The Solid Solution between NaClO 3 and NaBrO 3 Revisited

: NaClO 3 and NaBrO 3 are believed to form a complete solid solution from RT to fusion. The unique solid phase can thus be


General Concept of Solid Solutions and Motivations for the Choice of NaClO3-NaBrO3 System
Solid solutions by substitution are often encountered among inorganic and metallic systems as soon as the change between the two species can be simply the replacement of an atom by another atom with a close atomic radius. Often, in such a system, the lattice parameters of the solid solution for a given composition follow Vegard's rule, i.e., there is a linear relation between the lattice parameters of the crystal structure with the concentration of each of the components of the system [1]. However, this empirical law was not the general case, especially for metals [2,3].

NaClO3(x)BrO3(1−x) Mixed Crystals: Kinetic Ordering and Deviation from Ideal Cubic Structure
In 1984, Raja et al. [35] analyzed a crystal with Na(ClO3)0.7(BrO3)0.3 composition and found that it crystallizes, also, in the cubic P213 space group, confirming the existence of the solid solution. Nevertheless, many other works indicated different results. Since the 19th century, polarized optical observations on NaClO3(x)BrO3(1−x) solid solution crystals revealed birefringence that is physically not allowed in isotropic cubic structures [18,20,36]. The authors found that apparent single crystals exhibited sectors of different compositions (in both constituents of the solid solution), reducing the overall symmetry of these crystals to the monoclinic P21 or triclinic P1 space groups due to kinetic ordering of atoms during crystallization by water evaporation. Authors of the same research group also indicated the potential segregation of BrO3 − and ClO3 − microscale domains within an apparent 'single' crystal through alternating dextro-and levorotary domains [22]. A complete study of the NaClO3(x)BrO3(1−x) system was also recently carried out by Su et al. [37]-crystals of the solid solution between 14 to 81 wt.% NaBrO3 were found to be of worse optical quality, i.e., opaque with cracks contrary to crystals obtained outside this composition range. For these compositions, a very weak diffraction peak at 13.3° 2θ was found and could be indexed as the (100) peak, forbidden for the P213 space group (h00; h = 2n), indicating a slight deviation from an ideal cubic structure.
Nevertheless, it was reported that annealing close to the melting temperature leads to a homogenization of the crystals and a return to the P213 crystal structure [22,37,38]. It was not reported whether the NaClO3(x)BrO3(1−x) solid solution crystals return to equilibrium or if a transition towards the cubic structure is only possible at higher temperatures and could remain metastable at room temperature.

Heterogeneous Equilibria
The first study of the NaClO3-NaBrO3 system was performed in 1939 by Swenson and Ricci [26], regardless of the chirality. They published two isothermal studies of the ternary system NaClO3-NaBrO3-H2O at 25 °C and 50 °C (Figure 1a,c). They showed that there might be a complete solid solution between the two components at 50 °C (Figure 1d), but the possible existence of two solid solutions at 25 °C was envisaged ( Figure 1b). Indeed, careful examinations of (i) the solubility curve versus composition and (ii) the tie lines in this isotherm at 25 °C reveal an odd behavior. In this work, the authors suggested that the solid solution was discontinuous. The solubility curve appeared to be divided into three parts corresponding to the following solid phases: (i) a solid solution very close to pure NaBrO3, (ii) a bromate solid solution containing few ppm to 5-10% of chlorate, and (iii) a chlorate solid solution containing up to 60-65% chlorate solid solution. Another author suggested that this solid solution was continuous but with a marked tendency toward forming an isothermal invariant point.  Table I in ref. [26]) in the ternary isotherm at 25 °C. At this temperature and in accordance with the Gibbs phase rule, those three biphasic domains need to be interlaced by triphasic domains (triangular white domains). The three regions are represented with different colors. The green domain encompasses, by far, the largest part of the two-phase region from 0% bromate (100% chlorate) to ca. 60-65 wt.% NaBrO3 in the solid phase and from A to B for the corresponding saturated solution. The rather thin region, colored in grey, corresponds to a possible ordered domain ranging from 90-95% to almost 100% bromate in the solid phase and from B to C for the corresponding saturated solution. The last region colored in blue behaves like a pure component; the tie lines are likely to converge to nearly pure sodium bromate and the corresponding saturated solution ranging from C to D. (d) Swenson and Ricci's proposal for equilibria at 50 °C showing a continuous solid solution.

