miissbauer spectra of biotite from metapelites

American Mineralogist, Volume 75, pages 65M66, 1990

Miissbauer spectra of biotite from metapelites

M. Dlnsv Dv.q.nDepartment of Geological Sciences, University of Oregon, Eugene, Oregon 97403, U.S.A.

Ansrru,cr

Mdssbauer spectra of 52 biotite samples from metapelites of the Rangeley, Oquossoc,and Bryant Pond Quadrangles of northwestern Maine are examined to evaluate the oxi-dation state and site occupancies ofFe atoms. Spectra show nearly invariant peak positionsfor the cis-M2 site (D: l.l3 mm/s and A:2.60 mm/s) in the octahedral sheet of thebiotite but greater variation in Fe2+ trans-Ml doublets (6 : I . l3 mm/s and L : |.95-2.60mm/s), which reflects a range of distortion in that site. Tetrahedral Fe3+ is present inamounts ranging from 4-l4o/o of the total iron in all samples studied, roughly equivalentto the number of octahedral vacancies. Two octahedral Fe3+ doublets can be resolved inspectra for samples from the more oxidized assemblages studied. In more reduced rocksthese doublets are poorly resolved and can be best represented by a single doublet withaveraged hyperflne parameters. Fe3+/2Fe remains constant within analytical errors amongsamples from lower garnet to potassium feldspar + sillimanite metamorphic grades aslong as the oxide mineral phase does not change. Howevero octahedral Fe3+ content in-creases from a baseline of 40lo of the total Fe in graphite-ilmenite-bearing assemblages upto 400/o in biotites from hematite-bearing assemblages. The octahedral Fe,+ M2lMl ratiothat reflects ordering typically falls in a range from2 to 3.5, but extremes of I.14 up to5.20 are observed in samples from sulfide-bearing or highly-oxidized assemblages.

INrnonucrroN range of assemblages (graphite-ilmenite, magnetite, andhematite-bearing) and metamorphic grades (lower garnet

Historically the work of crystallographers and miner- through potassium feldspar * sillimanite zone) mappedalogists has been oriented toward the analysis of individ- in northwestern Maine. This group of samples was spe-ual samples considered to be representative of their lo- cifically chosen because ofexisting detailed petrologic andcalities. For this reason, too little is known about the major-elementcharactenzationbyGuidottiandcowork-correlation of crystallographic measurements (such as ers (Guidotti,1984; Guidotti etal., 1977,1988; Henry,those gleaned from X-ray structure refinements or from l98l; Wood, 1981). This paper discusses the spectro-Raman, rrrn, or Mcissbauer spectroscopy) with petrologic scopic data for these samples. The petrologic implicationsdata. Fortunately, there is some evidence to suggest links are considered in Dyar et al. (1986, 1987) and in Guidottibetween petrogenetic context and crystal structure. For and Dyar (in preparation).example, a revent review of Mdssbauer studies of trioc-tahedral micas (Dyar, 1987) found that only 55 of l5l Expnnrn'mNTAL DETAILSspectra reported in the literature noted some vague sam- The 52 biotite samples used in this study are from theple provenance (igneous or low grade were common de- Rangeley, Oquossoc, and Bryant Pond quadrangles ofscriptions); from these data a range of Fe site occupancies Northwestern Maine. They range through lower, middle,was noted, which appeared to be grossly related to sample and upper garnet grades, lower and upper staurolite grades,locality and paragenesis. These observations prompted a transition (staurolite + sillimanite) zone, lower and up-the author to undertake the present study to investigate per sillimanite grades, and potassium feldspar + silli-the answers to the following questions: manite-grade metamorphic rocks (Guidotti et al., lggg).

l. How do oxidation state and site occupancies of Fe All rocks contain quartz and plagioclase and appear toatoms in biotite vary on regional, local, and hand-sample have approached equilibrium. The majority of the sam-scales? To what extent can a single specimen be consid- ples (46) are from graphite-ilmenite-bearing assemblagesered representative of a region or a locality? (designated as Areas A and B; see Table l), five samples

2. If Fe3+ content and cation ordering prove to dem- contain magnetite (Area C), and one sample contains he-onstrate meaningful variability on any scale, can the vari- matite (OB- l6).ation be related to oxide assemblage, pressure, and tem- Major- and minor-element compositions of the biotiteperature conditions in any systematic way? were determined by electron microprobe analysis at the

To address these issues, Mossbauer spectra were ob- University of Wisconsin-Madison, the University of Or-tained on 52 biotite samples from pelitic schists over a egon, and the Smithsonian Astrophysical Observatory in0003404v90/0506-o656$02.00 6s6

DYAR: MOSSBAUER SPECTRA OF BIOTITE 657

Cambridge, Massachusetts. Either Bence-Albee (1968) orAlbee and Ray (1970) correction procedures were used.The acceleration potential was 15 keV, and sample cur-rents ranged from l0 to 40 nA. Beam size was defocusedto l0 pm to minimize alkali diffusion. Natural and syn-thetic minerals were used as standards for major ele-ments, whereas synthetic glasses were used as standardsfor some minor elements. Duplicate analyses confirmedconsistency among the instruments.

Biotite grains for Mdssbauer analysis were separatedfrom bulk crushed rocks using magnetic separation sup-plemented by hand-picking if needed. Approximately 200mg of each pure separate were finely ground under ace-tone (to minimize possible oxidation of Fe) and mixedwith sugar and acetone. Such a preparation coats indi-vidual grains with sugar to produce well-rounded parti-cles, thereby alleviating preferred orientation in the plexi-glass sample holder. For samples with finely intergrowncoexisting phases such as ilmenite and garnet, it was oftennecessary to make successive M0ssbauer measurementsto ensure purity. Ultimately, pure separates were ob-tained from all the rocks studied.

Mdssbauer spectra were recorded in 512 channels oftwo constant-acceleration Austin Science Associatesspectrometers, one located in the Mineral SpectroscopyLaboratory at the Massachusetts Institute of Technol-ogy and one at the University of Oregon. Sources of 70-20 mCi sTCo in Rh and Pd and gradients of approxi-mately 0.01545 mm/s/channel were used. More than halfthe samples were run on both spectrometers, and thereare no differences between spectra from the two instru-ments. At both locations results were calibrated againstan a-Fe foil of 6 pcm thickness and99.99o/o purity. Spectrawere fitted using different versions ofthe program sroNE(Stone et al., 1984) on DEC r-sr-1r/73, MicroVax II, andIBM ps/z MoDEL 80 computers; no variation in the fits wasobtained with the different computers. All three versionsof the sroNe program use a Gaussian, nonlinear regres-sion procedure with the capability ofconstraining any setof parameters or linear combination of parameters. Lo-rentzian line shapes were used for resolving peaks, be-cause there is no statistical justification for the additionof a Gaussian component to the curve shape used.

Procedures used to fit individual samples are describedin detail in Dyar and Burns (1986). In brief, two-, three-,four-, and five-doublet fits (the last-named with doubletsrepresenting Feffi, Fefrr+r, Feflol, Fefln+r, and Fel"i) were at-tempted on each spectrum with gradually released con-straints on widths and areas ofpeaks. Close attention waspaid to the location and intensity of Fe3+ p€aks. In somesamples 30-50 different models were tested to ensure cor-rect assigrrment of the smaller Fe3+ doublets. A statisticalbest fit was obtained for each sample using the y2 andMrsFrr parameters (Ruby, 1973). The practical applica-tion of these statistical parameters is discussed elsewhere(Dyar, 1984). At best, the precision of the Mdssbauerspectrometer and the fitting procedure is approximately+0.02 mm/s for isomer shift (6) and quadrupole splitting

(A) and +1.5o/o per peak for the area of well-resolved,distinct peaks (Dyar, 1984). However, the mica spectrapresented here contain several substantially overlappingpeaks. For this reason the errors are estimated to be ashigh as +0.05 mm/s and +0.03 mm/s for isomer shiftand quadrupole splitting of Fe3+ and Fe2* doublets, ro-spectively. The errors for the areas ofindividual doubletsare estimated to be + 3-4010, based on replicate analysesand on comparisons of Mdssbauer and wet chemicalanalyses of biotite samples from the Sierra Nevadabatholith in central California (Dodge et al., 1969; Ban-croft and Brown, 1975;Dyar and Burns, 1986).

