Introduction
Cancrinite-group minerals are microporous trigonal or hexagonal aluminosilicates. Their frameworks consist of layers composed by six-membered rings of Si- and Al-centred tetrahedra perpendicular to the c axis. The rings around the [0 0 z], [⅔ ⅓ z] and [⅓ ⅔ z] axes, as well as the layers formed by these rings are denoted by the letters A, B and C, respectively (Rinaldi and Wenk, Reference Rinaldi and Wenk1979; Ballirano et al., Reference Ballirano, Maras and Buseck1996).
Bystrite was first described as a trigonal (space group P31c) cancrinite-related mineral with the unit-cell parameters a = 12.855 Å and c = 10.700 Å, and the simplified formula Ca(Na,K)7(Al6Si6O24)(S32–)1.5⋅H2O (Sapozhnikov et al., Reference Sapozhnikov, Ivanov, Piskunova, Kashaev, Terentieva and Pobedimskaya1991; Pobedimskaya et al., Reference Pobedimskaya, Terentieva, Sapozhnikov, Kashaev and Dorokhova1991). Initially, two bystrite varieties have been distinguished: (1) K-rich and Cl-deficient and (2) Cl-rich and K-poor varieties. The former mineral has been approved as the separate mineral species sulfhydrylbystrite with the formula Na5K2Ca(Al6Si6O24)S52–(HS)– derived on the basis of chemical data and X-ray structural analyses (Sapozhnikov et al., Reference Sapozhnikov, Kaneva, Suvorova, Levitsky and Ivanova2017). The crystal structure of sulfhydrylbystrite is based on a four-layer Losod-type framework (the ABAC stacking sequence: see the review by Chukanov et al., Reference Chukanov, Aksenov and Rastsvetaeva2021) hosting small cancrinite (CAN) cages and large Losod (LOS) cages.
According to the International Union of Pure and Applied Chemistry (IUPAC) nomenclature (McCusker et al., Reference McCusker, Liebau and Engelhardt2001), the CAN cage [4665] is limited by six four-membered and five six-membered rings and the LOS cage [46611] is limited by six four-membered and eleven six-membered rings. The columns of CAN cages run along the [0 0 z] axis. The columns around the [⅓ ⅔ z] and [⅔ ⅓ z] axes consist of LOS cages.
In sulfhydrylbystrite, LOS cages contain Na+- and K+-dominant sites, S52– is the dominant anion in the LOS cage and (HS)– ions dominate over Cl– in columns of cancrinite cages. These ions are considered as the species-defining components of sulfhydrylbystrite. It is noteworthy that the S52− anion is thermodynamically favoured among Sn 2− anions for n = 2 to 8: the order of decreasing stability in aqueous solution is S52− >> S62− > S42− >> S72− > S32−>> S82− > S22− (Steudel and Chivers, Reference Steudel and Chivers2019).
The crystal structure of the presumed Cl-rich and K-deficient bystrite variety was investigated by Kaneva et al. (Reference Kaneva, Sapozhnikov and Suvorova2017). It was shown that this mineral is isostructural with sulfhydrylbystrite and is its analogue with Cl– rather than (HS)– in cancrinite cages, and that is has Na+ at the site occupied predominantly by K+ in sulfhydrylbystrite. It has been suggested (Sapozhnikov et al., Reference Sapozhnikov, Kaneva, Suvorova, Levitsky and Ivanova2017) that the presumed Cl-rich and K-deficient bystrite variety should be considered as bystrite s.s. However, this suggestion has not been discussed in the Commission on New Minerals, Nomenclature and Classification of the International Mineralogical Association (IMA–CNMNC) and the status of bystrite as a mineral species has remained uncertain.
On the basis of data from wet chemical analyses, 96% of total sulfur in bystrite occurs in the sulfide form (Sapozhnikov et al., Reference Sapozhnikov, Ivanov, Piskunova, Kashaev, Terentieva and Pobedimskaya1991), however the X-ray structural analysis shows that all sulfur belongs to the S52– anion. Minor sulfate sulfur could be partly formed as a result of partial sulfur oxidation during the analysis.
