The Solid Solution between NaClO₃ and NaBrO₃ Revisited

Abstract

NaClO₃ and NaBrO₃ are believed to form a complete solid solution from RT to fusion. The unique solid phase can thus be written: NaClO_3(1−x)-NaBrO_3(x) with: 0 ≤ x ≤ 1. This study shows that at high temperatures, this statement might be valid. Nevertheless, up to 50 °C, probably up to 160 °C, and even higher temperatures, this is not true when the system is in thermodynamic equilibrium. A large miscibility gap exists at room temperature (RT). This gap could be reduced up to a complete disappearance by fast crystallization, for instance, spray-drying. The necessary conditions to access equilibrium, including homochirality, are also discussed.

1. Introduction

1.1. 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].

Among inorganic solid solutions described in the literature, the NaClO₃-NaBrO₃ binary system (or NaClO₃-NaBrO₃-H₂O ternary system) has been widely investigated, especially as a relevant model for the study of deracemization [4,5,6,7,8,9,10,11,12,13,14,15], as well as an interesting case for the study crystal habit modification in the presence of impurities [16,17,18,19,20,21,22,23]. This may be partly due to the high crystallogenic character of the two components and their mixtures. Moreover, NaClO₃, NaBrO₃ and NaClO_3(x)BrO_3(1−x) intermediate compositions have low melting points [24,25] and high (ClO₃-rich) to medium (BrO₃-rich) solubilities in water [20,26]. NaClO₃ has a melting point of 262 °C and a solubility in water of 790 g.L⁻¹. NaBrO₃ has a melting point of 347 °C and a solubility in water of 275 g.L⁻¹.

1.2. Crystal Structure of NaClO₃ and NaBrO₃ Pure Compounds and Chirality in the Solid State

Sodium chlorate (NaClO₃, MW = 106.44 g.mol⁻¹, T_fus = 262 °C) and sodium bromate (NaBrO₃, MW = 150.89 g.mol⁻¹, T_fus = 347 °C) have isomorphous crystal structures that were formerly investigated by Dickinson in 1921 [27]. Both halates crystallize in the chiral space group P2₁3 (tetartohedral group in the cubic system), with crystallographic parameters a = 6.57584(5) and a = 6.70717(10) Angströms, respectively, at RT [28]. The space group P2₁3 is chiral thus, noncentrosymmetric in adequation with the property to exhibit second-order nonlinear optical phenomena [29,30] (Savage, 1962; Simon, 1968).

NaClO₃ and NaBrO₃ belong to achiral ionic compounds that exhibit supramolecular chirality in the solid state [31]. It was also found that optically dextro (D) rotatory NaClO₃ has the opposite configuration than the optically dextro (D) rotatory NaBrO₃, or, in other words, (NaClO₃)_D has the same structural configuration as (NaBrO₃)_L [32,33,34]. Thus, the (NaClO₃)_L/(NaClO₃)_D/(NaBrO₃)_L/(NaBrO₃)_D system is quaternary when the solid phases are considered.

1.3. NaClO_3(x)BrO_3(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 NaClO_3(x)BrO_3(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 NaClO_3(x)BrO_3(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 NaClO_3(x)BrO_3(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.

1.4. 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.

1.5. 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 NaClO_3(x)BrO_3(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 NaClO_3(x)BrO_3(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.

2. Material and Methods

2.1. Commercial Material

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

2.2. 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.

2.3. 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. Spray-drying parameters for two sets of experiments.

Parameters	Protocol 1	Protocol 2
Concentration of the NaClO3(1−x)-NaBrO3(x) aqueous solutions	2 g/150 mL water	6 g/50 mL water
The inlet temperature of the drying nitrogen	220 °C	220 °C
Feed flow rate (peristaltic pump)	360 mL/h	300 mL/h
The temperature of the feed solution	Room temperature	Ca. 60 °C
Atomizing airflow rate	473 L/h	500 L/h
Aspiration	40 m3/h	40 m3/h
Location of the powder at the end of the process	Mainly in the bowl collector	In the drying chamber, the cyclone and bowl collector

).

2.4. XRPD

XRPD patterns were obtained either with a Da Vinci D8 Bruker X-ray diffractometer with a Bragg-Brentano geometry (θ//θ) or a D8 diffractometer (Bruker, Germany) equipped with a modified goniometer of reverse-geometry (−θ/−θ) [39] (Coquerel, 2015). For both instruments, the incident X-ray beam consisted of the Cu Kα (λ = 1.5418 Å) with a tube voltage and amperage set at 40 kV and 40 mA, respectively. The diffraction patterns were collected with a LynxEye^® linear detector (Bruker, Germany). Routine XRPD analyses were performed with a step of 0.04° (2θ) and a 12 s per step counting time from 3° to 30° (2θ). The system is monitored with DiffracPlus XRD commander software version 2.6.1. Data were processed with DIFFRAC.EVA software release 2018 (version 4.3.0.1).

2.5. 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).

2.6. 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.

2.7. 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).

2.8. 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.

2.9. 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).

3. Results and Discussion

3.1. 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.

Large crystals were obtained by water evaporation at 20 °C and submitted to a polarized microscopic inspection (Polarized Optical Microscopy: POM hereafter) (Figure 4) and SEM-EDX analysis (Figure 5).

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,b.

3.2. 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 NaClO_3(x)BrO_3(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).

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.

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.

4. 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?