Solubility equilibria in the ternary systems Rb2SO4 – CaSO4 – H2O and Cs2SO4 – CaSO4 – H2O at ambient temperature, and the crystal structure of rubidium syngenite

For the ternary systems, Rb2SO4 – CaSO4 – H2O and Cs2SO4 – CaSO4 – H2O, solubility data have been determined at ambient temperatures (22 and 25 °C). In both systems, gypsum is found as the stable CaSO4-phase in the presence of low alkali metal sulfate concentrations. In addition, two double salts exist in the Rb-system: Rb-disalt (Rb2SO4·2CaSO4) and Rb-syngenite (Rb2SO4·CaSO4·H2O) in addition to gypsum and Rb2SO4. For Rb-syngenite, the crystal structure was determined by single crystal structure analysis (space group P21/m(11), a = 6.323 Å, b = 7.223 Å, c = 10.044 Å, β = 102.96°, Z = 2, V = 447.03 Å3), which is isomorphous to potassium syngenite (K2SO4·CaSO4·H2O). In the Cs-system, besides gypsum and Cs2SO4, only the double salt Cs2SO4·2CaSO4 exists.


Introduction
Rubidium and caesium are low in crustal abundance, about 100 and 5 ppm, respectively [1]. They accompany rocks and minerals enriched in alkali metals, especially those enriched in lithium and potassium. Lepidolite, the Li-rich mica is the chief ore mineral of rubidium. Caesium is extracted from pollucite, a caesium aluminosilicate. Extraction of lithium from the mica, zinnwaldite, in the deposit in Erzgebirge, Germany, also produces Rb e.g. as sulfate, for instance.
Naturally occurring rubidium consists of the isotopes 37 85 Rb (72%) and 37 87 Rb (28%), a beta emitter. The shortlived nuclides 37 83 Rb and 37 86 Rb are suitable for tracer experiments. Naturally occurring caesium consists of the nuclide 55 133 Cs (100%). The long-lived nuclide 55 135 Cs (half-life time 2.3 Ma) is radioactive and is formed during uranium fission and therefore as a long-term radioactive component. For long-term safety assessments of radioactive waste storage in deep geological formations, it is important to know how these isotopes will distribute on exposure to groundwater or moisture. If salt rocks are chosen as host 1 3 rock, calcium sulfate is present in the form of anhydrite (CaSO 4 ) or gypsum (CaSO 4 ·2H 2 O).
Anhydrite and gypsum, are important in the building materials industry. Anhydrite is used as floor screed, where potassium sulfate is added to accelerate the setting reaction. To understand, and hence control this effect, studies have been carried out with the alkali metal sulfates [2]. To enhance process understanding, rubidium sulfate and caesium sulfate have also been studied with the hope to find some correlation between phase formation and acceleration effect. Considering these objectives, the solubilities of the ternary systems Rb 2 SO 4 -CaSO 4 -H 2 O and Cs 2 SO 4 -CaSO 4 -H 2 O were investigated.
In the literature, only a few solubility data are known for both systems. D'Ans and co-workers [3] synthesized the double salts Rb 2 SO 4 ·CaSO 4 ·H 2 O (Rb-syngenite), Rb 2 SO 4 ·2CaSO 4 (Rb-disalt) [4], and Cs 2 SO 4 ·2CaSO 4 (Cs-disalt) [5] and subsequently determined the two-salt points (invariant points, IP) in a temperature range from 0 to 100 °C. In the Rb system, they found the IP between gypsum and Rb-disalt at 25 °C in a 1.04 molal Rb 2 SO 4 solution (without specifying CaSO 4 concentration), and the IP between Rb-disalt and Rb-syngenite in a solution of 1.42 molal Rb 2 SO 4 and 0.027 molal CaSO 4 . For the Cs-system, D'Ans and co-workers only determined the IP between gypsum and Cs-disalt in 0.78 molal Cs 2 SO 4 solution (without specifying the CaSO 4 concentration). A Cs-syngenite (as found for potassium and rubidium) does not exist according to the authors [3,5].
Smirnova & Skiba [6] and Israel [7] both determined solubility data in the system Rb 2 SO 4 -CaSO 4 -H 2 O at 25 °C, however, their results disagree (Fig. 1). The CaSO 4 concentrations given by Smirnova & Skiba are up to twice as high as Israel's and scatter considerably. The Rb-disalt is not found in contrast to D'Ans & Schreiner [3] and Israel [7].
For the system, Cs 2 SO 4 -CaSO 4 -H 2 O, currently only the solubility data of Israel [7] are available (Fig. 5). In accordance with D'Ans & Schreiner [3], Israel observed the double salt Cs 2 SO 4 ·2CaSO 4 as well as the single phases gypsum and Cs 2 SO 4 in this system.  [7] shows a comparable trend to this data but with a systematic deviation to lower CaSO 4 solution concentrations with increasing Rb 2 SO 4 content. Furthermore, the saturation concentration for Rb 2 SO 4 of 1.8 molal given by Israel [7] is too low and can be assumed to be a systematic analytical error or miscalculation by Israel when determining concentrations, however, it cannot be explained in detail.

