Hydrobiotite, a Regular 1:1 Interstratification of Biotite and Vermiculite Layers

American Mineralogist, Volume 68, pages 420425, 1983

Hydrobiotite, a regular 1:1 interstratification of biotite and vermiculite layers

G. W. BnrNorpv Materials Research Laboratory and Department of Geosciences The PennsylvaniaState University University Park, Pennsylvania 16802

Pernrcre E. ZN-sxr Materials ResearchLaboratory The Pennsylvania State Universiry University Park, Pennsylvania 16802

eNo Cnerc M. BBrure2 Department of Geosciences The Pennsylvania State University University Park, Pennsylvania 16802

Abstract

Well-ordered hydrobiotites consist of a regular alternation of biotite and vermiculite layers. Current requirementsfor use of a specialname for an interstratified mineral specify that the coefficient of variation, CV, of the basal spacings obtained from ten or more reflections should be less than 0.75. Data for three hydrobiotites from different sourcesgive mean basal spacingsof 24.514 measuredto 0.0.31for individual flakes and 0.0.22for orientedpowder layers. The CV valuesfor these samplesare 1.25,0.35,and 0.98 when equal weight is given to all reflections. The large values for the first and third samplesarise. primarily from the first-order difractions which are least accurately measurable;when these are omitted, the CV values are 0.67,0.29, and 0.20. Observed and calculated structure factors for fi)/ diffractions up to / = 3l vary similarly with the index /. Values of d(00/), calculated by Reynolds' method and by a modified Mering method for ordered mixed-layer sequencesfrom 40-&Vo biotite layers, show that the three samplesstudied have 45Vo,53Vo,and 49Vobiotitelayers. The calculationsalso show that ifthe percentageof biotite layers falls outside the range45-55% biotite, variable basal spacingswith CV > 0.75 will be obtained.

