Anton R. Chakhmouradian

Lueshite, pyrochlore and monazite-(Ce) from apatite-dolomite carbonatite, Lesnaya Varaka complex, Kola Peninsula, Russia

Abstract Apatite-dolomite carbonatite at Lesnaya Varaka, Kola Peninsula, Russia, hosts intricate mineral intergrowths composed of lueshite in the core and pyrochlore-group minerals in the rim. Lueshite is a primary Nb-bearing phase in the carbonatite and ranges in composition from cerian lueshite to almost pure NaNbO3. For comparison, the compositional variation of lueshite from the Kovdor and Sallanlatvi carbonatites is described. At Lesnaya Varaka, lueshite is replaced by nearly stoichiometric Na-Ca pyrochlore due to late-stage re-equilibration in the carbonatite system. X-ray powder diffraction data for both minerals are presented. Barian strontiopyrochlore, occurring as replacement mantles on Na-Ca pyrochlore, contains up to 43% Sr and 8-18% Ba at the A-site, and shows a high degree of hydration and strong ionic deficiency at the A- and Y-sites. This mineral is metamict and, upon heating, recrystallises to an aeschynite-type structure. Monazite-(Ce) found as minute crystals in fractures, represents the solid solution between monazite-(Ce) CePO4, brabantite CaTh(PO4)2 and SrTh(PO4)2. Our data indicate the high capacity of the monazite structure for Th and accompanying divalent cations at low temperatures and pressures that has a direct relevance to solving the problem of long-term conservation of radioactive wastes. Monazite-(Ce) and barian strontiopyrochlore are products of low-temperature hydrothermal or secondary (hypergene) alteration of the primary mineral assemblage of the carbonatite.

Three compositional varieties of perovskite from kimberlites of the Lac de Gras field (Northwest Territories, Canada)

Abstract In hypabyssal and crater-facies kimberlites of the Lac de Gras kimberlite field, perovskite occurs as reaction-induced rims on earlier-crystallized Ti-bearing minerals (magnesian ilmenite and priderite), inclusions in atoll spinels and discrete crystals in a serpentine-calcite mesostasis. The mineral is associated with spinels, apatite, monticellite, phlogopite, baryte, Fe-Ni sulphides, ilmenite, diopside and zircon. Uncommon accessory phases found in an assemblage with perovskite include titanite, monazite- (Ce), witherite, strontium-apatite, khibinskite, djerfisherite, wollastonite, pectolite, suolunite, hydroxyapophyllite and bultfonteinite. Three types of perovskite can be distinguished on the basis of composition: (I) REE-Nb-Al-poor perovskite with relatively high Sr and K contents (up to 2.2 and 0.6 wt.% oxides, respectively) occurring as mantles on priderite and inclusions in atoll spinels; (II) perovskite with elevated Al, Fe, Nb and LREE (up to 1.4, 8.3, 9.1 and 17.0 wt.% oxides, respectively) found as discrete crystals and rims on macrocrystic ilmenite; (III) perovskite significantly enriched in Na, Sr, Nb and LREE (up to 3.3, 3.4, 13.0 and 22.6 wt.% oxides, respectively) found as rims on perovskite I and II. The overwhelming majority of perovskite is represented by discrete crystals of type II. In some occurrences, this type of perovskite also has high Th contents (up to 5.5 wt.% ThO2) and Zr contents (up to 3.7 wt.% ZrO2). Textural evidence indicates that perovskite shows an overall evolutionary trend from the most primitive type I towards type III showing the highest Na, Nb and LREE contents. Perovskite of type I probably crystallized under relatively high pressures prior to the precipitation of MUM spinels. Perovskite II crystallized after magnesiochromite, pleonaste and MUM (magnesian ulvöspinel-magnetite) spinels, under increasing fO₂. The most compositionally evolved type III formed during near-solidus re-equilibration of the earlier-crystallized perovskite. The compositional variation of the Lac de Gras perovskite can be adequately characterized in terms of five major end-members: CaTiO3 (perovskite), CeFeO3, NaNbO3 (lueshite), Na0.5LREE0.5TiO3 (loparite), and CaFe0.5Nb0.5O3 (latrappite).

