Expanding Family of Litharge-Derived Sulfate Minerals and Synthetic Compounds: Preparation and Crystal Structures of [Bi 2 CuO 3 ]SO 4 and [ Ln 2 O 2 ]SO 4 ( Ln = Dy and Ho)

: During the last decades, layered structures have attracted particular and increasing interest due to the multitude of outstanding properties exhibited by their representatives. Particularly common among their archetypes, with a signiﬁcant number of mineral and synthetic species structural derivatives, is that of litharge. In the current paper, we report the structural studies of two later rare-earth oxysulfates, [ Ln 2 O 2 ]SO 4 ( Ln = Dy, Ho), which belong indeed to the grandreeﬁte family, and a novel compound [Bi 2 CuO 3 ]SO 4 , which belongs to a new structure type and demonstrates the second example of Cu 2 + incorporation into litharge-type slabs. Crystals of [Bi 2 CuO 3 ]SO 4 were obtained under high-pressure / high-temperature (HP / HT) conditions, whereas polycrystalline samples of [ Ln 2 O 2 ]SO 4 ( Ln = Dy, Ho) compounds were prepared via an exchange solid-state reaction. The crystal structure of [Bi 2 CuO 3 ]SO 4 is based on alternation of continuous [Bi 2 CuO 3 ] 2 + layers of edge-sharing OBi 2 Cu 2 and OBi 3 Cu tetrahedra and sheets of sulfate groups. Cu 2 + cations are in cis position in O5Bi 2 Cu 2 and O6Bi 2 Cu 2 oxocentered tetrahedra in litharge slab. The crystal structure of [ Ln 2 O 2 ]SO 4 ( Ln = Dy, Ho) is completely analogous to those of grandreeﬁte and oxysulfates of La, Sm, Eu, and Bi.


Introduction
Litharge-derived architectures are widely represented by both mineral and synthetic species exhibiting exceptional structural and chemical diversity, as well as a variety of properties. While the numerical majority of representatives belongs to the compounds of f -metals, its structural diversity is provided mostly by just two neighbor elements in the Periodic system, lead and bismuth. The easy formation of layered structures is commonly attributed to the "lone-pair" stereochemical activity of Pb 2+ and Bi 3+ , which favors their "one-sided" coordination. The majority of both synthetic and mineral contributions come from the chemistry of oxides [1][2][3] and oxyhalides [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18], most commonly the representatives of the so-called Sillén family. The latter generally correspond to ordered alternations of litharge-derived slabs and single or double sheets of monoatomic anions of  of litharge-derived slabs and single or double sheets of monoatomic anions of  For the majority of these architectures, interlayer charge balance requires partial aliovalent substitution for Pb 2+ and Bi 3+ . Overall, the initially neutral [PbO] litharge slabs are essentially more tolerable to the chemical nature of such substituents compared to charged [BiO] + . For instance, leadbased litharge slabs can accommodate various transition metal-based species (vanadate, chromate, molybdate, tungstate, etc.) while oxides and oxyhalides of bismuth are totally resistant to such substitution (Cd 2+ (4d 10 ), as a post-transition element [19], is not considered).

Synthesis
Crystals (Figure 2a) of novel [Bi 2 CuO 3 ]SO 4 were obtained under high-pressure/high-temperature (HP/HT) conditions. The synthesis was performed using the piston cylinder module of a Voggenreiter LP 1000-540/50 system installed at the Institute of Geosciences, University of Kiel, Kiel, Germany. CuSO 4 (Aldrich ≥99.0%, 0.119 g) and BiOCl (Aldrich ≥99.0%, 0.260 g) were weighed, mixed, and finely ground. The mixture was placed into a platinum capsule (outer diameter = 3 mm, wall thickness = 0.2 mm, length = 12 mm). The capsule was sealed on both sides and placed into the center of a 1/2-inch piston cylinder talc−Pyrex assembly. The pressure increased for 5 min at a rate of 0.2 GPa/min, until a working pressure of 1 GPa was reached, whereupon the temperature program was started at a rate of 60 • C/min up to the operating temperature of 600 • C, which was maintained at the set pressure for 6 h. The cooling time was 10 h (cooling rate ≈ 60 • C/h). Simultaneously with cooling, the pressure was released at a rate of 0.1 GPa/h. After room temperature had been reached, the experiment was decompressed during 20 min. The capsule was extracted from the high-pressure assembly and cut for further investigations. The product consisted of grass-green transparent [Bi 2 CuO 3 ]SO 4 crystals in association with unreacted BiOCl.    [35]. Rare-earth oxychlorides, LnOCl, prepared by thermal hydrolysis of LnCl 3 ·6H 2 O, were mixed with potassium sulfate (pre-dried at 140 • C for 6 h) in 2:1.1 ratio, thoroughly ground and placed in silica-jacketed alumina crucibles. The silica tubes were vacuum-sealed and annealed at 825 • C for 48 h (heating rate 50 • C/h, cooling rate 5 • C/h to 650 • C, after which the furnace was switched off. The products were washed several times with distilled water to remove the KCl by-product and excess K 2 SO 4 , and air dried.

