WIXLIB

lG3The Symmetry and Crvsta] StructureManganite, M (OH)O.By M. J. Buerger,"Vlinera logica] Laboratory. 1Iassachusctts InstituteCambridge, 1Iassachusetts(USA.).ofof Technology,A bstrac1.A rough structurefor manganitehas already been proposed by Garrido.Although unrecogni7.cd, this is a marcasite-likcstructure.From its formula, manganite ought rather to have an arsenopyrite-likestructure, which is a superstructurebased upon the marcasite type. A new structural investigationshows that this isthe case.YIanganite is definitely monoeIinic, not orthorhombic" as stated in all referencehooks. This lowered symmetry is cvidenced by the po]ysyuthetictwinning on thesupposed symmetry plane, (100), by inclined extinction in systems of twiu bandsin eleavage flakes, and by the diffraction symmetry. 21m.The new structural investigation has heen made by the equi-irwlination VVeissen berg method and inspeetive interpretationof films. The rnanganite structurehas the tion):,Space group:UnitEquipoints:H2r/daH.HU Acell :bm general;.raYi!all atomsParameters:j}InD.24r - : .70fJ 0 90 1O2. o.119.119position.Z, o.258.122.1220.500The formula .Mn(OH)O corrcsponds geometricallypyrite formula PeAsS; the 01 atoms are almostand physically with the arscnoecrtainly (011) groups having afunction similar to As and Sb in arsenopyrite and gudnnmdite. The (OH) grouphas a radius ahout .07 A greater than the 0 atom.The structure, although resembling the arsenopyritestructure.is. in greaterdetail, a layer structure:1I1n is in square eoordinationbetwccn 2(OH) and 20 inthe samc (010) sheet with one more (OH) and onc more 0 above and below the centerof the square and in adjacent sheets in such a way as to approach octahedm] coordination for the Yln. The hydrogen bonding and the mild layering of the structure into (010) sheets account perfectly for the outstanding(010) ekavage and the(010) translation-glidingplane.Introduction.A erystal structure of manganite has recently been proposed hy C arrid(1).Although recognizing that certain faint refleetions required doubling of his cellaxes in two directions, Carridoderived a rough structure with the following characteristics:u.1) Garrido.(1935) 224-241.Zeitschr.4.42 Ab 5.23 A1\-1.,r., Structuref. Kristallographic'.I);). Ed.de In manganite,c 2.87 A.Bull. Soc. Fran.". Mineral. 5811

1642\1. .J. BuergerFormulaSpaceweightsgroup:per cellvJ,2.1}!n at ()OO;oat u/)O;whcre ur 2JJnO(OH) } .17,'uO;}-! v.II.25.:{3R1.12'"2'it, 1\-J.Although Garridoapparcntly did not recognize the fad, this is a marcasitetype structure, a correlation which becomes apparent if manganiteis referred tothe marcasitc orientation.This is done by means of the following eydieal permutation ofax('s :rnanganitenlarcasitea-- Iib.- cc-- aGa rrido' s manganite and the marcasikthe marcasite orientation:mineralstlH'n comparernanganitpabFormula weightsper :{74.445.3U2.R74.42,).2:{2FeSS21IIn(OH)OPrnnnSpace group:Mctal position:1\on-metal position:as follows. usmg 2FcAsAsV)/000, ,\,\}Ouv;parameters:nr(hID; , -/1t,}r;},.21).203.33R.375u.- '/"-.175.361From this point on, the standard manganite orientation wil! be cmployed indiscussing 111anga.nite.1\ow, the formula type of manganite, 1J,ln(OH)O, is not the A B2 type requiredfor thc eustomary marcasite structure.It eorresponds, rather, with the arsenopyritetype of formula, A B' Bn, and therefore manganite should be investigated for arsenopyrite-like characteristics.Thesc have been listed in the arsenopyritepaperl), towhich tabulationthe following refcrenees apply.Garridohas already prcsented a ccrtain amount of evidenee tending to corroborate a eonjeeture that manganite has the arsenopyritetype crystal structure.Thus, hc found faint reflections indicating that the true eell should have doublethe a and e axcs of the onc he chose for thc rough structure.This eonfirrns eharacteristie 4. (Sec tabulation in the arsenopyritepaper.) A preliminary investigationby the writer developed the fact that the 0.0.4 series of speetra eonstitutes a graduallydeclining2) intensity serics rathcr than a regu lar ly dcclining series whichcharacterizesthe pure marcasite type of structure.i\langanite therefore has certain arsenopyrite-likecharacteristics.-----1) Buerger,the arsenopyrite2) Buerger,2\1. .J., The symmetry and crystal strueture of the mineralsgroup. Z. Kristallogr.(A) 95 (1936) 112-113.:VI. .J., op. cit.,91-92.of