Rrsur,rs AND DlscussroN

Mineral compositions reflecting both electron-micro-probe and Mdssbauer results are presented in Table lMclssbauer results for all samples are given in Table 2.In both tables, metamorphic zones and geographical areas(A, B, and C) are given following the designations of Gui-dorti er al. (1988).

Consistency of peak positions

The most striking characteristic of the Mdssbauer re-sults is the overall consistency ofpeak positions (Fig. 1),given the error estimates above and the standard devia-tions given in Table 2. In particular, the positions of theFe2+ cis-M2 peaks are remarkably constant (D : l. I 3 mm/sand A : 2.60 mm/s), although the corresponding Fe2*trans-Ml peaks vary over a wider range (6 : 1.13 mm/s, A : 1.95-2.60 mm/s). Standard deviations of param-eters for the two doublets (0.02 mm/s for D and 0. l3 and0.12 for A) are considerably smaller than the estimatederrors of the measurements.

The nearly constant parameters of the cls-M2 doubletssuggest that the Fe2* site in different biotite samples isidentical with respect to the s-electron density and elec-tric-field gradient. Therefore, the electronic structure ofthe Fe atoms in the cls-M2 site (and the geometry of theO atoms surrounding them) is the same from sample tosample in this study (within the sensitivity of the Mdss-bauer technique). This confirms the conclusion drawn byDyar (1987) from a literature survey. Apparently the en-vironments around the iron atoms in the cr's-M2 site arerelatively invariant and independent of composition. Incontrast, the greater variation in M0ssbauer parametersof the Ml site indicates that Fe2+ in the Ml site residesin a more variable oxygen environment. Perhaps the /raru-Ml site is more vulnerable to distortions caused by near-est neighbors owing to Ihe trans configuration of its OHbonds.

Where the concentration of Fe3+ is sufficiently high,peak positions for those doublets are also quite consis-tent. For example, in the upper sillimanite grade samplesfrom the Oquossoc quadrangle, Fe3+ doublets corre-sponding to the cls-M2 and trans-Ml sites are resolved,with parameters of 6 = 0.42 and 0.43 mm/s and A =

0.48 and 1.09 mm/s, respectively. In some cases the twodoublets cannot be distinctly resolved and can only be

658

Taeue 1. Biotite compositions recalculated with Fe3*

DYAR: MOSSBAUER SPECTRA OF BIOTITE

Samole name RA- RA- RA- RA- RA- RA- RA-D75-66 D72-66 D9-66 D28-66 C95-66 D86-66 D11-66

RA- RA- RA- RA. RA- RA-D37-66 D12-66 D14-66 D60-66 A65-66 C1-66

A A A A A A3 3 3 4 4 4

AreaZone

A A2 3

A A A A A1 1 2 2 2

sio,AlrosFeOFerO"MgoTio,MnOZnOKrONaroBaOHrO

Total

34.9 35 4 35.419.7 19.4 19.619.7 19.9 19.33.57 3.31 2.938.00 6.77 7.701.66 1 61 1 .700 j2 0 .21 0 .100.00 0.11 0.097.92 8.20 8.380.11 0 .18 0 .160.00 0.00 0.003.95 3.92 3.94

99.63 99.01 99.30

5.30 5.42 5.392 41 2.32 2.400 29 0.26 0.218.00 8.00 8.001 . 1 2 1 . 1 9 1 . 1 12.51 2.55 2.460 1 2 0 j 2 0 . 1 21.81 1 .55 1 .750.19 0 .19 0 .190.02 0 03 0.010.00 0.01 0.015.77 5.64 5.651.54 1 .60 1 .630.03 0.05 0.050.00 0.00 0.001.57 1 .65 1 .68

15.34 15.29 15.33

34.8 35.219.9 19.720.0 21.23.62 2.057.46 8.021.59 1 .66o.12 0 .120.00 0.008.41 8.060.17 0 .150.00 0.083.95 3.95

100 02 100.19

5.29 5.332.44 2.550.27 0j2I00 8 .001.12 0 .972.55 2.690.15 0 j21 .69 1 .810.18 0 .190.02 0.020.00 0.005.71 5.801.63 1 .560.05 0.040.00 0.011.68 1 .61

15.39 15.41

rttlo/c

35.5 35.719.6 19 .919.8 17 .53.00 4.557.38 8.241.63 1 .620.13 0.080.07 0.097.97 8.030.14 0.230.00 0.003.93 3.99

99.15 99.93Number of atoms

5.41 5.362.42 2.320.17 0.328.00 8.001 .09 1.212.53 2.200.17 0 .191.68 1 .850.19 0 .180.02 0.010.01 0.015.69 5.651.55 1 .540.04 0.070.00 0.001.59 1 .61

15.28 15.26

36.6 35.6 35.441 9.9 20j 19.6117 .8 19 .1 17 .773.76 3.17 3.217.96 9.65 8.751 .48 1.49 1 .460.11 0.07 0.040.00 0.11 0.077.22 8.18 8.070.29 0.30 0.200j2 0.00 0.003.99 4.05 3.94

99.23 101 .82 98.56

5.50 5.28 5.392.21 2.49 2.400.29 0.23 0.218.00 8.00 8.001 .30 1 .01 1 .112.23 2.36 2.260.13 0.05 0.161.78 2.13 1 .980.17 0 .17 0 .170.01 0.01 0.010.00 0.01 0.015.62 5.74 5.701 .38 1.55 '1.57

0.08 0.09 0.060.01 0.00 0.001.47 1.64 1.63

1 5.09 15.38 15.33

35.88 35.6119.45 19.7218.24 17.212.25 3.389.30 9.581.60 1 .530.09 0.060.o7 0.008.30 8.420.29 0.220.00 0.003.98 4.30

99.45 100.03

5.41 5.362.39 2.400.20 0.248.00 8.001.O7 1 .092.30 2.160.05 0.132.09 2.'t40 .18 0 .170.01 0.010.01 0.005.71 5.701.60 1 .610.09 0.060.00 0.001.69 1 .67

15.40 15.37

36.419.816.43.739.611.550.180.098.200.210.004.O4

100.21

5.412.440.158.001.032.04o.272.130.170.o20.015.671.550.060.001 . 6 1

15.28

D I

Al (tet)Fa3+ /tat\

Total (tet)Al (oct)FeetF€P+ (oct)MgTiMnZn

Total (oct)KNaBa

Total (int)Total

Samole name RA- RA- RA- RA- RA-c4-66 A73-66 A14-66 A69-66 A33-66

RA. RA- RA. RA- RA-853-66 848-66 A57-66 A96-66 A93-66

A A A A A6 6 6 6 6

RA-94-66

A

RA. RA-841-66 A95-66

A AAreaZone

A A A A A4 5 5 5 5

SiAlrosFeOFerO.MgoTio,MnOZnOKrONaroBaOHrO

Total

35.3 36.1 35.6 35.620.2 20.0 20.1 20.218.6 17.1 17 .6 18.12.82 2.34 2.41 1.999.03 10.1 9.65 9.21 .60 1 .50 1 .49 1.430.11 0.09 0.07 0.060.00 0 .13 0 .11 0 .108.27 8.30 8.18 8.220.32 0.30 0.30 0.330.01 0.00 0.00 0.003 99 4.03 3.99 3.97

