Experimental Study of Hydration/Dehydration Behaviors of Metal Sulfates M2(SO4)3 (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In) in Search of New Low-Temperature Thermochemical Heat Storage Materials.

To identify potential low-temperature thermochemical heat storage (TCHS) materials, hydration/dehydration reactions of M2(SO4)3 (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In) are investigated by thermogravimetry (TG). These materials have the same rhombohedral crystal structure, and one of them, rhombohedral Y2(SO4)3, has been recently proposed as a promising material. All M2(SO4)3·xH2O hydrate/dehydrate reversibly between 30 and 200 °C at a relatively low p H2O (=0.02 atm). Among them, rare-earth (RE) sulfates RE2(SO4)3·xH2O (RE = Sc, Yb, Y, Dy) show narrower thermal hystereses (less than 50 °C), indicating that they have faster reaction rates than the other sulfates M2(SO4)3·xH2O (M = Al, Ga, Fe, In). As for the heat storage density, Y2(SO4)3·xH2O is most promising due to the largest mass change (>10 mass % anhydrous basis) during the reactions. This is larger than that of the existing candidate CaSO4·0.5H2O (6.6 mass % anhydrous basis). Regarding the reaction temperature of the water insertion into rhombohedral RE2(SO4)3 (RE = Yb, Y, Dy) to form RE2(SO4)3·H2O, it increases as the ionic radius of RE3+ becomes larger. Since such a relationship is also observed in β-RE2(SO4)3·xH2O, RE(OH)3, and REPO4·xH2O, this empirical knowledge should be useful to expect the dehydration/hydration reaction temperatures of the RE compounds.

Introduction

Various industries generate low-temperature waste heat of around 250 °C or below. However, thermal utilization technologies in this temperature range are rudimentary. A promising technology utilizing low-temperature waste heat is thermochemical heat storage (TCHS) among other thermal energy storage technologies (sensible heat storage and latent heat storage). Compared with sensible and latent heat storage, TCHS is advantageous with regard to heat storage capacity and long-term storage ability.[1−3] However, although several applications such as mobilized thermal energy storage are proposed,[4,5] TCHS is still in the research stage and no practical material that meets all required criteria (i.e., heat storage capacity, reaction temperature, reaction rate, reaction reversibility, etc.) has been realized.

The hydration/dehydration reactions of some sulfates such as CaSO4, β-La2(SO4)3, and Y2(SO4)3 have advantages in reaction temperature and reaction rate. They proceed rapidly and reversibly below 250 °C and even in low water vapor pressure pH such as that in air.[6−9] These characteristics can be attributed to their crystal structures where the insertion/deinsertion of water molecules occurs.[7,9,10] Furthermore, regarding Y2(SO4)3, a larger heat storage capacity is expected through the characteristic multistep hydration/dehydration reaction.[9]

Elemental substitution is a common method effective to modify the reaction behaviors of chemical compounds. For example, regarding β-type rare-earth sulfates β-RE2(SO4)3 (rare-earth (RE) = La, Ce, Pr, Nd, Sm, Eu), the reaction temperatures and kinetics depend on the rare-earth elements.[7,11] β-RE2(SO4)3 has a crystal structure belonging to a monoclinic system in the no. 15 space group.[7] See the Supporting Information SI-1 for the crystal structures and dehydration/hydration behaviors of β-RE2(SO4)3.

Therefore, there is a possibility to improve the reaction behaviors of Y2(SO4)3 by substituting other cations for Y. In this study, we explore new materials for TCHS in rhombohedral trivalent-metal sulfates M2(SO4)3 (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In) with ionic radii of M varying from 0.535 to 0.912 Å. These seven rhombohedral sulfates except for Yb2(SO4)3 have the same crystal structure belonging to the space group no. 146.[12,13] Although the crystal structure of Yb2(SO4)3 belongs to the space group no. 161, it is similar to that of the other sulfates.[14] The Supporting Information SI-2 shows the crystal structures of rhombohedral sulfates (space group no. 148 or 161) and ionic radii of M3+.

