American Mineralogist, Volume 69, pages 4724E0, 194

Donpeacorite, (MnrMg)MgSi2O61 r n€\il orthopyroxene and its proposed phaserelations in the system MnSiO3r-MgSiO3-FeSiO3'

Enrcn U. PBrr,nseN.2 LewneNce M. ANovrrz, nNo Enrc J. EsseNn

Department of Geological SciencesUniversity of Michigan, Ann Arbor, Michigan 48109

Abstract

A Mn-rich orthopyroxene Mgl.alMns.5oCao.orSizOe, occurs in a manganiferous pod in themarble units near Balmat, N. Y. The pyroxene coexists with triodite, tourmaline, ferrianbraunite, manganoan dolomite , and hedyphane. Refinement of the crystal structure of thepyroxene (space group Pbca) shows that as expected virtually all of the Mn is located in theM2 site and the general formula is therefore (Mn,Mg)MgSi2O6. Because Mn is completelyordered into the M2 site and comprises over 50 percent of that site, the mineral is a neworthopyroxene. It has been named donpeacorite after Donald R. Peacor in recognition ofhis work on pyroxenes and manganese minerals. The unit cell parameters are a =

18.384(ll), b : 8.87E(7), c : 5.226(3),2 = 8,D (meas.) : 3.36(l). Donpeacorite is biaxialnegative with 2V* : 88', Alc, a = 1.677(2), B : 1.684(2), y = 1.692(2). It is yellow-orangewith a vitreous luster and has an approximate hardness of5-6 with perfect (ll0) cleavages.

The occurrence of donpeacorite and published data for metamorphic pyroxenes andpyroxenoids permit modeling of a phase diagram for the system MnSiOrMgSiO3-FeSiO3.The diagram is dominated by extensive fields for solid solutions of orthopyroxene and ofpyroxmangite, with a small field inferred for kanoite (MnMgSizOo, C2lc clinopyroxene) anda three-phase field for ferroan kanoite, manganoan hypersthene, and magnesian pyroxman-gite.

Introduction

Several Mn-bearing minerals have been described fromthe manganoan pods in the Balmat, New York areaincluding tirodite (Ross et al.,1969), magnesian rhodonite(Peacor et al., 1978), calcian kanoite, braunite, and hol-landite (Brown et al., 1979; Gordon et al., l98l). Typical-ly these minerals at Balmat form pink- to buff-coloredrocks which contrast strikingly with the enclosing rocks.They may fluoresce a deep red when exposed to ultravio-let light and thus have received attention well beyondtheir abundance. Spectacular yellow-orange rocks con-taining an unusual Mn-rich pyroxene were collected onthe 2500 level of the Balmat No. 4 Mine and provided forour study by John T. Johnson and William delorraine ofthe St. Joe Zinc Company, Balmat, N. Y. X-ray difrac-tion studies and a structure refinement of the pyroxeneindicate that it is orthorhombic and completely orderedwith over half of the M2 site occupied by Mn. Thisorthopyroxene therefore qualifies as a new mineral andhas been named donpeacorite after Dr. Donald R. Peacor

t Contribution No. 3E9 from the Mineralogical Laboratory ofthe University of Michigan, Ann Arbor, MI 48109.

2 Present address: Department of Geology and Geophysics,The University of Utah, Salr Lake City, Utah 84112.

of the University of Michigan Department of GeologicalSciences, in recognition of his work with manganeseminerals, as well as with pyroxenes and pyroxenoids. Thename donpeacorite applies to the ordered orthopyroxeneof endmember composition MnMgSi2O6. Thus the miner-al described herein is DpseEnar. Both the mineral and thename donpeacorite has been approved by the I.M.A.Commission on New Minerals and Mineral Names (A.Kato, written connunication, 1982). Type material ispreserved at the Smithsonian Institution, WashingtonD. C.

Mn-rich pyroxenes and pyroxenoids have recently re-ceived much attention (Peters et aL, 1977; Peacor et al',1978; Albrecht, 1980; Albrecht and Peters, 19E0; Brownet al., 1980: Brown and Huebner, l98l; Gordon et al.,l98l). Brown et al. (1980) assembled the then availablecompositional data and inferred phase equilibria at meta-morphic temperatures corresponding to the amphibolitefacies for three faces of the RSiO3 tetrahedron (R = Ca-Mg-Fe-Mn). With the recent discovery of the unusuallyMn-rich orthopyroxene donpeacorite at Balmat, N. Y. itis now possible to infer the general topology of a phasediagram for the remaining Mn-Mg-Fe face of the RSiO3tetrahedron. In this paper we describe this mineral andevaluate the pertinent metamorphic phase equilibria.

0003--0Mx84/0506-0472$02. 00 472

Occurrence and associations

The manganoan orthopyroxene described in this paperoccurs in manganese-rich siliceous marbles. Manganese-rich pods are scattered within several of the mappablesiliceous marble units at Balmat, N. Y. (Brown et al.,1980). These marbles, which are common throughout theAdirondack Lowlands (Engel and Engel, 1953) weremetamorphosed to the upper amphibolite facies (Bud-dington, 1939) during the 1.0 b.y. old Grenville orogeny.Recent estimates of peak metamorphic conditions atBalmat yield 6.5-rl kbar and 650-+30'C (Brown et al.,1978; Bohlen er al., 1980).

Donpeacorite occurs as l-3 mm interlocking grainsmaking up over 50 percent (modal abundance) of amassive coarse-grained yellow-orange rock. The ground-mass consists mainly of the fibrous Mn-Mg amphiboletirodite (40 percent modal abundance), with minor tour-maline, ferrian braunite, manganoan dolomite, hedy-phaneJike apatite and anhydrite. The tirodite, which isclosely associated with the orthopyroxene, is faint orangein hand specimen and colorless in thin-section. Thesubhedral orthopyroxene is pale buff in hand specimenand faintly pink in thin section. It shows no evidence oftwinning or exsolution. Refractive indices were deter-mined in oil on a crystal grain mount (Table l). The 2Vwas measured on a universal stage using hemisphereswith refractive index equal to 1.65 and making no tiltcorr6ctions (Table 1).