100.25 99.99 99.50 99.20

5.30 5.37 5.35 5.372.46 2.45 2.38 2.510.24 0.18 0.27 0.12I00 8.00 8.00 8.001 . 1 1 1 . 0 7 1 . 1 8 1 . 0 92.34 2.12 2.21 2.280.08 0.10 0.00 0.102.02 2.24 2.16 2.070.18 0 .17 0 .17 0 .160.01 0.01 0.01 0.010 00 0.01 0.01 0.015.74 5.72 5.74 5.721 5 8 1 5 8 1 . 5 7 1 . 5 80.09 0.09 0.09 0.100.00 0.00 0.00 0.001 .67 1.67 1 .66 1 .68

15.41 15.39 1 5.40 1 5.40

rttlc/o

36.0 36.1 35.720.0 19.7 19.516.8 16.2 17 .72.08 2.46 1.48

10.2 10.4 9.851.48 1 .59 1.820.07 0.08 0.100.11 0 j2 0 .118.41 8.44 8.460.33 0.26 0.340.00 0.00 0.004.00 4.00 3.97

99.48 99.35 99.03Number of atoms

5.39 5.40 5.392.47 2.44 2.440.14 0 .16 0 .178.00 8.00 8.001.05 1 .04 1 .032.11 2.03 2.240.09 012 0.002.28 2.32 2.22o.17 0.18 0.210.01 0.01 0.010.01 0.01 0.015.72 5.71 5.671 .61 1.61 1 .630.10 0.08 0.100.00 0.00 0.001.7't 1.69 1.73

1 5.43 15.40 1 5.45

36.1 36.3 35.920.o 20.o 20.216.3 1 8.8 17 .82.70 't.82 3.50

10.1 8.59 8.731.64 1.58 1 .430.08 0.04 0.020.10 0.00 0.008.33 8.44 7.680.34 0.30 0.310.00 0.08 0.094.02 3.99 4.00

99.71 99.94 99.66

5.39 5.44 5.382.40 2.35 2.41o.21 0.21 0.218.00 8.00 8.001 . 1 1 1 . 1 9 1 . 1 62.O3 2.36 2.230.09 0.00 0.182.24 ' t .92 1.950.18 0 .18 0 .160.01 0.01 0.000.01 0.00 0.005.66 5.68 5.821 .58 1.62 1.470.10 0.09 0.090.00 0.01 0.011.68 1 .72 1 .57

15.35 15.38 15.25

35.2 35.8 36.120.1 20.2 20.419.3 18 .8 19 .11 .37 1 .10 1 .359.64 9.23 9.451 .94 1.75 1.780.06 0.08 0.000.00 0.12 0.008.23 8.35 8.21o.23 0.32 0.250j2 0.00 0.113.99 3.99 4.04

100.18 99.74 100.79

s.28 5.38 5.372.56 2.50 2.480.16 0. ' t2 0.158.00 8.00 8.001.00 1 .08 1 .092.43 2.36 2.370.00 0.00 0.002.16 2.07 2.090.22 0.20 0.200.01 0.01 0.000.00 0.01 0.005.73 5.75 5.001.58 1 .60 1 .560.07 0.09 0.070.01 0.00 0.011.66 1 .69 1 .64

15.48 15.42 15.39

SiAl (tet)Fe'5* (tet)

Total (tet)Al (oct)Fe+FeF+ (ocDMgTiMnZn

Total (oct)KNABa

Total (int)Total

Note; Weight percentages of HrO were calculated on the basis of 2(OH) and normalized to 22 O. Tetrahedral Fep* was determined from the area ofthe tetrahedral Fe3* doublet (peak A0 on Table 2). Microprobe data represent averages of 10-20 analyses per biotite, depending on the instrumentand the analyst. Standard deviations varied but are approximately 0.30 for SiO,, 0.20 for AlaO3, 0.40 for FeO, 0.25 for MgO, 0.20 for TiO,, 0.05 for

Ttate 1-Continued


o-K-53 0-K-1s O-L-10 O-J-65 0-K-8' O-K-8 0-K-9 0-C-30 0-c-26 0-c-14 0-c-18 0-K-10 s-8/7/63A A A A7' 7', 7', 7'

Sample nameAreaZone

A A B8 8 8 ' ,

A A A A A A8 - 8 - 8 ' 8 8 8

si 35.8 35.6 35.2 35.2At2o! 19.9 19.8 19.6 20.1FeO 19.0 19.6 19.7 19.6Fe,O, 1.84 2.15 2.71 2.42MgO 8.79 8.17 7.78 7 .55Tio, 1 72 1 .80 1.73 2.03MnO 0.19 O 21 0.05 0.06ZnO 0.09 0.10 0.10 0.00K,O 8.46 8.23 8.27 8 60Na2O 0.29 0.27 0.27 O.22BaO 0.00 0.00 0.00 0.07H,O 3.99 3.97 3.94 3.95

Totar 100.07 99 90 99.35 99.80

5.38 5.38 5.36 5 332.52 2.54 2.47 2530.1 0 0.08 0.17 0.14I 00 8.00 8.00 8.001 .01 0.98 1.05 1 .062.39 2 48 2.51 2.490.10 0 .16 0 .14 0 .141.97 1 .84 1 .76 1 .710.19 0.20 0.20 0.230.02 0.03 0.01 0.010.01 0.01 0.01 0.005.69 5.70 5.68 5.641.62 1 .59 1 .61 1 .660.09 0.08 0.08 0.070.00 0.00 0.00 0.001 .71 1.67 1 .69 1 .73

15.40 15.37 15.37 1 5.37

5l

Al (tet)Fe3* (tet)

Total (tet)Al (oct)Fe2*F€F+MgTiMnZn

Total (oct)KNABa

Total (in0Total

rttlc/c

37.3 37.'t 37.91 8.1 19.3 19.612.6 12.6 9.72.00 2.29 0.81

13.1 12.9 15.62 .18 2 .20 1 .970.32 0.32 0.480.10 0.00 0.109.63 9.62 9.300.09 0.07 0.170.00 0.08 0.004.06 4 .10 4 .15

99.48 100.58 99.78Number of atoms

5.51 5.42 5.482.35 2.38 2.430.14 0.20 0.098.00 8.00 8.000.80 0.94 0.901 .56 1 .55 1 .170.09 0.05 0.002.90 2.81 3.350.24 0.24 0.210.04 0.04 0.060.01 0.00 0.015.64 5.63 6.701.82 1.80 1.720.03 0.02 0.050.00 0.01 0.001.85 1 .83 1 .77

1 5.49 15.46 't5.47

35.3 35.5 35.219.3 19.3 20.11 8.6 18.7 18.81 .79 1.80 1 .819.11 8.86 8.542.34 2.46 2.33017 0.20 0.130j4 0.08 0.008.94 8.84 8.810.23 0.28 0.250.00 0.00 0.o23.97 3.97 3.97

99.89 99.99 99.96

5.34 5.36 5.302.51 2.44 2.540.15 0.20 0.168.00 8.00 8.000.93 0.99 1.032.35 2.36 2.370.05 0.00 0.052.O5 1.99 1.920.27 0.28 0.260.02 0.03 0.02o.02 0.01 0.005.69 5.66 5.651 .72 1.70 1 .700.07 0.08 0.070.00 0.00 0.001.79 1 .78 1 .77