Introduction with a 24A basal spacing and also more complex sequences. in which "the sequence The name hydrobiotite has been in the mineralogical He found two samples . adefinite ratioof l:l biotite: literature for at least a century. It was used by Schrauf of layers . .may approach vermiculite Boettcher (1966,p. 284), discussing (1882,p. 3E1-383)for a mineralfrom Kremze resembling units." the origin of vermiculites and hydrobiotites, stated that biotite in its chemical composition but containing consid- "the hydrobiotite has been used in a variety of erably more water. It was mentioned by Dana (1892, p. name ways and should be defined wherever used"; he specifi- 632). Gruner (1934)recognized that the X-ray diffraction cally defined hydrobiotite as "a regular l: I heteropolyty- data of hydrobiotites "show clearly that [they] are made pic (mixedJayer) biotite:vermiculite" and (p. 286) stated up of interstratified single or double layers of mica and that "the sharp X-ray peak at 25.5A and the high order of vermiculite." He consideredan alternating1:1 sequence rational basal peaks obtainable in the X-ray patterns of I On leave of absence from: Comisi6n de Investigaciones hydrobiotite attest to the regularity of spacing." Cientificasde la Provincia de Buenos Aires - CETMIC, Calle 47 Foster (1963) made a critical review of the available y l15. Dept. Tecnologia, 1900LaPlata, Argentina. chemical data for vermiculites and hydrobiotites and, 2 Current address: Department of Geology, University of despite inherent difficulties in arriving at structural for- Illinois, Urbana, Illinois 61801. mulae in the absence of cation exchange data, showed m03-004)v8310304-0420$02.00 420 BRINDLEY ET AL.: HYDROBIOTITE. A REGULAR I :1 INTERSTRATIFICATION 421 that the formulae of hydrobiotites are intermediate be- Optical properties of the north-western Transvaal ma- tween those of biotite (or femrginous phlogopite) and terial were as follows: F = 1.560-1.562,y : 1.565-1.567, vermiculite, and therefore, are consistent with an inter- a biaxial negative interference figure, 2Y" - 19". stratification of layers. Hey (1962)recognized hydrobio- X-ray difraction data were obtained using a Philips tite as "a mixed-layer mineral, an interstratification of diffractometer with Ni-filtered CuKa radiation and were vermiculite and biotite," but did not mention regularity of recorded at l" 29lmin and with chart recording on a scale the mixed layering. of 2" Z?linch. When samplesin the form of single flakes Despite these studies, the name hydrobiotite seemsnot were used, the incident X-ray beam divergence was to have been fully accepted as a valid mineral species. choosenso that the radiation was confined to the surface The name is not included by Fleischer (1980) in his of the flake but this condition was not wholly satisfiedat "Glossary of Mineral Species" among a long series of 2e < l0'. To record diffracted intensities. it was neces- hydro-minerals.In a critical review of interstratified layer sary to make changesin the silt system, the X-ray tube silicates, Buley et al. (1982)made no mention of hydro- operating conditions, and the recording conditions. Sev- biotite. They specifiedthe following requirementsfor the eral diffractions were recorded before and after each use of special names. An interstratification of two layer change of experimental conditions so that all recorded types, A and B, should have sufficient regularity to give a intensities could be placed on a common basis. The well-defined series of at least ten essentially integral diffractometer was calibrated with respect to a silicon orders of reflection from the summation spacing, d4s = standard and a secondary tetradecanol standard (Brind- d^ + dB, and the coefficient of variation of das should be ley, l98l). less than 0.75 to demonstrateadequate regularity oflayer Attention has been given particularly to the evaluation alternation. The diffraction breadths of odd and even of basal X-ray difractions. Imperfectly interstratified order reflections should be closely similar. If odd orders layer sequences give diffraction peaks which do not are absent,their absenceshould be shown to be consis- follow precisely the Bragg reflection equation. The appar- tent with calculated intensities. ent spacingsd(00I) multiplied by the apparentorder / give The presentstudy was undertakento establishwhether a variable product /d(00D = d1and the degreeof order in interstratified biotite/vermiculites exist that satisfv these the layer sequenceis evaluatedby a statistical treatment requirementsfor use of the name hydrobiotite. of values of dr. If d is the arithmetical mean value, the mean deviation, AZ, is (2"ldrt1l)ln, and the coefficient of Materials and methods variation, CV, is [I"(dr- A2l@ - l)]t/2 x lfi)/Z, where n is the number of terms in the summations. Vermiculite ores from the Enoree area of S. Carolina. made available many years ago to one of us (G.W.B.) by Results the Zonolite division of W. R. Grace & Company, measurements showed a varied mineralogy that included vermiculite, Basal spacing biotite, and variously ordered mica./vermiculites.From Table I summarizesthe results for two samplesexam- material describedas a Cooper concentrate, small black- ined in flake form and one sample as a thin oriented ish flakes were selected under a binocular microscope. powder layer. The two samplesexamined as flakes yield- They were lightly ground in water with a small pestle and ed diffractions up to 0.0.31 and the powder data were mortar. The dispersedmaterial was allowed to dry slowly measuredas far as 0.0.22. The mean basal spacingsare in air to form thin layers on glass slides. Only basal respectively 24.46t0.17, 24.5I -+0.05 and 24.45* 0. 13A difractions appearedin the recorded patterns. X-ray data where the -r values are mean deviations. The CV values for this material are presentedin section 3 of Table l. are respectively1.25,0.35 and 0.95when equalweight is A number of sampleswere made availableby Dr. S. W. given to all reflections. The secondsample easily satisfies Bailey and from among these, a well-developed flaky the condition CV < 0.75 to justify the acceptanceof a material from north-western Transvaal provided thin specific mineral name. flakes up to about 8 x 8 mm size, which could be The large CV values for the first and third samplesarise examineddirectly without reduction to powder. Many of largely from the first order diffractions which are the least the flakes were contaminatedwith apatite and, to a lesser accurately measurable because of the small 20 angles, extent, with zircon. Uncontaminatedflakes were selected 3.6-3.8', at which they occur. When the first order under a microscope and were mounted directly on glass diffraction data are omitted, the mean spacings(as given slides with a trace of adhesive. A sample supplied by in Table l) are respectively24.5110.10, 24.5110.04, and Wards Natural ScienceEstablishment. labelled vermicu- 24.51'+0.044and the CV values are 0.67,0.29,and 0.20, lite from north-easternTransvaal, proved to be a regular- all of which satisfy the condition of CV < 0.75. Although ly interstratified biotite/vermiculite. Golden yellow and we cannot give an unequivocal explanation for the low pinkish flakes were selected for study but were smaller values of the first order spacings for the first and third than those from north-western Transvaal and were less samples, the balance of evidence appears in favor of suitable for X-ray intensity measurements. retaining hydrobiotite as a specific mineral name. 422 BRINDLEY ET AL,: HYDROBIOTITE, A REGULAR I : I INTERSTRATIFICATION