Chemistry

The pyroxene, amphibole, tourmaline, braunite andhedyphane were analyzed using the University of Michi-gan ARL-EMX electron microprobe utilizing wavelengthdispersive techiques. The operating conditions were 15kV, 150 plA emission current and 0.015 pA samplecurrent. The standards used were Marjalahti olivine for Si

Table l. Optical and unit cell parameters ofdonpeacorite

DonDeacorite Enstati te*

PETERSEN ET AL.: DONPEACORITE

Sl02AlZ03

l,ln0il90Fe04

C!0ilrA0Totrl

Sl0ZAlA03

iln0il90Fe04

Ca0lla20Total

Sl02Tl0241203

reoatl90tiln0

Cr0

ilazoKzo8a0

H2Q5Fc l0.F' ClTotal

473

Table 2. Electron microprobe analysis of donpeacorite andtirodite

orld. t|llcht P$c.nt FonnulrorthoPJ'roranaSr

55 .12

0 .2318.f826 .310 .140.69

0.03100.00

54.640.16

20,2224,12

nd0.82

nd99.96

55 .20nd

L.47

0.3022 .729.474 .77

1 .18ndnd

L.26

0.690.01

-0. ?9

t6.?8

Tlrodl te2

stAIl,lnil9FeCrl{a0riln

stAIilnl,tgF.

Calla0xiln3

srTI

Al lv

Alv lFe2+Irtgtiln

Ca

l{aKBa

OHFct0

1.983

0.0100.5631 .4110.0040.027

0.0035.989

0.28

2.0090.0070.630t.322

0.032

6.013

0 .32

7.9{8

0.052

0 .1970.0364.876l . 155

0.736

0,330

1.210

0.3140.003

22.479

*All optlcal drta fron Deer, Howle rnd Zussman (1978).**All unlt cell paramters from Hawthorne and lto (1977)

nd, not detected: lllorilrllzed to 4 catlons;2ronnallzed as x+Y+2.15 rlth all l{a ln A sltei3xm+tn/(ttn++tg); 4All Fe assuned as Fe2+;scalculated.

and Mg, Ingamells almandine for Al, synthetic rhodonitefor Mn, ANU wollastonite for Ca, Irving kaersutite for Feand plagioclase (Anaj for Na. Corrections for ZAFeffects were made using the program EMrADR vII (Ruck-lidge and Gasparrini, 1969). Two chemical analyses aregiven in Table 2. The orthopyroxene is rich in manganese,sample I having composition Mg1.a1Mnn.56Cao.orAlo.orSi1.e6O5.ee and sample 2 Mg1.32Mns.63Caa.s3Alo.e1Si2.s1O6.s1 (Fig. 1). The tirodite has significant Ca replacing Mnlike those described by Ross et al. (1969). The tourmalineis intermediate between uvite and dravite (Uva6Dr5a) andis much like the specimen whose structure was refined byBuerger et al. (1962). Semiquantitative analyses of braun-

B

Y

Y-q2Vr(neas. )

a(8)b(ff)ctff)v(R3)D(ca lc )

space group

optical Propert'lesr .677(2)1.684( 2)1.692(2)0.01588( s)

Unlt Cell Paramtersr8.384( 1 l )8.879( 7)5.226(3')

853.12 ( 70)3.403(2)Pbca

1.649

1.653

1.657

0.008

r25

18.214(4)n

8 .818( 2 )

5 . t t 1 ( ? \

831.483

3 . 2 1

Pbca

474 PETERSEN ET AL.: DONPEACORITE

MgMnSirO6 FeMnSi205

lMg2 Si2 06 Fe2Si205

Fig. l. Orthopyroxene quadrilateral illustrating the donpea-corite compositional field and plot of donpeacorite analyses.

ite and hedyphane show that the braunite is ferrian andthe hedyphane is phosphorian.

X-ray crystallography

Donpeacorite and tirodite crystals were studied usingsingle-crystal X-ray techniques. Weissenberg and precis-sion photographs indicated that donpeacorite is ortho-rhombic. The pattern of extinctions is consistent withspace group Pbca, the most commonly reported spacegroup for orthopyroxenes (Cameron and Papike, 1980,l98l), and thus it is isostructural with enstatite andhypersthene. Lattice parameters for donpeacorite, re-fined by least-squares utilizing data from a lolmin. pow-der diffractometer pattern internally standardized withquartz, are given in Table 1 and the diffraction data inTable 3. Some of the tirodite displays a faint, light-blueschiller of the kind commonly caused by submicroscopicexsolution lamellae. Exsolution relations have been de-scribed in the manganoan amphiboles from the Balmatarea by Ross et al. (1969). The presence of submicroscop-ic exsolution lamellae in our samples was confirmed byWeissenberg photographs which showed a predominantphase having space group P2lm intergrown with a muchless abundant amphibole, presumably with space groupAlm.

Since the standard chemical analysis does not permitan unambiguous structural formula to be calculated, acomplete pyroxene structure refinement was undertakento determine if Mn is ordered into the M2 site. Intensitydata were obtained from a cleavage fragment measuringapproximately 0.08 x 0.09 x 0.19 mm mounted forrotation about the c-axis on a Weissenberg-geometrydifiractometer. A Super-Pace automated diffractometersystem was used with MoKa radiation monochromatedby a flat graphite crystal and measured with a scintillationcounter. A scan was made across each reflection andbackground were measured on each side. The intensitiesof 1096 reflections were measured, of which 295 hadintensities below the minimum observable values, up to asine theta limit of 0.461 , for /r = 0-24, k : 0-12 and / = 0-5. All intensities were corrected for Lorenz-polarizationand absorption efects (p(Moka) : 28.810 cm-r). Amodified version of the program ABSRr written by C. W.Burnham was used for the absorption corrections.