15.48 15.44 15.42

35.4 35.1 36.6't9.7 19.7 18.517.9 20.3 15.94.36 2.23 2.886.93 6.93 8.962.62 2.62 1.850.13 0j2 0.080.16 0.16 0.008.47 8.71 8.37o.25 0.22 0.250.00 0.00 0.003.97 3.94 3.93

99.89 100.03 97 .32

5.37 5.33 5.592.64 2.47 2.130.19 0.20 0.288.00 8.00 8.001 .05 1 .06 1 .192.26 2.58 2.030.30 0.06 0.051 .56 1.57 2.O40.30 0.30 0.210.02 0.02 0.010.02 0.02 0.005.51 5.61 5.531 .63 1 .69 1.630.07 0.07 0.070.00 0.00 0.001.70 1 .76 1 .70

15.21 15.37 15.23

Sample name 16- 4- 16- 15- 14- 5-oH72

cor1

8',8'AreaZone

8t23t60 7t26t59 7/18/60 8/18/59 712160 9120161 0-b-41B B B B B B B

oG92 0H49P OH56c c c

oB16

sio,Atro3FeOFerO.MgoTio,MnOZnOKrONa2OBaOH.O

Total

36.3 34.5 34.4 35.9187 18 .6 19 .9 19 .516.7 17 .6 17 .2 15 .13.54 3.18 2.36 1.467.95 8.16 8.40 10.62.06 2.44 2.94 3.240.20 0.29 0.16 0.320.00 0.00 0.00 0.068.45 9.08 9.44 9.380.36 0.27 0.16 0.100.00 0.00 0.00 0.143.94 3.88 3.93 4.O2

98.20 98.00 98.89 99.82

5.52 5.34 5.25 5.342.12 2.48 2.48 2.600.36 0.18 0.27 0.068.00 8.00 8.00 8.001 24 0.91 1.09 0.83213 2.28 2. ' t9 1.890.05 0.19 0.00 0.101.81 1 88 1.91 2.360.24 0.28 0.34 0.360.03 0.04 0.02 0.040.00 0.00 0.00 0.015.50 5.58 5.55 5.591 .64 1.79 1.84 1.780.11 0.08 0.05 0.030.00 0.00 0.00 0.011.75 1 .87 1 .89 1 .82

15.25 15.45 15.44 15.41

wlch35.3 34.6 36.819.1 19 .0 17 .518.0 17 .0 20.62.73 3.33 3.426.49 6.61 6.153.58 3.07 2.260.19 0.21 1.030.00 0.00 0.009.15 8.70 9.460.12 0.21 0.020.00 0.00 0.063.92 3.85 3.96

98.58 96.58 101 .26Number of atoms

5.41 5.40 5.582.41 2.27 2.130.18 0.33 0.298.00 8.00 8.001.O4 1.22 0.992.30 2.21 2.610.13 0 .05 0 .101 .48 1.54 1 .390.41 0.36 0.260.03 0.03 0.130.00 0.00 0.005.39 5.41 5.481 .79 1.73 1.830.04 0.06 0.010.00 0.00 0.001 .83 1.79 1 .84

15.22 15.20 15.32

35.4 35.3 35.519.7 19.9 19.315.0 14.9 14.84.17 3.89 4.909.61 9.97 9.931.57 1 .45 1 .780.22 0.19 0.240.07 0.07 0.088.90 9.00 8.970.20 0.21 0.250.10 0 .1 1 0 .123.97 3.98 4.00

98.91 98,97 99.87

5.35 5.33 5.322.46 2.51 2.460.19 0.16 0.228.00 8.00 8.001.04 1.02 0.951.90 1 .88 1 .850.28 0.28 0.342.16 2.24 2.220.18 0.16 0.200.03 0.02 0.030.01 0.01 0.015.60 5.61 5.601.72 1.73 1.720.06 0.06 0.070.01 0.01 0.011 .79 1.80 1.80

1 5.39 15.41 15.40

35.0 35.2 36.719.2 18.6 19-214.1 15.5 7 .455.79 5.13 7.059.72 9.17 12.92.03 2.31 1 1.32o.21 0.30 i 0.230 . 0 6 0 . 0 3 ' 0 . 1 08.96 9.29 7.340.18 0 j4 0 .1 8o.22 0.'t7 0.053.98 3.97 4.03

99.45 99.81 96.55

5.27 5.33 5.472.49 2.37 2.430.24 0.30 0.108.00 8.00 8.000.93 0.94 0.931.78 1.95 0.930.41 0.28 0.69218 2.07 2.860.23 0.26 0.150.03 0.04 0.030.01 0.01 0.015.57 5.55 5.601 .72 1.79 1 .390.05 0.04 0.050.01 0.01 0.001.78 1.84 1.44

15.35 15.39 15.04

SiAl (tet)pss* (tet)

Total (tet)Al (oct)Fe.-

Fe3* (oct)Mgl l

MnZn

Total (oct)KNaBa

Total (int)Total

MnO, 0.05 lor ZnO,0.15 for KrO, 0.06 for NarO, and 0.05 for BaO. Significant digits are tabulated using the general convention of two significant digitsafter the decimal point for wt% oxides <10.0 and one digit for >10.0. Relative atom numbers are presented with two places after the decimal forconsistency; it is impossible to accurately estimate their accuracy given the uncertainties in H and O content of these samples.

660

Taeue 2. Mossbauer data


Samole Zone lS 0 Q S O A O t s 1 Q s 1 A 1 I S 2 Q S 2 A 2 r s 3 Q S 3 A 3

Ra-d75-66Ra-d72-66Ra-d9-66Ra-d28-66Ra-c95-66Ra-d86-66Ra-d1 1-66Ra-d37-66Ra-d1 2-66Ra-d1 4-66Ra-d60-66Ra-a65-66Ra-c1 -66Ra-c4-66Ra-a73-66Ra-a1 4-66Ra-a69-66Ra-a33-66Ra-b53-66Ra-b48-66Ra-a57-66Ra-a96-66Ra-a93-66Ra-84-66Ra-841-66Ra-A95-66o-K-53o-K-15o-L-10o-J-65o-K-8',o-K-8o-K-9o-c-30o-c-26o-c-14o-c-18o-K-109-8t7t631 6-8/23/604-71261591 6-711 8/601 5-8/1 8/5914-7121605-9t20161o-b-41or1oG92oH49',oH56oH72oB16Ave. RangeleySD RangeleyAve. all samplesSD all samples

112222a

a

2

e

4444

oo

77

8-8-8-Ie

n888',8',8',Io

o

0 1 4 0 . 1 70 1 4 0 . 1 80.15 0.200.07 0.310.14 0.260.21 0.20014 0 .210.18 0 .680.09 0.200.11 0 .250.15 0.210.13 0 .180.14 0 .190.09 0.230.19 0 .160.08 0.210.23 0.190.21 0.180 15 0.350.15 0 .160 13 0.220.13 0.220.16 0.210.05 0.290.19 0.270.05 0.280.20 0.450.20 0.490.21 0 .160.16 0.250.19 0.230.07 0.190.09 0.320.11 0.35u, | / u.co0.16 0 .130.15 0.340.19 0 .140.13 0 .160.12 0 .180.20 0.140.13 0 .160.16 0 .280.15 0.340.14 0 .180.13 0.330.13 0.220.15 0 .150.13 0 .190.13 0.240.16 0 .190 16 0.640.15 0 .250.04 0.110.15 0.250,04 0j2