Summary of the Literature Review and Aim of That Study
This literature review emphasizes the complex crystallization behavior by water evaporation for the NaClO3-NaBrO3 system, especially when the NaClO3(x)BrO3(1−x) compositions are far from the respective pure phases. The problem is that crystallization is most often kinetically driven by the ability of ClO3 −/ BrO3 − ions ordering and chirality ordering. Notably, both phenomena have been observed within an apparent well-shaped single crystal.
The question of whether the NaClO3(x)BrO3(1−x) mixed crystals belong to a complete solution at room temperature (e.g., 25 °C) is finally still open. The reason might be that equilibrium seems to be challenging to reach. Before examining the system regarding the chirality and deracemization of mixed compositions (in the next study), it appeared necessary to clarify the existence of heterogeneous equilibria between these two components.

Commercial Material
Sodium chlorate was purchased from Alfa Aesar with a purity of 99%. Sodium bromate was purchased from Merck with purity of 99%.

Crystallization by Solvent Evaporation
NaClO3 and NaBrO3 commercial powders were dissolved in water (40 mg/mL) to make undersaturated solutions of different NaClO3/NaBrO3 ratios. The solutions were transferred to individual opened large crystallization beakers for evaporation at RT or in a temperature-regulated ventilated oven. The crystals were collected at the end of evaporation to recover the same composition in NaClO3 and NaBrO3 that had been formerly introduced for the preparation of the samples. For some experiments at elevated temperatures, crystallizations were performed in opened vials under stirring.

Crystallization by Spray-Drying
The spray dryer used in this study is a Büchi B290 (conventional laboratory dryer designed with a co-current configuration). The compressed gas used is nitrogen, and the feeding solutions were aqueous. The laboratory-scale spray dryer was equipped with a nozzle 0.7 mm in diameter (Tip = 0.7 mm; Cap = 1.4 mm).
Several mixtures were dissolved in water, with the following parameters (Table 1). In the drying chamber, the cyclone and bowl collector

TR-XRPD
Solid crystalline phases were analyzed by means of X-ray powder diffraction (XRPD) on a D8 advance series II diffractometer (Bruker Analytic X-ray Systems, Germany). The instrument has an X-ray tube containing a copper anticathode (40 kV, 40 mA) and a LynxEye ® linear detector for every XRD scan. The step size was fixed at 0.04° with a counting time of 0.5 s/step over an angular range of 10°-40° (2-theta). The temperature was controlled and monitored using a TTK 450 heating stage (Anton Paar GmbH, Austria). Data were processed using DIFFRAC.EVA software release 2018 (version 4.3.0.1).

DSC
DSC experiments in the temperature range 0 °C to 400 °C were performed using a Netzsch DSC 204 F1 apparatus equipped with an intracooler. Each DSC run was performed with ca. 5-15 mg of a powdered sample in aluminum pans with pierced lids at 2 or 5 K/min heating rates. The atmosphere of the analyses was regulated by a helium flux (40 mL/min). Onset (respectively endset) temperatures are calculated from the intersection between the baseline and the slope of the first (respectively last) part of the endotherm.

SEM
Scanning electron microscopy (SEM), pictures were obtained with a JEOL JCM-5000 NeoScope instrument (JEOL Ltd., Tokyo, Japan) (secondary scattering electron) at an accelerated voltage of 10 kV. Samples were stuck on an SEM stub with gloss carbon and coated with gold to reduce electric charges induced during analysis with a NeoCoater MP-19020NCTR (JEOL Ltd., Tokyo, Japan).