The structure refinement was initiated using parame-ters for the structure of orthoenstatite eiven bv Haw-

Table 3. Principal diffraction lines for donpeacorite

l l lo d ( obs . ) d ( ca l c . ) hk l

10

100l010

DU9994

108J

25

596

111 l

4 .033 .243 . 1 83 .092.96L

2.8962.8472.7242 .5512 .531

2 .5102 .487 ,2 .490

2.4042.0222.04L

2.0717.99471.74201.49501.47941.4019

4.01 ,4 .053,223 . 1 93.0602.962

2.8962.8482.7252.5512,533

220, 2ll*411

420, 221500'3?t*

510*511*42L*131*6l l *

2.516 202*2.488, 2.490, 2.493 Ltz, 23t, szl

2.406 3022,025 422

2,040, 2.041 141, 820

2.073, 2.076 sr?, 721

CuKa radiat jon; *Peaks used in ref inemnt.

thorne and lto (1977). The program RFrNE 2 (Finger andPrince, 1975) was employed using the scattering factors ofDoyle and Turner (1968), and the weighting scheme ofCruickshank (1965). During the early stages ofthe refine-ment the scale factor and atom coordinates were refined,while the occupancies of Ml and M2 were held constantwith 100% Mg. In later cycles, refinement of occupancyfactors for Ml and M2 was alternated with cycles inwhich coordinates and/or anisotropic temperature factorswere refined. The final unweighted R-value was 5.6percent for all reflections, and 5.0 percent for all unreject-ed reflections. Reflections with intensities below theminimum observable value, and those with large individ-ual R-values (R > 0.5) were rejected. Final atom positionsand temperature factors are listed in Table 4; selectedinteratomic distances and angles are listed in Table 5; andmagnitudes and orientations of the thermal ellipsoids aregiven in Table 6. The calculated and observed structurefactors are listed in Table 7.3

The final refinement showed an Ml site occupancy of1.02(l) Mg (and -0.02(l) Mn) and an M2 site occupancyof 0.47(l) Mg and 0.53(l) Mn. Thus the manganeseappears to be completely ordered into the M2 site withinthe sensitivity of the method. Comparison of the suggest-ed composition with that obtained by electron micro-probe analysis (first analysis in Table 2) shows that thisresult is in very good agreement. Analysis 2, whichcontains slightly more Mn, was obtained on anothersample from the same locality. Small amounts of Ca andFe in the structure are approximately accounted for bythe Mn form factor and Al is accounted for by the Mgform factor. The differences from the actual analyses are

3 To receive a copy of Table 7 order Document AM-84-241from the business office, Mineralogical Society of America, 200Florida Avenue, NW, Washington D. C. 20009. Please remit$1.00 in advance for the microfiche.

PETERSEN ET AL.: DONPEACORITE

Table 4. Final atom positions and anisotropic temperature factors (x l0-5)

475

BzsBta\zBgsBzzBtt

M1n2s iAs i B

0rA02A034

01B028038

0.37s1( 10 .3776( 10 .27L?(L0.4749( 1

0.6543 ( 2 )0.4801 ( 1 )0 .3413( 1 )0 .3382( 1 )

0 .3379(3)0 . s022(3)0 .22s2(4)

0.3402 ( 3 )0.4872(4)0.2050 ( 4 )

0.872r (3)0.3684( 2 )0.0484( 3)0.7e4e( 3)

0.0407( 7 )0.0467 ( 6 )0.822?(7)

0 .798r ( 7 )0.7028(7')0 .s902(7)

2s0( 20)340( 10)220( l0)220( r0)

360( 40)330(40)430( 40)

550( 70)630( 50)520( 60)600( 60)

610( r40)870( 140)7r0( 140)

3( 13)10( l0)4( 14)

-30(20)-30( 10)30(20)20(20)

-80(40)-10(50)-30(s0)

3(4s)ls0( 50)

-2r0(60)

s l (5 )66(3)s6(4)45( 3)

l (s) -10(10)-11(4) - rs(7)-7(4) -16(e)

-10(4) -13(e)

20(20)l0( 20)40(2)

300(40)280( 30)4r0(40)

0 .1833(2)0.309e ( 2)0 .3017( 2 )

0. s632( 2)0.4349( 2)o.447s(z',)

68( e)48(8)72(8)

s8(8)77(s l64(8)

4r0( 140)1190( rs0)820( r30)

s(12) -20(20't20(10) lo(30)

-20(10) -7(24')

probably somewhat greater than the occupancy factorssuggest, but therefore well within the reasonable range ofcompositional variation in this sample.

Comparison of the detailed structure data obtained fordonpeacorite with that for other (Mn,Mg) pyroxenes

Table 5. Selected interatomic distances and anqles

(Hawthorne and Ito, 1977;Gordon et al.,l98l) allows theeffects of this ordering to be observed. The (Ml-O) bondlength of orthoenstatite is 2.076A, while that for donpea-corite is 2.0844. If Mn is indeed completely ordered intoM2 (and therefore Mg in Ml), the (Ml-O) distances ofdonpeacorite and orthoenstatite should be quite similar,as indeed they are. As several authors (Hawthorn and lto,1977; Cameron and Papike, 1980) have noted that the(Ml-O) bond lengXh is strongly a function of the cationradius, this comparison suggests strongly that Mg is thedominant cation in the Ml site. The slightly larger (Ml-O)distance may, however, indicate that some Mn or Fe ispresent in this site.

Comparison of the (M2-O) distances shows significantdifferences between donpeacorite and other pyroxenes.

Table 6. Magnitude and orientation of the principal axes of thethermal elipsoids

siA-01AsiA-02AsiA-03Asi A-034'(siA-o)

r,t1-0lAr,r1-0lA'titl-02A]it1-0lBitl-0lB'!,t1-028(lr-o)

0lB-0280lB-0380lB-038,028-038028-038,038-038,(o-o)01A-0lA0lA-02A01A-0lB01A-028018-0lA018-02A018-18018-02801A-02A02A-02502B-0lB018-01A(o-o)01A-01801A-02801A-02A01A-03A018-0280lB-02A018-038028-03A028-038,02A-03A02A-03802A-03803A-038,03A-03803A-038,(o-o)

03A-03A-03A

1 .516 (4 )1.500( 3)l .642 ( 4)1.5se( 4 )L.629

srB-olB t.624(4)siB-028 1.s86(4)srB-038 1.670(4)s tB-038 ' 1 .671(4)(sia-o) r.638'

