Document 3Q4K6BaEKJopbMb1RbmQ3NQYn

Canadian Mineralogist Vol. 18, pp. 101-107 (1980) PLAINTIFFS EXHIBIT WCD-221 CHARACTERISTICS OF A POTASSIAN WINCHITE - ASBESTOS FROM THE ALLAMOORE TALC DISTRICT, TEXAS ANN G. WYLIE Department of Geology, University of Maryland, College Park, Maryland 20742, U.S.A. CHARLES W. HUGGINS US. Bureau of Mines, Avondale Metallurgy Research Center, 4900 LaSalle Road, Avondale, Maryland 20782, US.A. Abstract Asbestos from the Diablo prospect, Allamoore talc district (Texas) has cell dimensions a 9.944(5), b 17.951(6), c 5.271(4) , P 104.41(5) (powder-diffraction data); the data are consistent with space groups C2, Cm or C2/m. Chemical anal yses indicate a potassian winchite-asbestos, (K0.43 N a o .0 2 ) ( N a ^ js C a n .n u M g o .z o ) ( M g 4 .8 2 A lo .1 1 F c 3 0.07) (Si^oALo.) 021.118 (OH1.88F0 .8i). Low indices of re fraction, no' = 1.576-1.596, ny' = 1.590-1.600, are probably related to the presence of fluorine and the low Ca and Fe contents. EDAX indicates that their variability is related to Na, K, Ca and Mg con tents. In composition, this asbestos probably ranges from a potassian richterite to a potassian winchite. The asbestos habit is here well developed: fibres are composed of fibrils 200--400 wide, oriented with c axes parallel, but with no preferred orienta tion of a and b. This fibril-bundle structure explains the optical properties of parallel extinction and anomalous interference figures. This asbestos is con sidered a product of low-temperature metasomatism of a siliceous dolomite (Rohrbacher 1973). The chemical and physical characteristics described here are consistent with this interpretation. Keywords: Potassian winchite-asbestos, Allamoore district, Texas, potassian richterite, fibril-bundle structure, low-temperature metasomatism, siliceous dolomite. Sommaire L'asbeste de Diablo, district de table d'Allamoore (Texas) a la maille suivante: a 9.944(5), b 17.951 (6), c 5.271(4) , P 104.41(5); son groupe spatial est C2, Cm ou C2/m. L'analyse chimique indique une winchite potassique asbestiforme, (Ko. Nao.02) (Na1.12Ca0.6sMg0.20) (Mg4 .82Alo.nFe 0.07) (Si7.oeAlo.01) 0 2 1 .8 8 (OHi.asFo.si). Les faibles va leurs des indices de rfraction, no' 1.576--1.596, nt' 1.590-1.600, reflte probablement la prsen ce du fluor et les basses teneurs en Ca et Fe; la variabilit des indices dpend des teneurs en Na, K, Ca et Mg. En composition, cette asbeste varie pro bablement d'une richtrite potassique une win chite potassique. Le facies typique de l'asbcste est ici bien dvelopp: les fibres consistent en fibrilles, larges de 200 400 avec c parallles, mais a et b quelconques. La structure en faisceau de fibrilles explique l'extinction parallle des fibres et les figures d'interfrence anomales. Rohrbacher (1973) consi dre l'asbeste comme produit du mtasomatisme de basse temprature d'une dolomie siliceuse; les caractres chimiques et physiques de l'asbeste de Diablo sont compatibles avec cette interprtation. (Traduit par la Rdaction) Mots-cls: winchite potassique asbestiforme, district d'Allamoore, Texas, richtrite potassique, fibrilles en faisceau, mtasomatisme de basse temprature, do lomie siliceuse. Introduction In 1971 a new deposit of amphibole-asbestos was discovered in the Allamoore talc district, Sierra Diablo, Texas. The asbestos, described by Rohrbacher (1973) as a potassian richterite, oc curs interlayered with talc in beds 1.2 to 3 m thick in dolomitic rocks of the Precambrian Allamoore Formation; the regional geology has been described by King & Flawn (1953) and King (1965). The asbestos content of the talc beds varies from a trace to about 75%. The occurrence of minor quantities of crocidolite is also noted at this locality. The origin of the asbestos and talc is attributed to low-temperature metasomatism of a siliceous dolomite (Rohr bacher 1973). The asbestos is light grey; individual fibres longer than five centimetres are common, and the fibres are of sufficient flexibility to be woven easily. Microscopic examination reveals that the ends of some of the fibres have been altered to talc, forming the pseudomorphic "fibrous talc" similar to that associated with tremolite in the Gouverneur district, New York (Stemple & Brindley 1960). Small amounts of quartz, dolo- 101 102 THE CANADIAN MINERALOGIST Crystallography and M orphology Fig. 1. Zero-level Weissenberg photograph of a potassian winchite fibre. The axis of rotation of the fibre is the c crystallographic axis. Ni filtered Cu radiation. 30 kV, 20 mA. mite and opaque minerals also are distributed throughout the samples but represent < 2 % of their total weight. Unit-cell dimensions were determined by leastsquares refinement of 41 reflections from the powder-diffraction record (Table 1) using the computer program of Finger (1969). Table 1 containing the indexed diffraction data is on file with the Depository of Unpublished Data, Na tional Science Library, National Research Coun cil of Canada, Ottawa. These dimensions are: a 9.944(5), b 17.951(6), c 5.271(4) A, /? 104.41 (5). The data are consistent with space groups C2/m, C2 and Cm. A zero-level Weissenberg photograph taken with Ni-filtered Cu radiation produced at 30 kV and 20 mA is shown in Figure 1. The fibre studied is approximately 0.08 mm in width and 2 mm in length. The c crystallographic axis parallel to the fibre length was the axis of rota tion. The presence of lines indicates that the fibre is composed of a large number of indivi dual crystals, all of which have their c axes parallel but which are randomly oriented with respect to the directions of their a and b axes. Fig. 2. Scanning electron-micrograph of a potassian winchite fibre. POTASSIAN WINCHITE-ASBESTOS FROM TEXAS 103 Fig. 3. Transmission electron-micrograph showing potassian winchite fibrils. The ribbon-shaped fibrils are between 2 0 0 and 400 A in cross-section. Some are slightly wrinkled, and some portions appear as dark streaks because they are in the correct position to produce Bragg reflections. Figure 2, a scanning electron-micrograph taken of what appeared macroscopically to be a single fibre, clearly shows the composite nature of the fibres. Examination of the fine size-fraction of the sample using a transmission electron-micro scope shows that long, flexible ribbons have nearly uniform cross-section measurements of 200 - 400 (Fig. 3). Electron-diffraction pat terns obtained on these single fibres show rows of spots having the 5.3 spacing typical of amphiboles. No streaking or other evidence of structural disorder were observed. We suggest that the word fibril be used to describe these flexible ribbon-like crystals. This term is commonly accepted in describing the single chrysotile fibre and may be equally ap propriate for amphibole-asbestos unit fibres. Throughout this paper, the term fibre implies a bundle of fibrils. Chemical Analysis and Structural F ormula Chemical analyses, performed on the asbestos by two laboratories, are given in Table 2. Both analyses show a greater loss on ignition than can be accommodated into the model formula for TABLE 2. CHEMICAL ANALYSES OF POTASSIAN WINCHITE-ASBESTOS Constituent 5102 A1z 3 Fe23 FeO CaO MgO K2 Na2 0 Ti2 F 1 58.0 .74 .63 " 4.44 23.9 2.58 4.45 .06 1.22 -** 57.5 .72 .71 " 4.53 25.0 2.19 3.87 " 1.34 Loss on Ignition 4.35 4.48 Oxygen of F Total equivalent 100.37 - .51 99.86 100.34 - .56 99.78 * Analysis by A. Dlnsdale, British Ceramic Research Association, Great Britain ** Analysis by Wayne Lowry, Bureau of Mines 1 0 4 THE CANADIAN MINERALOGIST per minute. The analysis was performed in a N2 atmosphere. the amphiboles as HaO. Therefore, weight-loss measurements were