Results

Dehydrated Phases of M2(SO4)3·nH2O (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In)

Figure shows the high-temperature X-ray diffraction (XRD) patterns of the dehydrated phases of the starting materials M2(SO4)3·nH2O (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In) at 300 °C (see Section ). The dehydrated phases of five sulfates with M = Sc, Yb, Al, Ga, Fe are single-phase rhombohedral M2(SO4)3, whereas those of the other three sulfates with M = Y, Dy, In contain secondary phases. Note that the dehydration reaction conditions to obtain a single phase of rhombohedral M2(SO4)3 (M = Y, Dy, In) have yet to be identified, although the dehydration reactions of M2(SO4)3·nH2O were performed under several temperatures and pH. Nonetheless, these results confirm that rhombohedral M2(SO4)3 (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In), which are materials to be investigated, are obtained by heating M2(SO4)3·nH2O to 300 °C.

Figure 1

High-temperature XRD patterns of the dehydrated phases at 300 °C in humidified oxygen (pH = 0.03 atm) formed from M2(SO4)3·nH2O ((a) Sc2(SO4)3·5H2O, (b) Yb2(SO4)3·8H2O, (c) Y2(SO4)3·8H2O, (d) Dy2(SO4)3·8H2O, (e) Al2(SO4)3·nH2O, (f) Ga2(SO4)3·nH2O, (g) Fe2(SO4)3·nH2O, (h) In2(SO4)3·nH2O). Reference patterns (PDF cards) are cited from the ICDD database. Formation of monoclinic In2(SO4)3 (PDF#00-042-0227) and orthorhombic Y2(SO4)3 (PDF#04-009-9561) are confirmed as the secondary phases in the XRD patterns. Dehydrated Dy2(SO4)3 is not a single phase of the rhombohedral Dy2(SO4)3. However, the secondary phase cannot be identified. Peaks of alumina in (b) are attributed to the sample holder in the XRD apparatus.

Cyclic Dehydration/Hydration Behavior of Various M2(SO4)3·xH2O (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In)

Figure shows cyclic thermogravimetry (TG) curves of M2(SO4)3·xH2O on the second heating–cooling cycle between 30 and 200 °C. In all TG curves, the sample mass decreases on heating, and the mass increases on subsequent cooling, suggesting that the dehydration/hydration reaction of M2(SO4)3·xH2O proceeds reversibly below 200 °C.

Figure 2

TG curves of M2(SO4)3·xH2O (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In) undergoing dehydration/hydration reactions. Initial samples are M2(SO4)3·nH2O. TG curves are collected on the second heating–cooling cycle (see Section ). Left graph (A) and right graph (B) show the TG curves of four rare-earth sulfates and four nonrare-earth metal sulfates, respectively. Vertical axis shows the hydration number x estimated from the mass changes of the samples. (Note: TG curve of Y2(SO4)3 was already reported.[9]) Based on high-temperature XRD (Figure ), the TG curves of M2(SO4)3·xH2O (M = Y, Dy, In) may not simply represent the hydration/dehydration behaviors of rhombohedral M2(SO4)3. Nevertheless, since Y2(SO4)3 contains a relatively small amount of orthorhombic Y2(SO4)3 as a secondary phase (Figure (c)), the TG curve of Y2(SO4)3·xH2O mainly represents the hydration/dehydration reaction behavior of the rhombohedral phase rather than the orthorhombic phase.