A revised formula of bystrite, Na7Ca(Al6Si6O24)S52–Cl–, has been approved by the IMA–CNMNC (Nomenclature Voting proposal 22-H, Miyawaki et al., Reference Miyawaki, Hatert, Pasero and Mills2023). A four-layer aluminosilicate framework with the ABAC stacking sequence, the predominance of S52– and Cl– extra-framework anions in the LOS cages and in the column of CAN cages, respectively, Ca2+ cations at the centres of the bases of LOS cages and the absence of K-dominant sites are the species-defining features distinguishing bystrite from other cancrinite-group minerals.
The holotype specimen of bystrite (Sample 1 in this paper) is deposited in the Fersman Mineralogical Museum (Moscow) with the catalogue number 92390. The cotype specimen of bystrite (Sample 2 in this paper) is deposited in the collection of the Sidorov Mineralogical Museum (INRTU), Irkutsk, Russia, registration number MMU/MF 28069.
In this paper, we provide data on bystrite which is a chloride and K-poor analogue of the K- and (HS)–-dominant mineral sulfhydrylbystrite, Na5K2Ca(Al6Si6O24)S52–(HS)– that are required to define this mineral as an individual species, as well as data on isomorphism of four-layer cancrinite-group minerals with the ABAC stacking sequence of layers of tetrahedra.
Samples
Data on four samples are presented. Samples 1 and 2 are bystrite. Both samples originate from the Malo–Bystrinskoe gem lazurite deposit, Baikal Lake area, Siberia, Russia. They form yellow anhedral equant grains up to 1 mm across in partly recrystallised lazurite calciphyre (Figs 1a and 2a). The associated minerals are calcite, lazurite, sodalite, fluorapatite, phlogopite, diopside, dolomite and plagioclase Pl34–43 (Fig. 1a). Bystrite grains are embedded in calcite granular aggregate. Lazurite and bystrite show no reaction relations. Sodalite forms inclusions in bystrite and lazurite, and dolomite occurs as inclusions in calcite. Locally, bystrite occurs as a component of fine-grained polymineral aggregates in which grains of earlier minerals are partly substituted by phlogopite and plagioclase (Fig. 1b). Bystrite-bearing assemblages have a metasomatic origin (Sapozhnikov et al., Reference Sapozhnikov, Ivanov, Piskunova, Kashaev, Terentieva and Pobedimskaya1991).
Sample 1 represents fragments of the bystrite holotype (Sapozhnikov et al., Reference Sapozhnikov, Ivanov, Piskunova, Kashaev, Terentieva and Pobedimskaya1991). The holotype specimen of bystrite is deposited in the Fersman Mineralogical Museum (Moscow) with the catalogue number 92390. New data on the chemical composition, infrared and Raman spectra and density (this work), powder and single-crystal X-ray diffraction, crystal structure (Kaneva et al., Reference Kaneva, Sapozhnikov and Suvorova2017), as well as optical data (Sapozhnikov et al., Reference Sapozhnikov, Ivanov, Piskunova, Kashaev, Terentieva and Pobedimskaya1991) were obtained for this sample. Additionally, for bystrite Sample 2, data on the chemical composition, infrared and Raman spectra, powder X-ray diffraction pattern and density are obtained. Sample 2 is considered as a cotype of bystrite.
Sample 3 is sulfhydrylbystrite previously described by Kaneva et al. (Reference Kaneva, Sapozhnikov and Suvorova2017). Its empirical formula is HxNa4.37K2.22Ca1.17(Si6.12Al5.87Fe0.01O24)S5.86Cl0.09, where S is total sulfur (Chukanov et al., Reference Chukanov, Shchipalkina, Shendrik, Vigasina, Tauson, Lipko, Varlamov, Shcherbakov, Sapozhnikov, Kasatkin, Zubkova and Pekov2022d). Sample 3 forms red–orange anhedral grains up to 0.4 mm across in a metasomatic rock mainly composed of calcite, diopside, nepheline, lazurite, afghanite, phlogopite and plagioclase, with Ca- and Ba-bearing stronalsite, Na2(Sr,Ca,Ba)(Si4Al4O16)⋅nH2O, occurring in the rock in different proportions (Figs 2b and 3). Sample 3 is a fragment of the material, another part of which is deposited as the holotype specimen of sulfhydrylbystrite at the Mineralogical Museum of Saint-Petersburg State University (catalogue no. 1/19636).