Results and discussion
In contrast, the solubility data of Smirnova & Skiba [6] for gypsum and Rb-syngenite are at higher CaSO 4 concentrations and, in part, scatter considerably. However, the invariant point given at 1.23 molal Rb 2 SO 4 is comparable to that determined in this work. The Rb-disalt, Rb 2 SO 4 ·2CaSO 4 , which is stable in this range, was not observed by Smirnova & Skiba [6] (Fig. 1).
Solid phases were characterized by X-ray powder diffraction patterns (Fig. 2) for the solids from the 25 °C samples ( Table 2). Figure 8 shows the Raman spectra for the 22 °C samples (Table 1).
To identify Rb-syngenite in the X-ray diffraction patterns, the reference pattern has been calculated from the crystal structure data (cf. analysis of a single crystal of Rb-syngenite as described below). The X-ray diffraction patterns of the separated solids, especially No. 13-15, showed Rb 2 SO 4 additional to the equilibrium phases because Rb 2 SO 4 crystallized from adhering mother liquid which dried during the measurement. Its ratio rises with increasing Rb 2 SO 4 concentration in the liquid phase. Reflections of gypsum of the solid mixture at the invariant point gypsum/Rb-disalt (No. 13, Table 4) could only be detected by high-definition diffraction patterns and therefore are not seen at this standard resolution. However, the SEM image in Fig. 3a clearly shows gypsum crystals with the Rb 2 SO 4 ·2CaSO 4 .

Structure of Rb-syngenite
From samples No. 9, 10, and 16, pure Rb-syngenite was obtained and the selection of a suitable crystal for structure determination by single crystal analysis was possible.
As expected, a structure analogous to potassium-and ammonium-syngenite was identified. The isomorphous phases crystalize in the monoclinic system of space group P2 1 /m (No. 11). Table 3 provides a comparison of the crystallographic data of the three double salt hydrates.

3
The crystal structure of Rb-syngenite, also consists of layers of oxygen-linked CaO 9 -polyhedra and sulfate tetrahedra in the ab direction. The Ca-polyhedra are alternating edge-linked in the b direction and via sulfate tetrahedra in the a direction. In the polyhedra, eight oxygen atoms belong to five sulfates and one water oxygen coordinates the calcium ion. In the b-direction, sulfate tetrahedra and water molecules orientate alternating away from the CaO 9 -polyhedron layers into the interlayer space. Together with the rubidium ions, they fill the interlayer space (Fig. 4).