SEM-EDX
Scanning Electron Microscopy (JEOL JSM 7200F, JEOL Ltd., Tokyo, Japan) coupled with an Energy Dispersive X-Ray spectrometer (EDX XFLASH 6160 from Bruker, Billerica, MA, USA) was used to collect images and for quantitative chemical analysis. EDS spectra were collected under a voltage of 10 kV with a working distance of 10 mm. Although no conductive coatings were deposited, the experimental conditions were satisfactory for recording BSE (Back Scattered Electrons) images and X-ray mapping experiments.

Optical Microscope
Polarized Optical Microscopy (POM) observation was conducted using a Nikon SMZ-10A microscope (Nikon Imaging Ltd., Tokyo, Japan.) coupled to a CCD camera connected to a computer. Data processing was run by using LEICA Microsystems Application Suite X software (version 5.0.2.24429).

Confirmation of Kinetic Ordering through Classical (Slow) Water-Evaporation Crystallization Process Evidence of a Miscibility Gap at 20 °C
To confirm the existence of a complete solid solution, or by contrast, the presence of two limiting solid solutions with a miscibility gap, several 50 wt.% NaClO3/NaBrO3 mixtures in aqueous solutions were dissolved and left under stagnant conditions to fully evaporate at 20 °C, 50 °C, 80 °C, 100 °C, 120 °C and 150 °C. Two closely related XRPRD patterns were obtained (Figure 2). XRPD patterns of the 50 wt.% powder samples obtained after evaporation at room temperature indicated a miscibility gap in the solid state ( Figure  2a), even though NaClO3 and NaBrO3 are isomorphous. A careful examination of the peaks of the powder samples crystallized at RT revealed that the two solid phases are not constituted by a physical mixture of pure NaClO3 and pure NaBrO3 but rather by a physical mixture of two solid solutions with a limiting domain of existence. How crystallization is conducted influences the apparent width in the composition of the solid solution. We suspect that rather robust metastable equilibria can be reached.
This miscibility gap was confirmed to exist (at least) up to 160 °C from the evaporation experiments conducted at elevated temperatures (Figure 2b). The crystallization pathway, from undersaturated solution towards total evaporation, was investigated on a 50 wt.% NaClO3/NaBrO3 mixture dissolved in water (Figure 3). During evaporation, several single particles were withdrawn at different times to investigate the solid phase. Results confirmed the solid-solid demixion (S.S.I + S.S.II + di-saturated saturated solution) during the last crystallization episodes. : 50:50 NaClO3/NaBrO3 physical mixture (reference); (blue XRPD pattern) particle withdrawn after one day after spontaneous nucleation; (orange XRPD pattern) after two days after spontaneous nucleation; (green XRPD pattern) at the end of the process (total evaporation). Only one particle was withdrawn at a time, so the overall composition of the system remained unchanged.  POM pictures reveal that what appears as single particles are a patchwork of domains of different sizes and geometries. POM also shows that the chirality of those domains can differ even within the same crystallographic sector (Figure 4).
EDX analysis on two spots belonging to two domains reveals an almost inverted composition between Cl and Br (8.6 Cl & 30.5 Br atom % for spot 1) and (24.8 Cl and 6.2 Br atom % for spot 2). It can be surmised that those particles are composed of four domains: supramolecular left and right domains enriched in bromate and supramolecular left and right domains enriched in chlorate. Due to the poor solubility of the bromate compared to the chlorate, it is logical that the center of the particle is enriched in bromate. POM and EDX analyses were per many observations for this system in the literature [18,20,22,36,37]. From the above data, we can conclude that when the thermodynamic equilibrium is attained, sodium chlorate and sodium bromate do not form a complete solid solution from room temperature up to 160 °C, but rather two partial solid solutions. The miscibility gap in the solid state likely extends to much higher temperatures.
To see if the miscibility gap in the solid state intersects the solid-liquid equilibria, temperature, resolved X-ray powder diffraction experiments (TR-XRPD) were run for the 50% (mass) composition ( Figure 6). Several heating ramps up to 250 °C were used on the same sample. With annealing for 11 h at high temperature, these thermal treatments were interspersed with a flattening of the powder to avoid the retreat effect of sintering.  Figure 6 shows the third analysis on heating. A careful inspection of the evolution of the diffraction peaks versus temperature indicates that the two patterns, related to the presence of a NaClO3-rich and a NaBrO3-rich solid solution, tend to coalesce into a single one. Even if the annealing temperature was close to that of the solidus (see experimental points (filled circles) in Figure 7), a genuine complete merge of the two patterns is not observed. Longer annealing has shown the beginning of decomposition and, therefore, could not be prolonged. The opposite supramolecular chirality of the particle may not help the sintering. Thus, this work leads to two hypotheses shown in Figure 7a  Note that homochirality is supposed to appear to give left-handed or right-handed crystals only. (b) Tentative homochiral binary phase diagram between NaClO3 and NaBrO3 showing a stable miscibility gap that does not close up at elevated temperature but intersects the liquid-solid equilibria, implying a three-phase invariant (peritectic transition). Note: for both (a,b), filled circles correspond to experimental points determined by DSC (present work). Question marks, on x axis, reflect the difficulties to find the precise limits of the miscibility gap for stable equilibria.