2.740(5)2.6s4( s )2.670( s)2.6s7(s)2. s83 ( s)2 .733 ( 5 )2.673

3.041 ( 3)2.e78(5)2.828( 6)2.806 ( 5 )2.839(5 )2 .841 (s )3 .061(3)3.030( 5 )3.001 ( s)2. e21 ( s)3 .232( s )2.828(6)2.95L

2.156( 3 )2. r04( 4)2.337 14)3. s42 ( 3)2 . l le ( 3 )2.044(4)2.540( 4)2.se4(4)2,278

r17.2(2)107.3 ( 2 )108.2( 2)10e.3 ( 2 )r04. s( 2 )109.7( 1 )109.4

2.036( 4 )2. r41 ( 4)2.202(4)2.068( 4)2.L87(412.049( 4)2.114

e3 .4 ( l )e l . 3 ( l )84.8( I )

176.9(2)84.4( l )

172 .e ( r )e2 .0 ( 1 )es.3( 2 )e5 .4( l )e l .8 ( 2 )94.8( 2 )81 .6( 2 )

104.6

82.6( 1 )8 3 . s ( 1 )8 7 . s ( 1 )

roe .1(1)s3 .6( 1 )84 .5( 1 )

123. 7( I )r14 .9( I )5 8 . 1 ( 1 )

r04.4( l )58 .8( r )

12e. e( r )8 4 . s ( l )6 9 . 4 ( l )74 .3( l )0 1 2

it2-0lAj42-02Ar42-034itz-03A,it2-0lB]42-028ti'|2-038r,r2-038,(r'rz-o)0lB-si B-0280lB-si B-0380lB-si 8-038,028-St B-038028-St 8-038'038-Si 8-038,(o-siB-o)

01A-Ml-01A01A-Ml-02A01A-Ml-01801A-Ml-028018-M1-0lA018-M1-02A018-Ml-018018-Ml-02801A-il1-02A02A-l,ll-028028-M1-018018-r.,rl0lA(o-n-o)

01A-fil2-01801A-r.,r2-02801A-M2-02A01A-M2-03A018-M2-0280lB-M2-02A0lB-it2-038'028-M2-03A028-til2-038'028-t42-03802A-M2-03A02A-142-038'02A-r,r2-03802A-li|2-038'03A-lil2-038(o-r'rz-o)038-038-038

RilS

Dlsplacilent A axls

Angle to:

B axls C axls

0.084(6) 76120) 7r(r2l 24(14)0.0e3(4) 162(221 7s(24) 81(m)0.0ee(4) 7e(22, 2s(1e) u1(13)

0.089(3) 72(61 7714) 2r(6)0.107(2) 154(7) 104(8) 69(6)0,117(2) 108(8) 19(7) e6(4)

siA

s tB

0.082( 4)0.093(3)0.102( 3)

0.085( 3)0.090( 3)0.099( 3)

0.086(8) 2r(14)0.107(7) L07l22l0.111(8) 7e(271

0.100(8) 51122)0.116(7) l3e(23)0.129(6) 78(l7l

0.074(12) 72114)0.103(7) 162( ls)0.120(7) 86(16)

0.104(7) l r3(2s)0.116(7) r54(e3)0.141(7) 77(Lr l

0.09s(8) 68(34) 60(8)0.10s(7) 157(33) 83(19)0.140(6) e5(8) 31(7)

7e(9) 1o7lr2l 2r(8)L22lr4l 146(13) ee(13)34(14) 117(14) 108(8)

32(15) 74(r7l 63(24)101(27) 12e(r6) 42(19)60(10) 136(16) l r8(r5)

6e(14) 25(17')84(57) 106(25)22(2ol ro9(22)

r1l (15) e2(15)146(60) 62(75)65(70) 28176)

86(12) 3e(2r)72(2rl 54122)le(2r) 104(16)

o.oe8(lo) 103(18)0.r08(7) 163(3e)0.113(7) 7e(s8)

2.828( 62.806( 52.952(53 .670 ( 53 .035 ( s2.841(s3 .032(s3 .697( s2 .583( 52.52s(s3.084( s

I23

t

23

I2?

18( 14)72(r4)92(8)

r12( 13)68(14)3r ( rr)3e(22168(2e)

r20(7)

89(7)85( r6)5(r6)

32(r4lro5( 24)62( e)

3. r38( 53.084( 52.9s0( s2 .733 (32.997

r63 .8( 1 ) 145.0( 3 )

476 PETERSEN ET AL.: DONPEACORITE

In order to limit problems due to variable M2 coordina-tion, only average values for the six closest oxygens(OlA, O2A, O3A, OlB, O2B, O3B) will be compared.The (M2-O) distance for orthoenstatite is 2.1494, thar fordonpeacorite is 2.2lEA, and kanoite and manganoandiopside yield 2.2994 and 2.495A, respectively. Thevalue for kanoite has been re-averaged from that inGordon et al. (1981) to account for only the six closestoxygens. As these last two pyroxenes differ structurallyfrom donpeacorite (space groups P2lc and CZc, respec-tively), they are not strictly comparable, but should showa similar trend. Comparison of the (M2-O) bond length oforthoenstatite with that for clinoenstatite (2.1424, spacegroup P21lc, Ohashi and Finger, 1976) suggests thatkanoite is a reasonable analog for a more manganoandonpeacorite in this respect. The occupancy reported forthe M2 site in a sample of clinopyroxene with exsolvedkanoite by Gordon et al. (1981) is 0.86 Mn, 0.14 Mg.Extrapolating linearly from orthoenstatite and this kan-oite to a stoichiometric -(1: l) Mn-Mg end-member yields(M2-O) equal to 2.3234. For the structural effects, wepredict a (M2-O) value for donpeacorite of 2.239A, whichis only slightly larger than that actually observed. If thekanoite value is corrected downward approximately0.014 to correct for the structural differences, the pre-dicted (M2-O) for donpeacorite is 2.n44, which is stilllarger than the measured value. Hawthorne and lto (1977)show that the relationship between (M2-O) and composi-tion may not be exactly linear, and suggest a conectionbased on the degree of bond-length distortion A:

l 6a: r ) t (n , -R)1R1,

6 , : ,where R is the mean bond length and R; is the individualbond length. The value calculated for donpeacorite is 5.86x l0-3 which would-give an approximate correction(their Fig. 5) of -0.08A. Using the above approximationcorrected for space group structural effects, we predict a(M2-O) bond length ot 2.2264 for donpeacorite, withinerror of the measured value.