made using a DuPont micro balance with a heating rate of 5C per minute in a Na atmosphere on a 20 mg sample of < 60 mesh material. Figure 4 shows a gradual weightloss of 2.2% between 400 and 925C and a more rapid loss of an additional 2.1% between 925 and 1180C. It is possible that the majority of the lower (1978), this amphibole is potassian winchite- asbestos, a member of the sodic-calcic amphi bole group. The chemical analyses given in Table 2 differ from that given by Rohrbacher (1973). His analysis shows more K2O (3.54 wt. % ) and Na26 (6.60 wt. % ) and less H2O (1.96 wt. % ), whereas fluorine is not reported. Potassium richterite-asbestos is the proper name for an am temperature 2.2% weight-loss represents ad phibole with the composition he reports. Since sorbed water. Although we would expect such there is no reason to doubt his analysis, it must water to be lost at temperatures below these, it be concluded that there is considerable varia is retained possibly because of the very small tion in the bulk composition of this asbestos. size of the fibrils. In addition, the higher tem perature 2.1% weight-loss closely approximates the water content expected for the ideal am Optical Properties phibole. Another possible source of excess water It is clear from the Weissenberg photograph is HsO+ in the A site, but this would probably be (Fig. 1) that fibres resolvable by the optical mi released as the structure is destroyed at the croscope are composed of tiny fibrils oriented in higher temperatures. In addition, there are only such a way that they share only a common 0.55 /4-site vacancies available. Some of this c-axis direction. Therefore, all fibres show par weight loss could be derived from the talc im allel extinction, and most small fibres show purity, but optical and powder X-ray-diffraction only two indices of refraction. Conoscopically, studies indicate that 2% by weight is probably both clearly defined, negative, obtuse, bisectrix the maximum talc content. Undetected multiple interference figures (2V 75) as well as optic chains such as those described by Veblen et al. normal or flash figures are common. The same (1977) in chesterite, jimthompsonite and clino- properties are common to other commercial jimthompsonite might also be present. Such asbestos: crocidolite, grunerite-asbestos (amo biopyribole-asbestos should contain more water site) and chrysotile. than an amphibole (Veblen & Burnham 1978) The indices of refraction were determined but none would approach 4.3%. Finally, it is for sodium light by standard immersion-tech possible that H + ions are related to anion sites niques employing Cargille oils calibrated in in other than those with which they are normally crements of 0.002. They have been designated associated. This hypothesis was proposed by ny' and n a '. There is considerable variability in Ernst (1968) to explain the excess water in an the magnitude of the refractive indices; na' varies amphibole of similar composition synthesized among fibres from 1.576(2) to 1.596(2); ny by Christophe-Michel-Levy (1957). This seems varies from 1.590(2) to 1.600(2); along-fibre the most likely explanation if the water is not variations of n a ' and n y ' may be as great as adsorbed. 0.006. All fibres have positive elongation. Based on the present study, it is not possible The indices of refraction are not only variable, to determine the exact nature of all the water. they are also unusually low for an amphibole. If 2.2 wt. % H20 is assumed to be adsorbed, Most natural amphiboles of similar composition the formula based on 2 4 (0 ,OH,F) is (Ko.43Nao.02) have an na > 1.600. However, Gibbs et al. (N a1.12Ca .0 68Mg0.20 (Mg4.82Alo.llFe3+0.07) (Sii.oo (1962) reported na 1.576 and try 1.595 for a Alo.oi) O21.55 (OHi .88F0.57). According to Leake synthetic