0.41 0.710.37 0.68

0.37 0.68

0.39 0.69

0.44 0.710.40 0.91

0.39 0.80

0.43 0.60

0.43 0.57

0.38 0.490.38 0.730.38 0.630.36 0.780.38 0.630.38 0.580.39 0.740.38 0.760.40 0.680.02 0.110.39 0.690.02 0.10

0.40 0.930.39 0.940.38 0.78

0.45 0.940.43 0.840.39 0.950.39 1.230.38 0.830.40 0.840.38 0.84

0.39 0.87

0.35 0.990.37 1 .100.38 0.92

0.39 0.90

0.38 1.29

o.971.060.87

0.821.01

0.38 1.21

0.41 0.930.35 0.850.34 0.860.39 0.920.47 1.010.37 0.90

0.38 0.78

0.39 1.340.39 1 .190.38 1.230.39 1.030.39 1.060.38 1.21o.41 1.270.39 0.940.03 0.130.39 0.990.03 0.16

1.07 2.29 231.07 2.23 191.08 2.30 221.09 2 .19 211.14 2 .26 311.09 2.36 241.07 2.32 301.09 2.24 211.06 2.30 201.08 2.27 221.07 2.26 171.09 2.25 201.02 2.23 211.08 2.26 241.09 2.21 221.08 2.23 201 . 1 0 2 . 1 9 2 11.09 2 .17 201.12 217 201.09 2.22 221.06 233 291.10 2.30 241.05 2.26 191.12 2 .29 431.10 2 .22 321.11 2.30 441.10 2 .04 171 . 1 0 2 . 1 1 2 81.09 2.22 251.06 2.20 241 09 2.29 331.09 2.3'l 331 . 1 1 2 . 1 3 1 51 . 1 1 2 . 1 3 2 21.09 2.08 251.09 2.21 231 .13 2.09 171.09 2.20 281.06 2.21 231.06 2.25 241 .11 2.21 241.07 2.21 251.09 2.22 251 .10 2.09 211.06 2.22 251.08 2.34 271.13 2 .59 641.12 2 .58 651.12 2 .60 581.12 2 .58 561.13 2 .58 561 . 1 2 1 . 9 5 91.09 2.22 24.20.02 0.07 5.91 .09 2.26 27.40.02 0.13 12.2

1 0I7I4o

1 26

1 11 18I

1 0q

1 1

o

oII65o43o5I

1 1

oa

o

1 21 4

1 15

1 2

1 ' l8

o1 01 2o7 .82.58 .02.5

2

2

34

4

J

44c

2b

452o

c

44

4

e

465

c

3

21 1222

c

2c

5o

1 04

1 24.32.04.72.4

0.430.430.40

0.380.39

I

227

I

7284.22 .06.45 .8

Note:Zone:Metamorphicgrade:1: lowergarnet ,2:middlegarnet ,S:uppergarnet !4: lowerstaurol i te,5:upperstaurol i te,6: t ransi t ionzone,7 : lower sillimanite (1-7 from Rangeley quadrangle), 7' : lower sillimanite, 8- : sulfide-bearing lower sillimanite, 8 : upper sillimanite (7'-8 fromOquossoc quadrangle), 8' : upper sillimanite, I : potassium teldspar + slllimanite (8' and g from Bryant Pond quadrangle). lS : isomer shift, QS :quadrupofe splitting,4 : peak arca(inah of total area). Peaks 0-4 represent Fep+ in tetrahedral, octahedral M2, and octahedral M1 sites and FeF+ inoctahedral Ml and octahedral M2 sites. % MIS - % MlSFlT, and % UN : % uncertainty in the fit (see Ruby, 1973).

fitted as one merged doublet with averaged parameters of6 = 0.42 mm/s and A = 0.85 mm/s.

M2lMl ratios

Because the higher velocity Fe2+ peaks are not over-lapped by Fsr+ peaks, it is possible to determine relativesite occupancies with some confidence (+0.4-0.6). Theratio of Fe2+in M2 to that in Ml (M2lMl) ranges from

a low of l.14 to a high of 5.20, with an average value of2.91. ln these samples, there are few biotites with the 2: Iratio that would be predicted by equipoint ranks for IMbiotite. There are 45 biotite samples in our data set thathave greater than twice the amount of Fe2+ in M2 thanin M l, but there are also seven samples for which thereverse is true. Previously published ranges of Fe2+ M2lMl site occupancy ratjos are similar to those of the Maine

DYAR: MOSSBAUER SPECTRA OF BIOTITE 66r

Taate 2-Continued

r s 4 0 s 4 A 4 o/o MIS % UN M2lM1 "hFe3'

1.12 2 .621.12 2 .601.13 2 .621 .13 2.601.14 2 .611 .14 2.691 .13 2.631.13 2 .611.12 2 .611.13 2 .621 . 1 2 2 . 6 11.13 2 .611 12 2 .591 . 1 2 2 . 6 11 . 1 2 2 5 91.12 2 .601 13 2.611.13 2 .611.13 2 .601.13 2 .591.13 2 .641 .13 2.661.12 2 .601.13 2 .631 .13 2.601.13 2 .621.11 2 .561 .10 2 .571.12 2 .601 .10 2 .571.13 2 .631.12 2 .641 12 2 .581.13 2 .571 . 1 1 2 . 5 51.12 2 .591 . 1 2 2 5 71.12 2 .591.12 2 .581.12 2 .591.14 2 .651.12 2 .581.13 2 .621.12 2 .571.12 2 .581.14 2 .641.06 2.241.08 2 .171.06 2.241.06 2.201.04 2.261.13 2 .541.12 2 .600.01 0.031.12 2 .570 02 0.12

0.09 0.010.02 0.010.03 0.010.11 0 .020.08 0.010 10 0.030.15 0 .02

-0.15 -O.O20.28 0.010.13 0.020.02 0.010.09 0.010.00 0.000.16 0.020.09 0.010.11 0.02

-0.06 -0.02-052 -0.07

0.27 0.01-0.06 -0.01

0.25 0.01-1 .08 -0 .15

0.18 0 .01-0.06 -o 02-0.02 -0.01-0.41 -0.07

0.20 0.030.09 0.020.15 0 .010.09 0.020.01 0.01

-0.03 -0.010.11 0 .030.10 0 .010.16 0.03

-0.22 -0.03-0 05 -0.02-0.05 -0.01

0.06 0.010.01 0.01

-0.01 -0.030.02 0.010.04 0.010.06 -0.020.04 0.010.04 0.010.04 0.010.11 0 .010.16 0.010.06 0.010.05 0.01o.27 0.02

2.70 14.03.58 13.02.95 12.03 .19 14 .01.94 8.02 66 12.01 .68 19 .02.95 17.03 .13 16 .02.90 13.04.06 14.03.50 10.03.05 15.02.66 12.O3.05 11 .03 .45 11 .03.33 9.03.50 10.03.40 12.O3.18 7 .02.00 13.02.83 8.03.47 15.01 .19 6 .01 .97 5 .01 . 1 4 6 . 04.41 8.02.25 9.02.50 1 1.02.75 10.01 .60 13.01 .61 14 .05.20 7.03 .18 8 .02.64 8.02.96 8.03.70 18.02.25 9.02.70 14.02.50 16.02.58 14.02 .56 11 .02 .44 11 .03.19 12.02.72 15.02 .11 15 .04.00 20.03.82 19.03.22 23.03.25 27.02.67 23.05.00 46.02.83 11.4o.78 3.42.91 13.20.82 6.5