A commonly discussed feature of the pyroxene struc-ture (Papike etal.,1973; Hawthorn and lto, 1977;Camer-on and Papike, 1980, t98l; Gordon et al., 19El) is the 03-O3-O3 angle, which describes the degree of rotation ofthe tetrahedral chains from the ideal 180'conformation ofa straight chain. As the two tetrahedral chains in Pbcapyroxenes are not symmetrically related, there are twosuch angles in donpeacorite, 163.8o for the A-chain and146,0'for the B-chain. Comparison with other pyroxenes(Cameron and Papike, l9E0) shows that the A-chainrotation falls near the average rotation reported, but theB-chain angle is at the extreme upper end of the reportedrange. The average cation radius in the Ml and M2 sites(Shannon and Prewitt, 1969) is approximatelv 0.75A.Comparing this with the plot of average cation ionicradius versus chain angle given by Cameron and Papike(19E0, Fig. 28) the A-chain angle agrees well while the B-chain is nearly 2o higher than the trend shown by other

orthopyroxenes. The scatter in the data suggests that 03-O3-O3 angles are not a simple function of cation ionicradius.

A similar approach cannot be used to study the rela-tionship between cation ordering and the O3-O3-O3angles. These angles are quite different in ortho- andclinoenstatite and thus kanoite cannot be used to predictthe values for donpeacorite. In addition, as the tetrahe-dral chains coordinate both octahedral sites, their orienta-tions should reflect an average cation property ratherthan properties of the individual octahedra.

The (Mg,Fe) orthopyroxenes may be used to model theeffect of cation ordering on the O3-O3-O3 angle. Burns(1970) has noted that Fe should order into the M2 site inorthopyroxene due to Jahn-Teller efects. The O3-O3-03 angle for the A and B chains may be plotted againstFe/(Fe + Mg) or average octahedral cation radius for theseries orthoenstatite-orthoferrosilite (Fie. 2). The A-chainangle increases continuously over this range, but the B-chain angle reaches a peak at about 50Vo Fe, and thenlevels out and may even decrease towards orthoferrosi-lite. This suggests that the B-chain angle is more sensitiveto the cation in M2 than the cation in Ml, and willtherefore reflect ordering into the M2 site.

The O3-O3-O3 angles of donpeacorite and a mangan-oan orthoenstatite synthesized by Hawthorne and Ito(1977) are shown as a function of average octahedralcation radius in Figure 2. While there are insufficient datato be sure whether the (Mn,Mg) orthopyroxenes follow atrend similar to that shown by the (Mg,Fe) orthopyrox-enes, the trend suggested by the three available points onthis join at least suggests that more than average cationradius affects the O3-O3-O3 angles.

Discussion

Donpeacorite at Balmat occurs in the amphibolitefacies, well down-grade from the orthopyroxene isograd

Mcon 0c tohadro l Co l ion Ro l io

o.6e oto o72 074o66t7o r

o 165

a rov

no

I r+sIn

t40B- Choin

t 35 o 6o o

T#fr.Fig. 2. O3-O3-O3 angle for (Fe,Mg) orthopyroxenes.

Sources of data: solid circles-Burnham et al. (1971); Dodd et al.(1975); Ghose (1965); Ghose and Wan (1973); Kosoi et al. (1974);Miyamoto et al. (1975); Smyth (1973); Sueno et al. (1976);Takeda (1972); Takeda and Ridley (1972). Open square-Hawthorne and Ito (1977). Open circle-This paper.

t oo 8o 4o 2

PETERSEN ET AL.: DONPEACORITE

Fig. 3. Two possible configurations of phase equilibria in the system MnSiO3-MgSiO3 consistent with lto's (1972) high-temperature phase equilibria and with our observations at metamorphic temperatures. Fig. 3a assumes that lto's manganoanclinopyroxene shows complete solid solution to kanoite. Fig. 3b is drawn assuming that Ito's clinopyroxene disproportionates toorthopyroxene and kanoite similar to pigeonite equilibria in the system CaSiQ-MgSiO3-FeSi03. The effect of Mn on theorthopyroxene polymorphs is hypothetical but is chosen to avoid interfering with the data of Ito and ours at metamorphictemperatures.

477

that marks the transition from the amphibolite facies tothe granulite facies (Engel and Engel, 1955). The presenceof donpeacorite in a marble, however, should not beexpected to correspond to the appearance of hypersthenein a metabasite; the substitution of manganese in enstatiteor dilution of the fluid phase by CO2 in the marble maywell have locally extended the stability of the orthopyrox-ene * fluid assemblage at Balmat. Even there the occur-rences of manganoan pyroxenes are extremely rare andlocalities for tirodite (and "hexagonite") are widespread,suggesting that only locally were bulk rock and fluidcompositions compatible with the generation of thesepyroxenes.

The distribution of MnlN4g between pyroxene andamphibole is of interest for reactions involving thesephases. Ignoring for the present intracrystalline cationdistributions, one may calculate the butk Kp:

(Mn/Mg)opx

Mr). Thus one would predict a Ki of 2.5 for a coexistingMnMgSi2O6 pyroxene and Mn2Mg5SiEO22(OH)2 tirodite.At Balmat, Ca is also involved in these minerals' andwhile the manganoan orthopyroxene takes little Ca, thetirodite contains some 37%o Ca in M4 (Table l). Thus Cadisplaces Mn in the M4 site of amphiboles while hardlyafecting the orthopyroxene and this increases the bulkKe in Mn:Mg. It therefore appears that the Kp of (Mn/

Mg) for pyroxene compared to amphibole is stronglydependent on the availability of calcium for the amphi-bole.