Figure shows the XRD patterns of M2(SO4)3·xH2O after the cyclic TG measurements. They indicate that the hydrated phases have similar crystal structure to that of anhydrous rhombohedral M2(SO4)3. This suggests that the rhombohedral M2(SO4)3 (M = Sc, Y, Yb, Dy, Al, Ga, Fe, In) is hydrated by H2O insertion. We have already inferred that the hydration of the rhombohedral Y2(SO4)3 proceeds H2O insertion into the crystal lattice and formation of the rhombohedral Y2(SO4)3·xH2O (x > ∼1) while almost maintaining the host structure.[9] Similarly, H2O insertion has also been confirmed in the hydration reaction of other compounds such as β-RE2(SO4)3 (RE = La, Ce, Pr, Nd, Sm, Eu), CaSO4, rabdphene-type REPO4 (RE = Y, La, Ce, Pr, Nd, Sm, Gd, Tb, Dy, Er), and BaH2(C2O4)2.[7,10,11,15,16] Such a mechanism can contribute to the high reversibility of the hydration/dehydration reaction, as mentioned in the Introduction section.

Figure 3

XRD patterns of M2(SO4)3·xH2O (M = (a) Sc, (b) Yb, (c) Y, (d) Dy, (e) Al, (f) Ga, (g) Fe, (h) In) at room temperature after TG measurements shown in Figure . Unidentified peaks in (a) may indicate that another scandium sulfate hydrate Sc2(SO4)3·nH2O is partially formed after the TG measurements because we stored Sc2(SO4)3·xH2O in TG under humidified argon (pH = 0.02 atm) around 30 °C for several tens of minutes before the XRD measurement.

As for the reaction rate, four rare-earth (RE) sulfates show relatively narrow thermal hystereses (less than 50 °C): Sc2(SO4)3·xH2O, Yb2(SO4)3·xH2O, Y2(SO4)3·xH2O, and Dy2(SO4)3·xH2O. The narrow hystereses correspond to the small driving force required to advance the reactions. This suggests that RE2(SO4)3 (RE = Sc, Yb, Y, Dy) have superior dehydration/hydration reaction rates compared with those of other M2(SO4)3·xH2O (M = Al, Ga, Fe, In).

In addition, as for the reversible hydration number change, Y2(SO4)3·xH2O or Yb2(SO4)3·xH2O is the most promising candidate among the four rare-earth sulfates. Note that Sc2(SO4)3·xH2O shows hydration number change that is significantly less than that of other RE2(SO4)3·xH2O and is unsuitable for TCHS. Table compares the maximum reversible mass changes and hydration number changes of M2(SO4)3·xH2O through the dehydration/hydration reactions as evaluated by the TG measurements. In terms of gravimetric heat storage capacity, Y2(SO4)3 shows a larger mass change than Yb2(SO4)3.

Table 1

Maximum Mass Changes (Anhydrous Basis) and Hydration Numbers Derived from the TG Results (Figure ) for the Dehydration/Hydration Reactions of M2(SO4)3·xH2O at 30–300 °C in Humidified Argon (pH = 0.02 atm)

sulfates	Shannon’s ionic radius[17]a (Å)	maximum mass change (mass %)	maximum hydration number, x
Al2(SO4)3	0.535	3.40	0.65
Ga2(SO4)3	0.620	5.62	1.3
Fe2(SO4)3	0.645b	4.73	1.0
Sc2(SO4)3	0.745	1.42	0.30
In2(SO4)3	0.800	10.1	2.9
Yb2(SO4)3	0.868	7.83	2.8
Y2(SO4)3	0.900	10.5	2.7
Dy2(SO4)3	0.912	5.33	1.9

Shannon’s ionic radius of M3+ with six coordination number.

The ionic radius of high-spin Fe3+.