6268oca l6064506263646970646467697070687058686651OJ

5Ut 56364oo5353787A6668636362606264616768C T

1 61 71 81 7214564.25.9

59.214 .9

I

,fl\

biotite samples (Dyar, 1987 for trioctahedral micas;DeGrave et al., 1987 for chlorites). Thus, the data sup-port the observation that Fe2+ is more commonly foundin M2 than in Ml sites. This does not necessarily implythat Fe2+ has a crs-M2 site preference. It is equally likelythat some other cation (or cations) has a site preferenceand that Fe2+ is randomly occupying the remaining avail-able sites. Additional support for this idea is the obser-vation (Tables I and 2) that ordering of Fe2* in M I andM2 does not appear to vary with metamorphic grade, Ticontent, or Fe2+/Mg ratio. The observed ordering is causedby some more complicated (or less obvious) cause. Un-

l 1

r r r l t l l l- 3 - 2 - 1 0 1 2 3 4

MM/SEC

Fig. l. Three nearly identical Mtissbauer spectra of biotitesamples from the upper sillimanite zone of the Oquossoc quad-rangle. Fits of these data are identical well within the knownprecision of the Mcissbauer technique.

fortunately the compositions of these biotite samples aretoo complex to allow assignment of site preferences toindividual cations.

Owing to the small areas and extensive overlap of thepeaks for Fe3+, quantitative conclusions about the distri-bution of Fe3+ between the cls-M2 and trans-Ml sites inthe octahedral sheet cannot be drawn.

Tetrahedral Fe3+

The presence of tetrahedral Fe3+ in amounts rangingfrom 0.09 to 0.37 cations per formula unit (based on 22O) was confirmed in all 52 of the samples studied (Fig.2B). This result is unexpected given that all of our sam-ples are saturated with respect to Al and therefore containsufficient Al3+ to fill the tetrahedral site. Because the Al3*

I

I

I

tiI

I-4

na nana nana nana na

662 DYAR: MOSSBAUER SPECTRA OF BIOTITE

oLUFl--H

azE.F

;'e

100 . 00

99 . B0

99 .60

99 .40

99 .20

99 .00

98 .80

98 .60

98 .40

98 .20

98 .00

97 . B0

97 .60

MM,/SEC

Fig.2. Two fits to the same Mdssbauer spectmm of biotite O-L-10. The fit shown above (A) has two doublets representing Fe'?+and two doublets representing Fe3*, all in the octahedral sheet. The fit at the right (B) has only one octahedral Fe3+ doublet andone tetrahedral Fe3+ doublet. Even by visual inspection it is clear that the latter fit is superior; this conclusion is borne out by thestatistics ofthe fit.

3I- I-2-3- 4

cation (0.39 A; is so much smaller than Fe3+ (0.49 A) intetrahedral coordination (Shannon, 1976), it seems im-probable that Fe3+ would preferentially substitute for A13+.However, the Miissbauer data presented here and in nu-merous examples in the literature (e.g., Ferrow, 1987) canconsistently be fitted with tetrahedral peaks. Either thereis a problem with the interpretation of the Mtissbauerdata, or some unconsidered factor is favoring Fe3+ oc-cupancy ofthe tetrahedral sites.

Previous work by Dyar and Burns (1986) has soughtto clarify the interpretation of Mdssbauer data for biotitethrough study of synthetic and naturally occurring end-member annites and ferriannites in which Al-deficientstoichiometries necessitate tetrahedral Fe3*. This pastwork,'coupled with a literature review of Mdssbauer data

for trioctahedral micas (Dyar, 1987) has established firm-ly the expected position oftetrahedral Fe3+ peaks in phyl-losilicate spectra. These peak positions are similar to thoseof four-coordinated Fe3+ in other silicates such as feld-spars (Levitz et al., 1980; Huggins, 1976) and sillimanite(Rossman et al., 1982). Therefore, it seems unlikely thatour interpretation of the Mdssbauer data is incorrect.

It seems equally unlikely that our well-developed un-derstanding of mica crystal chemistry is incorrect. Av-erage octahedral cation radii in these samples range from0.70 to 0.72 A (roughly the ionic radius of Mg2+), if va-cancies are considered to have ionic radii of about 0.80A (Guggenheim, 1984). This size range is roughly in themiddle of the range of that of all observed natural micacompositions and is similar to that of pure phlogopite.


oLrlFFH=azCE

;,e

100 .00

99 .80

99 .60

99 .40

99 .20

99 .00

98 . B0

98 .60

98 .40

98 .20

98 .00

97 .80

97 .60-4

Fig. 2.-Continued.

In this octahedral cation size range the lateral dimensionof the octahedral sheet is larger than that of the AlSi3tetrahedral sheet, and rotation of adjacent tetrahedrawithin the plane of the sheet must occur in order to re-duce the lateral dimensions of the tetrahedral sheet.Therefore, ifthe average octahedral ionic radius of0.7lA is used to predict site occupancies in these samples,tetrahedral Fe3+ would certainly not be expected becauseits presence would only enlarge the size ofthe tetrahedraand increase the size mismatch between the tetrahedraland octahedral sheets. Thus, arguments based on averageoctahedral ionic radii cannot explain the presence oftet-rahedral Fe3+.

However, because of the complex compositions of thesemicas, local structural environments may be very differ-ent from the average structure. Some local compositionsmay optimize the fit between tetrahedral and octahedralsheets. The number of tetrahedral Fe3+ cations per for-mula unit is relatively small (0.1-O.3 Fe3+ per eight tet-

0

MM/SEC

rahedral cations). Therefore a distortion of only one in27-80 sites would be required to accommodate the ob-served tetrahedral Fe3+ contents.

A likely candidate for the cause ofsuch a local distor-tion is a vacancy in the octahedral sheet. In 2M, mus-covite. for instance. there are two Al atoms and one va-cancy corresponding to the three octahedral sites (two M2and one Ml, respectively). Shared edges between occu-pied octahedra are shortened (Bailey, 1984). Thus, theentire sheet is thinned in the c direction and expanded inthe a-b plane.

In contrast, octahedral vacancies in trioctahedral micasare less commonly observed (0.24.4 vacancies per sixoctahedral cations), so overall structural adjustments ofthe type found in muscovite would not be expected.However, local scale adjustments probably occur becausevacant octahedra are large and edge lengths ofthe coor-dination polyhedra will therefore be longer. Local scaleadjustments of the tetrahedra in the vicinity of the en-

red

]-^-----

664

U A V E N U H B E R , . . - tt0go6

500 7go eoo l I a0 | 300 r 504HAVELENGTH, nm

IJAVENU|IBER, ".- l3taogg za8go tooag

orange

3SO SO0 7@ oO0 | tgo t300IIAVELENGTH, nrn

tfAVENUIIBER, cm-l3s9oo 2gooo tog9{,

green

s00 7A 9,gg I t@ t loe t6@

IIAVELENGTH, nm

larged octahedra will also be required. Perhaps local ad-justment of the geometry of the tetrahedral sites createsa larger or more distorted fourfold site that can as easilyaccommodate Fe3+ as Al3+. The striking similarity be-tween the concentration of tetrahedral Fe3+ and that ofoctahedral vacancies strongly suggests that the two arerelated.

