the anorthosite. The Duluth Gabbro Complex (Taylor, 1964) un- derlies more than 2,500 ka, is about 15,000 m thick, and consists, in large part, of anorthosite and anor- thositic gabbro. The oldest part of the complex is a coarse-grained anorthositic gabbro that intrudes Keweenawan flows and makes up the upper part of the complex. It contains 75 to 90 percent calcic labradorite and generally abundant titanaugite, oli- vine, and fine magnetite-ilmenite intergrowths. The anorthositic gabbro was intruded by basaltic magma of a second period of magmatic activity, which formed layered rocks consisting chiefly of troctolite, olivine gabbro, feldspathic gabbro, and syenogab- TITANIUM DEPOSITS IN ANORTHOSITE MASSIFS bro. These rocks are 4,500 m thick and show local rhythmic layering, fluxion structure, and gravity stratification. The entire series is cut by bodies of ferrogranodiorite and granophyre and late dikes of basalt and aplite. The iron-titanium oxide minerals are most abundant near the top of the series, in oli- vine metagabbro where distinct layers of “iron ores” are seen and in magnetite syenogabbro where mag- netite averages 12 percent by volume. The oxide minerals are apparently a product of the trend of iron-titanium enrichment during the course of dif- ferentiation and fractionation, as at Michigamau. Both the Michigamau and the Duluth complexes contain fairly large bodies that are high in TiOz, but no economic titanium deposits have been found in them despite intensive search. The fine-grained in- tergrowths of magnetite and ilmenite rather than discrete ilmenite grains (Grout, 1950; Grout and others, 1959) render the deposits an economically unimportant source of titanium at present. ANDESINE ANORTHOSITE MASSIFS The andesine anorthosites are the most wide- spread massifs and contain the largest known titani- um deposits (Isachsen, 1968). They generally occur as intrusives in granulite or upper amphibolite meta- morphic terranes and are, in turn, intruded by a suite of charnockitic rocks, which are characterized by perthitic feldspars and orthopyroxene rather than by two feldspars and hydrous mafic minerals. The anorthositic and charnockitic rocks may be inter- layered, as they are in the Lake Placid area of the Adirondack Mountains (Crosby, 1968) , or they may be gradational from anorthosite through noritic anorthosite, norite, and mangerite to charnockite, as they are in the Snowy Mountain Dome of the Ad- irondack Mountains (DeWaard and Romey, 1968). In many places, the charnockite suite only borders the anorthosite, and the exact relations are not too clear. ADIRONDACK MOUNTAINS, NEW YORK The Marcy Massif is the main anorthosite pluton in the Adirondack Mountains (Buddington, 1931) and consists predominantly of andesine anorthosite. Plagioclase averages An45 in composition and occurs as megacrysts and groundmass, both of which show evidence of cataclasis. Gabbroic or noritic anortho- site, in which are found the ilmenite-magnetite ore- bodies, forms a border facies against mangerite- charnockite, which surrounds the anorthosite massif for about three-quarters of its perimeter. Anortho- D3 site grades into and is cut by gabbro or norite and then by ilmenite-magnetite-rich gabbro and pyrox- enite. Although the initial composition of the materi- al that produced the anorthosite series is unknown, xenoliths and crosscutting relations show that the trend of the later mafic rocks was towards enrich- ment in iron and titanium and the eventual produc- tion of ilmenite-magnetite orebodies. The ilmenite deposits of the Sanford Lake (Ta- hawus) mining district in the Adirondack Mountains have produced more titanium ore than any anortho- site deposit in the world. Through 1964, 33,500,000 tons of ore had been produced from the South Ex- tension area of Sanford Lake alone (Gross, 1968) . Anorthosite of the Sanford Lake mining district grades into gabbroic anorthosite, and this, in turn, grades into anorthositic gabbro containing varying percentages of mafic minerals (Gross, 1968). The TiO2 content of anorthosite at the mine is 0 to 5.4 percent, that of gabbro is 5.5 to 9.4 percent, and that of the ores is 13.5 percent or more. Two types of orebodies are described: (1) “Anorthositic ore” as- sociated with anorthosite is found as irregular mas- sive lenses that show no flow structures, are coarse grained, and have sharp contacts with the anortho- site, and (2) “gabbroic ore” occurs as oxide-enriched layers within gabbro. These are fine to medium grained and have a well-defined flow structure simi- lar to that of the gabbro. The thickness of individual layers ranges from a few millimeters to several meters. Contacts are sharp or gradational with gab- bro. The anorthositic type forms a football orebody, and the gabbroic type a hanging wall separated by varying amounts of anorthosite and gabbro. The anorthositic ore and gabbroic ore are chemically dif- ferent; the anorthositic ore has an FezTiO2 ratio of 2: 1 or greater, whereas the ratio in gabbroic ore is invariably less than 2: 1. Discrete grains of ilmenite are present as well as exsolution intergrowths of ilmenite and magnetite containing as much as 35 percent FeTiO3 molecule and of ulvospinel and magnetite. Although there is no agreement on the origin of these ores, Gross (1968) stated that they were origi- nally formed by magmatic processes and reworked during later metamorphic events. ALLARD LAKE, QUEBEC, CANADA The largest operating ilmenite mine in Canada is in the Allard Lake anorthosite massif on the north- ern shore of the St. Lawrence River, about 800 km northeast of Quebec City (Hargraves, 1962). The D4 massif is andesine anorthosite surrounded by man- gerites; a mafic zone is in the contact area. Three main rock types are exposed in the northeastern border area of the massif: 1. Coarse-grained massive anorthosite, generally more than 95 percent calcic andesine, which shows considerable cataclasis. Plagioclase matrix grains 3 to 6 mm long and megacrysts 2 to 35 cm long have about the same composi- tion (An.,0,52). Bronzite is irregularly dis- tributed in the anorthosite as poikilitic crystals enclosing plagioclase or as alined aggregates that transgress primary foliation in the anorthosite. 2. Medium-grained oxide-rich norite consisting of protoclastically deformed plagioclase (Anions), pyroxenes with hypersthene>au- gite, apatite (8—10 percent), and two oxides with hemo—ilmenite>magnetite (20—54 per- cent). This rock occurs in sheets as much as 6 km long by 1 km thick that intrude the an- orthosite. The oxide minerals are interpreted by Hargraves as the product of an immiscible melt in the silicate liquid. 3. Pyroxene syenite gneiss that borders the anortho- site massif. The contact zone is characterized by conformable gneissosity and enrichment in mafic minerals. The ilmenite orebody at Lac Tio in the Allard Lake district occurs as a large sheet or lens of hemo- ilmenite located a few kilometers inside the margin of the anorthosite (Lister, 1966). The anorthosite near the deposit is coarse grained, unfoliated, and composed of andesine (An4o_4.,) containing less than 2 percent pyroxenes and ilmeno-hematite. The ore sheet is made up of massive hemo-ilmenite capped by layered hemo-ilmenite and anorthosite. The lower part of the sheet is over 90 m thick and contains blocks of anorthosite as large as 4.5 m in diameter. The upper part of the orebody is less than 60 m thick and consists of alternating layers of pure ilmeno-hematite and anorthosite containing dissemi- nated hemo-ilmenite; individual layers vary from a few centimeters to a few meters thick. Typical high-grade ore (table 1) consists of hemo- ilmenite with about 3 percent iron spinel, 0.5 per- cent pyrite, and a few feldspar remnants. The hemo- ilmenite consists of ilmenite with 20 to 22 percent titanium-bearing hematite disks exsolved in (0001) (Bergeron, 1972). The immiscible liquid said to have formed the de- posit must have had the general composition of the GEOLOGY AND RESOURCES OF TITANIUM TABLE 1.—Analysis of Lao Tio ilmenite ore (from Be’rgemn, 1972) Weight (percent) Ti02 _______________________________________ 34.3 FeO _______________________________________ 27.5 F9203 _______________________________________ 25.2 SiOa _______________________________________ 4.3 A1203 _______________________________________ 3.5 CaO ________________________________________ .9 MgO _______________________________________ 3.1 Cr203 _______________________________________ .10 205 ——————————————————————————————————————— .27 MnO _______________________________________ .16 S __________________________________________ .3 Na 0 and K O ______________________________ .35 P265 ______ 2. ________________________________ .015 orebody, and later hydrothermal replacement was minor (Lister, 1966, p. 285). The ore magma was injected into a dilation zone that formed within a shear zone in hot anorthosite. In the upper part, the ore sheet developed as the liquid invaded closely spaced shear planes and permeated fractures in crushed anorthosite to form the layered and dis— seminated deposit. ROSELAND, VIRGINIA The Roseland alkalic andesine-anorthosite massif is about 15 km long in a northeasterly direction and 4 km Wide (Herz, 1968). It occurs in country rock that had been metamorphosed to the pyroxene- granulite facies but later partly retrograded to greenschist facies during a Paleozoic metamorphism. Charnockite, noritic anorthosite, norite, and pyrox- ene-rich rocks are found largely in the border zone but are also scattered irregularly throughout the anorthosite mass-if. The anorthosite itself consists of large bluish-gray megacrysts, as long as 20 cm, of andesine antiperthite, Or25Ab51An23,5Celsian0‘5, and lighter colored granulated zones containing ande- sine-oligoclase and microcline. Pyroxenes, blue quartz (generally rutilated), clinozoisite, and mus- covite are common accessory minerals. This anortho— site and its associated ores are remarkably similar to those in Pluma Hidalgo (Paulson, 1964; T. P. Thayer, written commun., 1973) . Ilmenite and rutile occur in the border area of the alkalic anorthosite massif. Hemo-ilmenite is found disseminated in the contact rocks of the anorthosite, including mafic border phases and possible contact- metamorphosed granulites. Ilmenite also occurs to- gether with apatite in nelsonite dikes, which range as large as 600 m in length by 60 m in width (Wat- son and Taber, 1913). These dikes form the richest ilmenite deposits in the area and are always found TITANIUM DEPOSITS IN ANORTHOSITE MASSIFS outside the anorthosite body in charnockites and granulites of the border zone. The dikes appear to be later than either the anorthosite or the border rocks but were also deformed together with these rocks in later metamorphic events (Herz, 1968). The dikes approximate a possible eutectic composition in an iron-titanium oxide and phosphate system (Phil- potts, 1967) and may represent the later emplace- ment of an immiscible liquid, which formed in the silicate magma that produced the bordering rocks. Rutile occurs as disseminated deposits within the anorthosite and other rocks in the contact zone. Much of it is present where anorthosite is interlay— ered with granulite or is enriched in mafic dikes that cut both rocks. Some rutile appears to have been in- troduced in fractured zones, and the anorthosite is invariably saussuritized where rutile is abundant. In these zones, rutile grains range from 0.2 mm to 4 cm across and form deposits of an estimated grade of about 5 percent (D. N. Hillhouse, written com- mun., 1960). Rutile formed in the anorthosite ap- pears to have an unusually large amount of TiO2 (99.43 percent) (see also Ross, 1941) , which is high— er than any shown by Deer and others (1962, p. 36), whose TiO2 values in rutile range from 45.79 to 98.96 percent, and higher than samples from Magnet Cove, Ark., Trail Ridge, Fla., and Shooting Creek, NC. (table 2).. Saprolite resources over anorthosite, nelsonite, and adjacent rocks near Roseland were estimated by Fish (1962) as 20 million tons of ilmenite and rutile. PROSPECTING FOR TITANIUM DEPOSITS IN ANORTHOSITE Prospecting for titanium deposits in anorthosite should take into account the fact that different types of anorthosites have different types of titanium oc- currences. Titanium minerals are associated with all types of anorthosites, but, as of yet, economic min- D5 eral deposits have been found only in andesine anor- thosite massifs, not in labradorite anorthosite mas- sifs or in stratiform anorthosites. The economic massifs form batholithic plutons that appear to be concordant to country rock in Precambrian terranes. Andesine, both as deformed megacrysts as large as 2 m in diameter and as granulated groundmass, com- prises more than 90 percent of the rock. And-esine anorthosite massifs are associated with noritic gab- bros, other igneous rocks of the charnockite suite, and metamorphic rocks of the pyroxene granulite facies. Gravity measurements over these anorthosite massifs show a substantial negative Bouguer anoma- ly. Magnetic patterns are featureless. Departures from these norms, however, may occur in border areas. Radiometric absolute ages of the massifs and country rocks are about 1,300 i 200 my. Once an andesine anorthosite massif has been lo- cated, titanium deposits can be found by consider- ing that: 1. Ore deposits occur in peripheral areas of the massif, especially bordering mafic differenti- ates, and in adjacent country rock. 2. Ilmenite is associated with magnetite and apatite in gabbroic rocks. If plagioclase megacrysts are antiperthitic, rutile may be present as dis- seminations in anorthosite and country rock or in mafic veins and segregations. 3. Gravity and magnetic features may be anoma- lously high over iron-titanium ore deposits and associated mafic rocks, in contrast to the rest of the anorthosite (Rose, 1969, p. 30-40). 4. Detrital heavy minerals from small streams can serve as a guide to the distribution of ilmenite and rutile (Herz and others, 1970) . Prospecting by chemical analysis alone is decep- tive, since anorthosites themselves generally have very little Ti02, and, in many of the bordering mafic rocks, TiO2 is present in unusable mafic silicate min- TABLE 2.—Analyses of selected rutile samples (weight percent) [Analysts: A, A. C. Vlisidis: B—D, John Marinenko, U.S. Geological Survey, Reston, Va. ND, not detected] A B C D Sample Anorthosite Phonolite Detrital mixture Garnet schist (Roseland, Va.) (Magnet Cove, Ark.) (Trail Ridge, Fla.) (Shooting Creek, N.C.) Tl02 _________________________________ 99.43 93.9 98.3 86.1 FeO _________________________________ .55 m ‘1) 12.1 £8603 ________________________________ ND 1.2 .2 .9 2 s _________________________________ ND .1 <.1 <.1 V203 _________________________________ ‘3’ .9 .4 .4 Nb205 ________________________________ m 3.8 .4 .3 H20 _________________________________ ND .2 .2 ND 1 All Fe reported as F9203. 2Rmults of semiquantitative analysis in parts per million: Si=150, Al=100, M3220, Ca=15, B230, 341:3, Cr:200, M0215, Nb=200, V2500, Zr:300. Analysts: W. B. Crandell and Helen Worthing, U.S. Geological Survey, Reston, Va. D6 erals. Mineralogic studies are necessary to establish the mode of occurrence of titanium and the type of anorthosite. REFERENCES CITED Anderson, A. T., Jr., and Morin, M., 1968, Two types of massif anorthosites and their implications regarding the thermal history of the crust, in Isachsen, Y. W., ed., Origin of anorthosite and related rocks: New York State Mus. and Sci. Service Mem. 18, p. 57—69 [1969]. Bergeron, M., 1972, Quebec Iron and Titanium Corporation ore deposit at Lac Tio, Quebec: Internat. Geol. Cong, 24th, Montreal 1972, Guidebook excursion B—09: 8 p. Buddington, A. F., 1931, The Adirondack magmatic stem: Jour. Geology, v. 39, p. 240—263. Buddington, A. F., Fahey, Joseph, and Vlisidis, Angelina, 1963, Degree of oxidation of Adirondack iron oxide and iron-titanium oxide minerals in relation to petrogeny: Jour. Petrology, v. 4, p. 138-169. Crosby, Percy, 1968, Petrogenetic and statistical implications of modal studies in Adirondack anorthosite, in Isachsen, Y. W., ed., Origin of anorthosite and related rocks: New York State Mus. and Sci. Service Mem. 18, p. 289—303 [1969]. Deer, W. A., Howie, R. A., and Zussman, Jack, 1962, Rock- forming minerals, v. 5, Non-silicates: New York, John Wiley and Sons, 371 p. deWaard, Dirk, and Romey, W. D., 1968, Petrogenetic rela- tionships in the anorthosite charnockite series of Snowy Mountain Dome, south-central Adirondacks, in Isachsen, Y. W., ed., Origin of anorthosite and related rocks: New York State Mus. and Sci. Service Mem. 18, p. 307—315 [1969]. Emslie, R. F., 1970‘, The geology of the Michigamau intru- sion, Labrador (13L, 231): Canada Geol. Survey Paper 68—57, 85 p. Fish, G. E., Jr., 1962, Titanium resources of Nelson and Amherst Counties, Va., pt. 1, Saprolite ores: U.S. Bur. Mines Rept. Inv. 6094, 44 p. Green, T. H., 1969, High-pressure experimental studies on the origin of anorthosite: Canadian Jour. Earth Sci., v. 6, p. 427—440. Gross, S. 0., 1968, Titaniferous ores of the Sanford Lake dis- trict, New York, in Ridge, J. D., ed., Ore deposits of the United States 1933—1967, v. 1: New York, Am. Inst. Mining, Metall., and Petroleum Engineers, p. 140—153. GEOLOGY AND RESOURCES OF TITANIUM Grout, F. F., 1950, The titaniferous magnetites of Minnesota: St. Paul, Office of Comm. Iron Range Resources and Rehabilitation 1949—1950, 117 p. Grout, F. F., Sharp, R. P., and Schwartz, G. M., 1959, The geology of Cook County, Minnesota: Minnesota Geol. Sur- vey Bull. 39, 163 p. . Hargraves, R. B., 1962, Petrology of the Allard Lake anortho- site suite, Quebec, in Petrologic studies—A volume in honor of A. F. Buddington: New York, Geol. Soc. America, p. 163—189. Henrz, Norman, 1968, The Roseland alkalic anorthosite massif, Virginia, in Isachsen, Y. W., ed., Origin of anorthosite and related rocks: New York State Mus. and Sci. Service Mem. 18, p. 357—367 [1969]. 1969, Anorthosite belts, continental drift, and the anorthosite event: Science, v. 164, p. 944-947. Herz, Norman, Valentine, L. E., and Iberall, E. R., 1970, Ru‘tile and ilmenite placer deposits, Roseland district, Nelson and Amherst Counties, Virginia: U.S. Geol. Sur- vey Bull. 1312—F, 19 p. Isachsen, Y. W., 1968, Origin of anorthosite and related rocks—A summarizatio‘n, in Isachsen, Y. W., ed., Origin of anorthosite and related rocks: New York State Mus. and Sci. Service Mem. 18, p. 435—445 [1969]. Lister, G. F., 1966, The composition and origin of selected iron-titanium deposits: Econ. Geology, v. 61, p. 275—310. Paulson, E. G., 1964, Mineralogy and origin of the titanif- erous deposit at Pluma Hidalgo, Oaxaca, Mexico: Econ. Geology, v. 59, p. 753-767. Philpotts, A. R., 1967, Origin of certain iron-titanium oxide and apatite rocks: Econ. Geology, v. 62, p. 303—315. Rose, E. R., 1969, Geology of titanium and titaniferous de posits of Canada: Canada Geol. Survey Econ. Geology Rept. 25, 177 p. Ross, C. S., 1941, Occurrence and origin of the titanium de- posits of Nelson and Amherst Counties, Virginia: U.S. Geol. Survey Prof. Paper 198, 59 p. Taylor, R. B., 1964, Geology of the Duluth Gabbro Complex near Duluth, Minnesota: Minnesota Geol. Survey Bull. 44, 63 p. Verhoogen, John, 1962, Oxidation of iron-titanium oxides in igneous rocks: Jour. Geology, v. 70, p. 168—181. Watson, T. L., and Taber, Stephen, 1913, Geology of the ti- tanium and apatite deposits of Virginia: Virginia Geol. Survey Bull. 3A, 308 p. Yoder, H. S., Jr., 1968, Experimental studies bearing on the origin of anorthosite, in Isachsen, Y. W., ed., Origin of anorthosite and related rocks: New York State Mus. and Sci. Service Mem. 18, p. 13—22 [1969]. Titanium Deposits in Alkalic Igneous Rocks By NORMAN HERZ GEOLOGY AND RESOURCES OF TITANIUM GEOLOGICAL SURVEY PROFESSIONAL PAPER 959—E 15“"; CONTENTS Page Abstract ________________________________________________________________ E1 Introduction ______________________________________________________________ 1 Description of individual localities __________________________________________ 2 Magnet Cove, Arkansas ________________________________________________ 2 Kola Peninsula, U.S.S.R. ______________________________________________ 3 Tapira, Minas Gerais, Brazil ___________________________________________ 4 Outlook __________________________________________________________________ 6 References cited __________________________________________________________ 6 ILLUSTRATIONS Page FIGURE 1. Geologic map of the Magnet Cove area of Arkansas ____________ E2 2. Generalized geologic map of the Kola Peninsula and northern Karelia in the U.S.S.R __________________________________ 3 3. Location and ages of alkalic rocks from southern Brazil _________ 5 TABLE Page TABLE 1. Titanium deposits associated with alkalic complexes _____________ E1 III GEOLOGY AND RESOURCES OF TITANIUM TITANIUM DEPOSITS IN ALKALIC IGNEOUS ROCKS By NORMAN HERZ ABSTRACT Many types of alkalic rocks are enriched in Ti02 relative to the average content of the crust. Silicate minerals in alkalic rocks may have high Ti02 content, but the oxides il- menite, rutile, brookite, and perovskite are characteristic primary phases. These oxides tend to be higher in niobium than the same minerals from other rock suites. At Tapira, Brazil, titanium minerals in a carbonatite com- plex comprise tens of millions of tons contained Ti02. Anatase (as “leucoxene”) is present in laterite lying over parent rocks containing perovskite, ilmenite, and rutile, At Magnet Cove, Airk., and the Kola Peninsula in the U.S.S.R., production of titanium minerals is, at best, economically marginal. INTRODUCTION In many alkalic rocks, titanium is enriched sever- al times beyond the normal concentration in igneous rocks (the normal being about 0.7 to 1.0 percent TiOz). Nockolds (1954) listed the following average TiO2 contents of alkalic rock types, in percent: Alkali pyroxenite _________ 3.31 Nepheline monzonite ______ 2.49 Essexite _________________ 2.81 Ijolite ___________________ 1.41 Average subsilicic ________ 1.9 igneous rock (721 analyses). Several silicate minerals characteristic of alkalic rocks are high in TiOz. Except for sphene, melanite garnets from alkalic rocks contain the highest amount of TiO2 found in silicate minerals; melanite from ijolite in Iron Hill, 0010., has 5.08 percent Ti02; elsewhere, melanite contains as much as 17 percent (Deer and others, 1962, p. 91). Kaersutite, an am- phibole containing 5 to 10 percent TiOZ, and titan- augite, a pyroxene containing 3 to 6 percent Ti02, are common in, but not restricted to, alkalic rocks. The principal titanium minerals found in alkalic rocks, ilmenite, titano-magnetite, sphene, perovskite, rutile, anatase, and brookite, commonly contain sig- nificant amounts of niobium (Nb) and rare earths (Fleischer and others, 1952). Typical alkalic rocks and deposits are shown in table 1. Most titanium-oxide minerals form late in the crystallization history of alkalic complexes. Chemical activities of silicon, aluminum, and iron and the partial pressure of oxygen are especially important in determining whether titanium will enter silicate minerals or form independent oxides. Some dikes consisting of iron-titanium oxides and apatite in alkalic rocks have been ascribed to the formation of immiscible melts having the composition of iron and titanium oxides plus phosphate within a predomi- nately silicate liquid magma chamber (Philpotts, 1967 ). On the basis of extrapolation of experimental data, Philpotts deduced that liquids of magnetite- ilmenite—apatite composition can form immiscible melts in silicate magmas that are enriched in sodium, TABLE 1.