Primary, agpaitic and deuteric stages in the evolution of accessory Sr, REE, Ba and Nb-mineralization in nepheline-syenite pegmatites at Pegmatite Peak, Bearpaw Mts, Montana

SummaryThe pegmatites at Pegmatite Peak (Bearpaw Mts., Montana) crystallized from an evolved fraction of nepheline-syenitic melt enriched in Sr, Ba, light REE and Nb. These rocks are composed essentially of microcline (up to 1.1 wt.% Na2O and 1.0 wt.% BaO), altered nepheline (replaced by analcime, zeolites, muscovite and gibbsite), and prismatic aegirine set in an aggregate of fibrous and radial aegirine. The early accessory assemblage includes Mg-Fe mica, rutile, zircon, titaniferous magnetite and thorite. Precipitation of these phases was followed by crystallization of a plethora of rare minerals enriched in Sr, Ba, light REE and Nb. Three major stages are distinguished in the evolution of this mineralization: primary, agpaitic and deuteric. Primary repositories for Sr, REE and Nb included betafite, loparite-(Ce), crichtonite and ilmenite-group minerals. Betafite (Ta-poor, REE- and Th-rich) is present in very minor amounts and did not contribute significantly to the sequestration of incompatible elements from the nepheline-syenite melt. Loparite-(Ce) evolved predominantly by depletion in Sr and Ca and enrichment in Nb, Na and REE, i.e. from strontian niobian loparite (up to 22.0 wt.% SrO) to niobian loparite (up to 17.6 wt.% Nb2O5). Crichtonite contains minor Na, Ca and K, lacks detectable Ba and REE, and is unusually enriched in Mn (7.0–13.6 wt.% MnO). The ilmenite-group minerals evolved from manganoan ilmenite to ferroan pyrophanite, and have relatively low Nb contents (≤ 0.9 wt.% Nb2O5). During the agpaitic stage, the major repositories for incompatible elements were silicates, including lamprophyllite, titanite and chevkinite-group minerals. Lamprophyllite is generally poor in Ba, and contains relatively minor Ca and K; only few small crystals exhibit rims of barytolamprophyllite with up to 26.3 wt.% BaO. Titanite is devoid of Al and depleted in Fe, but significantly enriched in Nb, Sr, REE and Na: up to 6.4, 4.5, 4.4. and 2.9 wt.% oxides, respectively. The chemical complexity of titanite suggests involvement of several substitution mechanisms: Ca2++Ti4+⇐Na1++Nb5+, Ca2 ⇐Sr2+, 2Ca2+⇐Na1++REE3+, and Ca t++OZ-~--Nal+ + (OH)1−. Chevkinite group minerals evolved from Sr-rich (strontiochevkinite) to REE-rich compositions [chevkinite-(Ce)]. Strontiochevkinite from Pegmatite Peak is compositionally similar to the type material from Sarambi, and has high ZrO2 (up to 7.8 wt.%) and low FeOT (≤ 2.5 wt.%) contents. During the final stages of formation of the pegmatites, a deuteric F-bearing fluid enriched in Sr and REE precipitated carbonates and minor phosphates confined to fractures and cavities in the rock. In this youngest assemblage of minerals, ancylite-(Ce) is the most common Sr-REE host. Some discrete crystals of ancylite show significant enrichment in Th (up to 6.0 wt.% ThO2). Ancylite-(Ce) and bastnaesite associated with “metaloparite” and TiO2 (anatase?) comprise a replacement assemblage after primary loparite. The typical replacement pattern includes a loparite core with locally developed “metaloparite”, surrounded by a bastnaesite-anatase intermediate zone and an ancylite rim. Fluorapatite is rare, and has very high Sr, Na and REE contents, up to 21.4, 2.6 and 12.9 wt.% oxides, respectively. Compositionally, this mineral corresponds to the solid solution series between fluorapatite and belovite-(Ce). At this stage, hollandite-group minerals became a minor host for Ba; they demonstrate the evolutionary trend from priderite (5.2 wt. % K2O, 7.4 wt. % BaO) to Ba-Fe hollandite (19.2–21.4 wt. % BaO). Thus, the evolution of Sr, REE, Ba and Nb mineralization was a complex, multi-stage process, and involved primary crystallization, re-equilibration phenomena and late-stage deuteric alteration.ZusammenfassungDie Pegmatite von Pegmatite Peak (Bearpaw Mts., Montana) sind aus dem Restdifferentiat einer nephelinsyenitischen Schmelze, die an Sr, Ba, leichten SEE und Nb angereichert war, auskristallisiert. Diese Gesteine bestehen hauptsächlich aus Mikroklin (max. 1.1 Gew.