Table l. X-ray difraction data for basal reflections from hydrobiotites

(1) (2) (3) Flake, North-western Transvaal. Flake, North-eastern Transvaal. Oriented powder, South Carolina d(002) d(001) I(peak) d(002) d(001) I(peak) d(002) d(001) I(peak)

I 23.3 23.3 IO 24.3 24.3 7 23.6 23.6 20 t 12.30 24.60 80 LZ. Z5 24.46 60 L2.26 24.52 >100 J 1 7.99 23.97 8.27 24,8L ) 8. 18 Lq. )c ) 4 6.146 24,58 I 5 4.880 24.40 30 4. 909 L+. )) 30 4. 880 24.40 15 6 7 rlroo zi-.st 50 3.493 24.45 50 3. 500 24.50 50 8 3.035 24.28 I4 3. 071 24.)t t5 3. 051 24.4t 20 9 2.742 ,!_ut L2 2.725 24.52 20 2.733 '!_'o l0 _10

I1 2.261 24.87 0.5 L2 2,045 24.54 16 2.041 24.49 16 2.o43 24.52 I5 l3 r.881 24.45 1.0 1.889 24.56 1. 881 z4 .4) t L4 15 1.630 24.45 1.5 1.633 24.50 2 r.633 24.50 1.5

fo I7 t.oo, zi-.st ;- I.44L 24.50 1.443 24. )5 5 l8 19 L293 24.57 1.8 L.290 4 r.292 z4, J) 0.5 20 L.224 24.48 r.2 r.224 24.48 I r.227 24.54 I

2I L.Lt) aq.oJ 0.2 ta 1.115 24.53 0.3 1.114 24.5L 0.4 '!_tt 0.5 23 1_t* L.023 24.ss 1.0 L.O2r 24.50 25 o.979s 24.49 o.2 o.9773 24.43 o.2

zo 0.9455 24.58 0.3 0.9433 24.52 o.4 0.9080 24.52 L.4 0.9069 24,49 ? 28 29 0.8480 24.59 1.0 0.8445 24,49 2 30