Sample 4 is a K-rich analogue of bystrite. It occurs in a boudin 12 cm × 8 cm × 6 cm (Fig. 4), found in the year 2000 at the dump of the Malo–Bystrinskoe deposit. The mineral forms yellow grains up to 40 μm × 80 μm (Fig. 5) in an intermediate zone between the inner zone enriched in lazurite and a zone composed of diopside and calcite. The thin outer zone of the boudin is composed of a phlogopite–fluorapatite–calcite aggregate. The working number of Sample 4 in the Vinogradov Institute of Geochemistry, Siberian Branch of Russian Academy of Sciences is 1256.
Experimental methods and data processing
In order to obtain the infrared (IR) absorption spectra, powdered samples were mixed with dried KBr, pelletised, and analysed using an ALPHA FTIR spectrometer (Bruker, 2007) in the range 360–4000 cm–1 with a resolution of 4 cm–1. A total of 16 scans were collected for each spectrum. The IR spectrum of an analogous pellet of pure KBr was used as a reference.
Raman spectra of randomly oriented samples were obtained using an EnSpectr R532 spectrometer based on an OLYMPUS CX 41 microscope coupled with a diode laser (λ = 532 nm) at room temperature. The spectrum was recorded in the range of 100 to 4000 cm–1 with a diffraction grating of 1800 gr mm–1 and a spectral resolution of 6 cm–1. The radiation power at the output of the laser source was ~5 mW. The diameter of the focal spot on the sample was <5 μm. The back-scattered Raman signal was collected with a 40× objective; signal acquisition time for a single scan of the spectral range was 1 s; and the signal was averaged over 50 scans. Crystalline silicon was used as a calibration standard. The shape of the luminescent spectrum profile was corrected according to a standard procedure based on subtracting a profile of the same shape, constructed as a superposition of a set of Lorentz lines with half-widths exceeding a limiting value. The limiting half-width value was 100 cm–1.
Six spot analyses of Sample 1 were obtained in the Vinogradov Institute of Geochemistry, Irkutsk, Russia with a JXA_8200 Jeol electron microscope equipped with a wave dispersion spectrometer operated at an acceleration voltage of 20 kV, a current intensity of 10 nA and a counting time of 10 s. The beam was defocused to 20 μm to decrease the thermal effect on the sample. Under these conditions, the mineral was stable with respect to the beam effect. The following standards and analytical lines were used: pyrope (SiKα), albite (AlKα and NaKα), diopside (CaKα), orthoclase (KKα), baryte (SKα) and chlorapatite (ClKα).
Five electron microprobe analyses of Sample 2 were carried out in the Korzhinskii IInstitute of Experimental Mineralogy RAS, Chernogolovka, Russia, on an analytical suite including a digital scanning electron microscope (SEM) Tescan VEGA-II XMU (produced by Tescan Orsay Hld., Brno, Czech Republic) equipped with an energy-dispersive spectrometer (EDS) INCA Energy 450 with a wavelength dispersive spectrometer (WDS) Oxford INCA Wave 700, The EDS analyses were performed with an accelerating voltage of 20 kV, current of 65 to 80 pA, beam diameter of 120 nm and a counting time of 100 s. The beam was defocused to 20 μm to decrease the thermal effect on the sample. Under these conditions, the mineral was stable with respect to the beam effect. The following standards were used: CaF2 for F, albite for Na, synthetic Al2O3 for Al, wollastonite for Ca, potassium feldspar for K, SiO2 for Si, Fe metal for Fe, FeS2 for S and NaCl for Cl. Contents of other elements with atomic numbers > 6 are below their detection limits. The contents of all components obtained using test WDS analysis of Sample 2 (performed with an accelerating voltage of 20 kV and a current of 1–2 nA) coincide with the data obtained using the EDS-mode analysis within 1–2 rel.%.