—Titam'um deposits associated with alkalic complexes (modified from Routhier, 1963, p. 958—959) Type Locality Titanium mineralogy nlitifizfigfs Nepheline syenite Iveland-Evje region, south Nor- Nb-rutile, euxenite, and other Y-rich pegmatites. way; Kola Peninsula, U.S.S.R.; titanates. Ilmen Mountains (Urals) , U.S.S.R. Intermediate and basic Khibiny, Kola Peninsula, U.S.S.R.; In i jolite and pyroxenites, ititano- Apatite. alkalic rocks. Oka, Quebec, Canada; Jacupiranga, magnetite and ilmenite; in “silexo- sac Paulo, Brazil; Tapira, Minas carbonatites,” sphene and perovskite _ . _ Gerais, Brazil. with secondary anatase. Rutlle veins 1n carbonatite__ Magnet Cove. Ark. Nb-rutile withsecondary brookite Albite. and “leucoxene.” E1 E2 have a dioritic composition, and are undergoing strong differentiation. Some titanium orebodies found in alkalic rocks of the Kola Peninsula and Brazil may have formed by such a process. DESCRIPTION OF INDIVIDUAL LOCALITIES Titanium production of any significance has been attempted only in deposits in alkalic rocks at Magnet Cove, Ark.; the Kola Peninsula in the U.S.S.R.; and Tapira, Minas Gerais, Brazil. Potential domestic resources of titanium include occurrences at Iron Hill, Colo. (Larsen, 1942), and Lemhi County in Idaho (Anderson, 1960). ‘Perov- skite and ilmenite have been described in mafic alkalic rocks and carbonatite from Iron Hill. Niobium-rich ilmenite (as much as 1 percent Nb) and ilmeno-rutile (niobium-bearing rutile, Nb as much as 13 percent) are found in Lemhi County in a rock type that appears to be a carbonatite. MAGNET COVE, ARKANSAS The alkalic complex of Magnet Cove, Ark., con- sists of a variety of feldspathoid-rich rocks that have intruded sediments of Paleozoic age (Erickson and Blade, 1963). The complex has been dated as early Late Cretaceous (95:5 m.y.) (Zartman and others, 1965) , as have some alkalic rocks in Brazil. The intrusion is essentially a ring-dike complex (fig. 1) that has metamorphosed the country rock as far as 800 m. The core of the ring dike consists of ijolite and carbonatite. The ijolite is a feldspar—free rock, which averages 2.20 percent TiO2 and is com- posed of nepheline, sodium-rich pyroxene, ti- tanium-garnet, biotite, sphene, and magnetite; the carbonatite is a coarse-grained calcite rock contain- ing accessory sodium-rich perovskite and zirconium- rich garnet. The intermediate ring is composed of trachyte and phonolite, which average 1.8 percent TiO2 and contain feldspar, feldspathoids, aegirine, diopside, and sphene. The outer ring consists of sphene-bearing nepheline syenite, garnet pseudoleu- cite syenite, and jacupirangite. Smaller dikes of alkalic rocks are abundant, both within and outside the complex, as are veins of various types. Quartz- brookite-rutile veins are most common in recrystal- lized novaculite on the eastern edge of the complex, and feldspar-carbonate, feldspar, quartz-feldspar, and fluorite veins containing rutile occur within the complex. Titanium deposits include the Magnet Cove rutile deposit, the Christy brookite deposit, the Har- dy-Walsh brookite deposit, and the Moi-Ti brookite- molybdenite deposit. Of these, the only titanium GEOLOGY AND RESOURCES OF TITANIUM 92°53’ 52’ O 500 1000 1500 FEET HTH—th—Lfi—_%__l 0 500 1000 1500 METRES EXPLANATION Sedimentary rocks, metamorphosed near mtruswes I l Carbonatite PALEOZOIC B Ijolite 1% B m :3 Magnet Cove Phonolite 8 rutile deposit ii EDS E 2 x Garnet pseudoleucite Dd Christy brookite deposit syenite U 3 >6 Hardy-Walsh Nepheline syenite brookite deposit 4 X Mo-Ti brookite deposit FIGURE 1.—Geolog'ic map of the Magnet Cove area of Ar- kansas (after Erickson and Blade, 1963). Jacupirangite J deposit of any economic significance is the Magnet Cove deposit, which produced 5,000 tons of rutile concentrates from 1932 through 1944 (Erickson and Blade, 1963, p. 89). No production of brookite has been reported from any of the other deposits. The Magnet Cove rutile deposit consists largely of phonolite that has been intruded, brecciated, and hydrothermally altered by a number of veins of different types. Rutile is contained in feldspar-car- bonate-rutile veins and vein masses that range from several inches to tens of feet in thickness. About 15 to 20 percent of the TiO2 in the veins is» present as TITANIUM DEPOSITS IN ALKALIC IGNEOUS ROCKS “leucoxene” (Fryklund and Holbrook, 1950). The average grade of the deposit is 3 percent rutile. Most of the exposed material has been weathered to a clay-rich, highly friable material; information from drill cores shows that weathering has penetrated to a depth of at least 45 m. The estimated reserves of rutilebearing material at Magnet Cove are 7.75 million tons containing ap- proximately 100,000 tons of rutile concentrate con- taining 95 percent TiO2 (E. C. Toewe and others, written commun., 1971). Before the deposit can be economically exploited, however, two serious prob- lems must be solved: (1) The best present beneficia- tion techniques result in a recovery of less than 60 percent of the contained rutile, and (2) the rutile averages 2.2 percent Nb, 1.8 percent Fe, and 0.6 percent V (Herz, 1976, table 2; Erickson and Blade, 1963, p. 80—81)—-a composition that cannot be used for welding rod coatings or for the chloride process of pigment and metal manufacturing. KOLA PENINSULA, U.S.S.R. Important titanium deposits are found in alkalic rocks of the extreme northwestern part of the Soviet Union, in the Kola Peninsula and in the Karelian Autonomous S.S.R. Specific information on grade and tonnages is not available, although general de- scriptions of the geology of the mining areas do exist (Malyshev, 1957, p. 204—207; Vlasov and others, 1966; Yudin and Zak, 1971). The area is part of the Precambrian Baltic Shield and consists largely of Archean gneisses, granites, migmatites, and granu- lites and Proterozoic metasedimentary and volcanic formations, gabbro-anorthosites, basalts, granites, and alkalic gabbros and granites. These are overlain by upper Precambrian and Lower Cambrian plat- form sediments. Paleozoic alkalic rocks, including ultrabasic-alkalic rocks 340 to 590 my old and nepheline syenites about 290 my. old, intrude the Archean and Proterozoic rocks (fig. 2). The ultrabasic-alkaline group of plutons consists of an earlier phase of olivinite, pyroxenite, and nepheline pyroxenite and a later group of ijolite and related rocks. The earlier ultrabasic rocks include large stocklike bodies of late magmatic perovskite— titanomagnetite ores that are enriched in rare ele- ments (Yudin and Zak, 1971). In the largest deposit, Afrikanda, a pipelike orebody can be traced to a depth of more than 400 m. These ores are character- ized by high amounts of TiO2 (85-180 percent) and total iron (11—18 percent). The rocks of this phase are generally high in major elements Ca, Mg, Na, Fe, and Ti and in minor elements P, C, Sr, and E3 36° 38° 40° | | l <94? 04» 735* /,\|‘.\ mansk/I \ \/ \ " \/,:" ,/ 300 MILES 100 200 300 KILOMETRES EXPLANATION Hercynian nepheline syenite plutons: Kh = Khibiny L = Lovozero K = Kontozero Caledonian ultrabasic-alkaline plutons: 1. Mount Turley 5. Pesochnyy 11.Mavrugubinsk 2. Kovdozero 6. Ingozero 12. Kovdor 3. Vuorjarvl 7. Salmagorsk 13. Kurginsk 4. Sallanlatvlysk 8-10. Khabozero group 14. Seblyavr Cl Jotnian and Lower Cambrian subplatform and platform deposits i' -<./. .‘f,‘>-- ’.'L .>-. Proterozoic rocks including metasedimentary, metavolcanics, granites, alkaline gabbros, alkaline granites, basalts, and ultramafics Lower Proterozoic gabbro-anorthosite massifs \254 MogiGua§u+ + + ++++++++++++++++++ + + 1 + ,:++ T“ 1, ++Rio Bonito 59Casimiro 4. , + + + r + t 1119366. , +69+ +.d§Abreu Ipanema 123 24° - + -r + . + + 4- + Itapirapua + + 103 o + + v 4v 4- + v 4 t 1 + + Itanhaem 130 8° 1 Serfote 127 136 ’ .132 J * + + C a) ati ./ acupiranga 82 Cananéia a + + 0 100 O 100 ’I. Ilha Montao de Trige oCE AS 51 Cabo Frio ‘ 81 ‘ Sao Sebastiio EXPLANATION Basaltic rocks and later sedlmentary rocks [1 PALEOZOIC CRETACEOUS Sedimentary rocks } + + ,. + ng—T—Q 200 KILOMETRES Precambrian rocks 200 MILES E Alkaline rocks (and age in m.y.) FIGURE 3.—Location and ages of alkalic rocks from southern Brazil (after Amaral and others, 1967) . tonite, which forms part of a ring in the southern part of the complex; volcanic breccia, which forms, much of a ring to the southwest; and carbonatite, which is abundant throughout the central part of the complex. Large masses of other types occur, in- cluding an apatite rock of abundant perovskite and subordinate sphene and magnetite, which forms parts of two rings, an inner one, 2,000 m long and as high as 300 m on the western side, and another similar one on the eastern side. The Brazilian Ministry of Mines and Energy (Companhia de Pesquisa de Recursas Minerais, 1972) has extensively explored the area and esti- mated that it contains almost 132 million tons of ore reserves (measured, indicated, and inferred) aver- aging 21.6 percent Ti02. The greatest concentration of titanium minerals at Tapira is in an area about 3X4 km on which a lateritic cover about 100 m thick has formed. The primary titanium minerals of this deposit are perovskite, ilmenite, magnetite, and, rarely, rutile, but the bulk of the reserves are ana- tase and “leucoxene,” the weathering products of the primary minerals. At least 60 percent of the E6 total Ti02 measured in laterite samples is present as anatase. Perovskite occurs in veins and disseminations in the country rock; it and its weathered products form as much as 5 percent of the laterite. Magnetite and ilmenite also occur in veins and in the alkalic coun- try rock; they and their weathered products com- prise from 20 to 70 percent of some laterite. The Tapira deposit formed mainly as a result of two processes: (1) Development of a titanium-rich segregation in the carbonatite plug, presumably as an immiscible liquid in the melt, and (2) tropical weathering, which enriched the tenor of titanium in the protore minerals at the expense of calcium and iron and perhaps also niobium. Fresh rutile from Tapira has 0.5 percent Nb205 as well as 3.5 to 6.8 percent Fe (Alves, 1960, p. 16) , which renders fresh rutile from Tapira as economically handicapped as rutile from Magnet Cove, Ark. The composition of fresh perovskite is not known but is likely to be high in niobium also. OUTLOOK Titanium is enriched in many alkalic rocks. In many alkalic complexes, however, much of the titani- um is in silicates, especially garnet and clinopyrox- ene; in others, where titanium forms primary minerals like perovskite and rutile, the accessory elements niobium and iron have made the minerals unusable for chloride-process manufacturing of pig- ment and titanium metal. Titanium at present can be considered a resource in alkalic rocks in two cir- cumstances: (1) Where tropical weathering has produced an economic titanium product by leaching out the undesirable elements, as it has in Tapira and elsewhere in Brazil, and (2) where titanium can be obtained as a byproduct, as it can in mining for niobium and for phosphate, rare earths, and alumi- num in the Kola Peninsula. REFERENCES CITED Alves, B. P., 1960, Distrito niébio-titanifero de Tapira: Brazil Divisao de Fomento da Producao Mineral., Boletim 108, 48 p. [in Portuguese]. Amara], G., Cordani, U. G., Kawashita, K., and Reynolds, J. H., 1966, Potassium-argon dates of basaltic rocks from southern Brazil: Geochim. et Cosmochim. Acta, v. 30, p. 159—189. Amaral, G., Bushee, J., Cordani, U. G., Kawashita, K., and Reynolds, J. H., 1967, Potassium—argon dates of alkaline rocks from southern Brazil: Geochim. et Cosmochim. Acta, v. 31, p. 117—142. GEOLOGY AND RESOURCES OF TITANIUM Anderson, A. L., 1960, Genetic aspects of the monazite and columbium-bearing rutile deposits in northern Lemhi County, Idaho: Econ. Geology, v. 55, p. 1179—1201. Companhia de Pesquisa de Recursos Minerais, 1972, Jazidas de niquel e de alguns nao ferros em Minas Gerais: Belo Horizonte, Brazil Agéncia de Minas Gerais, open-file rept., p. 7—11 [in Portuguese]. Deer, W. A., Howie, R. A., and Zussman, Jack, 1962, Rock- forming minerals, v. 1, Ortho- and ring silicates: New York, John Wiley and Sons, 333 p. Erickson, R. L., and Blade, L. V., 1963, Geochemistry and petrology of the alkalic igneous complex at Magnet Cove, Arkansas: U.S. Geol. Survey Prof. Paper 425, 95 p. Fleischer, Michael, Murata, K. J ., Fletcher, J. D., and Narten, P. F., 1952, Geochemical association of niobium (colum- bium) and titanium and its geological and economic sig- nificance: U.S. Geol. Survey Circ. 225, 13 p. Fryklund, V. C., and Holbrook, D. F., 1950, Titanium ore de- posits of the Magnet Cove area, Hot Springs County, Arkansas: Arkansas Research Devel. Comm. Div. Geology Bull. 16, 173 p. Herz, Norman, 1966, Tholeiitic and alkalic volcanism in south- ern Brazil, in Internat. Field Inst., Brazil, 1966, Guide- book: Am. Geol. Inst., chapter 5, p. V-1—V-6. 1976, Titanium deposits in anorthosite massifs, in Geology and resources of titanium: U.S. Geol. Survey Prof. Paper 959—D, 6 p. Larsen, E. 8., Jr., 1942, Alkalic rocks of Iron Hill, Gunnison County, Colorado: U.S. Geol. Survey Prof. Paper 197—A, 64 p. Malyshev, I. I., 1957, Zakonomernosti obrazovaniya i raz- meshcheniya mestorozhdenii titanovykh rud [Regularities of formation and distribution of deposits of titaniferous ores]: Moscow, Gosudar. Nauch.—Tekh. Izd. Lit. Geol. i Okhrane Nedr., 272p. [in Russian]. Nockolds, S. R., 1954, Average chemical compositions of some igneous rocks: Geol. Soc. America Bull., v. 65, p. 1007— 1032. Parker, R. L., and Adams, J. W., 1973, Niobium (columbium) and tantalum, in Brobst, D. A., and Pratt, W. P., eds., United States mineral resources: U.S. Geol. Survey Prof. Paper 820, p. 443—454. Philpotts, A. R., 1967, Origin of certain iron-titanium oxide and apatite rocks: Econ. Geology, v. 62, p. 303—315. Routhier, Pierre, 1963, Les gisements métalliféres—Géologie let principes de recherche: Paris, Masson et Cie, 1295 p. [in French]. Turner, F. J., and Verhoogen, John, 1960, Igneous and meta- morphic petrology (2d ed.): New York, McGraw-Hill Book Co., 694 p. Vlasov, K. A., Kuz’menko, M. V., and Es’kova, E. M., 1966, The Lovozero alkali massif: New York, Hafner Pub- lishing Co., 627 p. Yudin, B. A., and Zak, S. I., 1971, Titanium deposits of north- western USSR (eastern part of Baltic Shield): Internat. Geol. Rev., v. 13, p. 864—872. Zartman, R. E., Marvin, R. F., Heyl, A. V., and Brock, Maurice, 1965, Age of alkalic intrusive rocks from east- ern and mid-continent States, in Geological Survey re- search 1965: U.S. Geol. Survey Prof. Paper 525—A, p. A162. Titanium Minerals in Deposits Of Other Minerals By E. R. FORCE GEOLOGY AND RESOURCES OF TITANIUM GEOLOGICAL SURVEY PROFESSIONAL PAPER 959-F CONTENTS Page Abstract _________________________________________________________________ F1 Magnitude of potential byproduct recovery ________________________________ 1 Sand and gravel mining operations _________________________________________ 1 Titanium minerals in bauxite ______________________________________________ 2 Rutile in porphyry copper and related deposits _______________________________ 3 Rutile in kyanite deposits __________________________________________________ 3 Titanium minerals in nepheline syenites ____________________________________ 3 Rutile in anorthositic feldspar deposits _____________________________________ 4 Titanium minerals in marine phosphorite ___________________________________ 4 Titanium minerals with vanadium, platinum, and chromium in stratiform mafic igneous .rocks ________________________________________________________ 4 References cited __________________________________________________________ 4 TABLE Page TABLE 1. Estimated resources and potential yearly production of titanium min- erals from some deposits of other minerals ____________________ F2 III GEOLOGY AND RESOURCES OF TITANIUM TITANIUM MINERALS IN DEPOSITS OF OTHER MINERALS By E. R. FORCE ABSTRACT Titanium minerals in quantities equivalent to a significant percentage of present world production are moved but not recovered from mines of other minerals. Resources of this type amount to millions of tons of contained Ti02. Some pos- sible sources of byproduct titanium are (1) ilmenite and minor rutile detrital concentrations in deposits of tin, mona- zite, gold, platinum, uranium, and sand and gravel, (2) leu- coxenitized ilmenite and, locally, rutile in some bauxites, (3) rutile in porphyry copper and related deposits, (4) rutile in kyanite deposits, (5) perovskite and other titanium minerals in deposits of niobium, apatite, and nepheline in nepheline syenites, (6) rutile in alkalic anorthosite deposits of feldspar, and (7) titanium minerals in marine phosphorite. MAGNITUDE OF POTENTIAL BYPRODUCT RECOVERY The recovery of titanium minerals as byproducts of other minerals is of interest because limited re sources are best conserved by recovery of byprod- ucts and because competing land use and environ- mental considerations limit new mining of known resources. The appraisal here is intended to include only those situations in which the major product is being mined or has recently been mined; thus, con- siderations of the major-product economics are eliminated. Information is available for only a small minority of specific deposits of the general types discussed. Therefore, these specific deposits are just examples. Even so, if all the ilmenite were recovered from these deposits, the present world production of i]- menite could be increased by 20 percent or more (table 1). The percentage for rutile recovery is poor- ly defined but is probably far greater than 7 percent; the values are limited partly by lack of major- product production figures for deposits in the U.S.S.R. and partly by lack of grade information in several other cases. Some mining operations are al- ready contributing, as byproducts, at least 5 percent of present world titanium production; in still others, titanium minerals are major coproducts. SAND AND GRAVEL MINING OPERATIONS Ilmenite and rutile are common minor minerals in a great many fluvial and marine sands and in sand- stones. If some other mineral could be mined from these sands and sandstones, ilmenite and (or) rutile could also be recovered; the cost of byproduct recov- ery, however, is not evaluated herein. Scant information is available about the presence of titanium minerals or the feasibility of their recov- ery for certain types of large sand-mining opera- tions, particularly gold and platinum placers and uranium mines in sandstone. Types of deposits for which more information is available are described below. Tin placers and their “amang” byproducts—In Malaya, and to a lesser extent in Thailand, alluvial placer deposits mined primarily for cassiterite (tin) are also the source of byproduct ilmenite and other minerals, which are collectively called “amang.” All the recovered minerals are apparently derived from granitic stocks. The ilmenite concentrate from amang is low in chromium and contains about 53 percent TiOz; thus, it is good raw material for the manufacture of pig- ment (Industrial Minerals, 1972). As a result of by- product recovery alone, Malaya currently ranks fifth in the world among ilmenite producers (Noe, 1973). Production of ilmenite from amang could increase about 30 percent if recovery were increased in Thai- land and Indonesia, if it is assumed that ilmenite contents are similar throughout the province (from values given by Noe (1973) and in Industrial Min- erals (1972) ). Hypothetical resources of ilmenite are on the order of 20 million tons (from figures F1 F2 GEOLOGY AND RESOURCES 0F TITANIUM TABLE 1,—Estimated resources and potential yearly production of titanium minerals from some deposits of other minerals [H, hypothetical; I, identified; R, reserves] Potential . Estimated earl Type Deposit Country Mineral resource priducdion Recovered? (106mm) (103 tons) Tin placers ________________________ Malaya, Thailand, Ilmenite _____________ 1 5 (H) 240 Mostly. Indonesia. Nigeria ___________ ____do ______________ .2 (H) 1 N04? Multiproduct Valley County, United States ______ Low TlOz ilmenite _ _ _ _ 1 (I) 30 When placers. Idaho. mined. Sand and gravel- _ ___________________ Southeastern United High TiOZ ilmenite ________ 186 No. States. Rutile _____________________ 17 No. Bauxite ______________________________________________ High Ti02 ilmenite -__ ,_____ 100 N0. Madras and Orissa- _ India _____________ Rutile _______________ .06 (R) _ _ - No. Porphyry copper- B agdad, Ariz ______ United States ______ _ _ _ -do ___________________ 3 N o. Kyanite ________ Quartzites _________ ____do ____________ ____do ______________ .4 (I) 2 No? Anorthosite _____ Montpelier, Va ______ _ _ _ _do ____________ _ _ _ _do ___________________ 2 Heavy tailings. Marine Aurora, N.C _______ ____do ____________ Ilmenite __________________ 10 No. phosphorite. Bone Valley Forma— _ _ _ _do ____________ High Ti02 ilmenite ________ 10 N0. tion, F‘la. Rutile ____________________ 1 No. given by Sainsbury and Reed (1973) and Noe (1973)). Alluvial cassiterite deposits of Nigeria also con- tain ilmenite and rare rutile; ilmenite recovery could amount to roughly one-third to one—sixth of the eas- siterite production (Mackay and others, 1949). Here, the titanium minerals and the cassiterite ap- parently have different source rocks, and some lean cassiterite resources have abundant ilmenite and rutile (Raeburn, 1926). Multiprodnot placers of Idaho.—Alluvial placers in and near Valley County in Idaho have been inter- mittently worked for monazite, euxenite, columbite, tantalite, zircon, ilmenite, and garnet. The deposits are in immature gravel derived from granitic rocks of the Idaho batholith; some of the minerals are localimed around specific sources. The placers contain more than 1 million tons of ilmenite (Savage, 1961, 1964), but the average grade is too low to justify mining for ilmenite alone. The “ilmenite” averages 45 percent Ti02 and has not been used for pigment (Storch and Holt, 1963). Sand and gravel pits in southeastern United States.—Davis and Sullivan (1971) showed that il- menite, rutile, and other heavy minerals could be recovered from sand and gravel mined mostly from Cretaceous sandstone and modern alluvium in the southeastern United States. Total heavy minerals average 1 to 2 percent; ilmenite is the most abundant heavy mineral. The TiO2 contents of ilmenite varied from 51 to 68 percent. The southeastern States produce only 2 percent of the Nation’s sand and gravel. The potential produc- tion of ilmenite and rutile from sand and gravel pits in other areas of the United States is unknown; the southeast, however, is probably the most favorable ‘ area because the intense weathering there produces high TiO2 ilmenite. Sand and gravel operations in some tropical countries would likewise be promising sources of titanium minerals. TITANIUM MINERALS IN BAUXITE The chemical enrichment of TiO2 and the residual enrichment of ilmenite and rutile during weathering are particularly evident in bauxitization, and many authors have discussed the possibility of recovering titanium minerals as well as alumina from bauxite (Calhoun, 1950; Patterson, 1967; Stamper, 1970, p. 449). According to Hartman (1959), titanium in bauxite commonly occurs both as grains of ilmenite greatly enriched in TiO2 (“le-ucoxene”) and as TiO2 polymorphs finer than 40 ,L. The fine-grained materi- al is difficult to separate, and perhaps separation would require chemical means. Some Hawaiian baux- ites (Sherman, 1952; Patterson, 1971) and some red mud waste from Indian bauxites (Patterson, 1967, p. 7) contain more than 20 percent Ti02. Grains of ilmenite or of “leucoxene” consisting of anatase and rutile are present in most bauxites, and, in bauxite over parent rocks containing significant amounts of ilmenite (such as volcanic rock, sand- stone, schist, and syenite), the grains commonly comprise more than 0.5 percent of the bauxite (Hartman, 1959). Such bauxites amount to about one-third of the total bauxite production (from data given by Patterson (1967) and Kurtz (1973, table 16) ). It seems plausible that 100,000 tons of ilmen- ite and “leucoxene” are being discarded each year from these bauxites. Bauxite in Madras and Orissa States in India (Roy Chowdhury, 1955, p. 210—211) is derived from TITANIUM MINERALS IN DEPOSITS OF OTHER MINERALS weathering of leptynite (quartz-garnet—feldspar gneiss) and khondalite (quartz-sillimanite-garnet- graphite gneiss). These rocks probably contain ru- tile, as do similar rocks nearby; if so, rutile is proba- bly present in the bauxite. On the basis of bauxite reserve estimates (Patterson, 1967, p. 10a) and the assumption that fresh rock contains 0.5 percent ru- tile, the maximum amount of rutile present is about 60,000 tons. Rutile may also be present in bauxite over gneiss in Guinea in West Africa. RUTILE IN PORPHYRY COPPER AND RELATED DEPOSITS Rutile is a common product of hydrothermal al- teration in the porphyry copper deposits of the southwestern United States (Creasey, 1966; Schwartz, 1966). At Bagdad, Ariz., it is present in copper ore (Anderson and others, 1955). This rutile contains about 2 percent Nb. Creasey (1966) found that, in propylitic zones of alteration, rutile is pres- ent only as “leucoxene,” whereas single grains of rutile are present in the argillic zones. Thin sections collected by R. G. Schmidt and D. P. Cox (U.S. Geo- logical Survey) from San Manuel and Ajo porphyry copper deposits show that grains of rutile coarser than 20 M are limited to the zone of potash feldspar; in this zone, these grains are as coarse as 0.3 mm. Sphene is present with or in place of rutile in sever- al specimens containing carbonate from Ajo; in two such specimens, the weight percent of rutile was 0.08 and 0.10. Some other types of porphyry and related deposits also contain rutile. A quartz dioritic porphyry cop- per deposit in Puerto Rico contains rutile (Cox and others, 1973, fig. 4). Rutile, high in niobium, is pres- ent in the Climax, Colo., molybdenum porphyry de- posit (Wallace and others, 1968, p. 626) and the Urad, Colo., tungsten molybdenum porphyry deposit (P. K. Theobald, oral commun., 1974). Rutile is be- ing produced (A. L. Clark, oral commun., 1973), along with copper, apatite, vermiculite, baddeleyite, monazite, Ti-magnetite, and pyrite, from a mafic alkalic deposit at Palabora, South Africa (South African Mining and Engineering Journal, 1970) . Possibly two reactions are responsible for rutile in hydrothermally altered porphyries. The first in- volves the introduction of sulfur. Kullerud and Yoder (1963) found that mafic minerals, such as biotite and hornblende, break down in the presence of sulfur to form pyrite, magnetite, and stripped silicates. Although Kullerud and Yoder did not ob- serve titanium minerals among the reaction prod- F3 ucts, a balanced equation having biotite or horn- blende among the reactants would imply their pres- ence. In porphyry copper deposits, high--TiO2 mag- matic biotite