JI 0.792L 24.56 0.1 0.7900 24.49 nt

o (001) 24.sI 24.5r zq. )L

Ad I o.ro 0.04 0. 04

Two flakes treated with ethylene glycol yielded the structure factor) were calculated from observed relative following re-sults:mean basal spacings24.45-+0.llA and intensities,taken as (peak intensities) x (breadthsat half- 24.50-+0.05A,CV values0.60 and 0.30 based on 20 and 16 maximum intensity), by the relation diffractions with d(00/) omitted. The negligible changeof Iou" = Foo" (l + cos2zrylsinze basal spacingwith glycolation is consistentwith vermiculite rather than smectite lavers. Calculated values of F2 have been derived using z parameters in A, for the hydrobiotite layer structure Dffiaction intensities taken as follows: The peak intensities recorded in Table I take account I K at 0.fi), 3 O at !1.67,2(Si,Al)at !2.28,3 O at !3.94, of the various conditions of divergence and receiving 3(Me,Fe)at t5.(X),3 O at 16.06,2(Si,Al) at -+7.72,3O slits, X-ray tube operating conditions and recording con- at -+8.33,3 H2Oat -r11.07,0.22Mg at +12.24. ditions. Most attention will be given to the data in section I of Table I for material from north-western Transvaal, These parameters are based on data tabulated by becausethese were obtained with a flake of the largest Bailey (1980, p. 56-7) for trioctahedral micas and given surface area, about 8 x 8 mm. The least satisfactory by Shirozu and Bailey (1966)for Me-vermiculite. Calcula- intensities are those for I : 1,2,3 for which an uncertain tions were made for trioctahedral cations ranging from allowance was made for the incident beam not falling Mg3to Mg1.5Fe1.5 and all octahedralsheets were assumed entirely on the flake surface.Relative values ofF2o6"(F : to have the same cation compositions. No attempt has BRINDLEY ET AL.: HYDROBIOTITE. A REGT]LAR I :1 INTERSTRATIFICATION 423

been made to obtain an optimum agreement between observed and calculated F values. The principal objec- tive has been to show that unobserved basal reflections are consistentwith the assumedstructure of hydrobiotite and this is clearly establishedby the comparisonshown in Figure l The observed F" values have been placed on a scale to make them generally comparablewith the calculated values, which are shown for two octahedral cation compositions, Mg3 and Mg2.75Fes 25. There is a good generalconcordance between strong, medium, weak and absent reflections. The results support the second main requirement for the use of hydrobiotite as a valid speciesname.

Calculated dffiaction patterns Reynolds (1980) has shown that computer generated diffraction patterns can be used to study the statistics of Fig. 1. Comparison of F2 values derived from experimental mixed-layer systems. Table 2 summarizesdata used for data (full line) and calculated from theoretical compositions computing such patterns for hydrobiotite with biotite/ver- basedon Mg3 and(Mg2.7sFeo zs) in octahedralpositions (dashed miculite ratios ranging from 60/40 to 40/60. Values of Iine. dotted line). d(Nl\, Ad, and CV have been calculatedfrom the result-

Table 2. Numerical data for biotite/vermiculite mixedJayer sequences

A. Statlstlcal data used 1n computlng blotlEe-vemicullte dlffractlon Patterns

Ratlo of Layers Junction Probabilities

7"8 7"Y Pg.g Pa.v Pv.n d(00r) AO "v.v 60 40 0.333 0.667 1.000 0.000 + 0.29 1.84 55 45 0.183 0.818 1.000 0.000 24.48 0.L2 0.73 50 50 0.000 1.000 1.000 0.000 24.41 0.07 0,35 55 0.000 r.000 0.818 0.182 24.44 0. r-4 0. 85 40 60 0.000 1.000 0.667 0.333 z+,+J u. zo L.62

B. Peak nlgration data from calculated patterns

Calculated peak positi.ons, for CuKq L=2 3 7 8 L2 13

50 40 7.46 10.66 25.66 28.97 33,10 44.55 48.O7 55 45 t,50 10.80 25,55 to nq 33.03 44.47 48.25 50 50 7.26 10,92 25.45 29.20 32.94 44.40 48,35 45 55 7.r4 11.06 25.38 29.34 32.80 44.34 48.45 40 60 7.00 11.15 25.28 29.48 32.62 44.28 48.58

C. Peak migration data by a nodified llerlng nethod

ZR CalcuLated values of 20, wlth (00{) in parentheses

67 10.28 25.90 28.54 33.88 44.80 47.58 (003) (004) (00,10) (00,rr) (00,13) (00,17) (00,18)

7.24 10.88 25,55 29.28 33.04 44,s6 48.50 (002) (003) (007) (oo8) (009) (00,12) (00, 13)

6.83 11.40 z).5u 29.93 32.3L 44.35 49.33 (003) (005) (00,u.) (00,13) (00,14) (00, 19) (00, 2r) 424 BRINDLEY ET AL.: HYDROBIOTITE. A REGULAR I : I INTERSTRATIFICATION