fluor-magnesio-richterite. The presence POTASSIAN WINCHITE-ASBESTOS FROM TEXAS 105 of fluorine and the absence of calcium and iron these same factors also probably affect the in are probably responsible for its low indices, and dices of the potassian winchite. BDAX Study of C hemical Variation Variation in optical properties implies a varia tion in chemical composition. In order to cor relate these variations, a qualitative chemical study employing an AMR Model 1400 scan ning electron-microscope equipped with a spec trometer for energy-dispersive X-ray analysis (EDAX) was undertaken. Two aluminum tabs were prepared for exami nation. Randomly selected, dispersed fibres were mounted on one tab, and a fibre 2 cm in length was mounted on the other; n a ' and n y' of the single fibre had previously been determined. Each varies by 0.004 parallel to the length, but seems constant across the fibre. Both tabs were carbon coated. Counting for each analysis was done for 1 0 0 seconds; all data were in excess of 1000 counts per second. The total area under the spectrograph peaks for Na, Mg, Si, Ca and K was recorded after correction for background. The peak area/peak area Si was then calculated for each element. Figure 5 and Table 3 summarize the results of this study. The analyses are numbered in TABLE 3. SUMMARY OF EDAX STUDY OF POTASSIAN WINCHITE-ASBESTOS Oxide MgO Na20 Peak area element P e a k a r e a Si Mean Range .275 .240 - .340 .024 .015 - .033 Wt. % oxide* Wt. t SIO. .423 .072 P r e d ic te d wt. % range 21.3 - 30.2, 2.6 - 5.7 KjO .057 .045 - .068 .041 1.9 - 2.8 CaO .067 .050 - .083 .078 3.4 - 5.6 * From Table 2. order of increasing Mg/Si. It is evident from the data that Mg/Si and Na/Si are inversely proportional to K/Si and Ca/Si. The error in troduced by instrumentation should affect the magnitude of these ratios by no more than 0.004. The estimate of compositional variation among the fibres is based on the assumptions that ( 1 ) the ratio of peak areas is proportional to the ratios of wt. % oxides and (2 ) wt. % Si02 is constant. Whether or not these assump tions are valid, there seems little doubt that there is significant compositional variation among the fibres and within a single fibre with respect to Na, K, Mg and Ca. The EDAX was originally undertaken to relate composition and optical properties. Gladstone's Law (Larsen & Berman 1934) predicts an in crease in indices of refraction when K2Q and Predicted variation Predicted variation Predicted variation weight % MgO weight % Na20 weight % KjO Predicted variation weight % CaO 3.I_4___ ,____5I.6 to UJ C>O- 25 <-I <z 20 - ou. DC 15 - ffi 10 - 5- .2 Total counts Mg Total counts Na Total counts SI Total counts Si Total counts K Total counts SI Total counts Ca Total counts SI F ig . 5. Summary of the variations in peak area (element)/peak area(Si) determined in 30 analyses of potassian winchite-asbestos by a scanning electron-microscope equipped with an energy-dispersive X-ray spectro meter. From the spectrograph, total peak-areas of Mg, Na, K and Ca are compared to total peak-area of Si. The variations in weight percent of the oxides MgO, Na20, K20 and CaO, as predicted from the spectro meter data, are indicated. 106 THE CANADIAN MINERALOGIST CaO increase and MgO and Na20 decrease. Although an increase in Ca/Si and K/Si and an associated decrease in Mg/Si and Na/Si seem to be associated with increasing refractive in dices along the length of the single fibre, this can only be considered a tentative conclusion. Despite the apparent optical homogeneity, the variations in composition laterally across the fibre are so extensive that they preclude a definite correlation between optical properties and com position without many more analyses. D iscussion Richterite commonly occurs in metamor phosed dolomite. Although winchites