breaks down to low-TiO2 hydrothermal biotite (Moore and Czamanske, 1973), and mag- matic titaniferous magnetite breaks down to stoichi- ometric hydrothermal magnetite (Hamil and Nackowski, 1971). Since sulfides are present, the similarity to Kullerud and Yoder’s reactions is ap- parent; the liberated TiO2 apparently forms rutile. Probably rutile and copper ore’minerals are to some extent products of the same reaction. Rutile in al- tered mafic intrusives associated with replacement sulfide copper deposits at the Haile mine (Pardee and Park, 1948, p. 112; TiOZ polymorph is actually anatase, Norman Herz, written commun., 1970) and at the Fontana mine in the southeastern United States also indicates the importance of sulfur in re- actions producing rutile. The second reaction involves the introduction of C02. Schuili‘ng and Vink (1967 ) found that equilibri- um between rutile, Sphene, and carbonate is a func- tion of CO2 pressure; high CO2 pressure pushes the reaction toward rutile and carbonate. At Ajo, the coexistence of all three phases implies unusually high CO2 pressure. RUTILE IN KYANITE DEPOSITS Espenshade and Potter (1960) reported the pres- ence of rutile in kyanite quartzites of the southeast- ern United States. The rutile content averaged about 0.5 percent in the quartzites analyzed. Identified re- sources are on the order of 400,000 tons of rutile (from figures given by Espenshade and Potter (1960) and Espenshade (1973)). Rutile is present in the important kyanite deposits of northern India, but the grade has not been re- ported (Varley, 1965). Rutile of grades locally as high as 3 percent is commonly present in other de- posits of alumino—silicate minerals (Varley, 1965; Espenshade, 1973). TITANIUM MINERALS IN NEPHELINE SYENITES Alkalic rocks of the Kola Peninsula are mined for a variety of commodities by means of techniques not common outside the USSR. Herz (1976b) discusses a situation in which titanium minerals could be a byproduct of mining nepheline (aluminum) and apatite (phosphate and rare earth) from nepheline syenite at the Khibiny massif and of mining Nb- F4 perovskite (niobium) from nepheline syenite at the vLovozero massif. RUTILE IN ANORTHOSITIC FELDSPAR DEPOSITS Antiperthitic andesine feldspar is mined from a1- kalic anortho‘site in two areas of Piedmont Virginia. One is the Roseland district, a formerly important source of rutile (Herz, 1976a). Figures for the amount of rutile in the feldspar ore are not available, but, since the areas of the best feldspar and the most rutile are not coextensive, the amount is probably small. The other area is near Montpelier. Rutile was formerly mined here also, from the same pits in which feldspar is now mined. The rutile is coarse grained and occurs as rods in intensely lineated anorthosite. Rutile, sphene, apatite, and ilmenite are presently accumulating in a “heavy tailings” pond. TITANIUM MINERALS IN MARINE PHOSPHORITE Ilmenite, rutile, and other valuable heavy minerals are present in the waste products from phosphate mining in Florida (Stow, 1968). These minerals are present as clastic grains in phosphate pebble con- glomerate of the Bone Valley Formation, where they compose, on the average, 0.23 percent of the con- glomerate (Pirkle and others, 1967). The TiO2 con- tent of ilmenite is more than 60 percent (Stow, 1968), probably because of weathering alteration. Lamont and others (1972) found recovery of the heavy minerals to be limited by their fine grain size. Lewis (1974; written commun., 1975) found that phosphate at Aurora, N.C., contains about 0.1 per- cent ilmenite and that tailings from the flotation plant contain 2.5 percent ilmenite. The TiO2 content of the ilmenite is 52 percent. A channel sample taken by F. W. Whitmore (U.S. Geological Survey) across about 9 m of Pleistocene(?) silty sand overburden at the mine contains about 0.4 percent fine-grained ilmenite (overburden is not included in table 1). TITANIUM MINERALS WITH VANADIUM, PLATINUM, AND CHROMIUM IN STRATIFORM MAFIC IGNEOUS ROCKS Titanium minerals occur within the Bushveld com- plex of South Africa in some of the world’s most im- mrtant reserves of vanadium, platinum, and chromium. Vanadium ore occurs as magnetite in plugs and cumulative layers in the upper part of the complex. GEOLOGY AND RESOURCES OF TITANIUM The ore contains ilmenite as discrete cumulate grains and both ilmenite and ulvospinel as intergrowths in magnetite (Willemse, 1969). The coarser ilmenite comprises from 1 to 10 percent of the ore (Moly- neux, 1970). An average of 1 percent would imply about 2 million tons of ilmenite in the ore reserves. Platinum ore occurs in the Merensky Reef, down- section in the complex. Platinum is present as many compounds in pyroxenite. Other platinoids, gold, nickel, and copper are byproducts. Ilmenite and ru- tile are present as accessory minerals; a maximum amount of about 2 percent (Cousins, 1969) is set by TiOg content. Cumulate chromite forms chromitite in the lower part of the complex. Rutile is present in some seams as interstitial crystals, but a maximum of 0.4 to 0.7 percent is set by TiO2 content (Cameron and Emer- son, 1959). Because the chromite is smelted in bulk form, rutile cannot presently be recovered. Similar possibilities of titanium mineral recovery probably exist in other complexes. The chromite de- posits of Fiskenaesset, Greenland, contain rutile (Ghisler and Windley, 1967 ). In some deposits of ilmenite in mafic igneous rocks, vanadium is a by- product, as it is at Otanmaki, Finland; in other similar rocks, titanium could possibly be the by- product. REFERENCES CITED Anderson, C. A., Scholz, E. A., and Strobell, J. B., Jr., 1955, Geology and ore deposits of the Bagdad area, Yavapai County, Arizona: U.S. Geol. Survey Prof. Paper 278, 103 p. Calhoun, W. A., 1950, Titanium and iron minerals from black sands in bauxite: U.S. Bur. Mines Rept. Inv. 4621, 16 p. Cameron, E. N., and Emerson, M. E., 1959, The origin of certain chromite deposits of the eastern part of the Bushveld complex: Econ. Geology, v. 54, p. 1151—1213. Cousins, C. A., 1969, The Merensky Reef of the Bushveld igneous complex, in Wilson, H. D. B., ed., Magmatic ore deposits: Econ. Geology Mon. 4, p. 239—251. Cox, D. R, Larson, R. R., and Tripp, R. B., 1973, Hydro- thermal alteration in Puerto Rican porphyry copper de- posits: Econ. Geology, v. 68, p. 1329—1334. Creasey, S. C., 1966. Hydrothermal alteration, in Titley, S. R., and Hicks, C. L., eds, Geology of the porphyry copper deposits, southwestern North America: Tucson, Univ. of Arizona Press, p. 51—74. Davis, E. G., and Sullivan, G. V., 1971, Recovery of heavy‘ minerals from sand and gravel operations in the south- eastern United States: U.S. Bur. Mines Rept. Inv. 7517, 25 p. Espenshade, G. H., 1973, Kyanite and related minerals, in Brobst, D. A., and Pratt, W. P., eds., United States mineral resources: U.S. Geo-1. Survey Prof. Paper 820, p. 307—312. TITANIUM MINERALS IN DEPOSITS OF OTHER MINERALS Espenshade, G. H., and Potter, D. B., 1960, Kyanite, silli- manite, and andalusite deposits of the southeastern States: U.S. Geol. Survey Prof. Paper 336, 121 p. Ghisler, Martin, and Windley, B. F., 1967, The chromite deposits of the Fiskenaesset region, West Greenland: Gronlands Geol. Undersogelse Rap. 12, 39 p. Hamil, B. M., and Nackowski, M. P., 1971, Trace-element dis- tribution in accessory magnetite from quartz monzonite intrusives and its relation to sulfide mineralization in the Basin and Range province of Utah and Nevada—A pre- liminary report, in Boyle, R. W., and McGerrigle, J. 1., eds., Geochemical exploration: Canadian Inst. Mining and Metallurgy Spec. V. 11, p. 331—333. Hartman, J. A., 1959, The titanium mineralogy of certain bauxites and their parent materials: Econ. Geology, v. 54, p. 1380—1405. Herz, Norman, 1976a, Titanium deposits in anorthosite massifs, in Geology and resources of titanium: U.S. Geol. Survey Prof. Paper 959—D, 6 p. 1976b, Titanium deposits in alkalic igneous rocks, in Geology and resources of titanium: U.S. Geol. Survey Prof. Paper 959—E, 6 p. Industrial Minerals, 1972, By-product minerals from tin min- ing [Thailand]: no. 57, p. 35. Kullerud, Gunnar, and Yoder, H. S., J r., 1963, Sulfide-silicate relations: Carnegie Inst. Washington Year Book, v. 62, p. 215—218. Kurtz, H. F., 1973, Bauxite, in U.S. Bureau of Mines min- erals yearbook 1971; V. 1, p. 199—214. Lamont, W. E., Brooks, D. R., Feld, I. L., and McVay, T. N., 1972, Rutile and associated heavy minerals in Florida phosphate flotation plants: U.S. Bur. Mines open-file rept., 35 p. Lewis, R. M., 1974, Recovery of heavy minerals from North Carolina phosphate tailings: North Carolina State Univ. Minerals Research Lab. Rept. MRL—4, 10 p. Mackay, R. H., Greenwood, R., and Rockingham, J. E., 1949, The geology of the plateau tin fields, resurvey 1945—48: Nigeria Geol. Survey Bull. 19, 80 p. Molyneux, T. G., 1970, The geology of the area in the vicinity of Magnet Heights, eastern Transvaal, with special refer- ence to the magnetic iron ore, in Bushveld igneous com- plex and other layered intrusions, symposium: Geol. Soc. South Africa Spec. Pub. 1, p. 228—241. Moore, W. J., and Czamanske, G. K., 1973, Compositions of biotites from unaltered and altered monzonitic rocks in the Bingham mining district, Utah: Econ. Geology, v. 68, p. 269—274. Noe, F. E., 1973, Titanium, in U.S. Bureau of Mines minerals yearka 1971; v. 1, p. 1167—1180. Pardee, J. T., and Park, C. F., Jr., 1948, Gold deposits of the southern Piedmont: U.S. Geol. Survey Prof. Paper 213, 156 p. Patterson, S. H., 1967, Bauxite reserves and potential alumi- F5 num resources of the world: U.S. Geol. Survey Bull. 1228, 176 p. 1971, Investigations of ferruginous bauxite and other mineral resources on Kauai and a reconnaissance of fer- ruginous bauxite deposits on Maui, Hawaii: U.S. Geol. Survey Prof. Paper 656, 74 p. Pirkle, E. C., Yoho, W. H., and Webb, S. D., 1967, Sediments of the Bone Valley phosphate district of Florida: Econ. Geology, v. 62, p. 237-261. Raeburn, Colin, 1926, The geology of Mama, Nassarawa Province: Nigeria Geol. Survey Bull. 9, p. 9-19. Roy Chowdhury, M. K., 1955, Bauxite deposits of India and their utilization: Indian Minerals, v. 9, p. 195—221. Sainsbury, C. L., and Reed, B. L., 1973, Tin, in Brobst, D. A., and Pratt, W. P., eds., United States mineral resources: U.S. Geol. Survey Prof. Paper 820, p. 637—651. Savage, C. N., 1961, Economic geology of central Idaho blacksand placers: Idaho Bur. Mines and Geology Bull. 17, 160 p. 1964, Titanium, zirconium, and hafnium, in Mineral and water resources of Idaho: Rept. for U.S. Cong., Senate Comm. Interior and Insular Affairs, U.S. 88th Cong., 2d sess., p. 217—223. Schuiling, R. D., and Vink, B. W., 1967, Stability relations of some titanium minerals (sphene, perovskite, rutile, anatase): Geochim. et Cosmochim. Acta, v. 31, p. 2399— 2411. Schwartz, G. M., 1966, The nature of primary and secondary mineralization in porphyry copper deposits, in Titley, S. R., and Hicks, C. L., eds., Geology of the porphyry copper deposits, southwestern North America: Tucson, Univ. of Arizona Press, p. 41—50. Sherman, G. D., 1952, The titanium content of Hawaiian soils and its significance: Soil Sci. Soc. America Proc., v. 16, p. 15—18. South African Mining and Engineering Journal, 1970, On the trail of zirconium oxide: v. 81, pt. 1, no. 4043, p. 1085—1091. Stamper, J. W., 1970, Aluminum, in Mineral facts and prob- lems, 1970: U.S. Bur. Mines Bull. 650, p. 437—462. Storch, R. H., and Holt, D. C., 1963, Titanium placer deposits of Idaho: U.S. Bur. Mines Rept. Inv. 6319, 69 p. Stow, S. H., 1968, The heavy minerals of the Bone Valley Formation and their potential value: Econ. Geology, v. 63, p. 973—977. Varley, E. R., 1965, Sillimanite: London, Overseas Geol. Sur- veys, Mineral Resources Div., 165 p. Wallace, S. R., Muncaster, N. K., Jonson, D. C., MacKenzie, W. B., Bookstrom, A. A., and Surface, V. E., 1968, Multiple intrusion and mineralization at Climax, Colo- rado, in Ridge, J. D., ed., Ore deposits of the United States 1933—1967: New York, Am. Inst. Mining, Metal- lurgy, and Petroleum Engineers, v. 1, p. 605—640. Willemse, J ., 1969, The vanadiferous magnetic iron ore of the Bushveld igneous complex, in Wilson, H. D. B., ed., Mag- matic ore deposits: Econ. Geology Mon. 4, p. 187—208. itUS. GOVERNMENT PRINTING OFFICE: 1976 0—240-961/15 m. A»... « B4,; nxunvnxum mmmmmgmmmmmmnm m»fl¢;x\&t§ mm JWflufiflflm: W ‘ mnmumnwmm m WA W xQ-u: u .mmmum W mm. umwmnm m mm m. .3 v v! fluvmnjgnwhnmwflmg¢£mfihtupuma! COVER PHOTOGRAPHS l 2 3 4 5 6 7 8 9 10 ll 12 13 14 snowm— NO") . Asbestos ore . Lead ore, Balmat mine, N. Y. . Chromite‘chromium ore, Washington . Zinc ore, Friedensville, Pa. Banded iron-formation, Palmer, Mich. . Ribbon asbestos ore, Quebec, Canada . Manganese ore, banded rhodochrosite . Aluminum ore, bauxite, Georgia . Native copper ore, Keweenawan Peninsula, Mich. . Porphyry molybdenum ore, CoIorado . Zinc ore, Edwards, N. Y. . Manganese nodules, ocean floor . Botryoidal fluorite ore, Poncha Springs, Colo. . Tungsten ore, North Carolina Rutile in Precambrian Sillimanite-Quartz Gneiss and Related Rocks, East-Central Front Range, Colorado By SHERMAN P. MARSH and DOUGLAS M. SHERIDAN GEOLOGY AND RESOURCES OF TITANIUM IN THE UNITED STATES GEOLOGICAL SURVEY PROFESSIONAL PAPER 959—G Geologic and geochemical study of rutile-bearing rocks in Colorado and a description of their occurrence elsewhere UNITED STATES GOVERNMENT PRINTING OFFICE. WASHINGTON 21976 UNITED STATES DEPARTMENT OF THE INTERIOR THOMAS S. KLEPPE, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Director Library of Congress Cataloging in Publication Data Main entry under title: Geology and resources of titanium in the United States. (Geological Survey Professional Paper 959) CONTENTS: Force, E. R. Titanium contents and titanium partitioning in rocks—Force, E. R. Metamorphic source rocks of titanium placer deposits.—Black, M. C., Jr. and Morgan, B. A. Rutile and sphene in blueschist and related high-pressure facies rocks.—Herz, N. Titanium deposits in anorthosite massifs.—Herz, N. Titanium deposits in alkalic igneous rocks—Force, E. R. Titanium minerals in deposits of other minerals—Marsh, S. P. and Sheridan, D. M. Rutile in precambrian sillimanite-quartz gneiss and related rocks, east-central Front Range, Colorado. Includes bibliographies. Supt. of Docs. no.: 119.16z959 A—G 1. Titanium ores—United States. 2. Petrology—United States. I. Series: United States Geological Survey Professional Paper 959. QE390.2.T5G46 553’.462 75-619269 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock Number 024-001-02834-4 APPRAISAL OF MINERAL RESOURCES Continuing appraisal of the mineral resources of the United States is conducted by the US. Geological Survey in accordance with the provisions of the Mining and Minerals Policy Act of 1970 (Public Law 91-631, Dec. 31, 1970). Total resources for purposes of these appraisal estimates include currently minable resources (reserves) as well as those resources not yet discovered or not currently profitable to mine. The mining of mineral deposits, once discovered, depends on geologic, economic, and technologic factors; however, identification of many deposits yet to be discovered, owing to incomplete knowledge of their distribution in the Earth’s crust, depends greatly on geologic availability and man’s ingenuity. Consequently, appraisal of mineral resources results in approximations, subject to constant change as known deposits are depleted, new deposits are found, new extractive technology and uses are developed, and new geologic knowledge and theories indicate new areas favorable for exploration. This Professional Paper discusses aspects of the geology of titanium as a framework for appraising resources of this commodity in the light of today’s technology, economics, and geologic knowledge. Other Geological Survey publications relating to the appraisal of resources of specific mineral commodities include the following: Professional Paper 820—“United States Mineral Resources” Professional Paper 907—“Geology and Resources of Copper Deposits” Professional Paper 926—“Geology and Resources of Vanadium Deposits” Professional Paper 933—“Geology and Resources of Fluorine in the United States” CONTENTS Page Page Abstract ............................................. G1 Precambrian metamorphic rocks—Continued Introduction .......................................... 1 Origin of metamorphic rocks—Continued Acknowledgements ................................... 1 Rutile-bearing gneisses .......................... G13 Geologic setting ....................................... 2 Other localities of rock types similar to those in the Precambrian metamorphic rocks ........................ 2 east-central Front Range ............................. 13 Sillimanitic biotite gneiss (unit A) ................... 3 Farmville district, Virginia ......................... 14 Biotite gneiss (unit B) .............................. 3 Kings Mountain district, North Carolina- Interlayered feldspar-rich gneiss, hornblende gneiss, South Carolina .................................. 14 and amphibolite (unit C) .......................... 3 Graves Mountain, Georgia .......................... 14 Hornblende gneiss and amphibolite (unit D) ........... 3 White Mountain, California ......................... 14 Feldspar-rich gneiss (unit E) ........................ 3 Yuma County, Arizona ............................. 15 Rutile-bearing gneisses ............................ 4 Santa Cruz County, Arizona ........................ 15 Chemical composition .............................. 8 Other United States occurrences .................... 15 Geochemistry of rutile ............................. 10 World occurrences .............................. .. . 15 Origin of metamorphic rocks ........................ 12 Economic potential .................................... 16 Major lithologic groups ........................ 12 References cited ...................................... 17 ILLU STRATIONS Page FIGURE 1. Map showing distribution of Precambrian igneous and metamorphic rocks in east-central Front Range, Colo. .......... G2 2. Generalized geologic map showing distribution of Precambrian rutile-bearing gneiss, east-central Front Range ........ 4-5 3-8. Photomicrographs: 3. Rutile alined parallel to foliation in biotite-quartz-plagioclase gneiss ....................................... 7 4. Rutile alined parallel to foliation in sillimanite-quartz gneiss .............................................. 7 5. Fibrolitic sillimanite partly replaced by topaz and partly recrystallized to prismatic sillimanite ................ 7 6. Rutile associated with topaz .......................................................................... 7 7. Prismatic sillimanite intergrown with rutile ............................................................ 8 8. Rutile associated with corundum ...................................................................... 8 TABLE S Page TABLE 1. Analytical data for rutile-bearing gneiss and for other Precambrian metamorphic rocks, east-central Front Range, Colo .................................................................................................. G9 2. Semiquantitative spectrographic analyses of rutile ............................................................. 11 3. Standard of purity for rutile ................................................................................. 12 4. Ionic substitution factors ................................................................................... 12 METRIC-ENGLISH EQUIVALENTS Metric unit English equivalent centimetre (cm) metre (m) kilometre (km) 0.3937 inch (in) 3.28 feet (ft) .62 mile (mi) GEOLOGY AND RESOURCES OF TITANIUM IN THE UNITED STATES RUTILE IN PRECAMBRIAN SILLIMANITE-QUARTZ GNEISS AND RELATED ROCKS, EAST-CENTRAL FRONT RANGE, COLORADO By SHERMAN P. MARSH and DOUGLAS M. SHERIDAN ABSTRACT Unusual rutile-bearing gneisses are thinly interlayered with metasedimentary and metavolcanic gneisses of Precambrian age in the east-central part of Colorado’s Front Range. Layers and lenses of rutile-bearing gneiss range in thickness from 15 cm to 30 m and in rutile content from trace amounts to about 5 percent. Rutile is most abundant in a sillimanitic topaz-quartz gneiss but is also found in sillimanite-quartz gneiss and biotite-quartz-plagioclase gneiss. The rutile-bearing sillimanitic gneisses have a simple but unusual mineral and chemical composition because they contain little else but quartz, sillimanite, itopaz, and accessory rutile. Spectro- graphic analyses of rutile from the east-central Front Range and from eleven other localities in the United States indicate that the Front Range rutile is relatively pure, that is, it contains fewer metallic cations. We believe that the rutile-bearing gneisses originated from intermediate to basic tuffs and flows; intense weathering resulted in clays enriched in titania, alumina, and silica. These weathering products were reworked by surface waters, and fluorine from volcanic sources was adsorbed locally. Subsequently, the entire sequence of interlayered sedimentary and volcanic rocks was folded and metamorphosed to the sillimanite zone by regional metamorphism. Although the majority of known occurrences of rutile-bearing gneiss in the east-central Front Range may never be mined, owing to environmental and economic factors, this type of rutile occur- rence is also found elsewhere and may represent potential rutile resources. INTRODUCTION Rutile occurs in thin layers and lenses of light-colored sillimanite-quartz gneiss and related rocks within a thick succession of Precambrian metamorphic rocks in the east-central Front Range, Colo. The known occurrences are in a belt, 3 km wide and 26 km long, that extends east-southeast from the vicinity of Idaho Springs to 4 km south of Morrison (figs. 1, 2). Layers and lenses of rutile-bearing gneiss range in thickness from 15 cm to 30 m and in rutile content from trace amounts to 5 percent. The presence of rutile in an unusual sillimanitic topaz-quartz gneiss located 10 km northwest of Evergreen, 0010., was noted in March 1968 by D.M. Sheridan during petrographic studies related to a program of regional mapping. D. M. Sheridan, S. P. Marsh, and R. B. Taylor reported the results of a preliminary geologic investigation of this deposit in US. Geological Survey Circular 567 (Sheridan, Taylor and Marsh, 1968). Marsh and Sheridan continued searching for additional deposits of rutile during field studies in 1968-69. Because the rutile-bearing rocks described here are rather undistinguished in appearance, they might very easily be overlooked, even by a trained geologist. This report summarizes the results of field and laboratory examinations of rutile—bearing gneisses of the east-central Front Range. These gneisses constitute a large resource of rutile. Rocks such as the sillimanitic topaz-quartz gneiss could be mined principally for topaz and sillimanite, both of which can be used by high-alumina refractory industries for the manufacture of mullite; salable byproducts of such an operation would be rutile, crushed quartz, and fluorine. The fact that these gneisses have not as yet been mined may be attributed to the high cost of hard-rock mining and milling (compared to the low—cost methods used in obtaining rutile from Australian beach sands); rising land values due to the westward spreading of suburban Denver; and likely objections of local residents. Although the rutile deposits described in this report may never be mined, similar deposits may be found in other gneissic terranes where mining might be possible. ACKNOWLEDGMENTS A major part of this investigation was conducted under a comprehensive research program authorized G1 G2 GEOLOGY AND RESOURCES 0F TITANIUM 105°45’ 30’ 15' 105°oo’ 40° 00’ 39° 30’ 0 10 20 MILES O 10 20 KILOMETRES EXPLANATION CENOZOIC, [:1 mesozouc, AND Sedimentary rocks PALEOZOIC W Pikes Peak Granite 1,040 m.y. Silver Plume Granite 1,440 m.y. [:1 Boulder Creek Granite and related rocks 1,710 m.y. PRECAMBRIAN Metamorphic rocks J Contact FIGURE 1.