Crystallization of NaClO3/NaBrO3 Mixtures by Spray-Drying. Evidence of a Metastable Complete Solid Solution
Literature shows that homogeneity could be an issue in preparing samples with different compositions. Therefore, mixtures NaClO3(x)BrO3(1−x) 0 ≤ x ≤ 1 have been prepared by spray-drying according to Protocol 1.
It can be surmised that in every droplet, a Kondepudi-like crystallization occurs. Accordingly, every droplet should give homochiral particles on drying unless there is a bias in the nucleation process due to an active chiral impurity. The global result should be the statistical nucleation of right-handed and left-handed particles, in other words: a racemic mixture. A recent article discussed the relationship between size and homochirality [40].
The obtained particles exhibit morphologies more isotropic than crystallization from stagnant solution for which the habitus is flattened with characteristic defects (Figure 8). Figure 8. SEM pictures-with the same magnification-of spray-dried samples obtained by using Protocol 1 from different initial compositions (ranging from 0% NaBrO3 (100% NaClO3) to 100% NaBrO3 (0% NaClO3)). As NaBrO3 is less soluble than NaClO3, the particles were smaller.
XRPD patterns on fresh spray-dried samples recorded at room temperature are shown in Figure 9. Contrary to powder samples obtained by slow water evaporation, the inlet shows the shift of the X-ray diffraction peak versus composition in the full composition range 0 ≤ x ≤ 1. The evolution of the cubic unit-cell parameter at room temperature as a function of the mole fraction of NaBrO3 is represented in Figure 10. Although the maximum difference between NaClO3 and NaBrO3 lattice parameters is only 1.92%, a possible slight deviation from the Vegard law seems to be observable, which could be related to the metastable character of the complete solid solution. More accurate measurements would be necessary to confirm the non-Vegard trend close to the pure components in the domain of stable partial solid solutions.  Table  S1). A deviation from the Vegard law (apparent non-linearity) seems present for some compositions. The blue line is for guiding the eyes.
The solid mixtures obtained by spray-drying (Protocol 1) were analyzed by DSC at a 5 K/min heating rate. Figure 11a shows the profiles versus composition on heating, and Figure 11b collects the onset and endset of the melting peak (solid-filled circles). The resulting binary system includes a solidus and a liquidus which are rather flat up to 60 wt.% NaBrO3. This seems consistent with the tie-lines of the ternary system at 25 °C determined by Swenson and Ricci [26]. The binary system is consistent with a continuous solid solution by means of substitution. To increase spray-drying productivity, a mixture of 50% (mass) was used with protocol 2 (concentration eight times higher than protocol 1 and the temperature of the feeding liquid was 60 °C). With these new parameters, part of the solid was stuck on the upper part of the drying chamber. The rest was stuck on the outlet of the drying chamber or in the collector bowl.
XRPD analyses of the solid in the drying chamber and the collector bowl are displayed in Figure 12 (The solid stuck at the outlet of the drying chamber appeared identical to that in the collector bowl). The two XRPD patterns show a clear difference: (i) the powder stuck in the upper part of the drying chamber is duplicated, testifying the presence of two similar solid solutions (ii) the powder sampled from the collector bowl is a single phase with diffraction peaks having intermediate d-spacings in comparison with the two partial solid solutions. For the former, the full and fast dehydration of the droplets is aborted, resulting in two solid solutions. For the latter, a fast and completed dehydration leads to a unique solid solution with the same composition as the feeding solution. This solid is the same result as when using protocol 1.
If the powder obtained after spray-drying in the collector bowl is put for one day under 84% relative humidity at RT, the single XRPD signature spits into two resembling XRPD patterns, just like what is obtained by complete slow evaporation. These results demonstrate that, whatever the composition, the two components could be forced to mix completely if the kinetics of crystallization is fast enough and if the drying is completed in a single step, such as during spray-drying using protocol 1. Of course, under these conditions, homochirality cannot be controlled; the final product is a random distribution of chiral particles -probably close to a racemic mixture. Thus, the binary section presented in Figure 11b is merely a metastable racemic section of the quaternary system: NaClO3\\-NaClO3//-NaBrO3\\-NaBrO3//, The symbols: \\ and//refer to opposite supramolecular chirality. The crystallographic features of the two components: the same space group, and a small difference only of the cubic parameter (less than 2% at RT), render possible the existence of two mirror-image metastable complete solid solutions.