The distribution coefficient of (Mn/I\dg) between the M2site of pyroxene and the M4 site of amphibole is alsoinstructive. Assuming all Mn in M4 of amphibole and M2of pyroxene, as shown by the structural refinement,

(Mr/Me)S3.K^( lareesi te) : +=

(Mr/Mg)f;fr6

0.63t0.32= 0 . lE0.58i0.054

This calculation suggests that the M4 site of amphiboleprefers Mn over Mg compared to M2 of pyroxene but thesignificant Ca in the amphibole will probably also affectthis Kp (large site).

No experimental data are yet available on the MgSiO3-MnSiO3 binary at metamorphic temperatures, but Ito(1972) has reported on experiments at 13fi)-1500'C. Hisexperiments show successive fields of orthopyroxene,clinopyroxene, pyroxmangite and rhodonite from MgSiO3to MnSiOr, but the positions of gaps between these fields

KoGulk) =0.630/1.320= 2.015.1.15514.876(Mn/Mg)amph

The partitioning of Mn vs. Mg between orthopyroxeneand amphibole may be rationalized in terms of theircrystal structures. In the MrrMg system, Mn clearlyprefers the larger M2 site in pyroxene and the M4 site inamphibole, while Mg is partitioned into Ml in pyroxeneand Ml, M2, and M3 in amphibole. Thus the approximateupper limit of Mn:Mg expected in pyroxene is 1:l =M2:Ml and in amphibole it is 2:5 = 2M+:(2Mr + 2M2 +

fa\ ---:l:----

\i\ --_"€;r___

- \ \mss \ \| \ ' r \ \| \ ' \ \ \\ \ , . . \ \ R h s "

,, \, '\r'

| r \ . \ \r \

'r, .... r..x.).'\, t\ psss \

t. t{n)r',. tt

\ t.. t..ta

I \ '.... '.\..

tr. ta \

rzTr tt. 'tr.t..

\ \ \ ' \ ' \ ' \\ \ r \ . \ \

\ \ r , \ . \ \- \ \ L - - - - - - - - \ \ \

\ \ i ! \ , ' r \ . \\ \ ! t z a ! \ . r ' ,oens \)..i,",i\

'",. '"ll

Xunsioo

478 PETERSEN ET AL.: DONPEACORITE

Pyumongite ss

Pyqfer@il,E

present, and a three-phase field orthopyroxeness-pyrox-mangitess-kanoitess must be located somewhere insidethe diagram. Only two or three kanoite occurrences areknown, and kanoite-donpeacorite pairs have not yet beenreported but the manganese-rich pods of the Balmat areamay yet yield this assemblage.

The discovery of donpeacorite at Balmat requires someminor revisions of the Brown et al. (1980) phase diagramfor the subsystem MnSiO3-MgSiO3-CaSiO3 (their Figs. 2and 4) with respect to their inferred MnSiO3 solution oforthopyroxene. The field for orthopyroxene along theMgSiOrMnSiO3 join should be extended towards theclinopyroxene field such that the separation between thetwo fields is reduced to less than 20 mole Vo. Untike theCaMgSi2O6-Mg2Si2O6 system, the MnMgSi2O6-Mg2Si2O6binary must show extensive solution in the orthopyrox-ene and Mn must be much more easily accommodatedthan Ca in the M2 site of the orthopyroxene structure atmetamorphic temperatures.

Acknowledgments

We would like to thank the geological staff of St. Joe MineralsCorporation Balmat-Edwards Division for collecting and provid-ing the samples examined in this study. Microprobe funds wereprovided by the University of Michigan. Thanks are due to thestatr of the University of Michigan Microbeam Laboratory formaintenance of the electron microprobe facility. We gratefullyacknowledge the help of A. Kato in critically reviewing oursubmission of donpeacorite as a new mineral to the I.M.A. S. J.Huebner and J. M. Rice are thanked for their thoughtful reviews.

References

Albrecht, J. (1980) Stability relations in the system CaSiOrCaMnSi2O6-CaFeSi2O6. Contributions to Mineralogy and Pe-trology, 74,251-2ffi.

Albrecht, J. and Peters, T. (1980) The miscibility gap betweenrhodonite and bustamite along the join MnSiO3-Casa6Mn66eSiO3. Contributions to Mineralogy and Petrology, 74,261-269.

Bohlen, S.R., Essene, E.J., and Hofman, K.S. (1980) Feldsparand oxide thermometry in the Adirondacks: an update. Geo-logical Society of America Bulletin,9l, ll0-113.

Brown, P.E., Essene, E.J., and Kelly, W.C. (197E) Sphaleritegeobarometry in the Balmat-Edwards district, New York.American Mineralogist, 63, 250-257.

Brown, P.E., Essene, E.J., and Peacor, D.R. (1980) Phaserelations inferred from field data for Mn pyroxenes andpyroxenoids. Contributions to Mineralogy and Petrology, 74,417-425.

Brown, P.E. and Huebner, J.S. (1981) Geothermometry of(Ca,Fe,Mn)SiO3 minerals: the experimental approach. Geologi-cal Society of America Abstracts with Program, 13, 4lE.

Buddington, A.F. (1939) Adirondack Igneous Rocks and theirMetamorphism. Geological Society of America Memoir 7.

Buerger, M.J., Burnham, C.W., and Peacor, D.R. (1962) Assess-ment of the several structures proposed for tourmaline. ActaCrystallographica, 15, 583-590.

Burnham, C.W., Obashi, Y., Hafner, S.S., and Virgo, D. (1971)Cation distribution and atomic thermal vibrations in an iron-rich orthopyroxene. American Mineralogist, 56, 850-876.