—Distribution of Precambrian igneous and metamor- phic rocks in the east-central Front Range. Colo. (faults and Tertiary igneous rocks omitted). by the Office of Emergency Planning and undertaken by the Department of the Interior under the Defense Production Act for the purpose of developing a domestic source of rutile. U. S. Geological Survey colleagues W. N. Sharp, J. W. Adams, R. B. Taylor, Mary E. Mrose, and R. L. Erickson have helped with mineralogy and report preparation. GEOLOGIC SETTING The rutile-bearing gneisses lie within a broad area of Precambrian metamorphic and intrusive rocks in the east-central Front Range (fig. 1). Metamorphic rocks in this part of the Front Range are principally biotite gneisses and feldspathic gneisses, but amphibolite, hornblende gneiss, and calc—silicate gneiss are abundant in the east. All the area shown in figure 1 is in the sillimanite zone of regional metamorphism. Most of the pelitic rocks of the report area (fig. 2) are characterized by the stable pair sillimanite and potassic feldspar, but rocks of somewhat lower grade, characterized by the pair sillimanite/muscovite are common in areas to the northeast (Sheridan, Maxwell, and Albee, 1967, p. 37 -38) and southeast (Bryant and others, 1973, p. 3.) In the east-central Front Range this high-grade metamorphism accompanied a long period of deformation involving two stages of plastic folding, the first of which formed tight to open folds trending west to northwest; these trends predominate in the report area (fig. 2). Elsewhere, these folds were modified subsequently by cross folds trending north-northwest to north-northeast (R. B. Taylor, oral commun., 1964). Late in this long period of tectonism, bodies of Boulder Creek Granite (largely granodiorite) were intruded; isotopic dating by Rb/ Sr determina- tions indicates that the Boulder Creek Granite is about 1.710140 m.y. old (Hedge, 1969, p. 79). A third episode of folding, also accompanied by regional metamorphism and followed closely be cataclastic deformation, affected parts of the east-central Front Range at about the time the Silver Plume Granite was emplaced (R. B. Taylor, oral commun., 1967; Moench and others, 1962, p. 45—54; Sheridan, Maxwell, and Albee, 1967, p. 69-70); the age of the Silver Plume is about 1,400 i 40 m.y. (Hedge, 1969, p. 81). This late deformation has only minor expression in the area of figure 2 (see zone of cataclastically deformed gneisses, Sheridan, Reed, and Bryant, 1972). The Pikes Peak Granite, youngest of the major Precambrian intrusives, was emplaced about 1,040 i 10 m.y. ago (Hedge, 1969, p. 83). Late in Precambrian time extensive northwest-trending faults—the “breccia reefs” of Lovering and Goddard (1950, p. 79)—were formed (Tweto and Sims, 1963, p. 1001); some were reactivated during the Laramide orogeny. PRECAMBRIAN METAMORPHIC ROCKS In addition to the rutile-bearing gneisses, metasedimentary and metavolcanic gneisses occur in the report area and are divided herein into five principal units, units A to E of figure 2. Because RUTILE IN GNEISS AND RELATED ROCKS, FRONT RANGE, COLORADO primary sedimentary and volcanic features have been largely obliterated by high-grade metamorphism, stratigraphic relations among the five units are unknown. SILLIMANTIC BIOTITE GNEISS (UNIT A) Rocks of unit A are approximately coextensive with the rutile-bearing gneisses (fig. 2). Unit A is composed of gray fine- to medium-grained sillimanitic biotite gneiss containing abundant darker gray layers, 60 cm to 15 m thick, of garnetiferous sillimanitic biotite gneiss and of cordierite—bearing sillimanitic biotite gneiss. Lenses of dark-colored coarse-grained cordierite-gedrite-garnet gneiss, 60 cm to 15 m thick, are present locally. BIOTITE GNEISS (UNIT B) Thick sections of unit B, biotite gneiss, occur in the southwestern part of the area (fig. 2). There and in the northwestern part of the area the adjacent rocks are commonly hornblende gneiss and associated rocks of unit D. In general lithology, unit B is somewhat similar to unit A but lacks the abundance of dark—colored layers of garnet-rich gneiss and cordierite-bearing gneiss. Unit B is composed of fine— to medium-grained light- to medium-gray strongly foliated biotite gneiss containing 10—50 percent mica, with biotite generally predominant over muscovite. Two major varieties, commonly as alternating layers a few centimetres to 15 m thick, are biotite-quartz- plagioclase gneiss and sillimanitic biotite-quartz gneiss. Some of the rocks are microcline-bearing and some are sparsely garnetiferous. Lenses of calc-silicate rock occur locally. Thin conformable seams of granitic material are abundant locally but are usually erratically distributed in this map unit. Granitic material, however, forms 25 to 50 percent of an especially migmatitic variety of biotite gneiss in the northeastern part of the area. INTERLAYERED FELDSPAR-RICH GNEISS, HORNBLENDE GNEISS, AND AMPHIBOLITE (UNIT C) Unit C is composed of feldspar-rich gneiss, hornblende gneiss, and amphibolite similar to those of units D and E, with which unit C is intercalated. In unit C these rocks are interlayered in approximately equal thicknesses on a scale of several centimetres to 10 m or more; locally one or another rock type predominates. Interlayers of medium- to dark-gray biotite-quartz—plagioclase gneiss are locally abundant. Layers and lenses of the relatively dark-colored hornblende gneiss, amphibolite, and biotite gneiss are common in the feldspar-rich gneiss (unit E), but even G3 in unit E they form less than 25 percent of the total rock exposed. HORNBLENDE GNEISS AND AMPHIBOLITE (UNIT D) The rocks of Unit D are commonly interlayered on a scale ranging from 30 cm to 30 m. Hornblende gneiss and amphibolite are dominant, but interlayers of calc-silicate gneiss, impure marble, and quartz gneiss are common in the southern half of the area. Cordierite-bearing biotite gneiss is common as interlayers in’ unit D shown terminating westward against unit B (fig. 2). Feldspar-rich gneiss, especially the medium-gray plagioclase-rich variety, is also present in parts of unit D. Fine- to medium-grained medium- to dark-gray or greenish-gray hornblende gneiss contains, in addition to hornblende and plagioclase, 15 percent or more of other minerals—clinopyroxene, quartz, biotite, or epidote-group minerals. Compositional layers range in thickness from 6 mm to a metre or more; parallel planar alinement of hornblende defines the foliation. Fine- to medium-grained dark—green to black amphibolite, consisting predominantly of hornblende and plagioclase, is unlayered but variably foliated. Fine- to coarse-grained calc-silicate gneiss dis- plays a well—developed compositional layering on a scale of a millimetre to a metre. Colors of the calc-silicate gneiss are dark gray, light green, yellowish green, white, gray, pinkish, or black, depending upon the various proportions of plagio- clase, hornblende,epidote, clinopyroxene, microcline, quartz, sphene, garnet, calcite, vesuvianite, scapo- lite, cummingtonite, tremolite, and magnetite-ilmen- ite. The coarse-grained variety of calc-silicate gneiss is common as pods or lenses rich in garnet, epidote, and quartz. Lenses of medium-grained impure marble are abundant locally and contain garnet, epidote, clinopyroxene, and other minerals of the calc-silicate gneiss assemblages in addition to calcite. Lenses and layers of fine-grained quartz gneiss are common near contacts with other rock units, range in thickness from 0.3 to 6 m and contain variable amounts of magnetite-ilmenite and garnet. Fine- to medium—grained biotite gneiss contains cordierite, variable amounts of sillimanite, and pale- brown biotite. FELDSPAR-RICH GNEISS (UNIT E) Unit E is composed of fine- to medium-grained light- to medium-gray, tan, or pinkish-gray feldspar—rich gneiss and is characterized by the predominance of feldspar and quartz over biotite. A closely spaced foliation is formed by planar alinement of biotite, which commonly forms less than 10 percent of the rock G4 GEOLOGY AND RESOURCES 0F TITANIUM // r 'll 12 3 4 5 MILES | | | l 5 KILOMETRES N w 5_ FIGURE 2.—Generalized geologic map showing distribution of though locally as much as 15 percent. Light-colored A19-A24; Sims and Gable, 1967, p. E8-E11; Sheridan, granitic-appearing varieties of the rock contain as Maxwell, and Albee, 1967, p. 16-17). much as 50 percent microcline; somewhat darker plagioclase-rich varieties contain only traces of BUTEE‘BEAMNG GNEISSES microcline. Conformable dark layers and lenses of Several conformable layers and lenses of rutile- hornblende gneiss, amphibolite, or biotite-quartz- bearing gneisses are shown in figure 2. The plagioclase gneiss, 13 mm to a metre or more thick, rutile-bearing gneisses are most commonly associated are locally abundant but generally are not continuous with rocks of units A and D, and locally with unit C; for more than 100 m. Much of the feldspar-rich gneiss some are along contacts between the main rock units. is similartorocks mappedas microcline-quartz-plagio- The longest and most continuous layers are in the clase-biotite gneiss in nearby areas (Moench, 1964, p. western half of the area (fig. 2). The southwestern- RUTILE IN GNEISS AND RELATED ROCKS, FRONT RANGE, COLORADO G5 I 5": ' 105°22'30" , ° ° 39°45 EXPLANATION U U _ V - o O a N M 3.. Sedimentary rocks a < 3‘ Unconl'ormity E 5 ‘ \ >- lntruded about I05” 1 5' Leucosyenite 3 1,040 my. ago 2 Related to Pikes Peak Granite E an \ “.3; 1 > 5 \ . :1 j, v 3", S Intruded about Silver Plume 2" F 1: 7r g 1.440 my. 390 3 Granite VX" “- , + o Xbco 39°37'30” 1» o o ~ 4 Boulder Creek Granite 1 2 3 4 5 MILES and other plutonie rocks 0 1 2 3 4 SKILOMETRES Geology simplified from the following sources: 1. Geology of the Squaw Pass quadrangle Gneissic hornblende quartz monzonite Amphibolite Probably intruded prior to 1,710 my. ago Intruded about 1,710 my. ego \ 4; G; Mapped by ‘ _ _ > E :5 ‘ ~ L4 S Douglas M. Sheridan. 1961—70, 1972, unpublished information Rutilebearing light-colored gneisses 163‘. it! M Sherman P. Marsh, 1968—69, 1972, unpublished information \, f” - z x» ‘ 2. Sheridan. Reed, and Bryant (1972) ' ‘ ‘ 3. Gable (1968, Plate 1) ~ 1, ‘ L 4. Bryant and others (1973) up.“ E W N Feldspar-rich gneiss _ >< Unit D g Hornblende gneis: and amphibo/ite in terlayered E with Variable amounts of Cele-silicate gneiss, 5 ' Complexly cordiarite-beeriny biotite gneiss, impure mar» g interlayered 8m" ble, and quartz gneiss o. Intergradatlnnal ' units-no I,” stratigraphic A order implied. Unit C Deposited prior lnterlayared feldspar—rich gneiss, hornblende to 1.710 my. 390 gneiss, and amphiba/itc, with locally abun- dant bio rite yneiss Biatite yneiss 7// /A Unit A Sillimenitic biotite gneis: containing garnet- beariny layers and cordierite-baaring layers 4 J Contact Fault Precambrian rutile~bearing gneiss, east-central Front Range, Colo. most layer ranges in thickness from 30 cm to 21 m and has an outcrop length of about 13 km. About 0.15 km north of the eastern end of this layer is a rutile- bearing lens 3.5 km long and as much as 150 m thick; this lens is actually composed of numerous smaller lenses of rutile-bearing gneiss, 6 to 30 m thick, separated by variable amounts of sillimanite-rich biotite gneiss. In the area between the southwestern- most layer and Interstate Highway 70 are three other lenses of rutile-bearing gneiss, ranging in length from 2 to 3 km and in thickness from 30 cm to 9 m. Other Dashed where inferred . 9 . F Sample localities Numbers refer to sample numbers in tables I and 2‘ Letters designate localities from which two or more samples {tables I and 2l were taken: A-1, 11. 12 E-Z, 17 03, 16. 20 D—4. 13, 14. 21 E-22, 26. 27 F—23, 24 6—28, 30. 31 H—34, 35, 36, 37 lenses shown on figure 2 are short; they range in length from 60 to 780 m and in thickness from 15 cm to 30 m. The rutile-bearing gneisses are generally lighter in color than the adjacent rocks. Whereas this characteristic is a useful megascopic feature when doing field studies in this region, it is by no means infallible. Other light-colored Precambrian rocks which occur similarly as thin layers and lenses but which are generally devoid of rutile are numerous varieties of feldspar-rich gneiss, calc-silicate gneiss, G6 cordierite-bearing biotite gneiss, and quartz gneiss as well as pegmatite and aplite. A hand lens is nearly always necessary to check for the presence or absence of tiny splendent grains of rutile, which most commonly is red but which also can be black, orange, or pale orange-yellow; the rutile grains are generally anhedral and less than 0.5 mm in size. Sillimanite is commonly present in noteworthy amounts in several varieties of rutile-bearing gneiss. Although some of these gneisses resemble other layered quartz gneisses which are common in this region, they lack the magnetite, ilmenite, or garnet which usually occur in the local quartz gneisses. Where biotite is present, it commonly is pale brown under the hand lens in contrast to its black color in most biotite gneisses and feldspar—rich gneisses of the region. Generally, the biotite content of rutile-bearing gneiss is less than that of the more common biotite gneisses of the region. The rutile-bearing gneisses are generally fine- to medium-grained; some appear coarse grained owing to the presence of large crystals of prismatic sillimanite and, more rarely, corundum in a fine- to medium-grained matrix. Some of the lenses are composed of one variety of gneiss but most layers and lenses are composed of several interlayered and intergradational varieties. Most common among the varieties are biotite-quartz-plagioclase gneiss, silli- manite-quartz gneiss, and rocks gradational between these two. In these rocks rutile commonly occurs in amounts of 0.5 percent or less but locally as much as 2 percent. Rutile-bearing sillimanitic topaz—quartz gneiss is the principal variety in the eastern part of the longest rutile-bearing layer (from sample locality A to D, fig. 2), in short lenses at and near sample locality 33, and in two partially exposed layers or lenses at sample locality H. The rutile content of the sillimanitic topaz-quartz gneiss is commonly in the range 1 to 2.5 percent and locally is about 5 percent. Rutile—bearing biotite-quartzplagioclase gneiss is a light-gray moderately well-foliated rock that is rich in plagioclase (commonly 45 to 55 percent). The plagioclase is albite—oligoclase rather than the oligoclase-andesine usually found in other biotite gneisses and feldspar-rich gneisses of the region. Biotite, most often a pale-brown phlogopitic variety, generally comprises less than 15 percent of the gneiss. Locally the gneiss grades to a darker gray color, corresponding to the presence of some layers containing black biotite. Andalusite, in blocky grains of 1 mm or less, is present locally as a rare accessory mineral. As viewed in thin section (fig. 3) subparallel grains of biotite define the foliation in rutile-bearing biotite-quartz-plagioclase gneiss; elongate grains of rutile are alined parallel to the foliation. GEOLOGY AND RESOURCES OF TITANIUM Rutile-bearing sillimanite-quartz gneiss is a White to creamy-white rock consisting dominantly of quartz and typically 20 to 30 percent sillimanite. The rock might also be called a sillimanitic quartzite because feldspar is rarely present in noteworthy amounts except in rock gradational to biotite gneiss. Muscovite is more common than biotite but the total mica content is generally less than 2 percent. The sillimanite is in fibrolitic needles that form interlensing laminae between thin granular quartz septa, giving the rock a thinly layered and moderately well-foliated appear- ance. As viewed in thin section (fig. 4), elongate grains of rutile are alined parallel to the foliation defined by the fibrolitic sillimanite. Rutile-bearing gneisses gradational in composition between sillimanite—quartz gneiss and biotite-quartz- plagioclase gneiss are light gray to gray and moderately to well foliated. Although the principal minerals occur in variable amounts, most commonly quartz predominates greatly over plagioclase, sillimanite ranges from 5 to 20 percent, and biotite is' 15 percent or less. Rutile-bearing sillimanitic topaz-quartz gneiss is a white fine- to medium-grained rock composed chiefly of quartz and topaz and lesser amounts of sillimanite, rutile, and apatite; muscovite and zircon are accessory minerals. Medium-grained white or glassy quartz layers 2 mm to 4 cm thick alternate with discontinuous thinner layers of finer grained creamy granular topaz—rutile rock and with radiating coarse-grained aggregates of prismatic sillimanite. The rock is similar in appearance to a layered sillimanitic quartzite and is not readily distinguished except by its greater heft, owing to the high specific gravity of topaz (213.5). Rutile commonly is concentrated in thin layers colored pale reddish by the tiny splendent rutile crystals. Samples of the gneiss contain rutile in amounts ranging from 0.3 to 5.3 percent (Ti02 determined by X-ray fluorescence) and topaz from 22 to 67 percent (mineral separation by heavy liquid). Fibrolitic sillimanite forms a linear orientation in some sillimanitic quartz laminae. Prismatic crystals of sillimanite, as much as 2 square mm in cross-sectional area and 3 cm in length, are in irregular to radiating aggregates flattened parallel to the layering but they lack linear orientation. Locally the gneiss contains thin laminae as much as 5 mm thick composed of a green crhomium-bearing muscovite. Viewed microscop- ically, fibrolitic sillimanite is partly replaced by topaz and is partly recrystallized to large grains of prismatic sillimanite (fig. 5); rutile is molded around topaz grains (fig. 6) and is intergrown with prismatic sillimanite (fig. 7). Thin seams of kaolinite were seen in thin sections of gneiss obtained from sample locality RUTILE IN GNEISS AND RELATED ROCKS, FRONT RANGE, COLORADO FIGURE 3.—Photomicrograph of elongate grains of rutile (R) alined parallel to foliation defined by planar alinement of biotite (B) in biotite-quartz-plagioclase gneiss from sample locality F (fig. 2). Other minerals are plagioclase (P), quartz (Q), and apatite (ap). Plane-polarized light. H and from a small lens 1 km east-southeast of sample locality 33 (fig. 2). A local variant of the more common types of rutile- bearing gneiss is a corundum-bearing sillimanite- plagioclase gneiss which occurs as lenses in rutile- bearing biotite-quartzplagioclase gneiss at sample locality G and also 0.3 km northwest of sample locality 29 (fig. 2). The sillimanite-plagioclase gneiss is a white or light-gray fine- to coarse-grained rock that is poorly 4.—Photomicrograph of elongate grains of rutile (R) alined parallel to foliation defined by alinement of fibrolitic silli- manite (fs) in sillimanite—quartz gneiss from sample locality 25 (fig. 2). Other minerals are quartz (Q) and biotite (B). Plane- polarized light. F IGURI: FIGURE 5.—Photomicrograph of fibrolitic sillimanite (f5) partly replaced by topaz (T) and partly recrystallized to prismatic silli- manite (S) in sillimanitic topaz-quartz gneiss from sample locality D (fig. 2). Other minerals are quartz (Q) and the rutile (R). Plane- polarized light. foliated and texturally complex. Grains of pinkish to gray corundum, as long as 2.3 cm, and aggregates of prismatic sillimanite, as much as 12 cm long and 1 cm thick, are diversely oriented in a fine- to medium-grained matrix rich in plagioclase (albite, An6-8) and fibrolitic sillimanite. Rutile forms 1 to 2 percent of the gneiss (mineral separations and thin- section modal analyses, 1,000 points each). At locality G the corundum-bearing gneiss contains wagnerite, a rare magnesium fluophosphate (Sheridan, Marsh and others, 1971). Other constituents of the gneiss are muscovite, apatite, monazite, zircon, tourmaline, and FIGURE 6.—Photomicrograph of rutile (R) associated with topaz (T)in topaz-quartz gneiss from sample locality B (fig. 2). Other mineral is quartz (Q). Plane-polarized light. G8 GEOLOGY AND RESOURCES 0F TITANIUM 7.—Photomicrograph of prismatic sillimanite (S) inter- grown with rutile (R) in sillimanite-quartz gneiss that occurs as a layer in sillimanitic topaz-quartz gneiss at sample locality D (fig. 2). Other minerals are quartz (0), and biotite (B). Plane- polarized light. FIGURE locally, chlorite. The tourmaline is in small yellowish-brown to pale-greenish-blue grains having refractive indices that suggest the mineral to be dravite. Foliation in the sillimanite-plagioclase gneiss is commonly obscure in the outcrops, owing to the diversely oriented large grains of corundum and aggregates of prismatic sillimanite. The foliation, however, is well shown in some thin sections and is defined by elongate grains of plagioclase, aggregates of fibrolitic sillimanite, and oriented flakes of pale-brown biotite. Tiny anhedral grains of rutile are alined in rows parallel to this foliation. Larger grains of rutile are subhedral and are associated with corundum (fig. 8). Layered hornblende gneiss associated in many areas with the light-colored rutile-bearing gneisses typically contain sphene and magnetite—ilmenite but no rutile. One exception to this occurs at locality G, where the hornblende gneiss adjacent to rutile- bearing light-colored gneiss also contains rutile in addition to sphene and magnetite-ilmenite. CHEMICAL COMPOSITION Chemical and spectrographic analyses of four samples of rutile-bearing sillimanitic topaz-quartz gneiss and of six samples representing other lithologic types of Precambrian metamorphic rock are shown in table 1. The localities from which the samples were obtained are indicated in figure 2. The analyses of rutile-bearing sillimanitic topaz— quartz gneiss show this to be a rather unusual rock, FIGURE 8.—Photomicrograph of subhedral grains of rutile (R) associated with blocky grain of corundum (C) in sillimanite- plagioclase gneiss from sample locality G (fig. 2). Other minerals are plagioclase (P), fibrolitic sillimanite (f5), biotite (B), monazite (M), and wagnerite (W). Plane-polarized light. for it is rich in silica (63-76 percent), alumina (18-27 percent), and fluorine (4-7 .5 percent) and contains 1.47-2.76 percent titanium oxide. The rock contains little iron, alkali metals, and alkaline earths, reflecting the fact that, unlike many other quartz-rich rocks of the Front Range, it contains no magnetite, ilmentite, or garnet and very little biotite and feldspar. Small amounts of CaO and P205 are present as accessory apatite, and small amounts of zirconium are present as accessory zircon. Small amounts of rare earths may be in monazite. Although monazite was not identified in thin sections of topaz-quartz gneiss, it has been found as an accessory mineral in other types of rutile-bearing gneiss. Small amounts of chromium are in green chromium-bearing muscovite found as thin laminae in some of the sillimanitic topaz-quartz gneiss. A sample of material rich in this green mica from sample locality H (fig. 2) contains 0.2 percent Cr203, as determined by J. S. Wahlberg, US. Geological Survey, by X—ray fluorescence. The petrography of the rutile-bearing sillimanite- quartz gneiss indicates high silica and alumina, accessory titania, and very little of any other constituents. The rutile-bearing biotite-quartz-plagio- clase gneiss is somewhat richer in Na and poorer in Ca than other biotite gneisses of the east-central Front Range. Also, it probably is richer in Mg and poorer in Fe than most other biotite gneisses. Corundum- bearing sillimanite-plagioclase gneiss is an alumina- rich rock also characterized by high Na and Mg relative to Ca and Fe. RUTILE IN GNEISS AND RELATED ROCKS, FRONT RANGE, COLORADO G9 TABLE 1.—Ana1ytical data for rattle-bearing gneiss and for other Precaman metamorphic rocks, east-central Front Range, Colorado Rock type. . . . Rutile-bearing sillimanitic topaz-quartz g'neiss1 Other Precambrian metamorphic rocks 2 Sample No . 1(A) 2(B) 3(C) 4(D) 5 6 7 8 9 10 Laboratory serial no . . . D102280 D102281 D102282 D102283 D127511W D127512W D127516W D127517W D127518W D127519W Field no. . . . R-22A R-23A R-24A R-25A S-C-6 S-C-7 S-C-11 S-C-13 S-C-14 S-C-15a Chemical analyses (weight percent) [Analyses of sample 1-4 are standard rock analyses by Elaine L. Manson. Analyses of samples 5-10, except for fluorine determinations. are rapid-rock analyses by Paul Elmore, Samuel Botts. Gillison Chloe. John Glenn, Lowell Artis, and Hezekiah Smith. Fluorine determinations in samples 5—10 are by Johnnie Gardner and Adolph Haubert using specific ion electrode method. All analysts are affiliated with the U.S. Geological Survey. Dashed line in table indicates no data] SiOg ....... 67.36 76.02 73.31 78.6 78.3 74.0 44.3 44.2 48.1 A1203 ...... 23.57 18.01 20.65 26.99 13.9 12.8 13.1 13.7 17.8 15.0 Fe203 ..... .01 .13 .09 .22 .29 .68 1.0 4.8 2.4 2.0 FeO ....... .10 .08 .09 .12 1.1 3.2 1.6 2.4 21.8 8.2 M30 ....... .01 .01 .04 .11 2.7 1.5 .55 2.0 8.9 4.1 CaO ....... .28 .31 .31 1.10 .39 .09 1.8 21.0 .31 14.8 Na20 ...... .02 .07 .05 .11 .97 .13 3.7 1.5 .76 2.9 K20 ....... .03 .05 .12 .18 .80 1.6 3.1 2.8 .33 .85 H20+ ..... .41 .61 .61 .59 .85 .80 .42 .33 1.2 .68 H20 — ..... .04 .08 .05 .05 .15 .05 .05 .10 .10 .07 TiOg ....... 2.76 1.47 1.47 2.71 .05 .49 .21 1.9 1.5 1.8 P20 5 ...... .23 .26 .23 .87 .00 .00 .06 .82 .12 .35 MnO ....... .01 .02 .02 .02 .00 .00 .07 .28 .43 .13 C02 ....... .02 .01 .01 .01 <.05 <.05 .12 4.0 <.05 .92 C1 ......... 00 00 00 ()0 .................................... F .......... 7.50 4.00 4.35 5.80 .07 .15 .07 .08 .12 .09 Subtotal 102.35 101.13 101.40 102.25 Less 0 . . 3.16 1.68 1.83 2.44 Total . . . 