are not commonly associated with this paragenesis, some amphiboles previously identified as rich terite probably are winchites (see, for example, richterite analysis # 2 , Deer et al. 1963). In addition, many previously identified "winchites" are not winchite at all. Of the seven analyses of "winchite" listed by Nayak & Leake (1975), only one (anal. # 4 ) refers to a winchite as defined by Leake (1978). The occurrence of winchite in metamorphosed carbonates may therefore be more common than previously sup posed. The habit of the potassian winchite-asbestos has all the characteristics of commercial as bestos. The fibres are silky, flexible, and have good tensile strength. The fibril size and bundle structure are similar to those of crocidolite, grunerite-asbestos, tremolite-asbestos, actinolite-asbestos and chrysotile, all of which have fibril diameters ranging from 200 to 5000A. The parallel extinction and anomalous interfer ence-effects reflecting the fibril-bundle structure are also characteristic of this habit. There is some evidence of incipient alteration of the fibres: ( 1 ) there is an unusually high amount of water reported in the chemical anal ysis, which may or may not be adsorbed, and (2 ) fibrous talc is present in small amounts throughout the deposit. The asbestos fibres are chemically inhomogeneous on a microscopic scale as well as macroscopically within the Diablo prospect, ranging from potassium rich terite to potassian winchite. The composition is notable for its low calcium content and the presence of fluorine. All of these characteristics are consistent with the low-temperature metaso matic origin for the asbestos as proposed by Rohrbacher (1973). A cknowledgments The authors are indebted to Drs. Millard Maienthal of the Food and Drug Administra tion, Washington, D.C., for the thermal anal yses, Allan A. Hodgson of the Fibre Research Laboratory, Uxbridge, Middlesex, England, who had the chemical analysis run by the British Ceramic Research Association, G.V. Gibbs of V.P.I. and S.U., Blacksburg, Virginia, for the least-squares computer program used on the X-ray data, the Lewter Oil Company, Austin, Texas, for the asbestos fibres, and Eric Steel, Bureau of Mines, Avondale, Maryland, for his assistance on the scanning electronmicroscope. This work was financed in part by a grant from the Bureau of Mines to the Department of Geology, University of Mary land. R eferences Christophe-Michel-Lvy, M. (1957): Premiers stades du mtamorphisme artificiel d'une dolo mie siliceuse: formation de trmolite et de diopside. Soc. fran. Minral. Crist. Bull. 80, 297-302. Deer, W. A., Howie, R. A. & Zussman, J. (1963): Rock-Forming Minerais. 2. Chain Silicates. John Wiley and Sons, New York. Ernst, W. G. (1968): Amphiboles; Crystal Chem istry, Phase Relations and Occurrence. Springer Verlag, New York. Finger, L.W. (1969): RFINE. A Fortran IV computer program for structure factor calcula tion and least-squares refinement of crystal struc tures. Geophys. Lab. Carnegie Inst. Wash, (unpubl.). Gibbs, G.V., Miller, J.L. & Shell, H.R. (1962): Synthetic fluor-magnesio-richterite. Amer. Min eral. 47, 75-82. King, P.B. (1965): Geology of the Sierra Diablo Region, Texas. U.S. Geol. Surv. Prof. Pap. 480. --------- & Flawn, P.T. (1953): Geology and min eral deposits of Precambrian rocks of the Van Horn area, Texas. Univ. Texas Publ. 5301. Larsen, E.S. & Berman, H. (1934): The micro scopic determination of the nonopaque minerals. U.S. Geol. Surv. Bull. 848. Leake, B.E. (1978): Nomenclature of amphiboles. Can. Mineral. 16, 501-520. Nayak, V.K. & Leake, B.E. (1975): On "win chite" from the original locality at Kajlidongri, India. Mineral. Mag. 40, 395-399. Rohrbacher, R.G. (1973): Asbestos in the Allamore Talc District, Hudspeth and Culberson Counties, Texas. Univ. Texas Geol. Circ. 73-1.