System Cs 2 SO 4 -CaSO 4 -H 2 O
For the Cs system, solubility investigations have been carried out at 22 and 25 °C up to a Cs 2 SO 4 concentration of 2 molal. Again, no differences larger than the experimental scatter could be observed between the 22 and 25 °C series (Fig. 5, Tables 4, and 5). In the presence of low Cs 2 SO 4 concentrations (up to 0.7 molal), gypsum is found as the stable solid (its metastable occurrence was observed up to 1.0 molal). With increasing Cs 2 SO 4 concentration, gypsum is replaced by the Cs-disalt, Cs 2 SO 4 ·2CaSO 4 , as the stable solid and the only double salt in the system. The identification of the solid phases by X-ray diffraction and SEM (for the 25 °C series), and Raman spectroscopy (for the 22 °C  (Table 2) series) are shown in Figs. 6, 7, and 8, right. In accordance with D'Ans & Schreiner [3], no Cs-syngenite exists. A comparison with the solubility data of Israel [7] for gypsum and Cs-disalt shows an onset of deviation above 0.5 molal Cs 2 SO 4 from the data determined within this work. According to Israel, gypsum would be the stable phase up to 1 molal Cs 2 SO 4 . Moreover, the CaSO 4 concentration at the invariant point gypsum/Cs-disalt shows an unusual drop to around 0.016 molal CaSO 4 in a 1.03 molal Cs 2 SO 4 solution (Fig. 5). More consistent with this works data, D'Ans & Schreiner [3] gave the IP gypsum/Cs-disalt in 0.78 molal Cs 2 SO 4 solution.  [14]). For the Li system, solubility data can be found at 25 °C and 50 °C, but they do not provide a consistent, meaningful data picture (compiled in Sohr et al. 2017 [15]). No double salt formation is known between Li 2 SO 4 and CaSO 4 , whereas different stoichiometries of anhydrous solids or hydrates are reported with the other alkali metal sulfates (overview   . We assume (due to our own experiences from the investigations and analyses of the Na 2 SO 4 -CaSO 4 -H 2 O system) that together with the labile salt (= eugsterite, 2Na 2 SO 4 ·CaSO 4 ·2H 2 O) they are comparable compounds, the composition of which has always been analyzed somewhat differently due to degradation and subsequent secondary components such as glauberite, thernardite (Na 2 SO 4 ) or mirabilite (Na 2 SO 4 ·10H 2 O) depending on the climatic conditions at the site of discovery or the changes on the way and time to analysis.   sum and anhydrite and the latter series to disalt as the initial solid phases respectively (see Table 1 and  (Tables 1 and 4), which were carried out in 2017 as part of Losch's PhD thesis [2] in our research group, were agitated in an air-conditioned room (22 ± 0.5 °C). The subsequent separation of the solution and solid phase was carried out by centrifugation. It should be noted that the equilibrium temperature for the solubility measurements in Losch's thesis [2] was reported as 25 °C instead of the exact 22 ± 0.5 °C. The 25 °C samples (Tables 2 and  5) were stirred in a thermostated water bath at 25 ± 0.2 °C.

General overview of the alkali metal M systems M 2 SO 4 -CaSO 4 -H 2 O
For the subsequent analysis, solutions were separated from the solids by vacuum filtration.

Analysis
For analysis of solution compositions in the 22 °C samples, all ions were quantified separately (samples in Tables 1  and 4). The concentrations of Rb + or Cs + and SO 4 2− were determined by ion chromatography (Dionex ICS 1000, Thermo Scientific) with a relative error of around 2%. The Ca 2+ concentration was determined by complexometric titration with 0.01 M Na-EDTA at a pH of 12 to 13 and with ErioT as an indicator.
In the case of the 25 °C samples (Tables 2 and 5) only Ca 2+ was analysed separately (via complexometry, as previously described). The clear solutions were evaporated to dryness where the loss of mass (= mass of water), and the subtraction of the titrimetrically determined Ca 2+ as CaSO 4 content allowed the calculation of the Rb 2 SO 4 concentration.
The solid phases of the 22 °C samples (Tables 1 and 4) have been identified by Raman spectroscopy (FT Raman spectrometer RFS 100/S, Bruker, with Nd-YAG laser) and the 25 °C samples (Tables 2 and 5) by X-ray powder diffraction (D8 Discover, Bruker Cu-Kα, linear detector Vantec 1).

Single crystal analysis and refinement
For single crystal diffraction with STOE IPDE II (image plate diffraction system) a suitable Rb-syngenite crystal was selected (sample No. 9 and 10, Table 1) and fixed by highpurity silicon grease on a glass capillary. The measurements were performed at a temperature of 153 K. Information about the measurement details, data collection and structure refinement are given in Table 8.
Further details of the crystal structure investigations may be obtained from the joint CCDC/FIZ Karlsruhe online deposition service: https:// www. ccdc. cam. ac. uk/ struc tures/ by quoting the deposition number CSD-433878.  Data availability Data availability is stated by the given CSD number.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.