99.19 99.45 99.57 99.81 100 100 100 100 100 100 Semiqmntltntive apectrograpbic analyses (parts per million)3 [Analyses of samples 1—4 by R. H. Heidel. U.S. Geological Survey. Analyses of samples 5-10 by G. W. Sears, Jr.. U.S. Geological Survey] Ba 7 7 70 150 150 300 500 300 30 50 Be N N N N 2 N 1 N N N Ce N N N 300 N N N N N N Co N N N N N 10 N 20 15 50 Cr 70 15 7 10 1 30 1 150 30 150 Cu 70 3 3 7 <1 50 N 5 1 30 Ga N 5 5 5 20 30 20 20 20 20 La N N N 150 30 50 N 30 N N Nb L L 10 L N 10 10 10 N N Nd N N N 200 N N N N . . . . . . . . Ni N N N N N 10 N 50 15 30 Pb L N N N 15 N 15 10 N N Sc 10 5 15 30 N 15 7 20 30 30 Sr 7 N 7 15 70 7 70 300 5 300 V 70 30 30 70 N 70 N 150 200 300 Y 30 10 30 70 20 20 70 30 15 30 Yb 1 2 3 7 3 3 7 <3 <3 <3 Zn N N N N N N N N 300 N Zr 300 150 150 700 70 150 200 150 20 70 1A1] sam les are s lits of bulk samples, originally about 50 pounds (22.7 k ) in the table each. ach of t e bulk samples was obtained by taking a composite samp e across the exposed width of the rutile-bearing gneiss. Localities are shown in figure 2. These localities, indicated in the table by letter symbols, (A) through (D) following the sample numbers designate localities from which two or more samples were taken and are shown in gigure 2 by the letter symbols only. All sam les are grab samples. Localities are shown in figure 2. The lithology represente by each of these samples is as follows: Sample 5, Biotite-quartz-plagioclase eiss containing cordierite and sillimanite; 6, Sillimanitic biotite-quartz gneiss; 7. icrocline-quartz-plagioclase-biotite gneiss; 8, Cale-silicate gneiss; 9, Gedrite- cordierite-garnet gneiss; 10, Hornblende gneiss. Analyses of six other Precambrian metamorphic rocks are given in table 1 to illustrate the 3Minor elements were determined by semiquantitative spectrographic methods described by Myers and others (1961). Results of the spectrogragggc analyses are to be identified with geometric intervals having the boundaries 1200, , 560, 380, 260, 180, 120, etc.. in parts per million but are reported arbitrarily in the table b approximate geometric midpoints such as 1000, 700. 500, 300, 200, 150. 100, etc. l’recision of a reported value is approximately plus-or-minus one interval at 68-percent confidence. or plus-or-minus two intervals at 95-ercent confidence. N, not detected at limit of detection; L, detected but below limit of determination; . ., not looked for. composition of some of the other lithologies in the area of rutile-bearing gneisses. Samples 5, 6, and 7 represent, respectively, cordierite-bearing sillimanitic biotite gneiss, sillimanitic biotite gneiss, and feldspar- G10 rich gneiss; they contain only 0.05 to 0.49 percent Ti02. Samples 8, 9, and 10 represent, respectively, calcsilicate gneiss, gedrite—cordierite-garnet gneiss, and hornblende gneiss; they contain 1.5 to 1.9 percent TiO 2, amounts that are comparable to Ti02 in some rutile-bearing sillimanitic topaz—quartz gneiss. Except for a rare occurrence of rutile in hornblende gneiss at locality G, no rutile has been observed in these rocks. Petrographic studies of hornblende gneiss and calc-silicate gneiss indicate that sphene and magnetite-ilmenite are common accessory minerals and probably account for much of the Ti02 reported. Opaque iron oxides in gedrite-cordierite-garnet gneiss may be ilmenite or magnetite—ilmenite. Some of the titanium in these rocks probably is in biotite, amphibole, and other minerals but detailed analytical studies of minerals have not been made. The fluorine content of samples 5 through 10 ranges from 0.07 to 0.15 percent, or 700 to 1,500 ppm. This range falls within the range, 60 to 1,500 ppm, reported by Fleischer and Robinson (1963, p. 63) but is higher than the average, 380 ppm, of 69 analyses of metamorphic rocks compiled from the literature. GEOCHEMISTRY 0F RUTILE Stringent regulations imposed by the Office of Emergency Planning concerning chemical impurities in rutile and the lack of published information on trace elements in rutile prompted a study of rutile specimens from the Evergreen locality and several other localities where similar environments are found. Twenty-nine samples of rutile from the east—central part of the Front Range, and eleven samples from three other United States localities were anlayzed spectrographically and the results are reported in table 2. These analyses show the similarities of trace-element content in rutile samples from similar environments in several areas of the United States and also show chemical impurities. A table of impurities permitted in rutile is used to demonstrate the usability of rutile from high-grade metamorphic environments. Following is a discussion of factors influencing the substitution of these trace elements in the rutile structure. The rutile samples analyzed in this report, with the exception of crystals from Graves Mountain, Ga., were separated from whole rocks. Each Graves Mountain, Ga., sample is from large crystals that were sent to us by Mr. P. Bennett of Aluminum Silicates Inc., Washington, Ga. The rock samples were crushed, ground, washed, and sieved to —65, +150 mesh and then separated into light and heavy fractions by heavy liquids (bromoform). The heavy fraction was separated magnetically, and the GEOLOGY AND RESOURCES OF TITANIUM nonmagnetic fraction was examined for rutile. The heavy nonmagnetic fraction commonly contains kyanite, andalusite, sillimanite, and‘topaz as well as rutile. Rutile grains were carefully hand—picked under a microscope and analyzed spectrographically. Most of the rutile in the whole-rock samples was in small anhedral grains, less than 0.5 mm across, and the color varied from black to pale orange-yellow. A few euhedral doubly terminated stubby prisms were seen, some of which exhibited the typical “knee shaped” rutile twins. Color is not consistent with a given sample and two or more colors are often present. Impurities in the analyzed rutile ranged from iron, found in all samples, to cobalt, found only two. The maximum impurities permitted in rutile purchased by the United States Government are listed in table 3 as stipulated by the Office of Emergency Planning. Although a number of analyzed samples exceeded the limits for one or more of the listed impurities, many of these are due to sample contamination rather than impurities in the rutile. Calcium and magnesium are considered contaminants for two reasons: they do not fit into the rutile structure and they are not present in the single crystals from Graves Mountain, Ga. Zirconium was present in all samples analyzed except those from Graves Mountain. Zirconium can substitute for titanium but most zirconium contamination of the samples is probably due to zircon. Zircon was observed in most of the heavy, nonmagnetic fractions. Some of the samples from the east-central Front Range contained lanthanum. A comparison of samples 1(A) and 4(D) in tables 1 and 2 indicates that the lanthanum is an impurity probably caused by monazite contamination. Small amounts of monazite were observed in the rocks. Antimony, tungsten, colbalt, and bismuth were also found in the analyzed rutile samples, although no standards of purity for these elements are given. Most of the trace elements found in the analyzed samples can substitute into the rutile structure. These substitutions can be influenced by several factors (table 4). One element can substitute extensively for another, in this case titanium, if their atomic radii do not differ by more than 15 percent. If the radii differ by more than 15 percent, but the valence of the substituting ion is within one unit (Ti+4 ), then some substitution may occur (Krauskopf, 1967, p. 585). Substitution may also occur if the ratio of the substituting ion to the ion being substituted for (titanium) is within limits determined by the structure of the mineral involved (Pauling, 1929). This ratio is expressed by R A /R X where RA is the radius of the substituting metal ion and R X is the radius of the coordinating anion. Minerals having the rutile G11 RUTILE IN GNEISS AND RELATED ROCKS, FRONT RANGE, COLORADO 83 88 888 88 88H 88 88 z z z 2 88.8 88.H 8H 88 88 83 88.H 888 88.H 88.8 .8 88 88 88 8 z 8 8 88 2 z 2 8H 88 88 z 8 8H 88 . 8 88 88 . 8 88.8 888 88 888 88 88 88 88 88. 88 8H 8H 888 88 88 83 88 88 H 88 88 88 H B 88 88.H 88.H 88 88 88 88 888.8 888.8 888 88 88 88 88.H 888 88 88 88 88 88 88 > 88 888 8H 88 8H 8H 88 88.H 888 H 888 88 88 88 888 88 8H 8H 88 8H 8 88 =8 83 888 88 88.8 83 88.H 88.8 88 888 888 88 z z z z z . z z z z . z . 88 8...; 88 888 88 8 88 88 888 888 888 88 88.8). 88.8 88.8). 88 88 8 88 88 88H 88 H 88 8). :2 8H 88 88 88 8 88 8 8H 8H 2 2 8H 8H 2 z 2 8H 2 88 8H 2 o: z z z z z z z z z . z z 88 z z z 88 z z z z 2 8.H 8H 88H 88 88 8 8 88 888 88 H 8 8H 8 8 8 88.H 8 8H 2 88 88 88H 5 z z z z z z 2 8H 88 z z z z z z z z z z z oo 88 8 8 z z z z z z z z 2 8H 2 8H 88 z z z z z a 8 8 8 z z z z z z z z z 2 .H z z z z z z z 3 z z z 88 8 8 8 z z z z z z z z z z z z z z m z z 2 H H 8 8 8 2 .H z z z z z z z z z 8< .H .H .H 8H8 88.8 88.8 88.8 .H .H .H .H 88.8 H88 88.8 .H 8H8 88.8 88.8 a 88.8 H88 28:85 88.8 888.8 888.8 8.8 H88 88.8 8.8 H88 888.8 .H 8 88.8 H88 5.8 888.8 88.8 88.8 88.8 .H H88 88.8 28.28“: 8.H 8.8 8.8 8.8 8.H 8.8 8.H 8.H 8.H 8.H 8.8 8.8 8.8 8.8 88.8 8.8 8H8 88 H8 8H8 8.8 28.82 8.8 8.8 H85 2.088 5.888 2.885 88280 8 H 888 888 8.98 8.98 8.0.8 8.0.8 5.2. 8788 878 8.88 8.88 8.88 82 2.8.8 8H. 8H. 8H. 8H. 8H. 8H. 88 HH. 8 88 88 2:88 $888 E8888 $888 88 88 6:8 688 88 8888 M2 @— an 3.58 .3550 EEO gnaw 88.58 .8330 «E588, .aO .ESEEE 328.5 “—8.5 .8915. oEoMMWwEE 55:32 wk 358w 33:32 3.35 «52:3 .350 Ea...“ 233m nguH—cOId—oo .omcam EEK 18.559.88.88» .893 :wwpmhgm ES“ 233m 88.8? 88.8 888.H 88H 88H 88; 2 8H 88.8 8H 8H 88.H 88H 88; 88; 88.8 88.8 88.8 88H .8 8H 88 88 8H 88 888 8 8H 88 8 8 8 8 88 8H 88 88 8 888 8H 888 88 88H. .H 888 88 88 88 88 88 88H 2 2 88.8 2 z z z B 88 888 88 88 8 8 8H 88 83 88 888 88.H 888 888 88 88 8... 8H 88 88.H > 888 888 88 88 8H 888 8H 8 888 88H 8H 888 888 8H 8H 88 888 88 88 :8 z z . z z z z z z z z z z z z z z z z z 88 88.87. 88 H 88 88H 83 88.8 8 888 88 88 88.8 88.8). 88H 88 88.8 88 88 888 88 82 88 8H 2 8H 2 88 2 8H 8H 8H 88 88 88 8 z z 8 8 8 a: z 88H 88 z z z z z 88 z 88H 2 .H 8H 8H 88 .H 888 88.8 3 88 88 z 8 z 88 88 888 8H 88 88 8 888 88 8 8 8H 88 888 .8 z z z z z z z z z z z z z z z z z z z 88 z z z z z 2 8H 2 2 8H 2 z z z z z z z z a z z z z z z z z z 2 H z z z z z z z 2 am 2 z 8 z z z z z z z z z z z z z z z z m z z z z z 2 H 2 H H 2 H z 2 .H z z z z 8< .H 58.8 8.8 8H8.8 88.8 88.8 .H 8 8.81 8.8 88.8 .H 8.8 8 8 8.8). 8.8 8.H 8.H 8.H 28.3 80 888.8 88.8 8.8 H88 8H8.8 8888.8 88.8 888.8 H88 88 88.8 888.8 888.8 88.8 28.8 28.8 88.8 8H8.8 H88 852% 8.8 8.8 8.8 3 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.H 8.8 8.8 8.8 8.8 28.25 3H 8H.mmz HH.mmz 8.882 8.882 8.882 H.mmz 5.8.x 8.88.8 3.x 8H8 H; 8.x 8.x 8.x 8.x 8.x 8.88 88 SH .82 8.28 8888 .888 88 E8 E88 E88 5:. 3.: 5:8 688 8H 8H 88: 8:: 8H 58: 582 2:3 :3: M2 2 85 xoounuggm 5.82 33:» $883.25.... EH82 .280 .owgm 8:95 35:82.88. .89.» nounmhgm ES. 235m .mamnfisa .3 8.388885 93 33:53 .550 .35 £855.? .532 05 .3 N 0.53.8 no 53% can can :32 v.83 3383 8:2: .8 25 5E? Eob 82:188. cannummwc 9855:: 29:88 .8888 3.: smack: 28 £35.? .532 .3 3&3 2...— :o 8.33:8: 35:32 32;. .N 0.5M: =0 :32? 0.5 3.229% E 35:83 dofluifiuwewv 8o «HE: 323 a: .4 4.233% .«0 SE: 88 8.388% .80: . Z duguccov 8533.3 a» £52an 9.5 35:53.31 .8 6288.38.59 33.5 . @8932. no «a 3:35 28 8:55.883; 338.5888; 8 .98 8.35 @3832 a .0 57.395 .3» .88 .88 .SN don .25 .88. .83 ms :28 359826 3.585888% 38:889—858.. .3 2 S 2: 5 8:88.53; 888.8%: 0.3 :5 55:: .5; 383 5.38. .38 63 dew 6% den 63 .comH 80:89:82 2: 9:28: £8235 9.8.5858» 58? 8.28553 3 3 9:8 888.3858 «2 858.88% 2: no 33QO .28: 8.550 “Ea 8.5.82 .3 “89:88.83 £5me Pinfimofiuwam 3553553888 .3 855.53% 953 35:55 .8282 .O .m— .3 808.3858 .55.. =< 533m .4 .< .3 8.x can 9x .v.m 3 Tm 83:83 .8 83.255; 3.38.8 383888 9.588888888888888 88.8838888383288ml.m 8:858. G12 Table 3.—-Standa'rd of purity of rutile [From National stockpile purchase specifications P-49—R5 (National Materials Advisory Board, 1972)] Weight percent Impurity impurity permitted Fe 203 ........................................ 1 Nban ....................................... .4 ZrOz ......................................... 1 V20 5 ........................................ .75 Cr203 ........................................ .75 MgO+Ca0 .................................... .25 Sn ............................................ .1 Mn ........................................... .75 structure have radius ratios (R A /R X) ranging from 0.41 to 0.73 for the octahedrally coordinated cations. In the rutile structure each cation is coordinated to six anions, and each anion forms a ligand to three cations. In the case of rutile, R X (for oxygen with a coordination number of three) equals 1.36A (Shannon and Prewitt, 1969) and RA (for titanium with a coordination number of 6) equals 0.61A (Shannon and Prewitt, 1969). This gives a radius ratio of RA/RX=0.61/1.36=0.45. Other octahedrally coor— dinated cations whose radii fall between the limiting radius ratio values of 0.41 and 0.73 can also substitute into the rutile structure. All but one of the elements listed in table 3 have radius ratios within the limits of the rutile structure and probably substitute for Ti in rutile. A fourth factor than can affect substitution into the rutile structure is that most of the trace elements found in the analyzed rutile belong to the transition elements and are related to titanium. All the trace elements listed in table 3 satisfy one or more of the factors for substitution. Chrome, iron, molybdenum, niobium, and vanadium satisfy all four and all but chrome and molybdenum are found in every sample analyzed (table 2). These five elements also make up most of the trace elements found. A second group of elements found in trace amounts in rutile satisfy three of the four listed factors: tin, tungsten, yttrium, and zirconium. Elements satisfying Table 4.—Iom‘c subsitution factors [Elements are listed in order of decreasing likelihood of substitution] Valence Ionic radius Radius ratio Transition Element (n+4) i 15 percent = 0.52 to 0.70 RA/RX element Titanium Ti +4 0.61 0.45 Yes Iron Fe +3 .64 .47 Yes Niobium Nb +3 .70 .51 Yes Vanadium V +4 .59 .43 Yes Chrome Cr +3 .62 .45 Yes Molyb- denum2 Mo +4 .64 .48 Yes Tin Sn +4 .69 .51 No Tungsten W +4 .58 .43 Yes Yttrium Y +3 .89 .66 Yes Zirconium Y +4 .72 .53 Yes Antimony Sb +3 .77 .57 No Cobalt Co +2 .73 .54 Yes Bismuth Bi +3 1.02 .75 No GEOLOGY AND RESOURCES OF TITANIUM only two of the factors are much less common and are unevenly spread throughout the samples. Cobalt occurs in only two samples, whereas antimony occurs in the rutile from all localities except the east-central Front Range. Bismuth satisfies only the valence factor. ORIGIN OF METAMORPHIC ROCKS We believe that the thick section of metamorphic rocks in the east-central Front Range was originally an assemblage of interlayered shales and graywackes, and local carbonate layers, associated with abundant mafic and felsic volcanic rocks which probably accumulated near their source. Although the origin of the thin lenses and layers of rutile-bearing gneiss is debatable, we believe that most likely they were originally the products of intense subaerial weathering that occurred during the accumulation of the thick succession of sedimentary and volcanic rocks. MAJOR LITHOLOGIC GROUPS The numerous varieties of metamorphic rock in the five principal map units shown in figure 2 comprise three major lithologic groups: biotite gneiss (units A and B), hornblende gneiss and amphibolite (units C and D), and feldspar-rich gneiss (units C and E). Biotite gneisses and schists in the Front Range were considered to be metamorphosed shaly sediments (Ball, in Spurr, Garrey, and Ball, 1908, p. 44) and subsequent authors have reached the same conclusions. Features that correspond to many present-day successions of sedimentary rock are the wide areal extent of these rocks, the lithologic variation across layers, close interlayering, the appar- ently conformable contacts, and the composition. Relict sedimentary features are present in some areas, for conglomeratic lenses were noted in mica schist (a lithologic unit traceable laterally to biotite gneiss of this report) in the Ralston Buttes quadrangle (Sheridan, Maxwell, and Albee, 1967, p. 10), and conglomeratic quartzite is interlayered with quartzite and schist in the Coal Creek area (Wells and others, 1964, p. 7-8). In accordance with all these investigators, we interpret the biotite gneiss of the report area as metasedimentary. The sillimanitic biotite-quartz gneiss was an alumina-rich shale and the intimately interlayered biotite-quartz—plagioclase gneiss was a sandy shale or graywacke. The sillimanitic biotite gneiss containing cordierite- and garnet-bearing layers was an alumina-rich shale having some layers relatively rich in magnesium and iron. Gable and Sims (1969, p. 57-59) have also RUTILE IN GNEISS AND RELATED ROCKS, FRONT RANGE. COLORADO ascribed a metasedimentary origin to similar cordierite-bearing rocks in the Front Range. Hornblende gneiss and amphibolite very likely are metamorphosed intermediate to basic tuffs and flows, and the commonly associated calc-silicate gneiss and impure marble were once sedimentary rocks ranging from calcareous shale to argillaceous limestone. Quartz gneiss, commonly containing magnetite and garnet, is a metamorphic rock equivalent to one type of iron formation. The feldspar-rich gneisses are thought to have been felsic volcanics. Hedge (1969, p. 12, 120-123) noted that their chemical composition favors an origin as acid volcanic rocks similar to some found in island arc association with basalts. Some authors (Moench and others, 1962, p. 38; Sheridan, Maxwell, and Albee, 1967, p. 23) have postulated a metasedimentary origin. Sedimentary features and a rock of volcanic composition are not necessarily incompatible; parts of the feldspar-rich gneiss are probably metavolcanics, but other parts were formed by metamorphism of sedimentary tuffs and volcaniclastic rocks. RUTILE-BEARING GNEISS We believe that the light-colored rutile-bearing gneisses of the east-central Front Range originated by intense weathering of intermediate to basic tuffs and flows. This resulted in thin layers of bentonitic clays enriched in titania (leucoxene), alumina (clay). and silica (clay and quartz). These weathering products were reworked by surface waters, and fluorine, emanating from volcanic sources, was adsorbed local- ly. Clays produced by weathering are commonly low in Ca and high in A1 and Ti (Rankama and Sahama, 1950, p. 209, 563). Rankama and Sahama (1950, p. 764) stated that fluorine is strongly adsorbed in soils, bentonite, and bottom muds and that fluorine is added to the “exogenic cycle by volcanic processes.” Serdyuchenko (1968) believes that many schists and gneisses containing sillimanite, kyanite, or corundum in numerous occurrences around the earth are metamorphosed kaolinite-bauxite weathering crusts. Dunn (1929, p. 248) thought that some of the kyanite-quartz rock in India was originally a surface decomposition of basalt to bauxitic clay. Espenshade and Potter (1960, p. 2425) concluded that kyanite quartzite in the Farmville district, Virginia, and kyanite quartzite and sillimanite quartzite in the Kings Mountain district, North Carolina-South Carolina, originated from sediments composed of quartz sand containing clay or bauxite. In the east-central Front Range the intergrading and interlensing of several varieties of rutile-bearing gneiss and their variability in rutile content, from G13 trace amounts to over 5 percent, are best explained by reworking of the original weathered materials. We believe that fluorine from volcanic emanations or from fluorine-rich volcanic rocks was introduced locally into these weathered materials during or shortly after reworking. In a preliminary report (Sheridan, Taylor, and Marsh, 1968, p. 6), we postulated that rutile-bearing sillimanitic topaz-quartz gneiss was formed by the action of fluorine metasomatism on a gneissic progenitor, such as a sillimanitic quartz-rich gneiss. Such metasomatism seemed necessary to explain the high fluorine content of the gneiss and to account for textural relations indicating that topaz, prismatic sillimanite, and rutile formed at the expense of a previously formed regional metamorphic suite (see figs. 5, 6, and 7 of this report). However, as R. H. Moench (oral commun., 1973) observed, the textural relations do not demand a separate later metamorphic event, because the principal regional metamorphism may have continued long enough after the two early stages of folding to cause the observed textural features. The relations suggest redistribution of primary amounts of fluorine within the affected layers during regional metamorphism, not a metasomatic event related to some later plutonic episode. OTHER LOCALITIES OF ROCK TYPES SIMILAR TO THOSE IN THE EAST-CENTRAL FRONT RANGE Many terranes in the United States and throughout the world contain rock types similar to the Evergreen occurrence. Rutile is reported from many of these occurrences and, although they are low-grade deposits (1-4 percent rutile content), they form a potential resource. The rocks are reported as metasediments, metavolcanics, or metamorphosed weathering crusts, the metamorphism being region-a1 or contact. Although the age of the metamorphosed deposits ranges from Precambrian to Mesozoic, most of reported deposits are in Precambrian rocks. The mineralogy of these rocks is remarkably similar — chiefly quartz, aluminum silicate (andalusite, kyanite, and sillimanite) and a variety of accessory minerals. Rutile is present as disseminated fine grains and as scattered larger crystals. Topaz is common and is locally abundant, as in localities described in this report. Phosphate minerals such as apatite, monazite, lazulite, and wagnerite are commonly found. Other accessory minerals found but not universally present are pyrophyllite, ilmenite, magnetite, tourmaline, pyrite, chrome—bearing mica, and zircon. Pegmatites are common in these areas but are younger and do not G14 display the distinctive mineralogy of the rutile- bearing units. Several of the known occurrences of rutile-bearing rocks are described below. FARMVILLE DISTRICT, VIRGINIA In the Farmville district of central Virginia (Espenshade and Potter, 1960, p. 34-53), layers of rutile-bearing kyanite quartzite in Precambrian rocks occur in a northeast-trending belt about 8 km wide and 50 km long. Individual layers range in thickness from a few metres to as much as several hundred metres where repeated by folding. A percent or less rutile is disseminated in the kyanite quartzite as small grains 1 mm across or smaller. As much as 1 percent of topaz is present locally. Principal associated rock types are hornblende gneiss and biotite gneiss, both containing small amounts of rutile. The chemical nature of these units indicates that they probably were originally graywackes and clay-bearing sandstones interspersed with basaltic or andesitic lava flows. The Farmville district accounts for 50 percent of the known kyanite resources in the southeast United States and, at 0.5 to 1 percent rutile, represents a potential titanium resource. KINGS MOUNTAIN DISTRICT, NORTH CAROLINA-SOUTH CAROLINA In the Kings Mountain district located on the border between North Carolina and South Carolina about 40 km west of Charlotte, NC. (Espenshade and Potter, 1960, p. 64-94), rutile occurs as a common accessory in kyanite quartzite and sillimanite quartzite. These units form the upper part of a complexly folded series of schists and gneisses which are dominantly biotite schist and gneiss and hornblende gneiss. These schists and gneisses have been derived principally from volcanic rocks. In the kyanite quartzite the rutile is evenly disseminated as small grains and crystals. No topaz has been reported from this unit but lazulite and tourmaline are rare accessories. The kyanite quartzite layers range in thickness from 6 to 10 m but appear thicker where repeated by folding. Rutile is an abundant accessory in the sillimanite quartzite. This unit ranges in thickness from 3 to 9 m and where undeformed extends along strike as much as 900 m. Topaz is a rare accessory and lazulite is locally abundant. The sillimanite quartzite may represent a later more intense metamorphism of kyanite quartzite. Both units are sedimentary in origin; probably they were clayey sandstones. The Crowders Mountain-Henry Knob area and the Reese Mountain-Clubb Mountain area are the most important in the district and they account for 40 percent of the known kyanite resources in the GEOLOGY AND RESOURCES OF TITANIUM southeast United States and represent a potential rutile resource. GRAVES MOUNTAIN, GEORGIA Some of the finest rutile specimens in the world come from northeast Georgia at Graves Mountain (Hurst, 1959). The rutile occurs in sericite-kyanite- quartz rock and is most abundant in the coarse kyanite-quartz facies cropping out along the top of the mountain. The rutile occurs as small individual disseminated grains and occasional larger crystals and constitutes 0.5 to 1 percent of the rock. Topaz has not been reported from the kyanite-quartz rock, but pyrophyllite and lazulite are common accessories. Some zones contain as much as 15 percent lazulite. The kyanite-quartz rock crops out in large lenticular bodies as much as 120 m wide and 550 m long and is associated with quartz sericite schist and quartz conglomerate. The Graves Mountain area is underlain by a thick sequence of metamorphosed volcanic and sedimentary rocks. The kyanite quartz rock was probably volcanic in origin, possibly a tuff and it belongs to the Little River Series of Crickmay (1952) which may be of Paleozoic age. The unit was folded and regionally metamorphosed to quartz-sericite schists early in its history. A period of fracturing followed the metamorphism and kyanite, lazulite, and rutile were deposited. A late-stage period of cross fractures developed with attendant deposition of pyrophyllite. Most of the rutile is found in a zone 30 m wide and 150 m long and represents a small but definable resource. WHITE MOUNTAIN, CALIFORNIA A commercial deposit of andalusite in Precambrian rocks of the Inyo Range near White Mountain, Calif., contains