Pyrcxrcngil,o(Rhodoitel

TP

Enslolite

OrthopyrcxeE ss

oo o

o o.--"- -q.&:o'8-$----s-:J 3 o-9$ FeSiOs

Fig. 4. Compilation of published analyses for the MnSiO3-FeSiO3-MgSiO3 face of the RSiO3 pyroxene-pyroxenoidtetrahedron. P : 4-8 kbar, T = 500-700'C for most of the points.Sources of data: Ford and Bradley (1913); Henry (1935); Tilley(1937); Heitanen (193E); Yosimura (1939); Kuno (1947); Lee(1955); Howie (1964); Momoi (t96a); Ewart (1967); Mason(1973); Peters et al. (1973); Marek er al. (1975); Kobayashi(1977); Krogh (1977); Peters etal. (1977); €hopin (197E); Deer etal. (197E); Murthy (1978); Schreyer et aI. (197E); Jolnso-n andKnedler (i979); Pavelescu and Pavelescu (1980); $ivaprakash(1980); Fukuoka (1981). The solid circles are rhe analysesreported in this paper. At the MnSiO3 apex half shaded squares(!).represent rhodonite analyses in which Xsusio. < 0.07;open squares (-) represent pyroxmangite analyses. Coexistingminerals are connected by solid lines and those inferred ascoexisting are connected by dashed lines.

are considerably diferent from those inferred from natu-ral assemblages at much lower temperatures (Figs. 3a,3b). One may speculate that Ito's (1972) clinopyroxene(En67Rh13) is a Cllc type continuously changing composi-tion with decreasing temperature to C2lc kanoite(En5sRh5s) as shown in Figure 3a. Alternatively, Ito'sclinopyroxene might be a pigeonite-type structure dispro-portionating to manganoan enstatite and kanoite at lowertemperatures (Fig.3b). Reversed experiments and carefulcrystallographic observations of synthetic products onthe MgSiOrMnSiO3 binary between 6fi) and ll(X)'C areneeded to resolve these speculations.

Published data for pyroxenes and pyroxenoids wellrepresented by the subsystem MnSiO3-MgSiO3-FeSiO3have been plotted in Figure 4. Compositional data andphase identifications were taken directly from the litera-ture. All analyses with CaSiO3 component greater than7Vo have been excluded, and only metamorphic pyrox-enes and pyroxenoids are plotted. Solid tieJines connectcoexisting minerals while dashed lines connect inferredcompositions. From the available data one-, two-, andthree-phase regions may be inferred. Orthopyroxene andpyroxmangite solid solutions appear to occupy large one-phase areas of the diagram. Three two-phase fields are

PETERSEN ET AL,: DONPEACORITE 479

Burns, R.G. (1970) Mineralogical Applications of Crystal FieldTheory. Cambridge University Press, London.

Cameron, M. and Papike, J.J. (lgEl) Structural and chemicalvariations in pyroxenes. American Mineralogist, 66, l-50.

Cameron, M. and Papike, J.J. (1980) Crystal chemistry ofpyroxenes. In P.S. Ribbe, Ed., Pyroxenes, Vol. 7, p.5-92.Mineralogical Society of America Reviews in Mineralogy,Washington, D.C.

Chopin, C. (1976) Les Paragenese reduits ou oxides de concen-trations manganiferes des "schistes lustres" de Haute- Maur-ienne (Alpes Francaises). Bulletin de Min6ralogie, 101, 514-531 .

Cruickshank, D.W.S. (1965) Errors in least-squares methods. InJ.S. Rollett, Ed., Computing Methods in Crystallography, p.112-116. Pergamon Press, Oxford.

Deer, W.A., Howie, R.A., and Zussman, J. (197E) Rock-Form-ing Minerals, Volume 2A: Single-Chain Silicates. John Wileyand Sons, New York.

Dodd, R.T., Grover, S.E., and Brown, G.E. (1975) Pyroxenes inthe Shaw (l-7) chondrite. Geochimica et Cosmochimica Acta,39, 1585-1594.

Doyle, P.A. and Turner, P.S. (1968) Relativistic Hartree-FockX-ray and electron scattering factors. Acta Crystallographica,424,390-397.

Engel, A.E.J. and Engel, C.G. (1955) Grenville series in the NWAdirondack Mountains, New York: Part l--4eneral featuresof the Grenville series, and Part 2--Origin of the majorparagneiss. Geological Society of America Bulletin, 64, l0l0-1047.

Ewart, A. (1967) Pyroxene and magnetite phenocrysts from theTaupo quaternary rhyolitic pumice deposits, New Zealand.New Zealand Journal ofGeology and Geophysics, 180-194.

Finger, L.W. and Prinz, E. (1975) A system of Fortran IVcomputer programs for crystal structure computations. U.S.National Bureau of Standards. Technical Note E54,

Ford, W.E. and Bradley, W.M. (1913) Pyroxmangite, a newmember of the pyroxene group and its alteration productskemmatite. American Journal of Science, 36, 169-174.

Fukuoka, M. (1981) Mineralogical and genetical study of alaban-dite from the manganese deposits of Japan. !{emoirs of theFaculty of Science, Kyushu University, Series D, Geology,24,207-251.

Ghose, S. (1965) Mg2*-Fe2* order in an orthopyroxeneMge eFe 1.s7Si2Ou. Zeitschrift ftir Kristallographie, 122, 8l -99 .

Ghose, S. and Wan, C. (1973) Luna 20 pyroxenes: evidence for acomplex thermal history. Proceedings of the 4th Lunar Sci-ence Conference. l. 901-907.

Gordon, W.A., Peacor, D.R., Brown, P.E., Essene, E.J., andAllard, L.F. (1981) Exsolution relationships in a clinopyrox-ene of average composition Caa a3Mn6 6eMg6.s2Si2O6 from Bal-mat, New York: X-ray diffraction and scanning transmissionelectron microscopy. American Mineralogist, 66, 127-144.

Hawthorne, F.C. and Ito, J. (1977) Synthesis and crystal-structure refinement of transition metal orthopyroxenes l:orthoenstatite and (Mg,Mn,Co) orthopyroxene. CanadianMineralogist, 15, 321-33E.

Heitanen, A. (1938) On the petrology of Finnish quartzites.Bulletin de la Commission Geologique de Finlande, 122, l-l 19 .

Henry, N.M.F. (1935) Some data on the iron-rich hypersthenes.Mineralogical Magazine, 24, 221-226.

Howie, R.A. (1964) Some orthopyroxenes from the Scottishmetamorphic rocks. Mineralogical Magazine, 33, 903-91L

Ito, I. (1972) Rhodonite-pyroxmangite peritectic along the join

MnSiO3-MgSiO3 in air. American Mineralogist, 57, 865-876.Johnson, J.H. and Knedler, K.E. (1979) A Mossbauer spectro-

scopic study of the cooling history of hypersthenes fromselected members of Taupo Pumice Formation, New Zealand.Mineralogical Magazine, 43, 27 9-285.