% Na2O und max. 1.0 Gew.% BaO), alteriertem Nephelin (verdrängt durch Analcim, Zeolithe, Muscovit und Gibbsit) und prismatischem Agirin, welcher von einem Aggregat aus fasrigem und strahligem Ägirin umgeben ist. Als frühe akzessorische Mineralien sind Mg-Fe Glimmer, Rutil, Zirkon, titanführender Magnetit und Thorit auskristallisiert. Anschließend bildete sich eine Vielzahl seltener, Sr-, Ba, leichter SEE- und Nb-reicher Mineralien aus. In den Proben von Pegmatite Peak sind drei Hauptphasen in der Entwicklung der akzessorischen Sr-, Ba-, SEE- und Nb-Mineralisation zu unterscheiden: eine primäre, eine agpaitische und eine deuterische. Primär wurden Sr, SEE und Nb in Betafit, Loparit-(Ce), Crichtonit und Mineralien der Ilmenitgruppe eingebaut. Betafit (Ta-arm, SEE- und Th-reich) ist ein sehr seltenes Mineral in den Pegmatiten, und hat die inkompatiblen Elemente nur unbedeutend konzentriert. Loparit-(Ce) entsteht im wesentlichen durch den Austausch von Sr und Ca durch Nb, Na und SEE; d.h. durch Umwandlung von strontium- und niobhältigem Loparit (≤ 22.0 Gew.% SrO) zu niobhältigem Loparit (≤ 17.6 Gew.% Nb2O5). Crichtonit enthält eine geringe Menge Na, Ca und K, ist ohne feststellbare SEE und Ba und ist gewönlich Mn-reich (7.0-13.6 Gew.% MnO). Mineralien der Ilmenitgruppe entwickeln sich von manganfiihrendem Ilmenit hin zu eisenführendem Pyrophanit und haben relativ niedrige Nb-Gehalte (≤ 0.9 Gew.% Nb2O5). Während der agpaitischen Phase waren Silikate wie Lamprophyllit, Titanit und Mineralien der Tscheffkinitgruppe die wichtigsten Träger von inkompatiblen Elementen. Lamprophyllit ist generell Ba-arm und ist durch relativ niedrige Ca- und K-Gehalte charakterisiert. Nur wenige kleine Kristalle zeigen barytolamprophyllitische Ränder (< 26.3 Gew.% BaO). Fe ist im Titanit (Al-frei) abgereichert während Nb, Sr, SEE und Na (jeweils max. 6.4, 4.5, 4.4 und 2.9 Gew.% Oxid) angereichert wurden. Die chemische Zusammensetzung des Titanits kann durch mehrere Substituierungen erklärt werden: Ca l++Ti4+~Nal+-I-Nbs+, Ca2+ ⇐ Sr2+, 2Ca2+ ⇐Na1++REE3+, und Ca2+ +O2− ⇐Na1+ +(OH)1−. Mineralien der Tscheffkinitgruppe entwickeln sich aus Sr-reichen (Strontiotscheffkinit) hin zu SEE-reichen Gliedern [Tscheffkinit-(Ce)]. Strontiotscheffkinit von Pegmatite Peak mit hohem ZrO2-(< 7.8 Gew.%) und niedrigem FeOT-Gehalt (< 2.5 Gew.%) hat eine ähnliche Zusammensetzung wie der Holotyp von Sarambi. Während der letzten Phasen der Bildung der Pegmatite brachte ein deuterisches, F-haltiges, Sr- und SEE-reiches Fluid Karbonate und in geringer Mengen Phosphate in Spalten und Hohlräumen im Gestein zur Ausfällung. Ankylit-(Ce) ist das häufigste Sr- und SEE-führende Mineral dieser jüngsten Mineralassoziation. Manche einzelne Ankylitkristalle zeigen eine bedeutende Anreicherung von Th (< 6.0 Gew.% ThO2). Ankylit, Bastnäsit, “Metaloparit” und TiO2 (Anatas?) ersetzten den ursprünglichen Loparit. Typische Verdrängungen zeigen sich als Körner mit loparitischen Kernen, welche örtlich mit “Metaloparit” verwachsen sind, weiters einer Bastnäsit-Anatas Zwischenzone und einem ankylitischen Rand. Fluorapatit ist hier ein seltenes Mineral und hat sehr hohe Sr-, Na- und SEE-Gehalte (jeweils 21.4, 2.6 und 12.9 Gew.% Oxid). Von der chemischen Zusammensetzung aus gesehen gehört dieses Mineral zur Fluoapatit-Belovit-(Ce)-Mischkristallreiche. Während der deuterischen Phase dienten die Mineralien der Hollanditgruppe untergeordnet als Träger für Ba; sie legen die Entwicklung von Priderit (5.2 Gew.% K20, 7.4 Gew.% BaO) zu Ba-Fe-Hollandit (19.2–21.4 Gew.% BaO). Somit ist die Entwicklung der Sr-, SEE-, Ba- und Nb-Mineralisation ein komplexer mehrphasiger Prozeß und umfaßt die primäre Kristallisation, Reäquilibrierungsphänomene und eine späte deuterische Alteration.