Evaluation of the Amount of Heat Absorbed During the Dehydration Reaction of Y2(SO4)3·xH2O

Regarding Y2(SO4)3·xH2O, the amount of endothermic heat during the dehydration reaction was evaluated by differential scanning calorimetry (DSC). Figure shows the DSC curves in the dehydration reaction of Y2(SO4)3·xH2O (x = ∼2) compared with those of the typical TCHS materials: CaSO4·0.5H2O and MgSO4·6H2O. Based on the areas of endothermic peaks, the evaluated amount of heat absorbed by Y2(SO4)3·xH2O (x = ∼2), CaSO4·0.5H2O, and MgSO4·6H2O are 299, 183, and 1366 kJ (kg-hydrate)−1 respectively. They deviate by 8–21% from the literature values of the standard enthalpy changes of dehydration reactions, i.e., 231 kJ (kg-hydrate)−1 for CaSO4·0.5H2O[18] and 1491 kJ (kg-hydrate)−1 for MgSO4·6H2O[19] when it dehydrates to MgSO4·0.1H2O.[20] Nevertheless, it is reasonable to compare the amount of endothermic heat using those measured by the same DSC apparatus and methods. Thus, the gravimetric heat storage density of Y2(SO4)3·xH2O (x = ∼2) is expected to be about a quarter that of MgSO4·6H2O and about 1.6 times higher than that of CaSO4·0.5H2O.

Figure 4

DSC curves of Y2(SO4)3·xH2O (x = ∼2), CaSO4·0.5H2O, and MgSO4·6H2O undergoing dehydration reaction. Since Y2(SO4)3·xH2O is obtained by preheating–cooling Y2(SO4)3·8H2O through the same process as the first heating–cooling cycle in TG measurements (Section ), the hydration number x should be approximately 2, as shown in the TG curves (Figure ). The amount of endothermic heat is attributed to the area of the endothermic peak between 40 and 160 °C. Here, the mass of Y2(SO4)3·2H2O is used for standardizing the amount of absorbed heat. The amount of endothermic heat is attributed to the area of the DSC peak between 80 and 150 °C for CaSO4·0.5H2O and between 55 and 265 °C for MgSO4·6H2O.

The DSC curve of Y2(SO4)3·xH2O shows five endothermic peaks (a–e) during the dehydration reaction. High-temperature XRD measurements (Figure S6) suggest that peaks b (36 °C) and e (117 °C) are attributed to sudden changes in the lattice parameters and the symmetry of the host structure. However, the causes of DSC peaks a, c, and d are not identified since there are no changes in the XRD patterns at the corresponding temperatures. See the Supporting Information SI-4 for the high-temperature XRD patterns and detailed explanations.

Discussion

Evaluation of Y2(SO4)3·xH2O as a TCHS Material

This section discusses the advantages and disadvantages of Y2(SO4)3·xH2O as a TCHS material compared with those of other candidate materials. Table shows heat storage density, mass change, and potential drawbacks on the handling of Y2(SO4)3·xH2O and other TCHS materials available for low-temperature heat below 200 °C. Although Y2(SO4)3·xH2O shows a much smaller heat storage density than other materials except for CaSO4·0.5H2O and β-La2(SO4)3·H2O, it is advantageous in the ease of handling without problems related to deliquescence, melting, phase transformation, slow hydration kinetics, etc., as shown in Table . With respect to price, calcium and magnesium compounds should be much cheaper than Y2(SO4)3. However, yttrium oxide, a raw material of Y2(SO4)3, is not very expensive (3 USD kg–1 in 2018[31]) due to oversupply, and new applications for Y are in demand in rare-earth industry.[32]