Kobayashi, H. (1977) Kanoite (Mn* 2,Mg)SizOe, a new clinopy-roxene in the metamorphic rocks from Tatehira, OshimaPeninsula, Hokkaido, Japan. Journal of the Geological SocietyofJapan, E3,537-542.

Kosoi, A.L., Malkova, L.A., and Frank-Kamenetskii, V.A.(1974) Crystal-chemical characteristics of orthorhombic py-roxenes. Kristallografi ya, 19, 282-288 (transl. Soviet Physicsand Crystallography, 19, l7l-174).

Krogh, E.J. (1977) Origin and metamorphism of iron formationsand associated rocks, Lofoten-Vesteralen, N. Norway 1. TheVestpolltind Fe-Mn deposit. Lithos, 10, 243:255.

Kuno, H. (1947) Hypersthene in a rock of amphibolite faciesfrom Tanzawa Mountainland, Kanagawa perfecture, Japan.Proceedings of the Japan Academy, 23, 113-116.

Lee, D.E. (1955) Mineralogy of Some Japanese Ores. 5, l-65,Stanford University Press, Palo Alto.

Mason, B. (1973) Manganese silicate minerals from Broken Hill,New South Wales. Journal of the Geological Society of,dustralia. 20. 397 -404.

Miyamoto, M., Takeda, H., and Takano, Y. (1975) Crystallo-graphic studies of a bronzite in the Johnstown achondrite.Fortschritte der Mineralogie, 52, 389-397.

Marek, V., Povondra, P., and Zak, L. (1975) Pyroxmangite,rhodonite and tephroite from Chvaletice, Czechoslovakia.Acta Universitatis Carolinae Geologica, 3, lE7-198.

Momoi, H. (1964) Mineralogical study of rhodonites in Japan,with special reference to contact metamorphism. Memoirs ofthe Faculty of Science, Kyushu University, Series D, Geolo-gy, 15,39-63.

Murthy, D.S.N. (1976) Ortho- and clino-pyroxenes from thegranulites of Namakal, Tamil Nadu (Madras), India. Mineral-ogical Magazine, 40, 7E8-790.

Ohashi, Y. and Finger, L.W. (1976) The efect of Ca substitutionon the structure of clinoenstatite. Carnegie Institution ofWashington Year Book, 75,743:746.

Papike, J.J., Prewitt, C.T., Sueno, S., and Cameron, M. (1973)Pyroxenes: comparisons ofreal and ideal strutural topologies'Zeitschrift fiir Kristallographie, I 38, 254-27 3.

Pavelescu, L. and Pavelescu, M. (1980) Chemistry and opticalproperties of some manganese and iron silicates in the South-ern Carpathians. Reviews of Roumanian Geology, Geophysicsand Geography, 4, slJO.

Peacor, D.R., Essene, E.J., Brown, P.E., and Winter, G.A.(19?E) The crystal chemistry and petrogenesis of a magnesianrhodonite. American Mineralogist, 63, I lt1 -1142.

Peters, T., Schwander, H., and Trommsdorf, V. (1973) Assem-blages among tephroite, pyroxmangite, rhodochrosite, quartz:Experimental data and occurrences in the Rhetic Alps. Contri-butions to Mineralogy and Petrology, 42,325-332.

Peters, T., Valarelli, J.V., Coutinho, J.M.V., Sommeraur, J.,and Raumer, J. von (1977) The manganese deposits ofBuritir-ama (Para, Brazil). Schweizerische Mineralogische und Petro-graphische Mitteilungen, 57, 313-327.

4E0 PETERSEN ET AL.: DONPEACORITE

Ross, M., Papike, J.J., and Shaw, K.W. (1969) Exsolutiontextures in amphiboles as indicators of subsolidus thermalhistories. Pyroxenes and Amphiboles: Crystal Chemistry andPhase Petrology. Mineralogical Society of America SpecialPublication, No.2, 27 5-2D.

Rucklidge, J. and Gasparrini, E.L. (1969) Specifications of acomplete program for processing electron microprobe data:EMIADR vu. Department of Geology, University of Toronto,unpublished circular.

Schreyer, W., Stepto, D., Abraham, K., and Muller, W.F. (1978)Clinoeulite (magnesian clinoferrosilite) in a eulysite of a meta-morphosed iron formation in the Vredefort structure. SouthAfrica. Contributions to Mineralogy and petrology, 65, 351-3 6 1 .

Shannon, R.D. and prewitt, C.T. (1969) Effective ionic radii inoxides and fluorides. Acta Crystallographica, 825, 925-946.

Sivaprakash, C. (1980) Mineralogy of manganese deposits ofKoduru and Garbham, Andhra Pradesh, India. EconomicGeology, 75, 1083-1104.

Smyth, J.R. (1973) An orthopyroxene structure up to 850"C.American Mineralogist, 5E, 636-648.

Sueno, S., Cameron, M., and Prewitt, C.T. (1976) Orthoferrosi-lite: high-temperature crystal chemistry. American Mineral-ogist,6l, 38-53.

Takeda, H. (1972) Crystallographic studies of coexisting alu-minan orthopyroxene and augite of high pressure origin.Journal of Geophysical Research, 77, 5798-5811.

Takeda, H. and Ridley, W.1. (1972) Crystallography and chemi-cal trends of orthopyroxene-pigeonite from rock 14310 andcoarse fine 12033. Proceedings of the 3rd Lunar Conference, l,423430.

Tilley, C.E. (1973) Pyroxmangite from Inverness-Shire, Scot-land. American Mineralogist, 22, 720J27 .

Yosimura, T. (1939) Studies on the minerals from the mangan€sedeposit of the Kasa mine, Japan. Journal of the Faculty ofScience, Hokkaido Imperial University, Series IV, Geologyand Mineralogy, 4:313452.

Manuscript received, March 9, I9E3;acceptedfor publication, January 2, BA.