Powder X-ray diffraction data (Kaneva et al., Reference Kaneva, Sapozhnikov and Suvorova2017; Sapozhnikov et al., Reference Sapozhnikov, Kaneva, Suvorova, Levitsky and Ivanova2017) were collected using a D8 ADVANCE Bruker diffractometer equipped with a Göbel mirror and VÅNTEC-1 PSD detector with radial Soller slits on the diffraction beam. Data were recorded in step scan mode in the 2θ range from 5 to 70°, using CuKα radiation. The experimental conditions were as follows: voltage of 40 kV, current of 40 mA, time per step of 1 s, and 2θ step size of 0.02°. VESTA (version 4.3.0) software (Momma and Izumi, Reference Momma and Izumi2011) was used to simulate the X-ray diffraction pattern of bystrite using the crystal-structure model by Sapozhnikov et al. (Reference Sapozhnikov, Kaneva, Suvorova, Levitsky and Ivanova2017). Unit-cell parameters were refined from the powder data using TOPAS 4 software (Bruker, 2008).
Results
Physical properties
Bystrite Samples 1 and 2 are yellow with vitreous lustre and pale yellow streak. A weak luminescence is observed under a λ = 532 nm laser beam. Bystrite is brittle, with Mohs hardness of 5 and distinct cleavage: on {10$\bar{1}$0}. The fracture is uneven. Density measured by flotation in heavy liquids (bromoform + heptane) is equal to 2.43(1) g⋅cm–3 for Sample 1 (Sapozhnikov et al., Reference Sapozhnikov, Ivanov, Piskunova, Kashaev, Terentieva and Pobedimskaya1991) and to 2.42(1) g⋅cm–3 for Sample 2. Density calculated using the empirical formula and unit-cell volume refined from single-crystal XRD data equals 2.412 g cm–3 (for Sample 1) and 2.428 g cm–3 (for Sample 2).
In plane-polarised light (λ = 589 nm), bystrite is yellow. The mineral is uniaxial (+). The refractive indices for Sample 1 are: ɛ = 1.660(2) and ω = 1.584(2). Pleochroism is strong: deep yellow on Ne and colourless on No.
Infrared spectroscopy
The IR spectra of bystrite and sulfhydrylbystrite (Fig. 6) are similar. The distinctive features of sulfhydrylbystrite are the bands at 3565, 3460 and 1645 cm−1 corresponding to stretching and bending vibrations of H2O molecules as well as a weak band at 2556 cm−1 corresponding to stretching vibrations of the HS− anion. The incorporation of H2O molecules in the structure of sulfhydrylbystrite is possible because of a deficit of S52− anionic groups (0.86 groups per formula unit in the empirical formula: Sapozhnikov et al., Reference Sapozhnikov, Kaneva, Suvorova, Levitsky and Ivanova2017). Weak bands at 3430 and 1628 cm−1 in the IR spectrum of bystrite are related to water adsorbed by KBr.
Based on high-level ab initio calculations, the wavenumbers of fundamental S–S stretching vibrations predicted for S52− coordinated by Li+ are 471, 463 and 416 cm−1 (Steudel and Chivers, Reference Steudel and Chivers2019). Similar bands are observed in the IR spectra of bystrite and sulfhydrylbystrite in the ranges of 413–422 and 454–466 cm−1. No bands in these ranges are in the IR spectrum of carbobystrite, a carbonate cancrinite-group mineral with the bystrite-type framework (Chukanov, Reference Chukanov2014).
Raman spectroscopy
A weak luminescence of bystrite is observed under laser beam, unlike sulfhydrylbystrite that shows a strong luminescence under the same conditions (Fig. 7).
The strongest band of SO42− in Raman spectra of sulfate minerals belonging to the cancrinite and sodalite groups is observed in the range of 970–990 cm−1 (Chukanov et al., Reference Chukanov, Sapozhnikov, Shendrik, Vigasina and Steudel2020a, Reference Chukanov, Vigasina, Zubkova, Pekov, Schäfer, Kasatkin and Yapaskurt2020b, Reference Chukanov, Zubkova, Pekov, Shendrik, Varlamov, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022b; Sapozhnikov et al., Reference Sapozhnikov, Tauson, Lipko, Shendrik, Levitskii, Suvorova, Chukanov and Vigasina2021). This band is absent in the Raman spectra of bystrite Sample 2 (see chemical data below) and sulfhydrylbystrite Sample 3 (Fig. 8).