Table 2

Properties of Y2(SO4)3·xH2O and Other TCHS Materials

reaction couple or material	heat storage density (kJ (kg-hydrate)−1)	mass change (dehydrated or desorbed state basis) (mass %)	potential drawbacks on the handling
Y2(SO4)3·xH2O/Y2(SO4)3 (this work)	299 (x = 2)	10.5	no special drawbacks on the handling have been observed
LiCl (solution)[21]	4387[21]	226[21]	crystallization of salt solution causes deterioration of heat storage density[21]
LiBr (solution)[21]	2019[21]	170[21]
LiOH·H2O/LiOH[22]	1440[22]	75.2	pure LiOH exhibits low-hydration reaction rate which requires developing composite materials[23,24]
Na2S·5H2O/Na2S·0.5H2O[25]	1784[25]	93.1	generation of toxic byproduct H2S gas by contacting with water[26]
MgSO4·6H2O/MgSO4·0.1H2O[20]	1491[19]	87.0	skin formation covering the bulk particles by contacting with humid air, which decreases the reaction rate[27]
MgCl2·6H2O/MgCl2·2H2O[28]	1239[29]	54.9	agglomeration due to partial melting of the salt during the dehydration or a partial deliquescence attributed to the hygroscopicity slower the reactions[30]
CaCl2·6H2O/CaCl2[28]	1649[29]	97.3
CaSO4·0.5H2O/CaSO4[6]	240[18]	6.62	phase transformation from CaSO4 (III) to low-hydration-reactive CaSO4 (II) during repetitive reaction[6]
β-La2(SO4)3·H2O/β-La2(SO4)3[7]	156[7]	3.18	no special drawbacks on the handling have been observed

Effect of Ionic Radius of M3+ on the Dehydration/Hydration Reaction

RE2(SO4)3·xH2O with relatively large ionic radii of RE3+ (RE = Yb, Y, Dy) show similar TG curves, which represent the dehydration/hydration reactions proceeding in mainly two steps at high- and low-temperature ranges. Dy2(SO4)3·xH2O actually exhibits three reaction steps, but this may be due to another coexisting anhydrate phase with the rhombohedral one as shown in high-temperature XRD (Figure ), which may split the hydration reaction at between 130 and 90 °C into two steps.

As we discussed for rhombohedral Y2(SO4)3,[9] the apparent hydration number x varied from 0 to approximately 1 by water insertion reaction at the high-temperature range (above 70 °C). This high-temperature reaction mechanism is considered to be common to rhombohedral Yb2(SO4)3 and Dy2(SO4)3, and monohydrate RE2(SO4)3·H2O (RE = Yb, Y, Dy) might be the hydrated product of the rhombohedral RE2(SO4)3 in the high-temperature reaction, which is given as followsIn the low-temperature reaction, RE2(SO4)3·H2O hydrate additionally.

Regarding the high-temperature reaction, the reaction temperatures increase in the following order: Yb2(SO4)3, Y2(SO4)3, and Dy2(SO4)3. Hikichi et al.[15] pointed out a similar tendency in the dehydration temperature of the rhabdophane-type REPO4·xH2O. Figure shows the relationship between RE ionic radius of RE2(SO4)3·H2O (RE = Yb, Y, Dy), rhabdophane-type REPO4·xH2O (RE = La, Ce, Pr, Nd, Sm, Gd, Tb, Dy, Y, Er),[15] β-RE2(SO4)3·H2O (RE = La, Ce, Pr, Nd, Sm, Eu),[7] and RE(OH)3 (RE = La, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Yb, Tm).[33] They all show the same empirical relationship, where the dehydration temperature increases as the RE ionic radius becomes larger. Although the mechanisms that induce such relationships are unknown, this should be useful to control the reaction temperatures of the dehydration/hydration of RE compounds. Regarding the dehydration/hydration reactions of the other five sulfates M2(SO4)3·xH2O (M = Sc, Al, Ga, Fe, In), it is difficult to determine and compare the dehydration/hydration reaction temperatures thermodynamically due to large thermal hystereses or wide reaction temperature ranges.