Conclusions & Perspectives
Spray-drying is a technique which favorizes access to metastable states because the entire crystallization-including nucleation, growth, and complete drying-takes place in the order of one second. Using this energetic and swift process, the sodium chlorate and sodium bromate appear fully miscible from RT to fusion. However, the two-mirror-image complete solid solutions are actually in a metastable state. Strangely enough, the metastable part of the solid solution corresponds to a slightly denser crystal lattice than the mean of the two stable limiting solid solutions. When the relative humidity exceeds a threshold of ca. 75-84%, the two (lefthanded and righthanded) complete solid solutions spit into mirror-image partial solid solutions.
For larger particles, such as those obtained by slow evaporation without stirring, this work evidences the presence of domains rich in bromate and others rich in chlorate, which have epitaxial relationships. For solid solutions grown from stirred solutions, the composition ranges of the two limited solid solutions are difficult to assess. On the one hand, spray-drying experiments and, on the other hand, EDX analyses of composition in different sectors of single particles are consistent with these fluctuations.
The construction of the genuine stable phase diagram does not appear that straightforward. Indeed, in addition to the formations of the partial solid solutions, it is necessary to reach homochirality. This could be ensured by continuous grinding (but the pollution by material snatched off from the beads could be an issue) or temperature cycling assisted or not by ultrasounds. The kinetics of attaining the 'true' equilibrium depends on the doses of those treatments (i.e., intensity versus time).
Future work will focus on the deracemization of NaClO3-rich solid solution and NaBrO3-rich solid solution up to their limits in compositions. It will also be interesting to test the deracemization inside the miscibility gap of these two partial solid solutions. Is there an influence of the two ssNaClO3 rich and ssNaBrO3 rich to reach identical supramolecular chirality so: homo-homochirality or, by contrast: hetero-homochirality?
Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/min13081006/s1, Figure  S1: XRPD of ground NaClO3/NaBrO3 physical mixture (black) and XRPD of unground NaClO3/NaBrO3 physical mixture (blue); Table S1: Refined average parameter versus composition of the solid solution obtained by Spray-Drying with protocol 1. Measurements at RT.  Data Availability Statement: On request, raw data are available from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest and agree on the last version of the manuscript.