In the ranges of S–S stretching (410–540 cm−1) and S–S–S bending (170–240 cm−1) vibrations, both spectra contain a strong doublet and a strong single band, respectively. In the Raman spectra of minerals belonging to the sodalite–sapozhnikovite solid-solution series with the general formula Na8(Al6Si6O24)(Cl,HS)2, bands in the ranges of 459–464 and 254–260 cm−1 are related to stretching and bending vibrations of the [(Cl,HS)Na4]3+ clusters (Chukanov et al., Reference Chukanov, Zubkova, Pekov, Shendrik, Varlamov, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022b). Thus, the peak at 505 cm−1 in the Raman spectrum of bystrite and the peak at 508 cm−1 in the Raman spectrum of sulfhydrylbystrite are to be assigned to S–S stretching vibrations of the S52– anion occurring in the structures of these minerals. The bands at 442 and 447 cm−1 may be either resonance modes involving S–S and Na–(Cl,HS) stretching vibrations or a result of overlapping of S–S and Na–(Cl,HS) stretching bands. Correspondingly, the peaks at 187 and 189 cm−1 are due to S–S–S bending vibrations of S52–, whereas the peaks at 250 and 255 cm−1 are due to Na–(Cl,SH) stretching vibrations.
The strong band of H–S stretching vibrations observed in the Raman spectrum of sulfhydrylbystrite at 2562 cm−1 is close to the band at 2553 cm−1 corresponding to the of H–S stretching vibrations of the HS− anion in sapozhnikovite Na8(Al6Si6O24)(Cl,HS)2 (Chukanov et al., Reference Chukanov, Zubkova, Pekov, Shendrik, Varlamov, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022b). The analogous band in the Raman spectrum of bystrite Sample 2 (at 2557 cm−1) is much weaker, which is in agreement with the chemical composition of this mineral.
Most probably, the Raman band of bystrite at 845 cm−1 is an overtone of the band at 442 cm−1. All other (weak) bands in the Raman spectrum of bystrite correspond to vibrations of the aluminosilicate framework.
A specific feature of the red–orange sulfhydrylbystrite variety distinguishing it from yellow bystrite is a series of additional weak Raman bands at 327, 362, 395 and 667 cm−1. Similar bands (at 330, 373 and 674 cm−1) are predicted for the cis-S4 neutral molecule based on high-level ab initio calculations (Eckert and Steudel, Reference Eckert and Steudel2003). The presence of cis-S4 (planar non-cyclic C2v isomer), that is a strong red chromophore (Rejmak, Reference Rejmak2020; Chukanov et al., 2022) in red–orange sulfhydrylbystrite is in agreement with its colour, different from the yellow colour of bystrite. Note that the acyclic cis-S4 isomer is the most thermodynamically stable (Eckert and Steudel, Reference Eckert and Steudel2003; Wong and Steudel, Reference Wong and Steudel2003; Rejmak, Reference Rejmak2020). This molecule was also detected in a number of sodalite-group minerals from the Malo–Bystrinskoe deposit (Chukanov et al., Reference Chukanov, Sapozhnikov, Shendrik, Vigasina and Steudel2020a, Reference Chukanov, Shendrik, Vigasina, Pekov, Sapozhnikov, Shcherbakov and Varlamov2022a).
Another specific feature of the sulfhydrylbystrite studied is the weak Raman band at 607 cm−1 corresponding to a minor admixture of the S2•− radical anion that is the cause of strong luminescence of red–orange sulfhydrylbystrite and some sodalite-group minerals, including the recently approved species, sapozhnikovite (Chukanov et al., Reference Chukanov, Zubkova, Pekov, Shendrik, Varlamov, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022b) and bolotinaite (Chukanov et al., Reference Chukanov, Zubkova, Schäfer, Pekov, Shendrik, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022c). A weak luminescence of bystrite Sample 2 under the laser beam and the absence of a detectable band at ~607 cm−1 indicate that the S2•− radical anion may only occur in this mineral in trace amounts.