I t I I I I

o = Es0

\ I I

t0 26 '8 3o 32 44 46 48 ,eo Fig. 2. Migration curves for diffraction peaks from hydrobiotites, calculatedby Reynolds' method (full line) and by a modified Mering method (dashedline). Small circles show calculated values for 40-60% biotite layers by Reynolds' method.

ing patterns for direct comparison with the experimental Possible application of a modified Mering proce- : data. Values for / I have been included, for although dure for interpreting dffiaction by mixed-layer these values may be influenced by the Lorentz-polaiza- systems tion and structure factor variations, they will not be It was shown by Mering (1949)that in a system of subject to experimental errors. In these calculations, the mixed spacings d1 and d2, regions of strong diffraction same z parameters for the layer structures have been tend to occur when integral values of lldl and lld2 are employed as were used earlier for the calculation F2 of closely similar. In mixed systemsranging from entirely d1 values. spacingsto entirely d2 spacings,a migration from order n1 As expected,the calculatedvalues of AZ and CV are of d1 to order n2 of d2 would be expectedwhen n1/dr: smallest for the 50/50 biotite/vermiculite ratio, and they n2ld2.The modified system to be considered visualizes increase more or less symmetrically as the ratio is migration of di-ffractionsbetween adjacent ordered sys- changed to 40/60 and 60/40. It appears that within a tems with simple ratios. For example, if we wish to biotite/vermiculite range about 55/45 and 45155, the CV considerthe effect of deviations from a 50/50B/V system, satisfiesthe requirement of being less than 0.75 to merit i.e.. a l: I ratio. we could consider ordered 33167and the use of the name hydrobiotite. Conversely we may 67133systems, i.e., l:2 and2:1 systems.A systemwith conclude that a biotite/vermiculite will satisfy the criteri- 60Vobiotite layers could be consideredas falling between on of CV < 0.75 provided the ratio of the components lies the ordered systems with 50% and 67Vobiotite layers. within 5Voof the 50%l50Voratio. Physically, this model presupposes that a 60Vobiotite The theoretical patterns show that the basal diffraction systemwould havea sequenceofeither one biotite layer, maxima migrate to higher or lower 20 values as the B/V ratio changes(Fig. 2). Values of 20 for diffraction peaks which will be useful in estimating the B/V ratios of hydrobiotites and of less regular interstratifications are listed in part B of Table 2. The calculated data can be used in two ways. First, one can match observed 20 Az0 values againstthe migration curves, and second,one can take differences A(20) between adjacent pairs of reflections. Figure 3 illustrates this second method; A(20) is d plotted againstpercentage of biotite layers. Experimental b a values of A(20) for the three samplesused in Table I are 3-40s060 located on the curves and the percentagevalues of biotite layers are read directly. For each sample, four estimates % biotite are obtained,and the averagevalues are: sample1,45% Fig. 3. Values of LQ€) between adjacent basal reflections, biotite; sample2,53% biotite; sample3,49% biotite. The obtained from Reynolds'-type calculations, vs. Vobiotite layers, order in which these values depart from 50% biotite (a) 002-003,(b) 007-008,(c) 008-009,(d) 0.0.12-0.0.13.Samples correspondsto the sequenceof CV values given earlier. l, 2 and 3 of Table I gives data points shown by !, x, and o. BRINDLEY ET AL.: HYDROBIOTITE, A REGULAR I :1 INTERSTRATIFICATION 425 or two biotite layers between each consecutive pair of thank also Dr. S. W. Bailey and Dr. Michael Fleischerfor useful vermiculite layers; the proportion of B or BB units would comments on the first draft of the paper. One of us (C. M. B.) be governed by the overall B/V ratio. was supported by a National Science Foundation Graduate Figure 2 compares the migration curves calculated by Fellowship. the Reynolds method and by the modified Mering