The SymmetryNotesand CrystalStructureon the Symmetryof }langanite,Mn(OH)O.16;')of 3'Ianganite.The writer had access to a number of manganite specimens from Ilfeld, Germany, which is also the locality of Garrido'smaterial. A numberof analyses of manganitefrom this locality have been eolleeted byKoeehlin1).In the best of this material, from which small crystals were selectedfor X-ray investigation,it is quite apparent that there arc twin lamellaeparallel to (100) which show up plainly on the crystal terminations.Sinc:e(:100)would be a symmetry plane if manganite were orthorhombic,andsince twinning cannot oc:c:ur on a symmetry plane, it is elear that manganite cannot be orthorhombic:.as recorded in all reference books. Thesymmetry of rnanganite has been under suspicion before, but the anomalies have always been explained away. The subject is reviewed hyKoechlin2)and by Hintze3).Importantevidence. for a symmetrylower than orthorhombicisthe inclined extinctionof alternatesystemsof bands 4)5) seen bytransmitted polarized light in thin cleavage flakes. These bands arcevidently the cross-sectionsof the ('100) twin lamellae mentionedabove.To these evidences of lower symmetry, the writer adds the following:If a rnanganite cleavage splinter (bounded by the excellent (010) andperfect (110) eleavages) is set up on a two-cirele goniometerwith thecleavage zone taken as the equator,then the natural(001) surfacemay have a (! coordinateof almost 4 . This should be 0 if thecrystal is orthorhombic.and it is this deviationwhich permits thepolysynthetic twinningto be observed on the basal terminationofthe cry"tal.This face, however, is a vicinal one and not the structural pinacoid,for crystal settingsmade with its use give poorlyoriented X-ray photographs.1) Kocchlin,l{., in Doeltcr,C. and Leitmcier,R., Handbuch del' '\Tineralchemie, l1I2, HJ2G, pp. k47-k48,analyses 4, 5, (), 10, 13. 14. 1;';, 16, 17.2) Koechlin,R., op. cit., p. 847.3) Rinbe,Carl,especial1y footnote 1.Handbuchcler Mineralogic,vol. 1, part 2, 1\H5, p.lH78,4) !VIiigge, 0., Zur Kenntnis del' optischcll und KohiisionseigenschaftendesManganit. Zbl. }lineral., Geol., Palaont. 11122, 1-2.5) Gnl'l'ido,:\1. .J., Structure de la mangnnite, Bull. Noc. Fi'Hn ;. }lil1( ral. 58(1935) 225.11*