Additional corroboration for the assignment of Fe3+ totetrahedral coordination comes from the color of the mi-


2 5

Fig. 3. Visible-region absorption spectra of mica thin sec-tions (cut normal to the cleavage) showing absorption owing toO, Fe3+, Fe2*, and Tia+. The top spectrum of an Al-deficientFe3+-bearing clintonite shows absorption bands at 22700,20300,and 19200 cm-' (characteristic of tetrahedral Fe3*) superim-posed on the tail of a UV band, which is most likely owing toFe3+-O charge transfer (Rossman, 1984). The middle spectrumhas strong absorption in the 20000-24000 cm-' region, whichis characteristic ofTi4+-Fe2+ interactions. The bottom spectrumhas peaks at 9000, 11000, and 14000 cm-', which representcontributions from Fe2* and Fe3+-Fe2+ charge transfer. Samplethicknesses are 374,181, and 209 pm, respectively. Solid linesrepresent light polarized parallel to cleavage, dashed lines rep-resent light perpendicular to cleavage.

cas. Biotite from our graphite- and ilmenite-bearing rockstends to be red-orange in plane-polaized light, whereasthe biotite that coexists with magnetite and hematite (cor-responding to more highly oxidizing conditions) is green.For the most part these colors arise from interactionsamong O, Fe3*, Fe2+, and Tia+ ions. Typical optical ab-sorption spectra of micas are shown in Figure 3. Biotitewith a large Tia+ concentration absorbs strongly in the20000-24000 cm-r region of the visible region spectrum,transmitting the red-orange color commonly associatedwith Ti (Guidotti, 1984). This color is the result of inter-action between Fe2+ and Ti4+ (Faye, 1968) accompaniedby substitutional broadening ofthe Fe-O charge-transferband in the near-UV (Rossman. 1984).

Assemblage: Graphite/llmenite Magnet Hema--ite tite

0Grade' Gt St St+ Sil Ksp Mixed St+

Sil +Sil Sl l

Frg. 4. PIot of octahedral Fe3+ vs. assemblage and oxide,showing a minimum Feg content of about 40lo of the total ironfor graphite + ilmenite-bearing assemblages regardless of meta-morphic grade. Error bars are calculated with standard devia-tions except for hematite grade (represented by a single sample),where error bar shown is estimated precision. Most variation inoctahedral Fe3+ content above the 40lo level occurs as a result ofincreasing oxidation state.

IJ

2 2s

P r soc o t 0

lrJ(Jz@uoU'o

lrJc,zctuooo

0 5

t . 8t 6l 1

1 . 2l . os . 80 . 64 . 19 . 2

9 . 8

o . o

9 . 1

9 . 2

I soo

40

35

'or 30l'l-

G 2 5It

f ; ,0oo t ss

1 0

III

Absorption effects also arise from Fe3+ and Fe2+ in theoctahedral sheet. Octahedral Fe2+ alone gives rise to ab-sorption bands at 9000 and I 1000 cm ', and the inten-sity of those bands increases nonlinearly as octahedralFe3+ increases (Smith et al., 1980). An Fe3+-Fe2+ charge-transfer band occurs around 14000 cm '. All three Febands (9000, I1000, and 14000 cm-') are comrnonly ob-served in spectra of biotite with high concentrations ofFe3+, Fe2+, and Ti4+ (Rossman, 1984).

The biotite samples in the present study fall into twogroups: reddish orange samples containing only about 120loof the total iron as Fe3+, and greenish brown biotite withhigher total Fe3+ values. Any optical absorptions thatmight arise from tetrahedral Fe3+ in these samples areapparently small relative to the intensity of the Ti4+-Fe2+and Fe3+-Fe2+ charge-transfer bands (Fig. 3). However,the red-orange color of the low Fer+ biotites from themost reduced assemblages indicates the presence of Tia+-Fe2+ absorptions without any significant contribution fromoctahedral Fe3+-Fe2+. Further confirmation of this con-clusion is the observation that the color of the biotites inthe reduced rocks changes systematically as a function ofTi content. As Ti increases (with increasing metamorphicgrade) the color changes from reddish orange to reddishbrown to chestnut brown. Thus, the darkening reddishshades reflect the Ti increase (which results in increasedTi4+-Fe2+ charge transfer) in a straightforward (thoughnot necessarily linear) fashion.

The greenish brown color associated with octahedralFe3+-Fe2+ charge transfer is only observed in the biotitefrom the more oxidizing assemblages. This biotite con-tains just as much Ti as biotite from similar metamorphicgrades with more reduced assemblages. Within a givenmetamorphic grade, bulk elemental compositions of bio-tites do not vary with oxidation state; therefore, the colordifference can be directly related to the concentration andoccupancy of Fe3*. If Fe3* is present in the octahedralsheet at levels above -0. I atoms/22 O atoms, there issignificant Fe3+-Fe2+ charge transfer, and the efects ofthe Ti4+-Fe2+ interaction are no longer visible. Biotiteconspicuously transmits in the 14000 cm-t charge trans-fer region. Therefore, the color of the biotite with lowerFe3+ contents from reduced assemblages indicates a lackof significant octahedral Fe3+ in those samples. The as-signment of Fe3+ to tetrahedral coordination based onMdssbauer data is therefore supported.

Variation over a range of metamorphic grade

Within the known errors of the Mdssbauer technique,neither the Fe3+ contents nor the Fe2+ M2/Ml ratios ofour micas display systematic variation that can be cor-related with changing metamorphic conditions. Only onesmall group of samples appears to be distinctive. Onesubset of micas from the lower sillimanite grade in theOquossoc quadrangle coexists with >30/o modal pyrrho-tite (samples O-K-9, O-K-8, and O-K-8'). As discussedin Guidotti et al. (1988), these samples have small Fecontents due to reactions in the rocks that draw Fe into

665

pyrrhotite. Although the total amount of Fe in these sam-ples is low, the percentage of Fe3+ remains relatively con-stant. This is significant because it implies that the amountof tetrahedral Fe3+ per formula unit is not constant butvaries with composition. These three samples show ex-tremes of Fe2+ ordering that probably reflect the relatedincrease in octahedral Mg. The fact that Fe2+ occupancieschange significantly where Mg increases dramatically sug-gests that Mg may have a strong site preference, probablyfor M2 (M2/Ml ratios in O-K-8 and O-K-8' are 1.60 and1.61, respectively). The slightly different bulk composi-tion of O-K-9 yields a reverse M2/Ml ratio of 5.20. Re-grettably, our three-sample data set does not provide suf-ficient information to explain why these extremes exist.

The amount of tetrahedral Fe3+ is nearly constant atan average of 8 + 2.50/o of the total Fe over the range ofmetamorphic grade studied. In the most reduced rocks,those from ilmenite * graphite assemblages, 2/t of theFe3+ present accounts for the 80/o present in the tetrahe-dral sites. Even in biotite samples from the more oxidiz-ing environments in which > l5-20o/o of the total iron isFe3+, only 8o/o goes into tetrahedral coordination, withthe balance in octahedral coordination. This observationstrongly suggests that tetrahedral Fe3+ is present becausethe structure requires it for stability its existence is notrelated to the oxidation state of the assemblage. This ideais discussed in detail in the companion paper (Guidottiand Dyar, in preparation). Only the amount of octahedralFe3+ varies systematically with the oxidation state in therocks (Fig. 4).

Coxcr,usroNs

M0ssbauer spectra of 52 biotite samples from low gar-net to potassium feldspar + sillimanite-grade metapelitesin northwestern Maine show almost constant peak posi-tions of Fe2+ in cis-M2 sites but greater variation in hy-perfine parameters of Fe2* in trans-Ml. This observationsuggests that the geometry of the M2 site is relativelyconstant, but that the average trans-Ml site is distortedto different degrees in different samples. This distributionmay be caused by variation in the average number ofoctahedral vacancies or in the next-nearest-neighbor cat-ions. Fe3+ peaks can sometimes be resolved as two dis-tinct doublets corresponding to Fe3+ in the two types ofoctahedral sites. More frequently, the spectra are best fit-ted with a single combined octahedral Fe3* doublet withhyperfine parameters representing averages of cls-M2 andtrans-Ml and a total of about 40lo of the total peak area.An average of 80/o of the total iron occurs as tetrahedralFe3* in all of the samples studied. Its presence may becorrelated with the number of octahedral vacancies.