The assignment of Raman bands was carried out in accordance with data from Eckert and Steudel (Reference Eckert and Steudel2003), Steudel and Chivers (Reference Steudel and Chivers2019), Rejmak (Reference Rejmak2020) and Chukanov et al. (Reference Chukanov, Sapozhnikov, Shendrik, Vigasina and Steudel2020a, Reference Chukanov, Vigasina, Zubkova, Pekov, Schäfer, Kasatkin and Yapaskurt2020b; Reference Chukanov, Shendrik, Vigasina, Pekov, Sapozhnikov, Shcherbakov and Varlamov2022a, Reference Chukanov, Zubkova, Pekov, Shendrik, Varlamov, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022b, Reference Chukanov, Shchipalkina, Shendrik, Vigasina, Tauson, Lipko, Varlamov, Shcherbakov, Sapozhnikov, Kasatkin, Zubkova and Pekov2022d).
Chemical data
Analytical data are given in Table 1. Sample 1 is a S-deficient variety of bystrite. Taking into account that according to the wet chemical analyses 4% of total sulfur may occur in the sulfate form (Sapozhnikov et al., Reference Sapozhnikov, Ivanov, Piskunova, Kashaev, Terentieva and Pobedimskaya1991), total sulfur was divided among S52– and SO42– based on the charge-balance requirement. Sample 2 is more S-rich. According to the Raman spectrum, this sample contains minor HS– admixture.
* Based on the wet chemical analysis, up to 4% of total sulfur in Sample 1 may occur in the sulfate form (Sapozhnikov et al., Reference Sapozhnikov, Ivanov, Piskunova, Kashaev, Terentieva and Pobedimskaya1991). Sulfur is divided among S52 – and SO42 – taking into account the requirement of charge balance in the empirical formula (see text).
** The Raman spectrum shows the presence of minor HS– admixture in Sample 2. Bands of SO42– are not observed in the Raman spectrum of this sample. Sulfur is divided among S52– and HS– taking into account the charge-balance requirement in the empirical formula (see text).
The empirical formula of Sample 1 based on (Si,Al)12 is Na6.97K0.04Ca0.98(Si6.03Al5.97O24)(S52–)0.93[(SO42–)0.15Cl0.83].
The empirical formula of Sample 2 based on (Si,Al)12 is Na6.75K0.04Ca1.11(Si6.09Al5.91O24)(S52–)1.04[(HS–)0.17Cl0.85] (sulfur is divided among S52– and HS– taking into account the charge-balance requirement). Some excess of extra-framework anions (2.06 instead of the theoretical value of 2) may be due to a minor admixture of the S2•– radical anion and/or analytical error.
The idealised end-member formula of bystrite is Na7Ca(Al6Si6O24)S52–Cl–. It requires (wt.%): Na2O 19.53, CaO 5.05, Al2O3 27.53, SiO2 32.44, S 14.43, Cl 3.19, –O≡S52– –1.44, –O≡Cl– –0.73, total 100.00.
X-ray diffraction and crystal structure
Powder X-ray diffraction data for bystrite are given in Table 2. The unit-cell parameters of Sample 2 refined from the powder data are: a = 12.852(1) Å, c = 10.692(1) Å and V = 1529.39(1) Å3.
* For the calculated pattern, only reflections with intensities ≥1 are given.
** For the unit-cell parameters calculated from single-crystal data.
The strongest reflections are marked in bold type.
The crystal structure of bystrite was first solved on Sample 1 (Pobedimskaya et al., Reference Pobedimskaya, Terentieva, Sapozhnikov, Kashaev and Dorokhova1991) and then refined on Sample 2 (Sapozhnikov et al., Reference Sapozhnikov, Kaneva, Suvorova, Levitsky and Ivanova2017; Kaneva et al., Reference Kaneva, Sapozhnikov and Suvorova2017). The structure is based on a Losod-type aluminosilicate framework with the ABAC stacking sequence.