Crystal chemistry and paragenesis of compositionally unique (Al-, Fe-, Nb-, and Zr-rich) titanite from Afrikanda, Russia

Abstract Titanite is a common accessory mineral in silicocarbonatite from the Afrikanda alkaline-ultramafic complex, Kola Peninsula, Russia. In addition to large crystals (described elsewhere), this rock contains microscopic crystals and aggregates of titanite intimately associated with, or mantling, primary Ti minerals (perovskite, ilmenite, magnetite, and garnets). The microcrysts commonly exhibit complex zoning patterns that represent a combination of oscillatory, core-rim, and/or sectorial zoning. Eight varieties of microcrystic titanite, differing in chemical composition, mode of occurrence, and style of zoning, can be distinguished. Most of the compositions have >20% of the Ti site occupied by Al, Fe, Nb, Zr, or a combination thereof, whereas substitutions at the Ca site are limited to <2%. Analysis of element correlations and Raman spectra suggests that the compositional diversity of the titanite arises from the following substitutions: (Al,Fe3+)(OH)Ti-1O-1, (Al,Fe3+)NbTi-2, Al(OH)Zr-1O-1, and ZrTi-1. Using the end-member notation, different varieties of the microcrysts contain up to 20 mol% CaFeSiO4(OH), 37 mol% CaAlSiO4(OH), 35 mol% Ca(Al0.5Nb0.5)SiO5, and 26 mol% CaZrSiO5. This compositional diversity is unparalleled by titanite from any other locality or rock type, including material from three other Kola carbonatites examined in the present work (Kovdor, Turiy Mys, and Seblyavr). All compositional varieties of the microcrystic titanite crystallized at late evolutionary stages as products of reaction between the primary Ti minerals and a deuteric aqueous fluid at temperatures near 200 °C, weakly acidic pH, and a(H4SiO40) >10.4. Under these conditions, the earlier-formed Zr minerals zirconolite and zircon were unstable and underwent extensive re-equilibration with the fluid, involving partial removal of Zr. Implications of these data for the safe disposal of nuclear waste are discussed. The structure of zirconian titanite was examined using a synthetic sample of analogous composition (25 mol% CaZrSiO5). Its structure was refined by the Rietveld method from X-ray powder diffraction data in space group A2/a [a 7.1119(7), b 8.7724(8), and c 6.6007(6) Å, β 113.569(4)°]. Reflections violating A centering and/or indicative of Ti-Zr ordering were not observed, and attempts to refine the structure in two alternative space groups (P21/a and P21) were unsuccessful.

Calcite and dolomite in intrusive carbonatites. I. Textural variations

Carbonatites are nominally igneous rocks, whose evolution commonly involves also a variety of postmagmatic processes, including exsolution, subsolidus re-equilibration of igneous mineral assemblages with fluids of different provenance, hydrothermal crystallization, recrystallization and tectonic mobilization. Petrogenetic interpretation of carbonatites and assessment of their mineral potential are impossible without understanding the textural and compositional effects of both magmatic and postmagmatic processes on the principal constituents of these rocks. In the present work, we describe the major (micro)textural characteristics of carbonatitic calcite and dolomite in the context of magma evolution, fluid-rock interaction, or deformation, and provide information on the compositional variation of these minerals and its relation to specific evolutionary processes.