Viewed in a petrologic context, these data assumegreater significance. For the most part, Fe3+/)Fe and Fe2+M2/Ml ratios do not vary over the broad range of meta-morphic grade examined here, except in samples con-taining pyrrhotite. Octahedral Fe3+ content does increasein biotite from increasingly oxidized assemblages. Thefact that the concentration of tetrahedral Fe3+ remains


666

constant over the wide range in metamorphic grade stud-ied suggests that structural requirements within the bio-tite, rather than variation in oxidation state, are respon-sible for its presence.

AcxNowr,nncMENTS

The author thanks Charles Guidotti for generously making his samplesavailable for this study and Don Hickmott for making the suggestion thatgot this whole project started. Conversations with M. Reed, T. Grover,S. W. Bailey and S. Guggenheim and reviews by E. DeGrave, E. S. Grew,S. Guggenheim, R. K. Kirkpatrick and C. R. Rebbert are gratefully ac-knowledged. Thanks also to Roger Burns and George Rossman for use oftheir analytical facilities and for support of the author during the earlystages of this work. I am indebted to Chryl Perry, Carolyn Rebbert,Kathleen Ward, and Dick Ziegler for assistance in data reduction and toAnne V. McGuire and Beth Holmberg of the Smithsonian AstrophysicalObservatory, Cambridge, Massachusetts, for last-minute electron micro-probe analyses needed to complete this paper. This work was supportedthrough N.S.F grants EAR-8709359 and EAR-8816935. Acknowledg-ment is also made to the donors of the Petroleum Research Fund, ad-ministered by the A.C.S., for.additional support ofthis research through

lJant 19217-G2.

RnronrNcns crrnnAlbee, A.L, and Ray, L. (1970) Correction factors for electron probe

microanalysis of silicates, oxides, carbonates, phosphates, and sulfates.Analytical Chemistry, 42, 1408-1414.

Bailey, S.W. (1984) Micas. Mineralogical Society of America Reviews inMineralogy, 13, 584 p.

Bancroft, G.M., and Brown, J.R. (1975) A Mdssbauer study of coexistinghomblendes and biotites: Quantitative Fe3+/Fe2+ ratios AmericanMineralogist, 60, 265-272.

Bence, A.E., and Albee, A.L. (1968) Empirical correction factors for theelectron microanalysis ofsilicates and oxides. Journal ofGeology, 76,382403.

DeGrave, E., Vandenbruwaene, J., and Van Bockstael, M. (1987) tFe

Mdssbauer spectroscopic analysis of chlorite. Physics and Chemistryof Minerals. 1 5. 1 73-1 80.

Dodge, F.C.W., Smith, V.C., and Mays, R.E. (1969) Biotites from graniticrocks ofcentral Sierra Nevada batholith, California. Journal ofPetrol-ogy, 10,250-27 L

Dyar, M.D. (1984) Precision and interlaboratory reproducibility of mea-surements of the M6ssbauer effect in minerals. American Mineralogist,69, tt27-tt44.

-(1987) A review of Mdssbauer data on trioctahedral micas: Evi-dence for tetrahedral Fer+ and cation ordering. American Mineralogist,72, t02-112.

Dyar, M.D, and Burns, R.G (1986) Mdssbauer spectra study of femr-ginous one-layer trioctahedral micas. American Mineralogist, 7 1,951-9 6 1

Dyar, M.D, Grover, T.W., Rice, J.M., and Guidotti, C.V (1987) Pres-ence of ferric iron and octahedral ferrous ordering in biotites from


pelitic schists: Implications for garnet-biotite geothermometry. Geo-

logical Society of America Abstract with Programs, 19, 650.

Dyar, M.D., Hickmott, D., Guidotti, C.V., and Cheney, J.T' (1986)M2/

M I ordering ofiron in biotites from northwestern Maine' International

Mineralogical Association Abstracts, 14, 94.Faye, G.H. (1968) The optical absorption spectra of certain transition

metal ions in muscovite, lepidolite, and fuchsite. Canadian Journal of

Earth Sciences, 5, 3l-38.Ferrow, E. (1987) Mdssbauer effect and X-ray diffraction studies of syn-

thetic iron bearing trioctahedral micas. Physics and Chemistry of Min-

erals. 14.276-280.Guggenheim, S (1984) The brittle micas In S.W. Bailev, Ed., Micas.

Mineralogical Society of America Reviews in Mineralogy, 13' 6I-104'

Guidotti, C.V. (1984) Micas in metamorphic rocks. In S W' Bailey' Ed.'

Micas. Mineralogical Society of Amenca Reviews in Mineralogy, 13,

357467.Guidotti, C.V , Cheney, J.T., and Guggenheim, S. (1977) Distribution of

titanium between muscovite and biotite in metapelites from N'W'

Maine. American Mineralogist, 62, 438448.Guidotti, C.V., Cheney, J.T., and Henry, D.J. (1988) Compositional vari-

ation of biotite as a function of metamorphic reactions and mineral

assemblage in the pelitic schists of westem Maine. American Journalof Science. 288A, 270-292.

Henry, D.J. (1981) Sulfide-silicate relations of the staurolite-gade pelitic

schists, Rangeley quadrangle, Maine,773 p. Ph.D' thesis University

of Wisconsin, Madison, Wisconsin.Huggins, F.E. (1976) Mdssbauer studies ofiron minerals under pressures

ofup to 200 kilobars. In R A. Strens, Ed., The physics and chemistry

ofminerals and rocks, p.613-640. Wiley, LondonLevitz, P., Callas, G., Bonnin, D, and kgrand, A.P. (1980) Etude par

spectroscopie Mdssbauer du Fer(III) dans des verres silicates multi-

composants d'int6r€t g6ologique. Revues de Physique Applique, 15'

rr69-1t73.Rossman, G.R. (1984) Spectroscopy of micas. In S'W. Bailey, Ed., Micas.

Mineralogical Society of America Reviews in Mineralogy, 13, 61-104.

Rossman, G.R., Grew, E.S., and Dollase, W.A. (1982) The colors of sil-

limanite. American Mineralogist, 67 , 7 49-761.Ruby, S.L. (1973) Why misfit when you already have X'?? In I'J. Gruver-

man and C.W. Seidel, Eds., Mdssbauer effect methodolory, 8, 263-

276. Plenum Press, New York.Shannon, R.D. (l 976) Revised effective ionic radii and systematic studies

ofinteratomic distances in halides and chalcogenides. Acta Crystallo-graphica, 432, 7 5l-7 67 .

Smith, G., Howes, B., and Hasan, Z. (1980) M6ssbauer and optical spec-

tra ofbiotite: A case for Fe3+-Fe2+ interactions. Physica Status Solidi,

A57, Kl87-K192.Stone, A.J., Parkin, K.M., and Dyar, M.D (1984) sroNe A progam for

resolving Mdssbauer spectra: DEc Users' Society Program Library, no.

I l-720 Marlboro, Massachusetts.Wood, C. (1981) Chemical and textural zoning in metamorphic garnets,

Rangeley Area, Maine, 586 p. Ph.D. thesis. University of Wisconsin,

Madison, Wisconsin.

MeNuscnrrr RECETvED Jurv 5, 1989MeNuscnrsr AccEprED Fsgnuenv 28, 1990

miissbauer spectra of biotite from metapelites

Documents