rutile (Kerr, 1932). Rutile also occurs in the surrounding quartz-sericite schist and quartz—mica- tourmaline schist. These units are thought to be a series of trachytic flows with an interlayering of some highly aluminous material (volcanic or sedimentary). All the metamorphic units were formed by contact metamorphisms by a quartz monzonite porphyry intrusive. A second stage of hydrothermal metamorphism acted on the area during the late stages of the intrusive producing the late mineralization of pyrophyllite and sericite and altering the quartz-sericite schist in many places to diaspore- quartz-pyrophyllite rock. The main andalusite deposit contains large zones of topaz—quartz, pyrite—quartz- andalusite, and trolleite-scorzalite-augelite rock (Gross and Parwel, 1968). Rutile is disseminated throughout the andalusite body as well as the surrounding altered schists in small grains (0.1 to 3 RUTILE IN GNEISS AND RELATED ROCKS, FRONT RANGE, COLORADO mm) and scattered larger crystals and is most abundant in the diaspore-quartz-pyrophyllite rocks. A chemical analysis of rutile from this rock (Gross and Parwel, 1969, p. 495) shows that it is relatively pure containing minor iron and vanadium and trace amounts of tin, chrome, and antimony. Niobium was not detected in the sample. These trace elements were probably included in the rutile from a magmatic source. The scarcity of titanium minerals in the unaltered surrounding schists suggest that the source of rutile mineralization was probably the quartz monzonite porphyry. Deposition of rutile occurred after topaz and andalusite and probably continued throughout both stages of metamorphism. The quantity of rutile present is difficult to estimate, inasmuch as no percentages or tonnages have been given, but rutile is present throughout the metamorphosed area, which covers approximately 2.5 km2 of relatively high-relief terrane. YUMA COUNTY, ARIZONA A rutile-bearing kyanite quartzite on the west side of the Dome Rock Mountains in the NE1/4 of sec. 34, T. 4 N ., R. 21 W., in extreme western Yuma County, Ariz., was visited by Marsh in January 1970. The rutile occurs as small disseminated grains in kyanite quartzite and some larger crystals in small localized quartz veins. The rutile-bearing kyanite quartzite may be divided into two zones, a northerly belt of magnetite-rich quartzite and a southerly belt of rutile-bearing kyanite quartzite. These belts are not well defined and appear to grade one into the other. The rutile zone is approximately 900 m long and 150 to 300 m wide, strikes east, and has a northerly dip of about 60°. The rutile zone thickens somewhat to the east. The rutile occurs in clots of small grains. These clots have a roughly rectangular bladed shape and could possibly be relics from some former crystal, perhaps ilmenite. Lazulite is locally abundant and topaz and wagnerite are present. The texture of the rutile-bearing kyanite quartzite is porous and appears leached. The area is mapped by Wilson (1960) as Mesozoic schists, but similar rocks in the adjacent Quartzsite quadrangle have been mapped (Miller, 1968) as Precambrian. Rocks in the area apparently were contact metamorphosed by a quartz monzonite probably of Mesozoic age (Wilson, 1960) adjacent to the rutile-bearing kyanite quartzite. Several samples were collected from the rutile-bearing kyanite quart- zite, and four spectrographic analyses of rutile separated from these samples are listed in table 2. Estimates of the amount of rutile in the deposit have not been made, but the outcrop area is sufficiently large to make the area a potential resource. G15 SANTA CRUZ COUNTY, ARIZONA Rutile-bearing quartzite occurs in a contact metamorphosed zone of Triassic volcanics (dacitic and latitic flows) and sedimentary rocks (conglomerates, sandstones, and quartzites) (Drewes, 1971) in the Santa Rita Mountains on the east side of Madera Canyon in the NW1/4 sec. 1, T. 20 S., R. 14 E., in northern Santa Cruz County, Ariz. Three samples of quartzite from this locality collected by Ed Montgomery of Duval Corporation show that andalusite is the common aluminum-silicate mineral and topaz, lazulite, and rutile are accessory. One sample contained as much as 10 percent topaz, suggesting fluorine metasomatism of the quartzite. A large area of Cretaceous granodiorite immediately adjacent to the area was probably responsible for the contact metamorphism. Spectrographic analyses of rutile separates from the three samples are listed in table 2. No accurate estimates of the amount of rutile in this area can be made at present, but the area of contact metamorphism is about 3 km long and 1 km wide. OTHER UNITED STATES OCCURRENCES Rutile-bearing gneiss has been reported in several other areas, an indication that primary rutile is more widely distributed in these types of metamorphic rocks than previously recognized. Fine-grained and locally abundant rutile has been reported in an andalusite, kyanite quartzite of Precambrian age at Kiawa Mountain, N.Mex. (Heinrich and Corey, 1959). A kyanite mine on the western slope of the Cargo Muchacho Mountains, Imperial County, Calif, is reported to contain euhedral crystals of red rutile (Pemberton and Bideaux, 1968). The deposit occurs in metamorphosed quartz and arkosic sandstone of pre-Mesozoic, possibly Precambrian age. Mesozoic granitoid rocks metamorphosed the sediments. Local hydrothermal action followed intrustion (Henshaw, 1942). WORLD OCCURRENCES The occurrence of rutile-bearing aluminum silicate rocks in other parts of the world is difficult to document. As with the United States occurrences, this rock type is noted in detailed descriptions of aluminum-silicate deposits, with rutile being mentioned as a minor part of the mineral assemblage. A few examples of this type of occurrence are briefly described here. Australia is the world’s largest producer of rutile, and most of it comes from beach placers on the east coast. These deposits have developed through two G16 stages of weathering and concentration; detritus from the Precambrian shield rocks were reconsolidated as Mesozoic rocks and then were reworked into the beach placers being mined today. This reworking of sediments served to concentrate the heavy minerals into simpler mineral assemblages because of weathering and removal of less-stable minerals, resulting in minable placer deposits consisting mainly of rutile and zircon. The probable source for the rutile was the crystalline rocks of the interior where rutile occurs in Degmatitic veins and crystalline schists of Precambrian age (Whitworth, 1956, p. 34). Probable commercial amounts of rutile occur in the pegmatites of the Avon district in Western Australia and at Myponga, Parra, and Radium Hill in South Australia. Rutile in the crystalline schists occurs mainly in South Australia in Blumberg, Yankalilla, and Williamstown. The rutile occurs disseminated and along bedding planes in biotite and sillimanite schists. In several of these occurrences rutile is a byproduct or impurity in the mining of aluminous refractory material. As much as 4 percent rutile occurs in a chlorite-corundum rock from Mount Painter, South Australia (Oliver and Jones, 1965). The Precambrian shield area of northern India contains numerous occurrences of biotite and sillimanite schists that contain disseminated fine- grained rutile or small clots of rutile along bedding planes (Dunn, 1929). No report is made of the grade of rutile, but its occurrence is noted at many localities and it appears to be common in some. Accessory minerals are topaz, tourmaline, green micas, corundum, and magnetite. These rutile-bearing rocks are thought to be metavolcanic or metasedimentary— probably basic lavas or bauxitic clays—in origin (Dunn, 1929, p. 243—252). In Sweden rutile occurs in three areas of Precambrian metamorphic quartzite (Geijer, 1964). The largest is at Haallsjiiberget in southwest Sweden where rutile occurs as streaks and veinlets in kyanite- bearing quartzite. Lazulite, iron—rich wagnerite, and pyrophyllite are also present. No size or grade figures are given, but rutile is reported as abundant in some of the kyanite concentrations. A similar but smaller deposit occurs at Dicksberget, 60 km south of Haallsjiiberget. Here, quartzite with 15-25 percent kyanite has accessory rutile, ilmenite, lazulite, and pyrite. The third deposit is at Vélstanaa near the southern Baltic coast where Precambrian quartzite contains commercially worked hematite deposits. Here, quartzite is overlain by alumina—rich mica schist containing kyanite, andalusite, and sparse sillimanite. Rutile is widespread and amounts to about 1 percent GEOLOGY AND RESOURCES OF TITANIUM of the rock. Other accessory minerals include lazulite, zircon, and pyrophyllite, These occurrences are thought to be high-temperature replacement deposits of originally high-alumina and silica rocks. Serdyuchenko (1968), described several areas where rutile occurs with aluminum-silicate rocks. In south Norway at Arendal rutile occurs in amounts as great as 1 percent in corundum-bearing gneisses con- taining lazulite and sillimanite. In the Orange River Basin in Namaqualand, South-West Africa, rutile occurs as an accessory (1 percent) in sillimanite and sillimanite-corundum beds in metamorphic rocks. ECONOMIC POTENTIAL For reasons cited in the introduction, rutile deposits of the Evergreen area may never be mined. Data reported below illustrate resources that can be expected if continued search elsewhere discloses deposits of similar size and mineralogy. The sillimanitic topaz-quartz gneiss constitutes the most important type of rutile-bearing rock in the Evergreen area. The following computations are made from the layers of this rock from sample locality A to sample locality D (fig. 2), a strike length of 2,100 m. The unit is exposed to a depth of 73 m; the thickness ranges from 3 to 21 m and averages 12 m. Chemical analyses of four samples (1 to 4, table 2) were used to calculate average mineralogic grade: rutile, 2.1 percent; topaz, 28.5 percent; sillimanite, 12.3 percent. The indicated resources, using a rock density of 3.00, are 5.5 million metric tons of sillimanitic topaz-quartz gneiss. These contain 115,000 metric tons of rutile, 1,600,000 metric tons of topaz, 670,000 metric tons of sillimanite, and 3,115,000 metric tons of quartz. Each additional 30 m of depth contains the inferred re- sources: 47,000 metric tons of rutile, 646,000 metric tons of topaz, and 280,000 metric tons of sillimanite. Although the rutile is fine-grained (commonly 0.5 mm or less in size), laboratory tests show that much of it is freed in the crushing stage and much of the remainder is freed by grinding to 175M m. Extending west-northwestward from the west end of the deposit described above (sample locality A, fig. 2) is the Z-shaped trace of other rutile-bearing gneisses that contain 0.5 to 2.5 percent rutile. As measured along the trace, these are about 11 km in length and range in thickness from 30 cm to 9 m. In the area to the southeast are several lenses of sillimanitic topaz—quartz gneiss that range in length from 60 to 800 m and in thickness from 15 cm to 8 m; the rutile content ranges from 0.5 to 3 percent. Other rutile-bearing layers typically contain less than 1 percent rutile and are not considered to be a resource. RUTILE IN GNEISS AND RELATED ROCKS, FRONT RANGE, COLORADO REFERENCES CITED Bryant, Bruce, Miller. Rd. D., and Scott, G. R., 1973, Geologic map of the Indian Hills quadrangle, Jefferson County, Colorado: U.S. Geol. Survey Geol. Quad. Map GQ-1073 7 p. [1974]. Crickmay, G. W., 1952, Geology of the crystalline rocks of Georgia: Georgia Geol. Survey Bull. no. 58, 54 p. Drewes, Harold, 1971, Geologic map of the Mount Wrightson quad- rangle, southeast of Tucson, Santa Cruz and Pima Counties, Arizona: U.S. Geol. Survey Misc. Geol. Inv. Map I-614. Dunn, J. A., 1929, The aluminous refractory materials: kyanite, sillimanite, and corundum in northern India: India Geol. Survey, Mem. 52, pt. 2, p. 145—274. Espenshade, G. H., and Potter, D. B., 1960, Kyanite, sillimanite, and andalusite deposits of the southeastern States: U.S. Geol. Survey Prof. Paper 336, 121 p. Fleischer, Michael, and Robinson, W. 0., 1963, Some problems of the geochemistry of fluorine, in Studies in analytical geo- chemistry: Royal Soc. Canada Spec. Pub. 6, p. 58-75. Gable, D. J ., 1968, Geology of the crystalline rocks in the western part of the Morrison quadrangle, Jefferson County, Colorado: U.S. Geol. Survey Bull. 1251-E, p. E1-E45. Gable, D. J ., and Sims, P. K., 1969, Geology and regional metamorphism of some high-grade cordierite gneisses, Front Range, Colorado: Geol. Soc. America Spec. Paper 128, 87 p. Geijer, Per,1964,Genetic relationships of the paragenesis AleiO5- lazulite-rutile: Arkiv Mineralogi och Geologi, v. 3, no.24, p. 423-466. Gross, E. B., and Parwel, A., 1969, Rutile mineralization at the White Mountain andalusite deposits, California: Arkiv Mineralogi och Geologi, v. 4, no. 6, paper 29 p. 493-497. Hedge, C. E., 1969, A petrogenetic and geochronologic study of migmatites and pegmatites in the central Front Range: Colorado School Mines, unpub. PhD. thesis, 158 p. Heinrich, E. W., and Corey, A. F., 1959, Manganian andalusite from Kiawa Mountain, Rio Arriba County, New Mexico: Am. Mineralogist, v. 44, p. 1261-1271. Henshaw, P. C., 1942, Geology and mineral deposits of the Cargo Muchacho Mountains, Imperial County, California: California Jour. Mines and Geology, v. 38, no. 2, p. 147-196. Hurst, V. J ., 1959, The geology and mineralogy of Graves Mountain, Georgia: Georgia Geol. Survey Bull 68, 33 p. Kerr, P. F., 1932, The occurrence of andalusite and related minerals at White Mountain, California: Econ. Geology, v. 27, no. 7, p. 614-643. Krauskopf, K. B., 1967, Introduction to geochemistry: New York, McGraw-Hill Book Co., 721 p. Lovering, T. S., and Goddard, E. N., 1950, Geology and ore de- posits of the Front Range, Colorado: U.S. Geol. Survey Prof. Paper 223, 319 p. Miller, F. K., 1968, Geology of the Quartzsite quadrangle, Arizona: U.S. Geol. Survey Open-file rept. (Denver 0F Temp. 783). Moench, R. H., 1964, Geology of Precambrian rocks, Idaho Springs district, Colordo: U.S. Geol. Survey Bull. 1182-A, p. A1-A70. Moench, R. H., Harrison, J. E., and Sims, P. K., 1962, Precam- brian folding in the Idaho Springs-Central City area, Front Range, 0010.: Geol. Soc. America Bull., v. 73, no. 1, p. 35-58. G17 Myers, A. T., Havens, R. G., and Dunton, P. J., 1961, A spectrochemical method for the semiquantitative analysis of rocks, minerals, and ores: U.S. Geol. Survey Bull. 1084-I, p. 207-229. National Materials Advisory Board, 1972, Processes for rutile substitutes; appendix B, National stockpile purchase specifica- tions, P-49-R5, November 1, 1967: Washington, D.C., National Materials Advisory Board NMAB-293. Oliver, R. L., and Jones, J. B., 1965, A chlorite-corundum rock from Mount Painter, South Australia: Mineralogy Mag., v. 35, no. 269, p. 140-145. Pauling, Linus, 1929, The principles determinging the structure of complex ionic crystals: J our. Am. Chem. Soc. v. 51, p. 1010-1026. Pemberton, H. E., and Bideaux, R. A., 1968, An occurrence of svanbergite in California: Mineral Explorer, v. 3, no. 1, p. 11. Rankama, Kalervo, and Sahama, Th. G., 1950, Geochemistry: Chicago, Chicago Univ. Press, 912 p. Serdyuchenko, D. P., 1968, Metamorphosed weathering crusts of the Precambrian, their metallogenic and petrographic fea- tures: Internat. Geol. Cong., 23d, Prague, Czechoslovakia, 1968, sec. 4 Proc., 37-42. Shannon, R. D., and Prewitt, C. T., 1969, Effective ionic radii in oxides and fluorides: Acta Cryst., sect. B, v. 25, p. 925-946. Sheridan, D. M., Marsh, S. P., Mrose, M. E., and Taylor, R. B., 1971, Wagnerite from Santa Fe Mountain, Colorado — a new occurrence [abs.]: Canadian Mineralogist, v. 10, pt. 5, p. 919. Sheridan, D. M., Maxwell, C. H., and Albee, A. L., 1967, Geology and uranium deposits of the Ralston Buttes district, Jefferson county, Colorado, with sections on Paleozoic and younger sedi- mentary rocks, by Richard Van Horn: U.S. Geol. Survey Prof. Paper 520, 121 p. Sheridan, D. M., Reed, J. 0., Jr., and Bryant, Bruce, 1972, Geolo- gic map of the Evergreen quadrangle, Jefferson County, Colo- rado: U.S. Geol. Survey Misc. Geol. Inv. Map I-786-A (1973). Sheridan, D. M., Taylor, R. B., and Marsh, S. P., 1968, Rutile and topaz in Precambrian gneiss, Jefferson and Clear Creek Counties, Colorado: U.S. Geol. Survey Circ. 567, 7 p. Sims, P. K., and Gable, D. J ., 1967, Petrology and structure of Precambrian rocks, Central City quadrangle, Colorado: U.S. Geol. Survey Prof. Paper 554—E, p. E1-E56. Spurr, J. E., Garrey, G. H., and Ball, S. H., 1908, Economic geology of the Georgetown quadrangle (together with the Empire district), Colorado: U.S. Geol. Survey Prof. Paper 63, 422 p. Tweto, Ogden, and Sims, P. K., 1963, Precambrian ancestry of the Colorado mineral belt: Geol. Soc. America Bull., v. 74, no. 8, p. 991-1014. Wells, J. D., Sheridan, D. M., and Albee, A. L., 1964, Relationship of Precambrian quartzite—schist sequence along Coal Creek to Idaho Springs Formation, Front Range, Colorado: U.S. Geol. Survey Prof. Paper 454-0, 25 p. Whitworth, H. F., 1956, The zircon-rutile deposits on the beaches of the east coast of Australia with special reference to their mode of occurrence and the origin of the minerals, in Geological survey of New South Wales: New South Wales Dept. Mines Tech. Rept., v. 4, p. 7-60. Wilson, E. D., 1960, Geologic map of Yuma County, Arizona: Tuc- son, Arizona Bur. Mines. QUS. GOVERNMENT PRINTING OFFICE: 1976—677340/84 «S. 1. Aflmflaafifi 4%, . ,L . L . .. . .1... $11,932! 2. .K. 1.... .vlliz3Lv3LE KO . . . .5...” x . . r a . .. 3 . J; , ,, a... dumb»; Eaxéisg BR‘ARY H h‘fiRSiT‘flOF Z‘ALJFORNEAW EARTH LCJENCES “ CENTRAL VIRGINIA ‘ gu— ARY ' 7 . L Q‘ _ ht ‘ m ””1“ WM ‘3' “f ”“5"? W. ,m w an ‘ > .0 COVER PHOTOGRAPHS 1. Asbestos ore 8. Aluminum ore, bauxite, Georgia 1 2 3 4 2. Lead ore-Balmat Mine, N. Y, 9. Native copper ore. Keweenawan 5 6 3. Chromite-chromium ore. Wash. Peninsula, Mich, 4. Zinc ore»Friedensville. Pa. 10. Porphyry molybdenum ore. Colo. 7 8 5, Banded Iron formationPalmer. 11. Zinc ore. Edwards. N. Y. Michugan 12. Manganese nodules, ocean floor 9 10 6. Ribbon asbestos ore. Quebec. Canada 13. Botryoidal fluorite ore. 11 12 13 14 7. Manganese ore. banded Poncha Springs, Colo. rhodochrosite l4. Tungsten ore. North Carolma Alluvial Ilmenite Placer Deposits, Central Virginia By J. P. MINARD, E. R. FORCE, and G. W. HAYES GEOLOGY AND RESOURCES OF TITANIUM GEOLOGICAL SURVEY PROFESSIONAL PAPER 959-H A discussion of ilmenite placers in alluvial deposits along streams draining the Roseland Anorthosite and nearby areas, and their economic potential UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1976 UNITED STATES DEPARTMENT OF THE INTERIOR THOMAS S. KLEPPE, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Director Library of Congress Cataloging in Publication Data Minard, James P Alluvial ilmenite placer deposits, central Virginia. (Geology and resources of titanium) (Geological Survey professional paper ; 959—H) Bibliography: p. 1. Ilmenite—Virginia. 2. Placer deposits—Virginia. 1. Force, Eric R., joint author. II. Hayes, G. W., joint author. III. Title. IV. Series. V. Series: United States. Geological Survey. Professional paper ; 959—H. QE391.I44M56 1976 553'.462 76—15552 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock Number 024-001-02877-8 APPRAISAL OF MINERAL RESOURCES Continuing appraisal of the mineral resources of the United States is conducted by the US. Geological Survey in accordance with the provisions of the Mining and Minerals Policy Act of 1970 (Public Law 91—631, Dec. 31, 1970). Total resources for purposes of these appraisal estimates in- clude currently minable resources (reserves) as well as those resources not yet discovered or not presently profitable to mine. The mining of mineral deposits, once discovered, depends on geologic, economic, and technologic factors; however, identification of many de- posits yet to be discovered, owing to incomplete knowledge of their dis- tribution in the Earth’s crust, depends greatly on geologic availability and man’s ingenuity. Consequently, appraisal of mineral resources results in approximations, subject to constant change as known deposits are depleted, new deposits are found, new extractive technology and uses are developed, and new geologic knowledge and theories indicate new areas favorable for exploration. This professional paper discusses aspects of the geology of titanium as a framework for appraising resources of‘ this commodity in the light of to- day’s technology, economics, and geologic knowledge. Other Geological Survey publications relating to the appraisal of re- sources of specific mineral commodities include the following: Professional Paper 820—“United States Mineral Resources” Professional Paper 907—“Geology and Resources of Copper” Professional Paper 926—“Geology and Resources of Vanadium Deposits” Professional Paper 933—“Geology and Resources of Fluorine in the United States” CONTENTS Page Page Metric-English equivalents ________________________ VI Areas sampled ___________________________________ H5 Abstract _________________________________________ H1 Tye River ----------------------------------- 5 Introduction ______________________________________ 1 Mmerabgy """""""""" " """""" 5 Heavy-mineral contents ___________________ 7 General 390108? —————————————————————————————————— 1 Grain size ___________________________ 7 Physiography ———————————————————————————————————— 2 Transport distance ____________________ 9 Previous mining activities _________________________ 3 Buffalo River and South Fork Rockfish River ____ 12 Present study ____________________________________ 3 James River _________________________________ 13 Field methods ________________________________ 3 Conclusions ______________________________________ 13 Laboratory methods ___________________________ 4 References cited __________________________________ 14 ILLUSTRATIONS Page PLATE 1. Map of the central Virginia rutile district showing sites augured, locations and cross sections of auger traverses, and logs of holes with percentage of heavy minerals _______________________________ In pocket FIGURE 1. Index map showing location of study area ________________________________________________________ H2 2. Photograph of angering; photograph of hole dug through cobble layer ____________________________ 4 3. Photomicrograph of dioritic ilmenite-rich facies of gneiss from the Pedlar Formation of Bloomer and Werner (1955) ____________________________________________________________________________ 5 4. Histograms showing grain sizes of some typical alluvial samples and their heavy minerals _____________ 8 5. Graph showing heavy-mineral and mud contents of some Tye River and James River samples ________ 9 6. Graph showing decrease in heavy-mineral content of Tye River and James River deposits as a function of distance downstream _____________________________________________________________________ 9 7. Sketch map showing location of source rocks of ilmenite in the Tye River drainage basin _____________ 14 TABLES Page TABLE 1. Locations of samples and cross sections __________________________________________________________ H6 2. Mineralogy of heavy fractions of samples from the Tye and James Rivers in order of distance down- stream ____________________________________________________________________________________ 7 3. Ti02 content of ilmenite separates _______________________________________________________________ 7 4. Approximate ilmenite resources of some terrace deposits in the study area __________________________ 7 5. Grain sizes and heavy-mineral data for alluvial samples _____________________________________________ 10 6. Heavy-mineral content of alluvium exposed in James River bluffs downstream from Scottsville ____ 14 VI CONTENTS METRIC-ENGLISH EQUIVALENTS Metric unit English equivalent Metric unit English equivalent Length Specific combinations—Continued millimetre (mm) : 0.03937 inch (in) litre per second (l/s) : .0353 cubic foot per second metre (in) : 3.28 feet (ft) cubic metre per second kilometre (km) : .62 mile (mi) per square kilometre [(m3/s)/km2] : 91.47 cubic feet per second per Area square mile [(ft3/s)/mi2] metre per day (in/d) : 3.28 feet per day (hydraulic square metre (m9) : 10.76 square feet (ft?) . CODdUCfiVitY) (ft/d) square kilometre (ka) : .386 square mile (mi?) metre per kilometre _ 3 hectare (ha) : 2.47 acres m/km) : 0-25 fet‘t DBI‘ mile (ft/mi) kilogeti‘le) per hour 911 f t (1 (ft/ ) m/ = . 