Single-Crystal XRD Studies
A single crystal of [Bi 2 CuO 3 ]SO 4 was attached to glass fiber using an epoxy resin and mounted on a Bruker SMART APEX II DUO diffractometer (Bruker AXS, Karlsruhe, Germany) equipped with a micro-focus X-ray tube utilizing MoKα radiation. The experimental data set was collected at 150 K. Unit-cell parameters were calculated using least-squares fits. Structure factors were derived using APEX 2 after introducing the required corrections [46]; details on data collection are in Table 1. The structure was solved using direct methods and refined in SHELXL [47]. The data are deposited in CCDC under Entry No. 2021664.  14

Powder XRD Studies
High-resolution data sets were collected for [Ln 2 O 2 ]SO 4 (Ln = Dy, Ho) on a PANalytical-X'Pert diffractometer (Malvern Instruments, Malvern, UK) utilizing CuKa1,2 radiations. The refinement was done using the JANA2006 software (version 2014.11-0) [48]. As in the case of isostructural [Pb 2 F 2 ]SeO 4 [34] and Bi 2 O 2 SO 4 [36], indexing the powder patters was not straightforward as two sets of Miller indices are possible for the strongest reflections yielding two alternative unit cells which led to close residuals upon LeBail full-pattern decomposition. The correct ones, listed in Table 1 [31] taken as the initial model) were chemically sensible. Due to weak scattering from the oxygen atoms and low sensitivity of the residuals to their coordinates, a mild constraint was imposed on the S-O distances in the SO 4 2-

Crystal Structure of [Bi 2 CuO 3 ]SO 4
The structure of [Bi 2 CuO 3 ]SO 4 contains two symmetrically unique Bi positions and one Cu position. The Bi1 and Bi2 sites are coordinated by nine and ten O atoms each, respectively (Figure 2b). The general feature of the Bi 3+ coordination in [Bi 2 CuO 3 ]SO 4 is the presence of four short and very strong Bi-O bonds (2.23-2.37 Å) located in one coordination hemisphere of the Bi 3+ cations. In the opposite hemisphere, the Bi 3+ cations form from five to six longer Bi 3+ -O bonds. The distortion of the Bi 3+ coordination polyhedra is due to the stereoactivity of s 2 "lone-pairs". However, the distortion of bismuth coordination environments is not as strong as usually observed for Pb 2+ in litharge-derived structures.
The  Figure 2. Cu-Cu distance is 2.79 Å. Unfortunately, we were unable to measure magnetic properties of [Bi 2 CuO 3 ]SO 4 due to the insufficient amount of pure material.
One symmetrically independent S 6+ cation forms rather symmetrical SO 4 tetrahedra. The individual S-O distances are in the range of 1.470(4)-1.491(4) Å, which is in good agreement for well-refined sulfate structures [49].

Crystal Structure of [Ln 2 O 2 ]SO 4 (Ln = Dy, Ho)
The crystal structure of [Ln 2 O 2 ]SO 4 is completely analogous to those of grandreefite and oxysulfates of La, Sm, Eu, and Bi. It contains one symmetrically independent Ln 3+ cation (Figure 4a)