Based on the <T–O> distances in the Si- and Al-centred tetrahedra (<Si1–O> = 1.614(3) Å, <Si2–O> = 1.615(3) Å, <Al1–O> = 1.731(4) Å and <Al2–O> = 1.729(4) Å; Sapozhnikov et al., Reference Sapozhnikov, Kaneva, Suvorova, Levitsky and Ivanova2017) it is concluded that Si and Al are fully ordered in the crystal structure. The Si–Al framework contains two kinds of cages, small cancrinite (CAN) and larger Losod (LOS) ones. The crystal structure of bystrite is illustrated in Figs 9 and 10.
Two extra-framework cation sites (M1 and M2, Fig. 9a) are situated at the centres of the bases of the cages and are occupied by Ca and Na, respectively. The M3 and M4 positions within the LOS cages (Fig. 9a) are split into pairs of sub-sites which are partly occupied by Na (Fig. 10). The X6 site (Fig. 9a) located inside the cancrinite cage on the three-fold axis is occupied by Cl–. The LOS cage contains the S52– anion which is stretched along [001]. The S52– anion has a chain configuration. The sulfur atoms are disordered in such a way that trans- or cis-conformers of S52– alternate in the structure (Fig. 9b,c). The S atoms belonging to S52– occupy the X1–X5 sites and coordinate the M2–M4 sites (Fig. 9a). A small amount of Ca2+ may occur at the M2 site, as suggested by excess of Na+ occupancy (~1.1) and chemical data.
Isomorphism of cancrinite-group minerals with the bystrite-type framework
Chemical variations of S-bearing bystrite-type are mainly determined by the substitutions of Na+ vs. K+ at the M4 site and HS– vs. Cl– at the X6 site. All minerals belonging to the bystrite–sulfhydrylbystrite solid solution described previously are close to the bystrite or sulfhydrylbystrite end-members [Na7Ca(Al6Si6O24)S52–Cl– and Na5K2Ca(Al6Si6O24)S52–(HS)–, respectively].
Sample 4 is an exception to this regularity. Its chemical composition is given in Table 3. The empirical formula of Sample 4 is Na4.64K1.65Ca1.40(Si6.17Al5.81Fe0.02O24)S4.98Cl0.72 where S is total sulfur. By analogy with other bystrite-type minerals, one can suppose that a major part of sulfur in Sample 4 belongs to the S52– anion. Sample 4 is characterised by rather wide variations of the contents of extra-framework cations (Na, K and Ca) whereas the contents of S and Cl are more stable.
* Note: All sulfur is given as S52–.
Unfortunately, no other data could be obtained for Sample 4 because of very small sizes of its individual crystals and their intimate intergrowths with associated minerals.
Carbobystrite, Na8(Al6Si6O24)(CO3)⋅4H2O (Khomyakov et al., Reference Khomyakov, Cámara and Sokolova2010) crystallised in a late state of peralkaline pegmatite formation, at a high activity of CO2 and H2O. Among associated minerals, there are hydrous silicates (natrolite, umbite and members of the labuntsovite group). Sulfur is concentrated in associated sulfides of chalcophile elements (sphalerite and galena). As a result, carbobystrite does not contain Ca and S. There is no evidence of solid solution among carbobystrite and bystrite–sulfhydrylbystrite.
Comparative data for bystrite, sulfhydrylbystrite and carbobystrite are given in Table 4.
Discussion
Raman spectroscopy is a sensitive tool used to detect polysulfide species (Eckert and Steudel, Reference Eckert and Steudel2003; Chukanov et al., Reference Chukanov, Sapozhnikov, Shendrik, Vigasina and Steudel2020a, Reference Chukanov, Vigasina, Zubkova, Pekov, Schäfer, Kasatkin and Yapaskurt2020b; Sapozhnikov et al., Reference Sapozhnikov, Tauson, Lipko, Shendrik, Levitskii, Suvorova, Chukanov and Vigasina2021; Chukanov et al., Reference Chukanov, Shendrik, Vigasina, Pekov, Sapozhnikov, Shcherbakov and Varlamov2022a, Reference Chukanov, Vigasina, Zubkova, Pekov, Schäfer, Kasatkin and Yapaskurt2022b, Reference Chukanov, Zubkova, Schäfer, Pekov, Shendrik, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022c, Reference Chukanov, Shchipalkina, Shendrik, Vigasina, Tauson, Lipko, Varlamov, Shcherbakov, Sapozhnikov, Kasatkin, Zubkova and Pekov2022d). Both bystrite and sulfhydrylbystrite show strong Raman bands corresponding to the S52– anion which is a yellow chromophore. These minerals are the only mineral species containing S52– as a species-defining component. The lack of isomorphism in the bystrite/carbobystrite relationship may be due to the lack of available samples. Possibly, chemical differences between these minerals are due to geochemical rather than crystal-chemical factors. At least, in minerals belonging to the sodalite–sapozhnikovite solid-solution series a wide isomorphism involving Cl– and HS– is observed (Chukanov et al., Reference Chukanov, Zubkova, Pekov, Shendrik, Varlamov, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022b).