3 00 per secon s Volume metre per secdond (En/s) : 3.28 feet per second ~ 3 _ - 3 metre square per ay fig): (clentimetre (cm ) ; 69.8521 33:33 311311185111 ) (m2/d) : 10.764 feet squared per day (ft2/d) cubic metre (m3) = 35.31 cubic feet (ft3) (transmissmm cubic metre : .00081 acre—foot (acre-ft) cubicametre per second __ . cubic hectometre (hms) 2810.7 acre-feet (m /S) — 22-826 million gallons per day litre : 2.113 pints (pt) . (Meal/d) litre : 1.03 quarts (qt) cubic metre per minute litre : .26 gallon (gal) (m3/min) 2264.2 gallons per minute (gal/min) cubic metre : .00026 million gallons (Mgal or litre per second (NS) = 15.85 gallons per minute 103 gal) litre per second per _ cubic metre : 6.290 barrels (bbl) (1 bbl:42 gal) metre [(1/s)/ln] : 4.83 gallons per_mlnute per foot kil t h [(gal/mln) /ft] ‘ ome re per 0111' we‘ght (km/h) 2 .62 mile per hour (mi/Ii) gram (g) = 0.0359 ounce, avoirdupois (oz avdp) gggegeircsfifi’c’m ("1/5) : 2237 miles per hour 333; (t) 2 11902" €§R§dgh%l‘i‘i“2“88$sufiib Mp) centimetre (g/cmo = 62.43 pounds per cubic foot (lb/w) n = .98 t l 2 211 1b gram per Square to Be on, ong ( ' 0 ) centimetre (g/cmg) = 2.048 pounds per square foot (lb/ftz) ' ' ' gram per square Specific combinations centimetre .0142 pound per square inch (lb/inz) kilogram per square centimetre (kg/cm?) 2 0.96 atmosphere (atm) Temperature kilogram per square D . centimetre = .98 bar (0.9869 atm) degree CelSiUS ( C) = 1.8 degrees Fahrenheit (°F) cubic metre per second degrees Celsius (m3/s) : 35.3 cubic feet per second (ft3/s) (temperature) :[(1.8><°C) +32] degrees Fahrenheit GEOLOGY AND RESOURCES OF TITANIUM ALLUVIAL ILMENITE PLACER DEPOSITS, CENTRAL VIRGINIA By J. P. MINARD, E. R. FORCE, and G. W. HAYES ABSTRACT Point bars and flood plains along rivers draining the Roseland, Va., anorthosite body and nearby mountain areas in the Blue Ridge physiographic province contain placer de-~ posits in which there are significant amounts of rutile and ilmenite. Highest values generally are in deposits closest to the sources. Although high values (4 percent ilmenite with some rutile) seem to be associated with deposits along streams draining the area of anorthosite, equally high values are pres- ent along some streams where the source of titanium minerals is not anorthosite but hypersthene—bearing gneisses in the Pedlar Formation of Bloomer and Werner (1955). There- fore, sources of titanium minerals may be not only the small anorthosite body but also the much more widespread gneisses of the Virginia Blue Ridge. Values decrease from as much as 9 percent in alluvial placers on the anorthosite body along the Tye River to about 1 percent 80 km downstream along the James River. The higher values upstream are generally in small deposits, whereas the low values downstream are in large deposits. Downstream movement of heavy minerals over long periods of time may have resulted in further con- centrations in such Coastal Plain deposits as tidal deltas and beach ridges. INTRODUCTION The purpose of this paper is to report the result of a study of rutile and ilmenite alluvial placer deposits in, near, and downstream from the Roseland rutile district in central Virginia. The heart of the district is in Nelson and Amherst Counties (fig. 1); this area was one of the most important sources of il- menite and rutile for many years. Early in this cen- tury, the entire world supply of rutile came from this district, and it continued as an important con- tributor until 1949. Recently the plant ceased opera- tions. However, large resources of both rutile and ilmenite may still be present, and it is hoped that this report may help stimulate further interest in the potential of the area, as was suggested by Herz and others (1970). Herz and others (1970) reported ilmenite- and rutile-rich sediments in streams draining the Rose- land Anorthosite, with which the titanium minerals seem to be associated. Their study included analyses of 31 samples of sand and gravel collected from the upper 15—30 cm of riflie deposits in the present stream channels. Most of their samples were col- lected in streams on the anorthosite body or immedi- ately adjacent to it; two samples were collected in channels several kilometres downstream from the anorthosite (Herz and others, 1970, pl. 1). They con- cluded (p. F8) “that valuable deposits may have been created by stream action” and recommended that “To fully evaluate the available resources of il- menite and rutile, churn drilling and detailed map- ping in stream valleys will have to be carried out.” The present study was partly guided by these sug- gestions. The area studied is larger than that of Herz and others and includes the drainage basins of the South Fork Rockfish River, Tye River, and Buf- falo River, all tributaries to the James River, and along the James in the general area where these rivers flow into it. The area of this report includes parts of Nelson, Amherst, Albemarle, and Bucking- ham Counties (fig. 1). Some reconnaisance sampling was also done downstream from this area along the James River. GENERAL GEOLOGY The Roseland district is in the Blue Ridge prov- ince. The general geology in the area was described by Watson and Taber (1913), Bloomer and Werner (1955), and Herz (1968). Rock types presently drained by the local streams discussed in this report include Roseland Anorthosite, hypersthene gneiss and products of its incomplete retrograde meta- morphism, biotite gneiss, migmatite, schist, granitic igneous rocks, and greenstone. The upper reaches of the James River, however, which are well outside the study area, drain sedimentary rocks of the Val- H1 H2 100 MILES I f‘ ’ \ l \l i 80{ T 100 KlLOMETRES GEOLOGY AND RESOURCES 0F TITANIUM "’2. its h/\\ f . l Chm-1m A /\ 0 , ,7 fig ,, , , ,, 3 ALBEMARLE \1 ‘ . l ‘ - ‘ oDanville —~—_ 81" q. _ 80° 79“ FIGURE 1.—Index map showing Nelson, Amherst, Albemarle, and Buckingham Counties, and Roseland, Va., the center of the central Virginia rutile district. Area covered by plate 1 is shaded. ley and Ridge province. Titanium deposits are known to be associated with the Roseland Anorthosite, a northeast-trending body about 15 km long and 4 km wide (pl. 1) . The anorthosite consists largely of light- bluish-gray megacrysts of andesine antiperthite that are cut by zones of cream to white granulated feldspar (Ross, 1941). Charnockitic and mafic rocks are present as dikes and irregular patch-es and lenses throughout the anorthosite body but are more abun- dant in the border zone (Herz and others, 1970, p. F3, F4). Quartz, where present, is blue. Titanium minerals are ilmenite and rutile, both rimmed by “leucoxene” (Ross, 1941) . Ilmenite and apatite are present in the border zone, chiefly in nelsonite dikes. Some varieties of these dikes are rich in rutile, magnetite, biotite, and hornblende, or are gabbroic. The dikes range in width from several centimetres to 20 m, and are as much as 600 m long (Watson and Taber, 1913, p. 101, 102). The dikes are younger than the anortho- site and are the source of the richest saprolite de- posits of ilmenite (Fish and Swanson, 1964). At least one other formation in the area, the Pedlar of Bloomer and Werner (1955) , also contains high per- centages of ilmenite. The Pedlar Formation is a coarse-grained porphyroblastic gneiss which locally contains relict hypersthene. Ilmenite averages 1.5 percent (R. O. Bloomer, written commun., May 1973) but may be as much as 8 percent and is rim— med by “leucoxene.” PHYSIOGRAPHY A jumbled mass of mountains, which range in a1— titude from 600 to 1,200 m, trends from the north- ern part of the area westward and southward in an arc, nearly encircling a hilly erosional reentrant low- land of the headwaters of the Tye, Piney, and Buffa- lo Rivers. A series of northeast-trending linear ridges separates this intermontane hilly lowland from the James River valley in the southeast part of the area. Altitudes of these linear ridges range from 300 to 400 m. Altitudes of the hills within the moun- tain-locked lowland range from 250 to 300 no (pl. 1). The Roseland Anorthosite is a low plateau in the intermontane lowland; it has a relief of about 30 m. After draining the intermontane lowland, the Tye and Piney Rivers and, later, the Buffalo River join as they flow through a narrow gorge cut through the linear ridges in the southeast, before joining the James River at Norwood. The Tye River descends ALLUVIAL ILMENITE PLACER DEPOSITS, CENTRAL VIRGINIA about 700 m in the study area. The Rockfish River flows northeast, then southeast, to join the James at Howardsville, 29 km downstream from Norwood (pl. 1). The Rockfish descends about 850 m from its headwaters. The James River follows a wandering northeast course through rolling country along the southeast boundary of the area; the gradient aver— ages about 1 m/ km. The flood plain of the James River is as much as 1 km wide. The Tye River system has its widest flood plains on the anorthosite body, but they are mostly less than 400 m wide. The flood plain of the South Fork Rockfish River (in the upper Rockfish River Valley) is as much as 1 km wide. Thick saprolite blankets much of the area, but bedrock outcrops are common on steeper slopes and along streams. Most of the mountain areas are for- ested; typically, the only cleared areas of any extent are in the lowlands and stream valleys. PREVIOUS MINING ACTIVITIES The earliest mining activity apparently was in 1878 when minor investigations were undertaken by the Philadelphia and Reading Coal and Iron Co. in its exploration for iron deposits. Some subsequent activity was directed towards investigation of the phosphate content in nelsonite. The next significant activity was by the American Rutile Co. in 1900, in the first attempt to mine rutile, largely from bed- rock. In 1930, the Vanadium Co. of America began mining saprolite along the Piney River (Fish, 1962, p. 5). In 1944, these properties were acquired by the American Cyanamid Co., which mined titanium min- erals in the saprolite at several places in the area until 1971. A more detailed history of mining has been given by Fish (1962), Fish and Swanson (1964) , and Herz and others (1970) . PRESENT STUDY The present study began as an outgrowth of that by Herz and others (1970). Herz and Minard planned the sampling of flood-plain deposits in and near Herz’s study area in order to supplement his data. Sampling was done by Minard and Hayes, and analyses by Force and Hayes. Although this study is supplemental to that by Herz and others (197 0) , it differs in several ways: 1. Most samples were obtained by augering below the ground surface of stream terraces and point bars, instead of by shoveling a 10-quart bucket full of bottom material from riflies in present stream channels. H3 2. The area of sample collection was increased to in- clude downstream extensions of streams drain- ing anorthosite and nearby streams not drain- ing anorthosite. Large-volume deposits along the J am-es River, which drains the entire area, were also sampled. 3. Hurricane Camille occurred in August 1969 (Vir- ginia Division of Mineral Resources, 1969; Williams and Guy, 1973), after the sampling reported in Herz and others (1970) but before that done for this report. Flood waters associ- ated with the hurricane locally deposited sedi- ments containing high percentages of titanium oxides. 4. Analysis procedure has a different emphasis in this study. Size analyses were done on many samples in order to examine the influence of sorting on the heavy-mineral concentrations. Methylene iodide (specific gravity 3.3) was used as a separating medium in order to limit more closely the heavy fraction to minerals of economic interest (ilmenite and rutile) ; rela- tively little study of the mineralogy was done. 5. The number of samples collected was 260, as com- pared with 31 collected by Herz and others (1970). Cross sections constructed along auger traverses and logs of individual holes are shown on plate 1. Percentages of heavy minerals, mostly ilmenite, are also shown for those samples analyzed. FIELD METHODS The method used in most of the sampling pro-gram was to auger a series of holes on a line of traverse across the flood plain, terrace, or point bar normal to the stream channel. Generally two to four holes were augered along each traverse line. From one to as many as five or six lines were traversed across each terrace or bar, depending on its length. In some places only one hole was augered, usually because of the small area of the terrace or bar, the shallow depth to bedrock, or because of obstacles such as ditches and crop cover which prevented access by the truck-mounted auger. Each hole was augered to bedrock or, in some places, probably to a boulder layer. Samples were collected as channel samples from each 1.5-m auger length. The auger was rotated slowly to a depth of 1.5 m, rotation was stopped, and the auger was with- drawn from the hole. The outer surface of the ma- terial on the auger flight (spiral land) was scraped off, and the remaining material along the entire 1.5- m length was sampled continuously. The flight was H4 thoroughly cleaned, lowered in the hole until it touched bottom, another auger length added, rota- tion started, and penetration to 3 m achieved. Rota- tion was stopped, the auger string withdrawn from the hole, and the process repeated each 1.5 m or until further penetration was not possible. A truck- mounted power auger was used, having 1.5-m long auger lengths of 11-cm diameter in a continuous string (fig. 2A). Holes were dug by hand through surface cobble layers to enable augering below these layers (fig. 23‘). FIGURE 2.—A, Augering at cross section 27 across a point bar along the James River. About 2 m of the 3 m of ex- posed auger lengths contain a sample of silty sand. B, Hole dug through cobble layer so the auger could penetrate un- derlying pebbly sand. Contrast with silty sand of A. At cross section 15, South Fork Rockfish River. LABORATORY METHODS Although 260 samples were collected, only 148 were analyzed in the laboratory, and, of these, the GEOLOGY AND RESOURCES OF TITANIUM analyses for 122 are used in this study. Those sam- ples consisting totally or largely of silt and clay gen- erally were not analyzed, and some that were ana- lyzed Were not used. Some sample analyses were discarded because of faulty laboratory procedures. Each sample analyzed was dried, and the clumps were disaggregated by a rubber roller to ensure a more correct size distribution. A 100—300-g split of the dried disaggregated sample was made for analy- sis. After being weighed, the sample was placed in a 62/1. screen and first shaken dry to remove loose silt and clay and then washed to remove any silt and clay coatings on the sand grains. Any remaining muddy coating of the sand grains was removed by immersing the washed sample, on the 62,; sieve, into an ultrasonic cleaner for 15—20 minutes. No chemi- cal removal of grain coatings was necessary for pur- poses of mineral identification. The dried washed sample was again weighed, and the weight loss was entered as the silt and clay frac- tion. A RoTap 1 having 2-mm, l-mm, 500”, 250M, 125”, and 62,; screens (1 ¢ interval) was used to size-sort the sand and gravel. Each fraction was weighed, and heavy minerals were collected sepa- rately from the fractions. Methylene iodide, having a specific gravity slight- ly less than 3.3, was used as the heavy liquid in separations (rather than bromoform, which has a specific gravity of about 2.9). This was to reduce the amount of noneconomic heavy minerals, especial- ly sillimanite, hornblende, and biotite, in the concen- trate. Separations were done in a gravity funnel for each size fraction. These fractions were washed with acetone, dried, and weighed. For some samples, the minerals in the heavy frac- tion were separated and identified. Magnetite was removed from the concentrates by means of a hand magnet; further magnetic separations were made by using a Frantz isodynamic separator. Concentrates were separated on the‘Frantz at a final setting of 0.35 amperes (with forward and side slopes of 20°) , after which they were weighed. To avoid loss of sam- ple, amperage was progressively increased from 0.05 to 0.35 amperes on successive runs of the sample; the magnetic fraction of each run was caught in the same container. Magnetic separation at 0.35 am- peres was done to separate the rutile and other min- erals from the ilmenite. This amperage was deter- mined experimentally and appears to be mostly suc- cessful (Herz and others, 1970, used the same value). W names in this publication are used for descriptive pur- poses only and do not constitute endorsement by the U.S. Geological Survey. ALLUVIAL ILMENITE PLACER DEPOSITS, CENTRAL VIRGINIA For many samples, not all the above steps were necessary. The most common shortcut was to do heavy-mineral separations of only a few size frac- tions, which resulted in minimum heavy-mineral values. For many other samples, such as those for which no magnetic separation was done, complete analyses were not made. AREAS SAMPLED The first period of sampling was in June 1970, along the flood plain of the Tye River on the anortho- site body near its southern edge (pl. 1, table 1). Six point bars were sampled southeastward along the river to the town of Tye River. From here to Nor- wood, at the confluence of the Tye and the James River, only two small bars were sampled. This seg- ment of the Tye River is mostly in a narrow gorge having very few bars, most of Which are small (pl. 1). The river distance from the southern edge of the anorthosite body to Norwood is 32 km. Sampling was continued downstream along both sides of the James to beyond Scottsville, a river distance of about 53 km, a total distance of 85 km downstream from the anorthosite. Terraces and bars along the James River are many and large (pl. 1). During this period, 185 samples were collected; 177 were taken along the James and Tye Rivers. The remaining eight were collected from shallow holes near.Winter- green in the valley of the South Fork Ro‘ckfish River (pl. 1), 9 to 14 km north of the north end of the anorthosite body and in an entirely different drain- age basin having no streams draining from the anor- thosite. This was done to see if high values of titani- um minerals were present in deposits derived from areas other than the known ilmenite- and rutile— bearing anorthosite body. Seventy-five additional samples were obtained during March and April 1971; these were collected in scattered areas to fill in gaps from the earliest sampling. Of these samples, 10 more were collected near Wintergreen, and six additional, sites were sam- pled along the James River above the confluence with the Tye River, instead of below it, as had been done earlier. This was done to compare detritus part- ly contributed from the Roseland district (via the Tye River) with that which came from the upper James River and was not in any part derived from Roseland. Samples also were collected along the Buffalo River. These included detritus partly derived from the south end of the anorthos1te body and some that had its source entirely outside the body. Surprising- H5 ly high percentages of ilmenite are present in ter- race and bar deposits whose drainage has no con- tribution from the area of the anorthosite body, These areas are along the Buffalo River and Beaver Creek, at and above their junction 2 to 3 km south- west of the south end of the anorthosite body (pl. 1). A value of nearly 8 percent ilmenite was ob_ tained from a bedrock outcrop here also (sample 510, table 1, fig. 3). FIGURE 3.—Photomicrograph of dioritic ilmenite-rich facies of gneiss from the Pedlar Formation of Bloomer and Werner (1955) showing ilmenite (black) trimmed by sphene-ana— tase intergrowths. Plane light, X 20. Sample 510 from location of cross section 9 (table 1, pl. 1). TYE RIVER The Tye and Piney Rivers together drain most of the Roseland Anorthosite terrane, as well as an area of gneiss, some of which is altered granulite (Herz, 1968). No samples were collected from the Piney River valley for this study; 32 samples were collected and analyzed from the Tye River valley. The areal extent of alluvial deposits and the dis- tribution of samples in the Tye River valley are shown on plate 1 and in table 1. The total area of these deposits is 1.6 kmg. There are no alluvial de- posits of appreciable areal extent in the Tye River valley downstream from the junction with Piney River. The samples collected from this area were mostly poorly sorted pebbly sands from small ter- race deposits. MINERALOGY Ilmenite is predominant in the heavy-mineral con- centrates of the Tye River samples (tables 2, 4). Rutile is also present but in varying and much lesser H6 GEOLOGY AND RESOURCES OF TITANIUM TABLE 1,—Locations of samples and cross sections [Hmc, heavy-mineral content] 53:31: 8,331,231" 531:; qu 3.4311421 n gl e Latitude Longitude Remarks 1---_ 478 _______ Tye _________ Horseshoe 37°45’45” 78°59’30” Mountain. 2---- 314-321 --- ___-do ------ Arrington ______ 37°43’ 78°58’52” 3--_- 322—324 --- ___-do ______ ___-do _________ 37°42’50” 78°58’52" 334 was horizontal chan- 4-___ 333—336 --- ___-do ______ ___-do _________ 37°41'50" 78°58'20” ’s‘fiifjc‘glflfggrftfggss 71 percent. 5---- 325—331 _-- ___-do ______ ___-do _________ 37°41’15” 78°57’40” 6_--- 519—520 ___ ___-do ------ ___-do --------- 37°39’ 78°57’23” 7---- 337—340 _-- ___-do ______ Shipman _______ 37°40’45” 78°50’15” 8_--- 341—343 --_ ___-do ------ ___-do --------- 37°40'40” 78°49’05” l509 was sum of 2 hori- zontal channel samples across surface of stream 9---- 506—513 _-- Buffalo ______ Piney River ---- 37°39'05" 79°06’15” Eiésv’vi’é‘isiipifii’ét: rock; Hmc, 8 percent. 511 was horizontal sam- ple; Hmc, 29 percent. 10_-_- 527 ------- ___-do ------ -___do --------- 37°38’05” 79°03’50” 523 and 524 were hori- zontal channel samples egress stream bars. a , n o , n me, 9 percent and 13 11__-- 521—524 ___ ___-do ------ Amherst _______ 37 36 40 79 03 01 percent, respectively. Along Rutledge Creek near its confluence with Buffalo River. 12‘--_- 525—526 --- ___-do ------ ___-do _________ 37°35’05” 79°00’23” 13---- 518 _______ ___-do ------ Buflalo Ridge --- 37°36’42” 78°55’05” 14--_- 479 _______ Rockfish _____ Horseshoe 37°52’15” 78°54’57” 8 482 Mountain. 4 0— ——— 9 r n 0 I H 15-___ { 503_504 ___ ___-do ------ Sherando _______ 37 52 52 7s 53 53 16__-_ 483—484 ___ ___-do ______ Greenfield ------ 37°53'15” 78°52'15” 17__-- 485—486 ___ ___-do ______ ___-do _________ 37°53’37” 78°51’45” 18--_- 514-516 ___ James _______ Gladstone ------ 37°35’08” 78°49’45” 19___- 528-529 ___ ___-do ------ Shipman ------- 37°37’34" 78°49’20" 20‘_-__ 348—364 -_- ---_d0 ______ ___-d0 _________ 37°38’22” 78°48’38” 21-_-- 365—372 --- ___-do ______ Shipman ------- 37°38’40” 78°47’42” 22____ 373—383 --- ----do ------ ___-do _________ 37°38'40” 78°47'25” 23--_- 387—396 ___ ___-d0 ______ ___-do --------- 37°38’25” 78°46’32” 24____ 397—401 ___ ___-d0 ------ ___-do _________ 37°38’15” 78°45’27” 25-_-- 402—410 ___ ___-do ------ Howardsvillew ___ 37°38’15" 78°43’22” 26-__- 436—444 ___ ___-do ______ ___-do _________ 37°39’53” 78°43’10” 427 was horizontal chan- nel sample of loose sur- 27--__ 423—435 --- ___-do ------ __-_do --------- 37°40’06” 78°42’43” face sand on a narrow terrace near the river; Hmc>20 percent. 28_--- 445—457 ___ ___-do ------ ___-do _________ 37°40’25” 78°42’40" 29____ 412-418 --- ___-do ------ ___-do --------- 37°41'15” 78°41'33” 30--_- 466—471 ___ ___-do ------ Esmont ________ 37°45’30" 78°36’10” h - o I n a , Outside the area s own 513.1"- 488—493 ___ ___-do ______ Scottsv111e ------ 37 45 22 78 28 { on plate 1. o e A __-_ 344—347 --_ James _______ Shipman ------- 37°38’30” 78°48'30” B _-_- 384—386 ___ ___-do ______ ___-do --------- 37°38’36" 78°47’08” C__-- 419—422 ___ ___-do ______ Howardsville -- 37°41’30” 78°41’41” D --_- 458—460 --- ___-do ______ ___-do _________ 37°41’40” 78°39'06” E ____ 461—465 ___ ___-do ______ --__do _________ 37°43’52” 78°38’45" F___- 473—477 -_- ___-do ------ Esmont -------- 37°45’50” 78°33’20” . G --__ 494—495 ___ ___-do ______ Scottsvrille ______ 37°45'20" 78°27’43” { 01‘5“!" the area Sh°wn on plate 1. amounts (Herz and others, 1970). In contrast to “il- menite” from many placer mines, this ilmenite has a sharp pattern on an X-ray difl‘ractometer and has a chemical composition (table 3) near that of sto- ichiometric ilmenite. Ilmenite grains commonly are rimmed or veined by white fine-grained material (“leucoxene”) Which is poorly crystalline even to X—ray diffraction; it consists primarily of aluminous ALLUVIAL ILMENITE PLACER DEPOSITS, CENTRAL VIRGINIA TABLE 2,—Mineralogy of heavy (sp gr >33) fractions of samples from the Tye and James Rivers in order of dis- tance downstream [Only 125/4-250/L-size fractions examined. Opaque minerals were separated magnetically; ilmenite separates were verified by X-ray difl’ractometer. H7 HEAVY-MINERAL CONTENTS Plate 1 and table 1 show the distribution and tenor of heavy minerals in alluvial deposits in the Tye A, abundant (>10 percent); C, common (1—10 percent); P, present River vaHGY- several depOSibS appear to have mln' (<1 ”mm” eral concentrations high enough to be of economlc Sample No. interest. Volume of the deposits, however, is limited 317 339 374 426 (table 4) Drainage _______ Tye. Tye. James James , Rlver. River. River. Rlver. TABLE 4.—Approximate ilmenite resources of some terrace Cross_sect10n .---- 2 7 22 2'7 deposits in the study area Ilmemte (havmg' 76 69 76 65 “leucoxene” rims) Ilm,e- (percent) . Approxi- Aver- nite Magnetite P 1 3 10 Cross $3.3; SEE. €23; 13$? (percent) ' Drainage sectiorlis of of of re- “Leucoxene” _ _ _ _ A A C C (river) 1:31:13 9:5 terrace heavy con- sources Rutfle __________ C P P P (see pl. 1) allu- mm- can- (10* ' vmm erals trate tonnes) Epldote _________ C C C A (ma) (per- (pep. Kyanlte ________ C P C P cent) cent) Upper Tye __ 2,3 2><106 3.5 80 10 . . . Lower Tye __ 7, 8, A 1 X 10" 2.0 70 3 TABLE 3,—T102 content of selected rlmemte separates Buffalo _____ 9 1 x 106 3.0 80 7 F t 0.15—0.35 f t‘ f th 1 ‘ d‘d t t f x 6 [ Edi—2250,. size fragdltlfimsragnlelsse: $13.15;? xiii} 3.?