Clinohydroxylapatite: a new apatite-group mineral from northwestern Ontario (Canada), and new data on the extent of Na-S substitution in natural apatites

A monoclinic analogue of the mineral hydroxylapatite was found in altered leucogabbro making up one of the intrusive units in the Mesoproterozoic Seagull pluton in northwestern Ontario, Canada. The mineral is part of a replacement assemblage developed metasomatically after igneous plagioclase. It occurs as masses of coalescent spherulites < 30 μm in diameter, and is paragenetically associated with prehnite, hibschite, titanite, rutile and an unidentified Ca silicate. The new mineral is white and chalky in appearance, has a Mohs hardness of 5 and a measured density of 3.07(2) g/cm3 ( D calc 3.13 g/cm3), and shows little deviation from a uniaxial optical behavior (α 1.632, γ 1.649). Compositionally, it is an intermediate member of the ternary system Ca5(PO4)3(OH) - Ca5(PO4)3Cl - Na3Ca2(SO4)3(OH). The average content of the latter two end-members is ca. 20 mol. % each. The combination of high Na and S contents (up to 0.63 and 0.71 apfu, respectively) is unparalleled by any previously reported natural composition, and indicates significant solubility of these elements in natural apatites even at low crystallization temperatures (< 400 °C). The monoclinic symmetry of this mineral, determined from microbeam X-ray diffraction patterns, distinguishes it from hydroxylapatite. By analogy with synthetic Ca5(PO4)3OH, its space group is P 21/ b [ a 9.445(2) A, b 18.853(4) A, c 6.8783(6) A, γ 120.00(2)°], and the deviation of symmetry from the archetypal (space group P 63/ m , a ∼9.4 A, c ∼6.9 A) probably results from ordering of (OH) anions in [00z] anionic columns and consequent doubling of periodicity along [010]. In keeping with the nomenclature of apatites, this new mineral was named clinohydroxylapatite.

Rb-Cs-rich rasvumite and sector-zoned “loparite-(Ce)” from Mont Saint-Hilaire (Québec, Canada) and their petrologic significance

Rasvumite and “loparite-(Ce)” from the Mont Saint-Hilaire alkaline complex in Quebec were re-examined using a variety of analytical techniques. Rasvumite crystals from a marble xenolith and tawite (“sodalite xenolith”) entrained in nepheline syenite contain significant amounts of Rb and Cs (up to 7.2 and 2.6 wt.%, respectively). Our data indicate that these elements are more compatible with respect to rasvumite than sodalite, tainiolite, or perovskite-type phases. Cubo-octahedral crystals and penetration twins of “loparite-(Ce)” from the tawite comprise {100} growth sectors composed of loparite-(Ce) and {111} sectors composed of lueshite; the proportion of Na 0.5 Ce 0.5 TiO 3 and NaNbO 3 components varies by ≥15 mol.% between the sectors. In addition to the light rare-earth elements and Ti, the {100} sectors are enriched in K, Sr, Ba, Y, Th, U, Fe, Si and Zr with respect to the {111} sectors, which show higher levels of Na, Ca, Nb and Ta. Some elements (Ba, Th and U) exhibit a two-fold or greater difference in D between the sectors. Crystal-chemical analysis of the sector zoning indicates that higher-charged cations partition into surface protosites with fewer bonds satisfied (in agreement with Dowty’s model). Among isovalent A -site cations, the larger partition into the {100} sectors. This observation is at variance with Dowty’s predictions, but can be readily explained in terms of the relative differences in bond strength between large and small cations (estimated from their bond-valence parameters). The distribution of B -site cations is highly charge-dependent (but size-independent) and constrained mostly by heterovalent substitutions in the A site within a given sector. Comparison with the published data shows that the inter-sectorial distribution of cations in the perovskite structure is controlled not only by their charge, radius and involvement in coupled substitutions, but also by the chemistry of crystallization environment ( e.g. , availability of Nb). The implications of these data for the study of element partitioning in perovskites are discussed. The loparite-lueshite intergrowths and Rb-Cs-rich rasvumite in the tawite are interpreted to have crystallized in equilibrium with sodalite, aegirine and tainiolite from halogen-rich peralkaline magma. The tawite and its host nepheline syenite may have formed from cognate immiscible magmas, as proposed earlier by Piilonen, McDonald and Lalonde.