}J. .J. BuergerlUethod.The X-ray study has be('n carried on entirely by the equi-inclinationW cis.sen berg method ')2)") and a1l films have been interpreted by the inspeetivc methodwithout indexing procedure. The cell characteristicswere derived froll1 films madewith Pe K rildiation. The pamnwters were fjxed from reflection jntensities recordedon two complete sets of equatorial films, one set under-exposcd to differentiate thejntensiti('s of the stl'Ong ref]('etions and the other set oVf'r-exposed to a1low comparisons between weak rcfJeetions.(;ell eharacteristics.Diffraetionsyml!1etry.The equi-inclinationWeissenbergphotographs for rotations about each of the three customary manganite axesaJl display a position symmetrywhich cannot be distinguishedfrom-02/. The intensity distributions,however, are partly of ]owe1' symmetry:Rotations about the b axis give equi-inelinationWe is sen be r g filmsdisplayingthe symmetryO2 for aJl levels. Rotationsabout the customary manganit(, a amI c axes both display the symmetry C21 for the 1:erolevels and O{ for the II-levels. (See Figs. 1--5).:Fig. 1 3.Pl'Oof of the inferior symmetryof manganite ami of the l3-eenterednature of the nHtnganite lattic(' referred to the customary manganite axial orientation. ote that all photographs display the position symm('try 2/, with the symmetry lines in the positions indicated by both full and broken inked Jines. Theintensity distributions,however, display a lower symmetryin some eases: onlythe full lines are true syn))netr c lines when aceonnt is taken of the intensity distribution. The photograpllsindicate the following information:rOf1't', ;;pO!]dingorthorholtlld('p!aJlp lattif'('pattpl'lIFig.Fig.Fig.Fig.Fig.1:2:3:4:G:a axisa axisr; axisc axisb axisequator1st layerequator1st layerequatorThis information constitutesmanganite axial orientationthe 2-fold axis.11101110Rcorresl)OndingorUlorhombil' ta('hillgSC(jUf'Df'C10107proof that the cell is B-('elltered with the customaryand that manganite is monoclinic with the b axis asi) Buerger,;Vr. J., The \Veissenbergreeiprocallatticeprojection and theteehnique of interpreting\V e i ssell berg photographs:Z. Kristallogr. (A) 88 (1934)3iW-:3RO.2) Buerger,,\1. J., The application of plane groups to the interpretation ofWe issen berg photographs:Z. Kristallogr.(A) HI (19:35) 23fi-28!J.3) Buerger,M. J., An apparatusfor eonvenientlytaking equi-inelinationWeiss en berg photographs:Z. Kristallogl'. (A) !H (ilJ3G) R7-!JH.

The Symmdryand CrystalStruetureof '\IanW'onik,Mn(OH)O.,/.!!,I I'"1;1ig. L-;IFig. 2.;-/.//.1-/Fig. 3.Fig. 4.1 ()?

JI. .J. BuergerThe diffraction symmetry is thusplainly '2/rn, and the crystal system,./is plainly monoclinic,in harmony/I/with the symmetryobservations alIready made. The customary manganiI/ite b axis is the two fold axis; no/Ichange in the customarymanganite/Iorientationneed therefore be made/I//from a point of view of symmetry.,/The patSpace lattice type./terns of the equi-inc1inationphoto//graphs of the various levels for the b/axis rotation are all identical, indi/Icating that the primitive and not theIcenteredlIlonoe1inic lattice is theIIbasic group of translationsfor man/ganite.Theaandcaxesof thisIprimitivetranslationunit extend in//the [1OJ] and [1Oil directions referredto an appropriatece]] with edgesparallel to the customary manganiteaxes. The customarymanganiteaxes may he retained,however, bychoosing a doubly primitive, B-centered cell. This alternativechoice isdesirable for it brings out the pseudo-orthorhombiccharacterof thecrystal and also avoids the confusion of a new orientation.The B-centering of the manganite ce]] lIlay now be justified in somedetail, because this is contrary to Garrido'sindexing. A rigorous studyof the B-centcring was made by comparison of the several equi-inclinationphotographsfor the same axial rotation.All levels for rotation abouttIle b axis display the plane lattice typel) 8 (referring in this discussionto the eorrespondingorthorhombiclevel patterns because of the strongpseudo-orthorhombicdimensional aspect). For rotations about the a andc axes, the level patterns are type 1J for even levels and type 10 forodd level", (Figs. 1--5) i. e., the plane net "tacks arc both of type iO. Thelattiee is therefore B-eenteredwith the axes chosen.Ce]] dimensions.A study of the rotations about the three eonven-I///,// --tional manganite axes, together with a studyphotogra ph s indicates the following cell:of severa]W ei ssen berg1) Buerger,M. .1., The application of plane groups to the interpretationW pisspn hpJ'g photographs:Z. Kristallogr.(A) !n (HJ35) 2;';5-2StJ.of