In bystrite, the HS– anion is absent or occurs in minor amounts. However, its presence in sulfhydrylbystrite and some bystrite varieties is important because it is an indicator of highly reducing conditions. This matter was discussed in reference to the formation of the HS–-dominant sodalite-group mineral sapozhnikovite, Na8(Al6Si6O12)(HS)2, and associated oxalate-dominant cancrinite-group mineral kyanoxalite, Na7(Al5–6Si6–7O24)(C2O4)0.5–1.0⋅5H2O, under reducing conditions which appeared as a result of aegirine crystallisation (Chukanov et al., Reference Chukanov, Zubkova, Pekov, Shendrik, Varlamov, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022b).
It was also shown (Chukanov et al., Reference Chukanov, Shendrik, Vigasina, Pekov, Sapozhnikov, Shcherbakov and Varlamov2022a) that the conversion of SO42– and S3•– into HS– coupled with the conversion of CO2 into C2O42– takes place in slyudyankaite and other sodalite-group minerals from the Malo–Bystrinskoe deposit after their heating under reducing conditions (over the Fe–FeS buffer) at 700°C. The by-products of these transformations are S4, S4•– and S2•–.
Sapozhnikovite, sulfhydrylbystrite and bystrite are the only minerals in which the presence of HS– was reliably detected. These minerals are the only feldspathoids containing all or almost all sulfur in the sulfide form.
Unlike most sodalite-group minerals from the Malo–Bystrinskoe deposit, sapozhnikovite contains only trace amounts of CO2 molecules which were partly converted to COS molecules (Chukanov et al., Reference Chukanov, Zubkova, Pekov, Shendrik, Varlamov, Vigasina, Belakovskiy, Britvin, Yapaskurt and Pushcharovsky2022b). Similarly, sulfhydrylbystrite and bystrite do not contain CO2 molecules in amounts detectable by routine IR spectroscopy.
Marbles that are among the host rocks of lazurite-bearing bodies are locally enriched in graphite and pyrite (Ivanov and Sapozhnikov, Reference Ivanov and Sapozhnikov1985), which could be a cause of reducing conditions in the formation of bystrite.
As noted above, most available analyses of bystrite and sulfhydrylbystrite correspond to samples close to the end-members of these minerals, i.e. K-rich samples are Cl-depleted and Cl-rich samples are K-depleted. Most likely, this regularity has a geochemical rather than crystal-chemical cause because there are no significant differences between crystal-chemical characteristics (charges, effective radii and force constants) of Cl– and HS–. The existence of a K- and Cl-rich bystrite-type mineral with the idealised formula Na5K2Ca(Al6Si6O24)S52–Cl– and species-defining K, S52– and Cl– (Sample 4) confirms this assumption. Moreover, there is a complete solid-solution series between sapozhnikovite and its Cl– analogue sodalite.
Acknowledgements
The authors are grateful to Owen Missen, Fernando Cámara, Stefan Farsang and an anonymous reviewer for the useful discussion. The crystal-chemical analysis, Raman spectroscopic studies and crystal chemical analysis of bystrite and sulfhydrylbystrite by NVC and MVF were supported by the Russian Science Foundation, grant No. 22-17-00006. Data on infrared spectra and chemical data of associated minerals were obtained in accordance with the state task, state registration number ААAА-А19-119092390076-7.
Competing interests
The authors declare none.