Nomenclature of the perovskite supergroup: A hierarchical system of classification based on crystal structure and composition

Abstract On the basis of extensive studies of synthetic perovskite-structured compounds it is possible to derive a hierarchy of hettotype structures which are derivatives of the arisotypic cubic perovskite structure (ABX3), exemplified by SrTiO3 (tausonite) or KMgF3 (parascandolaite) by: (1) tilting and distortion of the BX6 octahedra; (2) ordering of A- and B-site cations; (3) formation of A-, B- or X-site vacancies. This hierarchical scheme can be applied to some naturally-occurring oxides, fluorides, hydroxides, chlorides, arsenides, intermetallic compounds and silicates which adopt such derivative crystal structures. Application of this hierarchical scheme to naturally-occurringminerals results in the recognition of a perovskite supergroupwhich is divided into stoichiometric and non-stoichiometric perovskite groups, with both groups further divided into single ABX3 or double A2BB′ X6 perovskites. Subgroups, and potential subgroups, of stoichiometric perovskites include: (1) silicate single perovskites of the bridgmanite subgroup; (2) oxide single perovskites of the perovskite subgroup (tausonite, perovskite, loparite, lueshite, isolueshite, lakargiite, megawite); (3) oxide single perovskites of the macedonite subgroup which exhibit second order Jahn-Teller distortions (macedonite, barioperovskite); (4) fluoride single perovskites of the neighborite subgroup (neighborite, parascandolaite); (5) chloride single perovskites of the chlorocalcite subgroup; (6) B-site cation ordered double fluoride perovskites of the cryolite subgroup (cryolite, elpasolite, simmonsite); (7) B-site cation ordered oxide double perovskites of the vapnikite subgroup [vapnikite, (?) latrappite]. Non-stoichiometric perovskites include: (1) A-site vacant double hydroxides, or hydroxide perovskites, belonging to the söhngeite, schoenfliesite and stottite subgroups; (2) Anion-deficient perovskites of the brownmillerite subgroup (srebrodolskite, shulamitite); (3) A-site vacant quadruple perovskites (skutterudite subgroup); (4) B-site vacant single perovskites of the oskarssonite subgroup [oskarssonite]; (5) B-site vacant inverse single perovskites of the cohenite and auricupride subgroups; (6) B-site vacant double perovskites of the diaboleite subgroup; (7) anion-deficient partly-inverse B-site quadruple perovskites of the hematophanite subgroup.

Fluorine-, yttrium- and lanthanide-rich cerianite-(Ce) from carbonatitic rocks of the Kerimasi volcano and surrounding explosion craters, Gregory Rift, northern Tanzania

Abstract Cerianite-(Ce), ideally CeO2, occurs as rounded grains up to 5 μm across in a block of highly altered calcite carbonatite lava from the Kerimasi volcano, and as euhedral crystals up to 200 μm across in carbonatite-derived eluvial deposits in the Kisete and Loluni explosion craters in the Gregory Rift, northern Tanzania. X-ray powder diffraction data (a = 5.434(5) Å ) and Raman spectroscopy (minor vibration modes at 184 and 571 cm-1 in addition to a strong signal at 449 cm-1) suggest the presence of essential amounts of large cations and oxygen vacancies in the Kisete material. Microprobe analyses reveal that the mineral contains both light and heavy trivalent rare earth elements (REE) (7.9-15.5 wt.% LREE2O3 and 4.9-9.7 wt.% HREE2O3), and that it is enriched in yttrium (7.1-14.5 wt.% Y2O3) and fluorine (2.2-3.5 wt.%). Single-crystal structure refinement of the mineral confirms a fluorite-type structure with a cation anion distance of 2.3471(6) Å. The cerianite-(Ce) is considered to be a late-stage secondary mineral in the carbonatitic rocks.

The crystal structure of synthetic simmonsite, Na2LiAlF6

Abstract The structure of the synthetic fluoroperovskite, Na2LiAlF6 (simmonsite), has been determined by powder X-ray diffraction using the Rietveld method of structure refinement. The compound adopts space group P21/n [#14; a=5.2842(1); b=5.3698(1); c=7.5063(2) A; β=89.98(1)°; Z=4), and is a member of the cryolite (Na2NaAlF6) structural group characterized by ordering of the B-site cations (Li, Al) and tilting of the BF6 octahedra according to the tilt scheme a−b−c+. Rotations of the B-site polyhedra are less (ΦLi=14.9°; ΦAl=17.0°) than those found in cryolite (ΦNa=18.6; ΦAl=23.5°) because of the larger difference in the ionic radii of the B-site cations in cryolite as compared to those in simmonsite. Na at the A-site is displaced from the special position resulting in 10- and 8-fold coordination in simmonsite and cryolite, respectively. By analogy with the synthetic compound, naturally occurring simmonsite is considered to adopt space group P21/n (#14) and not the P21(#4) or P21/m(#11) space groups.