Figure 5

Relationship between the dehydration temperature and RE ionic radius[17] of (A) RE2(SO4)3·H2O (this work), (B) RE(OH)3,[33] (C) β-RE2(SO4)3·xH2O,[7] and (D) rhabdophane-type REPO4·xH2O.[15] These dehydration reactions are as follows: (A) dehydration of RE2(SO4)3·H2O to rhombohedral RE2(SO4)3, (B) dehydration of RE(OH)3 to RE2O3, (C) dehydration of β-RE2(SO4)3·xH2O to β-RE2(SO4)3, and (D) dehydration of rhabdophane-type REPO4·xH2O to rhabdophane-type REPO4. Vertical axes in (A), (B), and (C) represent the assumed equilibrium temperature. Equilibrium temperatures of rhombohedral RE2(SO4)3·xH2O and β-RE2(SO4)3·xH2O are estimated from TG measurements. See the Supporting Information SI-5 for the detailed method for the estimation. Equilibrium temperatures of RE(OH)3 are calculated using the thermodynamic data summarized by Bernal et al.[33] assuming that RE(OH)3 and RE2O3 are virtually in equilibrium. Vertical axes in (D) represent the temperatures at which the dehydration of rhabdophane-type REPO4·xH2O occurs with the highest reaction rate based on differential thermal analysis (DTA).

Conclusions

We examine the hydration/dehydration reaction behaviors of M2(SO4)3·xH2O (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In) to find new candidates for low-temperature TCHS materials. TG measurements suggest that M2(SO4)3·xH2O hydrate/dehydrate reversibly between 30 and 200 °C at a relatively low pH (=0.02 atm). As for the reaction rate, RE2(SO4)3·xH2O (RE = Sc, Yb, Y, Dy) show relatively narrow thermal hystereses (less than 50 °C) between hydration and dehydration reactions. This suggests that they have superior reaction rates compared with those of nonrare-earth sulfates M2(SO4)3·xH2O (M = Al, Ga, Fe, In). Among them, Y2(SO4)3·xH2O shows the largest mass change during hydration/dehydration reaction (larger than 10 mass % anhydrous basis), which may lead to a high gravimetric heat storage capacity. The mass change of Y2(SO4)3 is larger than that of CaSO4 (6.6 mass % anhydrous basis). Therefore, among the eight sulfates, Y2(SO4)3·xH2O is the most promising in terms of both reaction rate and heat storage capacity. DSC measurements evaluated the amount of absorbed heat during the dehydration of Y2(SO4)3·xH2O (x = ∼2) to be approximately 299 kJ kg–1. While the heat storage density is much smaller than that of other candidate materials such as MgSO4·6H2O (1491 kJ kg–1), CaCl2·6H2O (1649 kJ kg–1), etc., Y2(SO4)3·xH2O has the advantage of the ease of handling.

RE2(SO4)3·xH2O (RE = Yb, Y, Dy) exhibit similar reaction behaviors, which consist of two reaction steps of the “high-temperature reaction” and the “low-temperature reaction”. In the high-temperature reaction (above 70 °C), the apparent hydration number x varied from 0 to approximately 1 with small thermal hysteresis of less than 30 °C.

The reaction temperature of the high-temperature reaction increases as the ionic radius of RE3+ becomes larger. Such a relationship between the reaction temperature and RE3+ ionic radius is also observed for the dehydration reactions of β-RE2(SO4)3·xH2O, RE(OH)3, and REPO4·xH2O. This empirical knowledge should be useful to estimate the reaction temperatures of the dehydration/hydration of RE compounds.

Experimental Section

Sample Preparation

Eight rhombohedral M2(SO4)3 (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In) samples were prepared by the dehydration of their hydrates M2(SO4)3·nH2O. Four of the sulfate hydrates were obtained by purchasing commercial reagents: Y2(SO4)3·8H2O (Strem Chemical, 99.9%), Al2(SO4)3·nH2O (Kanto Kagaku, 51.0–57.5% anhydrous basis), Fe2(SO4)3·nH2O (Nacalai Tesque, 76.1 mass % anhydrous basis), and In2(SO4)3·nH2O (Wako Pure Chemical Industries, 75.0–85.0% anhydrous basis). The other sulfate hydrates, Ga2(SO4)3·nH2O, Sc2(SO4)3·5H2O, Dy2(SO4)3·8H2O, and Yb2(SO4)3·8H2O, were produced by liquid-phase synthesis. Five millimol of a metal oxide M2O3 (M = Ga, Sc, Dy, Yb) was dissolved in 100 mL of 1 mol L–1 sulfuric acid aqueous solution. Then the solution was concentrated at 80 °C to precipitate a powder of a sulfate hydrate. Figure shows the XRD patterns of the precipitated products. The phases of three hydrated products with M = Sc, Dy, Yb are identified to be single-phase Sc2(SO4)3·5H2O, Dy2(SO4)3·8H2O, and Yb2(SO4)3·8H2O, respectively. Although the phases of the synthetic product of Ga2O3 and H2SO4 are unknown from the XRD pattern, we assume that they are some gallium sulfate hydrates Ga2(SO4)3·nH2O.

