American Mineralogist, Volume60, pages 717-718, lgTS

Brianite,NarCaMg[POa]r: A phosphateAnalog of Merwinite, CarCaMgISiO4],

Peur BnrnNMoonn Departmentof the GeophysicalSciences, The (lniuersity of Chicago,Chicago, Illinois 60637

Abstract Brianite,Na,CaMg[po,l,, a t3.36(5)A,t s.zl14e,, c 9.13(3)A,p 91.2(0.2)",space group P2r/a,is structurallyisomorphic to merwinite,CarCaMg[SiOn]r.

Fuchs,Olsen, and Henderson(1967) describeda calculationsalong [0] l], [03I], and [100]which are new species,brianite, which occurs as a rare con- comparedwith the earlier study. It is instructiveto stituent in small pocketsin the Dayton octahedrite note that Moore and Araki (1972)pointed out mis- from Montgomery County, Ohio. Brianite was settingsof the same kind in an earlier study on observedto possessfine lamellar structurewith ex- merwinite.The refinedcell data for brianite obtain tinction anglesbetween adjacent lamellae of 2-3". from an indexing of the original powder data in The authors concludedthat the compoundis almost Fuchsel al. The first thirty linesof thesepowder data ideally NarCaMg[POn],on the basis of detailed have been indexed (Table 2) on the basis of a electronmicroprobe study. calculatedpowder pattern derived from merwinite Rotation and Weissenbergphotographs led Fuchs crystal structure data. Brianite is polysynthetically et al to concludethat brianiteis orthorhombic,and twinnedon { 100},explaining the lamellarappearance their resultsare presentedin Table l. A marked of the grains. spatial similarity between brianite and bredigite, Moore (1973), in a topological-geometrical (Ca,Ba)Ca,rMgr[SiOn]r, wasnoted by theseauthors, analysisof alkali sulfate and calcium orthosilicate having been suggested by Dr. A. Kato as a private structures,noted three kinds of large cation communicationin their study. For this reason,I polyhedra:Xlt2-pt with (0 < p < 6) so asto definethe proposedto studybrianite greater in detailas part of pinwheel type; I/rlor polyhedra; and Mt"t, the oc- a broad programon the structuresystematics of slag- tahedron (p : 6). The ytror polyhedra, owing to relatedphases. shared faces with the tetrahedralanionic groups, A singlecrystal waskindly donatedby Dr. Louis most likely receivethe cationsof lowestnet charge. Fuchs of ArgonneNational Laboratory.Although Accordingly,it is proposedthat brianiteis isostruc- the crystallographicaxes were locatedand the axial turally relatedto merwiniteas follows: dimensionsmatched those of the earlierstudy, it was observedthat the intensitydistributions violated each TABLEl. Structure andrts Relation to of the three reflection planes expected for an ""t ?:1Ji;r:rianite orthorhombiccrystal. Accordingly, the photographs which constitutedthe original study were requested and these were presentedby Dr. Fuchs. These photographs,too, revealedthe sameviolations in in- "(i) 10.50 10.s3 r3.36(s) !3.25 b(i) tensitydistributions. More significant,however, was 18. 16 18.ls s .23(2) 5.29 (A) the observationthat the "a"-axisand ..c,'-axisrota- " 13.31 13.36 9.13(3) 9.33 gt.z(o.z)" tion photographsand their zero-layerWeissenberg B 90" 90.4o 9t.9oo grouP projectionsalmost exactly duplicatedphotographs sPace P2221 P21/a pT1/a about the [0l l]- and [00]-axes, respectively,of lFuchs et al. (1967) for brianite. merwinite. Reorientationof brianite forced me to 2cornputed = frorn (3) wirh a [0tt], b = 1O:11, c = [100]. concludethat the compound 3Brianite i's in fact structurally (this study). Results refined fron partly isomorphicwith merwinite.Table I liststhe new data indexed powder data. 4Merwinite on brianite,its comparisonwith merwiniteand the (Moore and Araki L972).

7t7 718 PAUL BRIAN MOORE

Tr.sn 2. Indexed Powder Data for Brianite* ist, such as the analogsof larnite, bredigite,and the "I-silicate" phase.Such compoundswould have I d(obs) d(calc) hkl I d(obs) d(calc) hk1 compositionsNaCa[POn], NaoBaCarMg[POa]4, and 2 9.1s 9.13 001 2 2.432 2.432 213 NarCasMg[POn]a,resPectivelY. r 5.37 5.44 201 1 2.382 2.358 221 1 4.55 4.54 011 6 2.3LL 2.sL3 5lI 3 4.27 4.28 111 3 2.284 2.291 511 Acknowledgments 2 4,75 4.72 270 3 2.254 2.269 022 Olsen 9 3.734 3.735 2LL 7 2.230 2.226 600 It is a pleasure to thank Drs. Louis Fuchs and Edward 7 3.344 3.338 400 3 2.166 2.156 222 who provided the crystal and encouraged the study of the subtle J. IOI 311 4 2.737 2.742 222 O J. 156 structure of brianite. Dr. Takaharu Araki assisted in computa- 1 2. 809 2.814 410 3 2.078 2.092 014 8 2.718 2.703 4r1 2 2.049 2.03r 322 tions. The investigation was supported by a Materials Research 2.O73 322 9 2.679 2.676 411 5D 2.O13 Grant (NSF) awarded to The University of Chicago' r0 2.62s 2.637 013 9 1.875 1.886 I 1. 783 7.?7r 205 6 2.596 2.606 020 References I Z.5JJ Z.)Ot L20 2 1.760 r.757 414 2 2.452 2.436 220 3 t.731 L.739 522 Fucus, L. H., E. Or-srn,ruo E' P. HrNpEnsor (1967)On the oc- The first thirty lines rePorted by Fuchs et al. (1967). currenceof brianiteand panethite,two newphosphate minerals from the Dayton meteorite.Geochim. Cosmochim Acta, 31, t711-17 19 . Brianire Na(l) Na(2) Ca Mg [POnlz MooRE, P. B (1973) Braceletsand pinwheels:a topological- Merwinite Ca(2) Ca(3) Ca(l) Mg [Sio4l, geometricalapproach to the calciumorthosilicate and alkalisul- PolYhedra ftlol flrol xtl2t 14ret lTatoe)z fate structures.Am. Mineral St, 32-42. (ideal) -, ANDT. At,qrt (1972)Atomic arrangementof merwinite, CarMg[SiOr]r, an unusual dense-packedstructure of It seemsreasonable that brianite,like merwinite,is geophysicalinterest. A m. M ineral' 51, 1355-13'14. a well-orderedand stoichiometriccompound. In ad- dition it is tempting to suggestthat other alkali- Manuscripl receiued,February 24, |975; *ccepted alkalineearth phosphate analogs of silicatesmay ex- for publication,APril I, 1975'