Painite, Cazrblllro,Rl: Its Crystal Structure and Relation To
Total Page:16
File Type:pdf, Size:1020Kb
American Mineralogist, Volume 61, pages 88-94, 1976 Painite, CaZrBlLlrO,rl: Its crystalstructure and relationto jeremejevite,Bu[[rAl6(OH)sO1r], and fluoborite, Br[Mgr(F, OH)rOr] Plur- BRre,uMooRE r.No TnreHnnu AnerI Department of the Geophysical Sciences, The Uniuersity of Chicago Chicago, Illinois 60637 Abstract Painite-CazrB[Al,O,s], hexagonalP6s, a :8.715(2), c : 8.472(2)A,Z : 2-possessesa rigid and dense [AlrO,r]'- octahedralframework, topologically identical to those found in jeremejevite, B'[IBAI6(OH)aO,,] and fluoborite, B,[Mg,(F,OH),O,]. R : 0.071 for l6l8 independent reflections. The octahedralframework is linked to [BOr]3 trianglesand [ZrOu]8-trigonal prismsatlf z and a largepipe at 0 0 z is cloggedwith compressed[CaO"]'0- octahedra. Average interatomic distancesareCa-O=2.398,2r-O:2.126,8-O:1.380,A1(l)-O: l.9l5,Al(2)-O: 1.918, and Al(3)-O : l.9l4A. The octahedral framework in painite, resistantto attack by acids and bases,suggests a highly refractoryphase and the possibilityof other equally resistantcompounds, hypothetical examplesbeing NaNbs+B[Alrors] and llUutB[Al"O,"]. ln the latter, thepipewould be lree from obstructions. Introduction Three-dimensionalsingle-crystal X-ray diffraction intensitiesabout the a2-rotation axis were collected Painite is a curious mineral species,first reported on a PnILnrn semi-automateddiffractometer with by Claringbull, Hey and Payne(1957) from the ruby graphite monochromatized MoKa, (tr : 0.70926A) minesof Mogok, Burma. It was found as a garnet-red radiation. With 2d-^* : 69.5", data were gathered 1.7 gram singlehexagonal crystal of hardness8. They through the k -- 0- to ll-levels. Other salientdetails: proposeda formula Ca,SiBALoOr.,the slight imbal- scanspeed lomin t, aperture2.0",20 secondback- ance in charge due either to non-integral oxygen or ground measurements,half-scan range of 2.0owiden- total cations. ing to 2.5" at higher levels.Reflection pairs of the We continue to expressinterest in dense oxide type l(hkl) and l(hkl) were collected.An absorption structures,especially those of the aluminates.In addi- correction was applied to the crystal using u : 22.8 tion, a structural relationship was suggestedin the cm 1, approximating the shapeof the chip by select- crystal cell data for jeremejevite, AI6(OH)r(BOr)u. ing seven facets, we employed the Gaussian in- Our findings reveal an elegantstructure for painite, tegration method, described by Burnham (1966). the aluminate framework of which is related to jere- This correction was small since maximum difference mejeviteand to the octahedralframework of fluobo- in transmission factors was about 57r. Additional rite, Mg'(F,OH)B(BO3).Finally, we demonstratethat correctionswere made for Lorentz and polarization the correct formula for painite is CaZrB[AIrO,r]. effects. Weights applied to the processedreflections included an uncertainty for the z-angle maximum Experimental mis-settingof 0.lo, tilting error of 0.05o, average A thin sawn plate of the type crystal (B.M. 1954, crystal size error of 0.008 mm, and the counting 192) irom near Ohngaing village, Mogok, upper statistics.Two setsof Fo were available:those which Burma, was kindly provided by Mr. PeterEmbrey of were based on an average of F(hkl) and F(hkl) (cen- the British Museum of Natural History. The chip trosymmetric case) and separate F(hkl) and F(hkl) selectedfor structure analysis was a flat fragment, (noncentrosymmetriccase). measuring0.31 X 0.22 X 0.10mm alonga,,a\, andc, During the crystal structure analysis,it was neces- respectively.Calibrated precessionphotographs es- sary to challengethe chemicalcomposition of painite tablished the cell data in Table l. proposed by Claringbull et al (1957), and we sub- 88 PAINITE,' CRYSTAL STRUCTURE AND RELATION TO JEREMEJEVITE AND FLUOBORITE mitted a fragment to electron probe analysis,per- TlsI-r [. Crystal Cell Parameters for Painite, formed by Dr. L M. Steele. The crystal was Fluoborite, and Jeremejevite* homogeneousthroughout and revealedmajor Ca,Zr, Painite Fluoborite Jerenejevite and Al with all other elements of atomic number B.irs(2) s.o6 greaterthan Na not detected.The standardsselected "tAl c(I) 8.4?2(2) 3.06 8 l8 were diopside (Ca), synthetic ZrO"(Zr), and corun- 222 2 space grcup P6g P63/n P6sln (?) dum (Al). Correction of the data utilizedthe AnpnN-z program of Hadidiacos, Finger, and Boyd (1969). fomula uit CaZrAleolslBOtl Me3(F,oH)rImr] Ar5 (oH)3 [Bo3] s The resultsin Table 2 depart seriouslyfrdm the origi- axial ratios o.97O (c/a) 1 0r3 (3cla) 0 9S5 (c/a) nal chemicalanalysis, with poor agreementfor all the specific gravityr 4.01 Z.ag 3.28 elements reported. We suspect that ZrO2 (not re- density, m cn-r 4.01 2.82 3-24 ported in the earlierstudy) was in part coprecipitated Cell paraneters for fluoborite fron Takeuchi (1950) and