Figure 6

XRD patterns of the products of liquid-phase synthesis between M2O3 and H2SO4 (M = (a) Ga, (b) Sc, (c) Dy, (d) Yb).

Thermogravimetry (TG)

To investigate the dehydration/hydration reaction behavior of M2(SO4)3 (M = Sc, Yb, Y, Dy, Al, Ga, Fe, In), TG was conducted using a Rigaku Thermo plus TG 8120. A sample of M2(SO4)3·nH2O was heated to about 400 °C with a heating rate of 20 °C min–1, which is the temperature at which dehydration to an anhydrate should be completed and subsequently cooled to about 30 °C in the TG apparatus (“first heating–cooling cycle”) under humidified argon gas (pH = 0.02 atm) flow. In the first cooling process, the cooling rates were 20 °C min–1 until the temperature reached 130 °C and 1 °C min–1 from 130 to 30 °C. At this point, M2(SO4)3·xH2O was obtained as a rehydrated phase of anhydrate M2(SO4)3. On the subsequent “second heating–cooling cycle” between 30 and 200 °C, the dehydration/hydration reaction behaviors of M2(SO4)3·xH2O were investigated.

In all of the TG measurements, the sample material was put in a platinum cylindrical pan with a diameter of 5 mm and a height of 5 mm, and the same empty pan was used for the reference.

X-ray Diffraction

X-ray diffraction (XRD) measurements were conducted on a PANalytical X’Pert-Pro MPD using Cu Kα radiation. High-temperature data collection was achieved using an Anton Paar HTK1200N high-temperature oven chamber under humidified oxygen.

Differential Scanning Calorimetry (DSC)

To evaluate the amount of absorbed heat during the dehydration reaction of Y2(SO4)3·xH2O, CaSO4·0.5H2O, and MgSO4·6H2O, differential scanning calorimetry (DSC) measurements were conducted using a Rigaku Thermo plus EVO2 DSC8231. Y2(SO4)3·8H2O (Strem Chemical, 99.9%), CaSO4·2H2O (Wako Pure Chemical Industries, ≥98.0%), and MgSO4·6H2O (Nacalai Tesque, ≥99.5%) were used as starting materials. They were put in an aluminum cylindrical pan with a diameter of 5 mm and a height of 2 mm and covered with a stainless mesh sheet lid. Then, the pan and the lid were compressed in a specialized sample crimper. The crimped blank pan with the lid was used for reference. Y2(SO4)3·xH2O, CaSO4·0.5H2O, and MgSO4·6H2O were synthesized by preheating–cooling of their starting materials using the DSC apparatus in humidified argon (pH = 0.02 atm). Y2(SO4)3·xH2O is obtained by preheating–cooling 5.0 mg of Y2(SO4)3·8H2O to 400 °C with a heating rate of 20 °C min–1, the same process as the first heating–cooling cycle in TG measurements (see Section ). CaSO4·0.5H2O is obtained by preheating–cooling 3.3 mg of CaSO4·2H2O between 30 and 200 °C with a heating rate of 20 °C min–1 and a cooling rate of 1 °C min–1. MgSO4·6H2O is obtained by preheating 3.0 mg of MgSO4·7H2O between 30 and 45 °C with a heating rate of 0.5 °C min–1.