Discussion
The tetrahedral sulfate, selenate, chromate, and molybdate anions are yet the largest species which can be accommodated in the space between metal-oxide or metal-fluoride litharge slabs. Their effective size essentially exceeds that of the largest monoatomic (Te 2-) species. This results in strong distortion (mostly stretching in the ab plane) of the [M2O2] 2+ layers. The derivatives of the smallest sulfate anion are the most numerous and include compounds of all rare earths, as well as bismuth. The dissimilarity in bonding to oxygen atoms results in a large distortion of the sulfate tetrahedra; it is their utmost chemical stability that makes these distortions tolerable. According to our results for the Dy and Ho compounds, all [Ln2O2]SO4 crystallize in the structural relationships between [Ln2O2]CrO4 and [Ba2F2](S2O3) suggest possible existence of some more isostructural compounds. However, while Pb contributes to the fluoride sulfate and selenate, barium contributes only to fluoride thiosulfate but not to derivatives of other tetrahedral anions. This may be caused by very low solubility of PbCrO4, BaSO4 and BaSO4 compared to PbF2 and BaF2, while those of BaF2 and BaSO4, as well as PbSO4 (PbSeO4) and PbF2 are of the same order [34]. Incorporation of transition metal cations in square oxygen nets with the O-O distances of 2.7-2.8 Å would suggest a M-O distance of 1.9-2Å which is slightly below the common range for Cu 2+ but essentially small for Co 2+ . Therefore, the M 2+ cations reside above the "liharge" oxygens in the structures of [Bi2MO3]SO4 as clearly seen in Figure 3 d,e. Their coordination polyhedron is expanded to a distorted square pyramid (or a particularly stretched octahedron) by the oxygen atoms of sulfate tetrahedra. This is not possible in the oxyhalide or oxychalcogenide strcutures shown in Figure 1a without very strong distortions of the anionic layer and coordination polyhedra of Bi 3+ or Pb 2+ . It is possibly the polyatomic nature of the sulfate anions which permits to satisfy the coordination requirements of both Bi 3+ and M 2+ . Note that both Co 2+ (3d 7 ) and Cu 2+ (3d 9 ) exhibit pronounced Jahn-Teller effect which permits gross distortions of ochahedral coordination. With other M 2+ cations like Ni 2+ (3d 8 ) this is expected to be essentialy less likely. Furthermore, the arrangement of sulfate anions in the structures of [Bi2MO3]SO4 results in very irregular, "one-sided" coordination of Bi 3+ , which is, however, rather common due to high stereochemical activity of its lone-pair. This irregularity would not be favored by rare-earth cations. Indeed, interaction of Ln2O3 oxides (Ln ≤ Ho) with CuSO4 (as well as other sulfates of divalent metals) yields only mixtures of [Ln2O2]SO4 and CuO (or other MO oxides) with no hint ath the intermediate [Ln2MO3]SO4 composition [50]. However, it seems rather likely that some other layered structures may be found among oxysulfates of Bi and Co or Cu.

Discussion
The tetrahedral sulfate, selenate, chromate, and molybdate anions are yet the largest species which can be accommodated in the space between metal-oxide or metal-fluoride litharge slabs. Their effective size essentially exceeds that of the largest monoatomic (Te 2-) species. This results in strong distortion (mostly stretching in the ab plane) of the [M 2 O 2 ] 2+ layers. The derivatives of the smallest sulfate anion are the most numerous and include compounds of all rare earths, as well as bismuth. The dissimilarity in bonding to oxygen atoms results in a large distortion of the sulfate tetrahedra; it is their utmost chemical stability that makes these distortions tolerable. According to our results for the Dy and Ho compounds, all [Ln 2 O 2 ]SO 4 crystallize in the monoclinic grandreefite structure (Table 1). In the predominantly ionic structure of [Pb 2 F 2 ]SO 4 with essentially larger [Pb 2 F 2 ] 2+ slabs, the configuration of the sulfate anion is close to regular. Selenate is more voluminous and less chemically stable so only a handful of grandreefite-type compounds is known, including those of the earliest rare-earths and the "direct" analog of grandreefite.  4 . Therefore, Bi 3+ behaves as a size analog of Eu 3+ , falling probably beyond the stability limit of oxide selenates. The structure of [La 2 O 2 ]CrO 4 (determined from powder neutron data [32]) exhibits almost regular chromate anions which suggests their relative rigidity. This may explain both why the grandreefite structure is adopted by an only compound of the earliest rare-earth element and why an alternative structure is formed for the compounds of Pr-Tb which is not formed for selenates while the size of CrO 4 2- [34]. Incorporation of transition metal cations in square oxygen nets with the O-O distances of 2.7-2.8 Å would suggest a M-O distance of 1.9-2Å which is slightly below the common range for Cu 2+ but essentially small for Co 2+ . Therefore, the M 2+ cations reside above the "liharge" oxygens in the structures of [Bi 2 MO 3 ]SO 4 as clearly seen in Figure 3d,e. Their coordination polyhedron is expanded to a distorted square pyramid (or a particularly stretched octahedron) by the oxygen atoms of sulfate tetrahedra. This is not possible in the oxyhalide or oxychalcogenide strcutures shown in Figure 1a without very strong distortions of the anionic layer and coordination polyhedra of Bi 3+ or Pb 2+ . It is possibly the polyatomic nature of the sulfate anions which permits to satisfy the coordination requirements of both Bi 3+ and M 2+ . Note that both Co 2+ (3d 7 ) and Cu 2+ (3d 9 ) exhibit pronounced Jahn-Teller effect which permits gross distortions of ochahedral coordination. With other M 2+ cations like Ni 2+ (3d 8 ) this is expected to be essentialy less likely. Furthermore, the arrangement of sulfate Minerals 2020, 10, 887 8 of 10 anions in the structures of [Bi 2 MO 3 ]SO 4 results in very irregular, "one-sided" coordination of Bi 3+ , which is, however, rather common due to high stereochemical activity of its lone-pair. This irregularity would not be favored by rare-earth cations. Indeed, interaction of  [50]. However, it seems rather likely that some other layered structures may be found among oxysulfates of Bi and Co or Cu.