for jerenejevite fron with CaO and AlrOr, and in part reported as the bolovastlKov eE aL. I 1955t . rclaringbull contribution to SiO, which we found to be absentin et a7,. (1957), for painite The data for fluoborite ancl jerenejev- ite are fron Geijei md Dmour respectively, quoted in Palache et al, (195I). the crystal(< 0.01%).Although we couldnot analyze boron, the structure analysis clearly revealed the presence of fully occupied [BOJ'- triangles. The convergedto R : 0.16 and did not reduce further, probe analysisis in satisfactoryagreement with the despitea rather sharp and cleanelectron density map. proposed formula CaZrBAlrO,r, which yields p : We noted,however, that the Alt+ temperaturefactors 4.01 gm cm-3, in excellentagreement with the specific were unexpectedlylarge (1.5 to 3.04-') and decided gravity of 4.01 reported by Claringbull et al (1957). that the structure was in fact noncentrosymmetric, based on spacegroup P6s, and that the actual struc- Structure analysis and refinement ture departed only slightly from centrosymmetry. One of the reasonswe investigatedpainite was due Since correlations in parameters were high, two to an apparent relationship with jeremejevite, blocked matrices were refined in alternate sequence. A16(0H)a(BOr)u.Golovastikov, Belova, and Belov We also concludedthat the formula unit was in fact (1955), who approximately determined the crystal CaZrBAlrO* and substitutedthe scatteringcurve for structure,established that for jeremejevite,Z : 2, a Zr in place of Ca at "Ca(2)". This conclusion was : 8.56,c : 8.18A, spacegroup P6r/m. We proposed forced on the basis of site multiplicity refinement and that painite and jeremejevitepossessed similar struc- inspection of interatomic distances.The parameters tures and that Al3+ in the former isomorphically re- refined smoothly until R became0.071 for all 1618 placed some of the B3+ in the latter compound. A non-equivalent reflections, including F(hkl) and three-dimensionalPatterson synthesisP(zuw), sup- F(hkl) pairs. Inversion of the atomic coordinatesdid ported the isomorphism for the octahedralAl'+ and not materially affect the reliability index and it was the anion frame, but the remaining cations were not not possibleto selectunambiguously the proper polar clearly deciphered. orientation of the structure.This study provides yet The N(z) test on special reflections(see Howells, another examplefor which a crystal structure analy- Phillips, and Rogers, 1950)and the P(y) test on gen- eral reflections(see Srikrishnan and Parthasarathy, TnsI-n 2. Chemical Composition 1970)supported the centric group P6"/m. A trial p- of Painite synthesis (Ramachandran and Srinivasan, 1970). 1 yielded unexpectedresults: no isomorphism between 83+ and Al3+ was noted and, in addition, a strong CaO 7.L7 r5.7 8.33 A1203 69.02 76.2 68.19 densityappeared at and a weaker densityat 000. 3 +* Bzos n.d, ) ) < 17 We tentativelyascribed Ca(l) 0 0 0; Ca(2)3tt Al(l)* ZrO2 t8.77 n.d. 18.31 0 l and Al(2) + + 0.09. The anion frame agreedwell Si02 absent 5.6 with the jeremejevitemodel. Further B'-synthesesled Hzo n.d. 0.7 : to R 0.33, and at this stage it was apparent that 94.96 100.4 r00.00 "Ca(2)" was some other more dense ionic species. IElectron probe analysis, this study (I. Allowing its occupancyto vary, a trial least-sqpares M. SteeIe, analyst), refinement led to R : 0.17, whereupon B3+ was 2claringbult et a7., (L957). 3conputed unambiguously located at * ? *. Further refinement for CaZrB[A190ts]. P. B. MOORE AND T. ARAKI Tlnu 3. Painite.Atomic Coordinatesand curves for Ca'+, Ztn+, Al3+,Bo, and 01- derive from Isotropic Thermal Vibration Parameters* Cromer and Mann (1968), and the anomalous dis- persion corrections for Ca, Zr and Al from Cromer B(i2) and Liberman (1970).The secondaryextinction cor- rection (Zachariasen, 1968) converged to co : 3.0(3) ZT 2/3 I 17 tl4 0.s7(1) X 10-6 and the scalefactor, s : 6.31(a)where Fo : 0 0 0 .492e(2) 1.23(2) JF". A1(1) 0.3329 (1) 0.5373(r) . 74e0(3) .6r(2) The polar nature of the structure must be accepted A1(2) .342s(2) .362s(2) .07e1(1) .63(2) as fact and obtains from slight deviations of the Ar (3) .3437(2) . 3s96(2) .4224(r) .s4(2) [AleOrE]e- framework from centrosymmetry. Golo- B 1/3 .24es(4) .84(8) vastikov et al. (1955) proposed the centric group, 0(1) . 3992(3) .ssrs (3) .249s(4) .81(3) P6"/m, for the structurally related jeremejevite, but refinementare crude. We suspectthat 0 (3) .40s2(3) .s237(3) . sele (5) .6e(4) their data and o (4) .4060(4) .s23s(4) . e070(3) .s4 (3) a more precise refinement of the jeremejevite struc- ture will revealthe spacegroup P6r; this is supported 0(2) . 328r(3) . 2r86(3) . 2s03(s ) .6e(3) 0(s) . 30le(3) . 193s(4) .s807(3) . s8(3) by a positive piezoelectric test reported in Palache, o (6) .2s83(4) .rer2(4) .9r6s(3) .73 (4) Berman, and Frondel (1951) obtained by the os- Estinated standard errors refer to the last digit.