“5‘2315331 gs: James ------ (5313113193 3 1° 2'5 65 1° M] 423-432) Percent Sample No. Drainage TiOz in ilmenite 323 ________ Tye River _____________ 51.7 343 ________ ____do _________________ 51.3 372 ________ James River ____________ 47.8 467 ________ ____do _________________ 45.6 481 ________ Rockfish River _________ 48.9 513 ________ Bufl'alo River __________ 52.0 sphene having subordinate anatase (M. L. Bird, oral commun, 1972). Cores of ilmenite grains are crys- talline and probably are relatively unaltered. Rims of grains commonly are broken, abraded, or both, indicating that they formed before transport. Mod- ern stream sediments here also contain rimmed il- menite. Ross (1941, pls. 18, 19) shows rims of “sphene leucoxene” around ilmenite grains in fresh specimens of Roseland Anorthosite and nelsonite; rims also occur on ilmenite grains in some gneiss from the Pedlar Formation and other hypersthene- bearing rocks (fig. 3). Herz (1968, p. 365) regards these rims as the result of Paleozoic retrograde metamorphism. Blue (rutilated; Ross, 1941) quartz is present in retrograded gneiss from the Pedlar Formation and in Roseland Anorthosite. It appears to be particular- ly abundant in the alluvial samples in which titani- um minerals also are abundant and is believed to have been derived from the same sources. Locally, it makes a helpful prospecting tool for titanium min- erals. Epidote is present in those samples gathered farthest downstream (va 341—3, 337—40) and indi- cates dilution by tributaries draining greenstone of the Catoctin Formation. Hurricane Camille, in August 1969, caused heavy damage in the Tye River valley and left flood-plain sand deposits that are markedly enriched in heavy minerals, compared with other Tye River valley de- posits. Although the cause of the enrichment is not definitely known, the enrichment probably resulted from stripping of heavy-mineral-rich saprolites from the source rocks and erosion, reworking, and concentration of older Tye River valley deposits. All these processes were on a large scale (Virginia Di- vision of Mineral Resources, 1969; Williams and Guy, 1973) . Samples from the Tye River valley show that some relationships exist among heavy-mineral concentra- tions, grain-size distribution, and distance down- stream from source. GRAIN SIZE Grain sizes of most samples are shown in table 5. Figure 4 shows size-distribution histograms of dif- ferent, but typical, samples that show several char- acteristics. Heavy minerals (specific gravity >33) commonly are finer than the mode of the entire sam- ple. Samples having the highest heavy-mineral con- centrations have modes in the medium- to coarsea sand range. Sorting is variable, and Trask sorting coefficients range from about 2 to 8 (precise values cannot be calculated because of the crudeness of the size sepa- ration). The heavy-mineral content seems to be weakly related to the sorting of the deposits (fig. H8 GEOLOGY AND RESOURCES OF TITANIUM 18-— 16— Sample No. 320 Tye River 14— 12— 10— — 3_ 6-— 4— 2_ 34— EXPLANATION 32— Heavy minerals (Sp. gr. > 3.3) Light and heavy minerals 30— 28— 26— 24— 22— 20— 18— 16— ' Sample No. 328 14— Tye River 12- 10— 8— 6— PERCENT OCCURRENCE (by weight) 4— 22m 20— 18— 16* Sample No. 336 Tye River | 2 1 .5 .25 .125 .06 GRAIN SIZE (mm) lgl Sample No. 383 James River lgl Sample No. 424 James River Sample No. 423 James River I 2 l 1 I .5 .25 .125 .06 FIGURE 4.—Histograms showing grain sizes of some typical alluvial samples and their heavy minerals from Tye River and James River deposits. Gravel contents have arbitrarily been divided into two grades and mud contents into six grades. ALLUVIAL ILMENITE PLACER DEPOSITS, CENTRAL VIRGINIA 10 I I I I I I I o 9 — _ ° Sample Nos. 314—324 (less 8 _ those not analyzed—Table 5) _ Upper Tye River 7 - _ o 6 — _ o 5 — _ 3: to >. o 0 n v 3 _ O O o _ ‘2 o E .2. 2 ‘ o ‘ E >- 1 F — > 4: Lu I 0 l I I I I I I P— z 8 E 6 I T I I I n. 5 —- ._ Sample Nos. 423—456 (less 4 _ those not analyzed—Table 5) _ James River 0 3 —- o _ ° 0 o 2 _ 8 _ O o 1 _ 0°C 0 _ o o o 0 o l J l | l I l 0 10 20 3O 4O 50 6O 70 80 PERCENT MUD (by weight) FIGURE 5.—Heavy-mineral and mud contents of some Tye River and James River samples. In each diagram, all sam- ples are from one terrace. 5) ; however, this may be no more than a diluting effect of the amount of mud in the sample. Within a sample series from the same auger hole, the percent- age of heavy minerals commonly increases as depth increases. This is apparently a consequence of lesser mud content at depth. H9 TRANSPORT DISTANCE Among samples having a given modal grain size and mud content, heavy-mineral concentration is an inverse function of distance from the source. Figure 6 shows the downstream decrease in heavy-mineral Confluence of James River and Confluence of Tye River and: Piney Buffalo James Rockfish River River River River 10 I I I I I o 9 .— o 8 — A, All samples " 7 — _ o 6 — 0° — o 5 — _ 3 00 .E .20 4 — e — “g o 3 ‘6 o o 0 V 3 _ 83° 0 o _ 3 Q) g 0 00 0 % Lu 2 _ o O €00 E O O _ E 8% o o 2 ° c‘0) o 1— 0 ° 0 0 —° E 06) 8 CE g 0 Lu 0 C8 0 o I 0 I I I I I I I .— Z 8 6 n: 0 Lu | l I l n. 5 — _ 4 — B, Muddy sands with — .06—.125 mm mode 3 — g o _ o o 2 ‘ 8 ° m — (Sb 0 ° 0 1 _ o <9 _ o o o I I I I I I I 0 10 20 30 40 50 6O 70 DISTANCE DOWNSTREAM FROM SAMPLE 314 (km) FIGURE 6.—Decrease in heavy-mineral content of Tye River and James River deposits as a function of increasing dis- tance downstream from sample 314 (cross section 2). H10 GEOLOGY AND RESOURCES OF TITANIUM TABLE 5.—Grain sizes (as determined by sieving) and heavy-mineral data for alluvial samples [All percentages are rounded to the nearest whole number. Ilmenite defined by magnetic properties. A, averaged on plate 1; 'I, questionable mode; n.d.. not determined] Per- P t P t P t Modal (16m: 55;; ercen ercen ercen i me- - Sample 22?: 3t sand mud . Nigdall heavy interval nite 1:11:11;- No. (>2 mm) (0.06 mm- (<0.06 "(1 "f minerals heavy in in 2.0 mm) mm) mm (sp gr >3.3) minerals heavy total (mm) min- 1 erals samp e 0 47 53 0.125—0.25 3.0 'I n.d. __ 0 44 56 .06—.125 2.6 '! n.d. _- 1 50 49 .125—.25 2.9 '! n.d. __ 21 51 28 .125-.25 5.4 012540.25 76 4.1 43 37 20 1’ 3.5 .125—.25 n.d. __ 11 58 31 .125-.25 9.3 .125—.25 86 8 23 56 21 .5—1 8.6 .25—.50 n.d. _- 30 54 16 .5—1 6.6 .25-.50 n.d. __ 22 52 26 .25-.5 3.0 'l 80 2.4 4 65 31 .2~.5 3.4 ? 82 2.8 28 61 11 .2—.5 1.9 ‘l 82 1 6 2 62 36 2- 5 3.3 'I 77 2 5 2 77 21 25* 5 1.7 7 n.d. __ 2 85 13 25- 5 4.2 .125v.25 n.d. .. 16 63 21 25—.5 4.4 .125—.25 n.d. __ 7 48 45 7 1.0 " n.d. 3 70 27 2- 5 4.2 'I 78 3 3 22 59 19 2— 5 6.0 '.’ n.d. __ 3 60 37 .25—.5 4.4 .125—.25 n.d. __ 0 96 4 .25—.5 71.0 .125—.25 n.d. _- 2 64 34 .125—.25 5.9 .125—.25 n.d. __ 3 51 46 .125—.25 3.0 .125—.25 n.d. __ 0 59 41 .125—.25 1.9 .125—.25 n.d. __ 0 69 31 .125—.25 1.4 '.’ n.d _ 5 59 36 .125—.25 2.0 ? 69 1 4 5 65 30 .125-.25 2.0 T n.d. __ 0 48 52 ? 1.9 T 71 1 3 0 43 57 9 1.7 '1 70 1 2 7 49 44 7 1.7 T 73 1 2 0 45 55 .125-.25 2.0 7 n.d. —- 0 43 57 .125—.25 1.7 7 n.d. —— 0 38 62 .125—.25 1.1 n.d. - 19 59 22 .5-1.0 3.7 .125-.25 n.d. -- 0 38 62 7 .9 T 67 6 Muddy; not analyzed 3 25 72 7 6 7 65 4 Muddy; not analyzed 0 43 57 ? 1.0 '1 64 6 0 33 67 '! 8 T n.d. -— 13 31 56 T 8 7 n.d. __ Muddy; not analyzed 0 31 69 7 .4 7 n.d. __ 2 28 70 9 .5 " n.d. -— Muddy; not analyzed 5 30 65 " .7 'l n.d __ 11 37 52 ? .8 " n.d __ Muddy; not analyzed 36 63 7 .9 '7 n d -— Muddy; not analyzed 0 54 46 .2—.5 1.6 " n.d __ 22 47 31 .2—.5 2.4 7 n.d. __ 0 43 5‘7 .125—.25 2.2 n.d. —- 0 79 21 .25~.50 3.2 .125—.25 76 2 4 0 57 43 .25—.50 1.6 n.d. -— Muddy; not analyzed 33 32 35 .125—.25 .8 7 n,d __ Muddy; not analyzed 6 49 45 .125A.25 1.9 7 ,d. _- 29 36 35 Several modes 1.6 .125‘25 n.d. __ Muddy; not analyzed 0 79 21 .2—.5 1.7 1' n.d. -— 17 72 11 .2-.5 1.7 7 n.d. -— Muddy; not analyzed 0 56 44 __________ 1.5 ’ n.d. -— 1 70 29 .2—.5 1.9 " n.d. _- Muddy: not analyzed 35 62 __________ 9 " n.d -_ Muddy; not analyzed 1 38 61 .125—.5 .4 '1 n.d. __ Muddy; not analyzed 3 73 24 .2—.5 2.9 7 n.d. -- Muddy; not analyzed 3 32 65 ? 1.0 T n.d. -_ Muddy; not analyzed 403 ___________ 0 65 ‘35 .2—.5 1.9 7 n.d. __ 404 ___________ 10 75 15 .2—.5 2.2 " n.d. ALLUVIAL ILMENITE PLACER DEPOSITS, CENTRAL VIRGINIA H11 TABLE 5.—Gram szzes (as determmed by szevmg) and heavy-mineral data for alluvuzl samples—Contmu‘ed Per; Per- Percent Percent Percent Modal 11:; cent Sample P613531: sand mud . Néodall heavy interval nite 11:}:- No. (£2 mm) (0.06 mm- (<0.06 "(1 9”? minerals heavy in in 2.0 mm) mm) mm (sp gr >3.3) minerals heavy total (mm) min- sample erals 26 60 14 0.2-0.5 1.5 '! n.d. -_ 06 __ Not analyzed 407, 408 _________ Muddy; not analyzed 2 46 52 .2—.5 1.0 ? n.d. -- 32 41 27 .5-.1 1.0 '.’ n.d. __ Not analyzed 0 43 57 7 .9 ? n.d. _- 9 42 49 .2—.5 .8 ? n.d. -- 415—418 Muddy; not analyzed 419, 420 Muddy; not analyzed 421 --- 2 38 60 7 .7 ? n.d. .. 422 — 16 28 56 ? .5 7 n.d. -— 423 ----------- 0 47 53 .125—.25 1.1 ? n.d. -— 0 73 27 .25—.5 2.5 .125—.25 n.d. -- 0 77 23 .25—.5 2.1 .125—.25 n.d. .— 6 78 16 .25—.5 3.4 .125—.25 65 2.2 Not analyzed Not analyzed 0 77 23 .2—.5 2.2 ? .d. -— 2 80 18 .2—.5 3.0 ? n.d. -— 5 80 15 .2~.5 2.6 ? n.d. -- Muddy; not analyzed 24 53 23 .5—1 .5 '1‘ n.d. .- 436—438 _________ Muddy; not analyzed A 1439 -- 10 51 39 .2—.5 2.3 '! n.d. -— 1440 -- 16 45 39 2....5 1.5 T n.d. -— 441. 442 Muddy; not analyzed 24 44 32 .2—.5 .8 '! n.d. -— 23 47 30 .06-.2 .8 ? n.d. -- Muddy; not analyzed 0 54 46 ? .9 'I n.d. -— 3 65 32 .2A.5 1.3 7 n.d. __ Muddy: not analyzed 4 47 49 ? 1.2 'l n.d. -— 7 43 50 '.’ .9 '! n.d. -- 453 ----------- Muddy; not analyzed 464 0 23 77 7 .4 7 n.d. -- Muddy; not analyzed 0 53 47 Y .5 7 n.d. __ 5 66 29 .2—.5 .8 ? n.d. -- 458, 459 ......... Muddy; not analyzed 460 ---- -- 48 21 31 '! .2 ? n.d. -_ 461-463 Muddy; not analyzed 464 -- o 34 66 2 .6 7 n.d. _- 24 45 31 ? 2.0 '! n.d. .. Muddy; not analyzed 3 84 63 ? 1.0 T 65 .65 Muddy; not analyzed 30 40 30 Y 1.2 ? n.d. .- Muddy; not analyzed 7 40 53 ? 1.1 ? n.d. -- 5 28 67 .2~.5 1.4 ? n.d. .. Error in analysis 2 37 61 '! .7 '! 77 .54 34 36 30 .2-.5 2.4 'I 75 1.8 Error in analysis 55 23 22 Several modes .8 ? n.d. __ Error in analysis 49 33 18 Several modes 2.2 ? n.d. ._ 34 25 41 7 1.4 'I n.d. -- Muddy; not analyzed 18 22 60 ? .4 ? n.d. -_ Muddy: not analyzed Not analyzed 9 66 25 .2—.5 1.9 7 n.d. __ Muddy; not analyzed 2 46 52 7 1.1 ? n.d. -- 4 35 61 7 1.8 7 n.d. __ 7 54 39 .2—.5 3.2 ? 74 2 4 56 39 5 ? 1.9 7 84 1.6 Error in analysis 6 43 51 .2—.5 3.3 '.’ n.d. __ 13 43 44 .2—.5 3.4 ? n.d. __ 2 38 60 '! 2.5 ? n.d. __ Muddy; not analyzed 19 35 46 .2-.5 1.2 7 n.d. __ 39 33 .2—.5 4.1 'I n.d. 28 Muddy; not analyzed H12 GEOLOGY AND RESOURCES 0F TITANIUM TABLE 5.—Grain sizes (as determined by sieving) and heavy-mineral data for alluvial samples—Continued Per— t Per- Percent Percent Percent Modal llfrli‘e- _cent Sample Percerit sand mud . L§°dall heavy interval nite 11:11,: No. (game ) (0.06 mm- (<0.06 "(1 €er minerals heavy in in mm 2.0 mm) mm) mm (sp gr >3.3) minerals heavy total (mm) mm- sample erals Not analyzed 8 38 54 0.2—0.5 2.9 7 n.d -- 1 93 6 2—.5 25.0 ? n.d. __ rock 8.0 ? n.d.. __ 0 40 60 T 1.2 ? n.d. _- 31 64 5 .2—.5 15.4 ? n.d. _- 40 57 3 .2‘.5 29.0 '! n.d. __ 3 43 54 ? 3.1 ? n.d. __ 18 68 14 .5—1 20.6 .2».5 n.d.. __ 7 73 20 .2—.5 4.4 ‘I 85 3 7 Muddy: not analyzed Error in analysis 515A, 5158 ________ Muddy; not analyzed. 5150 nu __ Error in analysis 515D 23 44 33 .27.5 1.9 '! n.d. __ 516A—5160 Muddy; not analyzed 51 _ 25 30 45 ‘I 1.9 “.7 n.d. __ 518A ____ Muddy; not analyzed 518B ____ 12 50 38 .2—.5 1.8 '.' n.d. __ 519A, 5198 Muddy; not analyzed 1 C ____ Not analyzed 520A, 5203 Muddy; not analyzed 23 43 34 " 2 4 ? n d __ Muddy; not analyzed Muddy; not analyzed 2 50 48 '! 3.5 ? .d. __ 28 43 29 'l 2.3 ? n.d. __ 1 95 4 27.5 9.0 ? n.d. __ 1 94 5 2—.5 13.0 '1 n.d. __ 525A, 5253 ______ Not analyzed 526A, 5268 _ Not analyzed 0 20 80 ? .8 n. __ 0 51 49 .2—.5 3.7 " n.d __ 5 74 21 .2~.5 8.7 " n. __ 15 57 28 2—.5 8.0 " n.d. -- Muddy; not analyzed Not analyzed 1 74 . 25 .2—.5 1.5 " n.d __ 12 75 13 .5—1 1.1 " n.d __ Muddy; not analyzed 39 1 S 43 '! .5 ‘7 n.d _ _ 1 Channel sample content in samples having nearly the same grain size characteristics. A few of the samples shown are from the James River, downstream from the mouth of the Tye River. The decrease probably is due pri- marily to dilution. Among the entering tributaries are the Piney River, having 18 percent of the drain- age area of the entire Tye River system; Brown Creek, having 4 percent; the Buffalo River, having 36 percent; and Rucker Run, having 9 percent. The James River, at its junction with the Tye, has 11 times the drainage area of the Tye River. Of these streams, other than the upper Tye River, only the Piney and Buffalo Rivers drain anorthosite, and the Buffalo drains only a minor area of it. Clearly, dilu- tion does not appear to be as rapid as would be ex— pected if the anorthosite were the only major source of heavy minerals (predominantly ilmenite). There- fore, ilmenite probably is being contributed from other sources. In the following descriptions of drain- age basins that contain no anorthosite, the contribu- tion of heavy minerals by the gneisses of the Pedlar Formation are discussed. BUFFALO RIVER AND SOUTH FORK ROCKFISH RIVER Nineteen samples were collected along the Buffalo River, a tributary to the Tye River (pl. 1). The up- stream samples (506—513) are from a part of the stream that drains no anorthosite. Gneisses, includ- ing those from the Pedlar Formation, are the pre— dominant rocks in all the drainage areas. Valley de- posits consist primarily of pebbly sands. The heavy- mineral concentrates from sediments sampled in the Buffalo River drainage basin, consist almost entirely of ilmenite (table 5), with some magnetite and zir- con. Ilmenite is commonly rimmed with “leucoxene” as in the samples from the Tye River drainage area. Plate 1 and table 5 show the distribution and grade of heavy minerals within the deposits of the Buffalo River valley. Generally, the percentage of heavy min- ALLUVIAL ILMENITE PLACER DEPOSITS, CENTRAL VIRGINIA erals is quite high. No relationship of heavy-mineral content to sorting was noted here. The highest group of values in samples 506—513 is believed to be re- lated to influx from Beaver Creek; sediment present- ly being transported in the headwaters of Beaver Creek is black because of the high ilmenite content. The source rock here is the Pedlar Formation, which in this area contains abundant ilmenite rimmed with “leucoxene” (fig. 3) ; two analyzed samples of bed- rock each show 8 percent ilmenite. The South Fork Rockfish River drains a small area of high relief north and east of the Roseland Anor- thosite (pl. 1). No anorthosite is known to occur in the drainage area; the source rocks are gneiss, in- cluding gneiss of the Pedlar Formation, and the greenstone in the Catoctin Formation. Valley-bottom deposits are predominantly cobbly sands. Nine sam- ples were analyzed from this area; heavy-mineral concentrates from these samples consist mostly of ilmenite (table 5). However, epidote is abundant in most concentrates; it was probably derived from weathering of greenstone. Ilmenite is rimmed by “leucoxene” as it is in the Tye River valley deposits. Plate 1 and table 5 show the distribution and grade of heavy minerals within alluvial deposits of the South Fork Rockfish River. A few values are mod- erately high, but none is of particular economic in- terest when the cobbly nature and small volume of the sediment are considered. No significant relation- ship of heavy-mineral content to sorting was noted. Nowhere in the drainage basins of the Buffalo River and South Fork Rockfish River are the alluvial deposits of sufficient volume to form an economically attractive deposit (table 4). The sediments are of interest in this study, however, because they are an example of ilmenite entering the system from non- anorthosite source areas. The conclusion reached in analyses of the Tye River samples, that the anor- thosite is not the only major source of alluvial il- menite in the area, is supported by samples taken from these streams. Gneiss of the Pedlar Formation and other hypersthene—bearing gneiss also appear to be significant sources of ilmenite in the study area. JAMES RIVER All the streams previously discussed are tribu- taries to the James River. By the time some of the waters of the James have arrived in the study area, they have passed through the Valley and Ridge and Blue Ridge provinces and have entered the Piedmont. Alluvial deposits along the James River are of much larger volume than those found along the other H13 rivers, as discussed earlier in this report. Terrace flood-plain widths of 800 m and thicknesses of 5 m are common. Samples were collected over a river distance of 58 km, of which 6 km are upstream from the mouth of the Tye River. Sixty-nine samples from the James River were analyzed. As is true in the samples previously discussed, i1- menite having rims of “leucoxene” is the predomi- nant heavy mineral (tables 2, 4). Average TiO2 con- tent of ilmenite is lower than in the Tye River sam- ples, probably because of the addition of ilmenite having iron oxide intergrowths (table 3). Also pres- ent as important constituents are magnetite and epi- dote. Biotite and frosted quartz grains are conspicu- ous among the light minerals. The presence of blue quartz in a sample rich in ilmenite was observed in only one sample (347), and this was from near the mouth of the Tye River, in which blue quartz is com- mon in samples rich in ilmenite. Plate 1 and table 5 show the distribution and grade of heavy minerals within alluvium along the James River. In general, concentrations of heavy minerals are less than 2.5 percent. Increase in heavy- mineral concentrations as depth below the surface of the ground increases, is evident at several sample sites. This in turn is again dependent on the grain- size distribution of the sample; the basal deposits are coarser, and the coarser fractions have higher con- centrations of heavy minerals. A few check samples were collected from bluffs along the James River downstream from Scottsville. These samples also show the lower average heavy- mineral values (table 6) characteristic of increased distance downstream. Ilmenite again is the predomi- nant heavy mineral; rutile is minor in all samples except those from farthest downstream near Gooch- land, where it is also present in the local gneiss bedrock. CONCLUSIONS 1. Small deposits of alluvium in the. Tye River drainage basin have high percentages of il- menite, Whereas nearby large deposits along the James River have low percentages. High- est values occur in coarse mud-free deposits which most often were at the base of the al- luVial deposits sampled during this study. Dilution occurs in a downstream direction and decreases both the proportion of the heavy- mineral concentrate in the samples and the value of the heavy-mineral concentrate by ad- mixing magnetite and epidote. Exceptionally H14 TABLE 6.—Heany-mineral content of alluviu'm exposed in James R1126?“ blufls downstream from Scottsville [Separation from channel samples 1-3 m long. C, common; P, present; N, not detected.] . Percent Dis- heavy- tance mineral 32:22.; 3...]... sag; Raffle from 63.01“" content. con- Scotts- 10“ pre— centrate VLlle domi- (km) nantly ilmenite 23.1___-Levee deposit (sand) _______ 1.9 P 24.4 ________ do _____________________ 1.0 P 26.8----Flood—plain deposit .3 N (silty sand). 28.1____Flood-plain(?) deposit .2 P (silty sand). 33.0----Levee deposit (silty sand) ___ .3 P 37.0____;Flood-plain deposit (sand)___ 2.3 N 39.4__.__Levee(?) deposit (sand) ___- .8 N 42.5----Flood-plain(?) deposit .9 N (silty sand). 44.3----Flood-plain deposit (sand) __ 1.4 P 45.9----Flood—plain deposit (silty 1.1 P sand). 48.9----Flood-plain(?) deposit .5 N (silty sand). 52.6----Flood-plain(?) deposit .4 N (silty sand). 53.5----Flood-plain(?) deposit .5 P (sand). 54.7____Flood-plain deposit (sand) __ 1.8 P 59.7----Flood-plain deposit (sand) __ .7 P 62.1__,__Flood-plain(?) deposit .3 N (silty sand). 69.2----Flood-plain deposit 1.3 N (silty sand). 71.6----Flood—plain(?) deposit .5 C (silty sand). 79.3----Flood—plain deposit (sand) __ 1.5 C 81.5____Flood-plain deposit .7 C (silty sand) . high values of titanium minerals shown in horizontal channel samples across present stream bars suggest that such deposits may be buried and easily missed by sparse sampling, but may be readily recoverable in standard mining operations. 2. As this study began, our belief was that the Rose- land Anorthosite was the only major source for high-grade alluvial placer deposits of i1- menite in the area. However, when we discov— ered that hypersthene-bearing gneisses also were important contributors, the existence of additional ilmenite placers in the Blue Ridge seemed possible. Figure 7 shows the location of the source rocks in and near the study area. 3. As redefined by this study, the area of the source: rocks favorable for formation of ilmenite plac- ers is at least 16.4 percent (746 kmz) of the James River drainage at the downstream end of the study area. At Richmond, where the James River enters the Coastal Plain province, its drainage basin is 1.48 times as large as it is GEOLOGY AND RESOURCES OF TITANIUM 79° 00’ l TYE RIVER /’\/) i» 37° 45’ 0 10 MlLES o 10 KILOMETRES E X PLANATI ON Subdiv—QiR-Jfi; River drainage basin E Hypersthene-bearing gneisses Tye River drainage basin Roseland Anorthosite FIGURE 7.—Sketch map showing location of source rocks of ilmenite in the Tye River drainage basin. Rock boundaries generalized from Bloomer and Werner (1955) and Herz (1968). in the study area. It seems plausible that titani- um minerals may be present in economic con- centrations in some of the Coastal Plain sediments. Long-range plans include a sampling program that continues downstream along the James River and into the Coastal Plain. Sampling traverses will be made across the inner Coastal Plain to explore and attempt to locate possible areas where ilmenite placers may have been concentrated in tidal deltas or bars, or in beach ridges. REFERENCES CITED Bloomer, R. 0., and Werner, H. J., 1955, Geology of the Blue Ridge region in central Virginia: Geol. Soc. America Bull., v. 66, no. 5, p. 579—606. ALLUVIAL ILMENITE PLACER DEPOSITS, CENTRAL VIRGINIA Fish, G. E., Jr., 1962, Titanium resources of Nelson and Amherst Counties, Virginia [Pt.] 1, Saprolite ores: U.S. Bur. Mines Rept. Inv. 6094, 44 p. Fish, G. E., Jr., and Swanson, V. E., 1964, Titanium re- sources of Nelson and Amherst Counties, Virginia [Pt.] 2, Nelsonite: U.S. Bur. Mines Rept. Inv. 6429, 25 p. Herz, Norman, 1968, The Roseland alkalic anorthosite massif, Virginia, in Isachsen, Y. W., ed., The origin of anorthosite and related rocks: New York State Mus. and Sci. Serv- ice Mem. 18, p. 357~367 [1969]. Herz, Norman, Valentine, L. D., and Iberall, E. R., 1970, Rutile and ilmenite placer deposits, Roseland district, Nelson and Amherst Counties, Virginia: U.S. Geol. Sur- vey Bull. 1312—F, 19 p. H15 Ross, C. S., 1941, Occurrence and origin of the titanium de- posits of Nelson and Amherst Counties, Virginia: U.S. Geol. Survey Prof. Paper 198, 59 p. Virginia Division of Mineral Resources, 1969, Natural fea- tures caused by a catastrophic storm in Nelson and Am- herst Counties, Virginia: Virginia Minerals, Spec. Issue, 20 p. Watson, T. L., and Taber, Stephen, 1913, Geology of the titanium and apatite deposits of Virginia: Virginia Geo]. Survey Bull. 3—A, 308 p. Williams, G. P., and Guy, H. P., 1973, Erosional and deposi- tional aspects of Hurricane Camille in Virginia, 1969: U.S. Geol. Survey Prof. Paper 804, 80 p. T UNITED STATES DEPARTMEI\T OF THE INTERIOR - PROFESSIONAL PAPER 959—H GEOLOGICAL SURVEY PLATE 1 37°5759' 00 55, _ 50' 45' 40’ 35' 78°3Q’ EXPLANATION ’ ' ; ‘ " ; 37 55 MAP Traverses Location 1—8 Tye River valley 9—13 Buffalo River valley 14—17 Upper Rockfish River valley 18—30 James River valley Cross section 31 and auger hole G are along the James River east of the area shown. They can be located using the coor— dinates given in table 1 11 Traverse. Number indicates associated cross section (9 Auger hole. Letter indicates log Roseland Anorthosite body (after Herz and others, 1970, pl. 1 V ‘) . g9} £3 / CROSS SECTIONS * — —: Clay—silt ' .-_' Sand 322:: Gravel 50, 50, ’ Schist pebble /'/ Bedrock Log of hole. Number indicates percent of heavy minerals by }1_9 weight $0 r‘fi’ "3 Auger sample number. Some samples that are listed in table 1 L/ are not shown in the cross sections. The reasons for these omissions are given in remarks column of table 1 Column Cross section (vert. exag. 10xsca1e of cross (vert. exag. 2xhor. scale ~ section and 200>0 1‘ . Mt Vince. l .5 0 1 2 3 4 5 KILOMETRES r—I H H I—I H I—————1 If . CONTOUR INTERVAL 20 OR 50 FEET DATUM Is MEAN SEA LEVEL V n \, . . . . c . H 79°07'30" 5' 79°00 55' 50' 4 ; , , , . " * ' _ 37°35' I te ~G l - A 7 Base from parts of US. Geological Survey 15’ Quadrangles 5 4O 35 n nor 80 oglcal Survey; Reston; Va. 1976 676099 78°30, Lovingston, Va.; 1943; Covesville; Va.. 1935; Amherst, Va.; 1950; Shipman, Va., 1961; and Buckingham; Va.; 1961 a? O .x‘ a? Q' / 2O MAP OF THE CENTRAL VIRGINIA RUTILE DISTRICT SHOWING SITES AUGERED, LOCATIONS AND CROSS SECTIONS OF AUGER TRAVERSES, AND LOGS OF HOLES WITH PERCENT OF HEAVY MINERALS