proce- References dure. Four and perhaps five of the peak migrations derived by the two methods agree very closely and only Bailey, S. W. (1980)Structures of Layer Silicates.In G. W. Brown, Eds., Crystal Structures of Clay two show appreciabledifferences. Here we wish to point Brindley and G. Minerals and their X-ray Identification, p. 1-123. Mineralogi- only to the potential applicability of the very simple cal Society, London. modified Mering method. Table 2C shows values of 20 Bailey, S. W., Brindley, G. W., Kodama,H. and Martin, R. T. calculatedfor regular2:1, l:1, and l:2 B/V ratiosand the (1982) Report of the Clay Minerals Society Nomenclature 00/ values used. These data were used in plotting the Committeefor1980-1981. Clays and Clay Minerals,30,76J8. modified Mering migration curves in Figure 2. In those Boettcher, A. L. (1966)Vermiculite, hydrobiotite, and biotite in cases where the curves calculated by the two methods the Rainy Creek igneouscomplex near Libby, Montana. Clay show appreciabledifferences, these differencesmight be Minerals.6.283-296. explained by an interaction of diffraction peaks of difer- Brindley, G. W. (1981) Long-spacing organics for calibrating ent intensities or of rapidly changingvalues of F2. long spacingsof interstratified clay minerals. Clays and Clay Minerals,29,67-68. Relation of hydrobiotite to biotite Dana,E. S. (1982)Descriptive Mineralogy, Sixth Edition, Wiley, New York. It is relevant to the discussionof hydrobiotite to draw Fleischer, M. (1980)Glossary of Mineral Species,19E0, Mineral- attention to conclusions reached by Boettcher (1966). ogicalRecord, P.O. Box 35565,Tucson, Arizona E5740. Discussing particularly the vermiculite and hydrobiotite Foster, Margaret D. (1963)Interpretation of the composition of found near Libby, Montana, he stated that "A compari- vermiculites and hydrobiotites. In Ada Swineford, Ed., Clays son of chemical analyses of the hydrobiotite and biotite and Clay Minerals. Proceedingsof the Tenth National Confer- suggestsno simple compositionalrelationship as might be ence on Clays and Clay Minerals, p. 70-89, Macmillan Com- expected if the hydrobiotite is a simple weathering prod- pany, New York. W. (1934)The structuresof vermiculitesand their uct." Furthermore, "the regular stacking sequenceex- Gruner, J. collapseby dehydration. American Mineralogist, 19, 557-57 5. hibited by the hydrobiotite is not inherited from the Hey, H. M. (1962)An Index of Mineral Speciesand Varieties parent biotite. Consequently, the hydrobiotite is consid- arrangedChemically. Second, ievised edition, p. 193.British ered to be a high-temperaturealteration product in which Museum,London. the l: I stacking sequencerepresents the attainment of Mering, J. (1949).L'interfdrence des rayons-X dansles systdmes homogeneousequilibrium at the time of alteration." (The i stratification d6sordonn6e.Acta Crystallographica2, 171- description "regular stacking sequence" relates only to 377. the alternation of layer types and carries no implication of Reynolds, R. C. (1980)lnterstratified Clay Minerals. In G. W. spacegroup symmetry.) Brindley and G. Brown, Eds., Crystal Structures of Clay There are therefore good reasonsfor regarding hydro- Minerals and their X-ray Identification, p. 249-303. Mineral- London. biotite as a valid mineral speciesbesides the evidencethat ogical Society, Schrauf, A. (1882) Beitrdge zur Kenntniss des Associationsk- the component layers are organized in a manner suffi- reises der Magnesiasilicate.Zeitschrift fiir Krystallographie ciently regular to meet the requirements set out for und Mineralogie,6, 321-388. allocation of namesto interstratified layer minerals. Shirozu,H. and Bailey, S. W. (1966)Crystal structure of a two- layer Mg-vermiculite, American Mineralogist 51, 1134-1143. Acknowledgments Weare indebted to Dr. S. W. Baileyand the Zonolite Division Manmcript received,May ll,1962 of W. R. Grace& Companyfor materialsused in this study.We acceptedforpublication, October 15, 1982.