Sustainable Breakthrough in Manganese Oxide Thermochemical Energy Storage: Advancing Efficient Solar Utilization and Clean Energy Development

Abstract

Solar power generation systems, recognized for their high energy quality and environmental benefits, require efficient energy storage to ensure stable grid integration and reduce reliance on fossil fuels. Thermochemical energy storage (TCS) using metal oxides, such as the Mn₂O₃/Mn₃O₄ redox system, offers advantages like high energy density, wide temperature range, and stability, making it ideal for solar power applications. This study investigates Mn₃O₄ and Mn₂O₃ as initial reactants, analyzing reaction temperature range, rate, conversion efficiency, and cyclic performance via synchronous thermal analysis. Microstructural characterization was performed using XRD, SEM, BET, XPS, nanoparticle size, and zeta potential measurements. The results show that Mn₃O₄ reversibly converts to Mn₂O₃ with over 100% conversion efficiency over five cycles with 3.3% weight loss, indicating stable performance. Mn₃O₄ oxidation follows Arrhenius’ Law below 700 °C but deviates at higher temperatures. The oxidation mechanism function is G(α) = α and f(α) = 1, with an activation energy of 20.47 kJ/mol and a pre-exponential factor of 0.268/s. Mn₂O₃ synthesized via ammonia precipitation exhibits reversible redox behavior with 3.3% weight loss but samples from low-concentration precursors show poor cyclic performance. The reduction reaction of Mn₂O₃ has an activation energy of 249.87 kJ/mol. By investigating the Mn₂O₃/Mn₃O₄ redox system for TCS, this study advances its practical integration into solar thermal power systems and offers critical guidance for developing scalable, low-carbon energy storage technologies. These findings can support Sustainable Development Goals (SDGs) by advancing renewable energy storage technologies, reducing carbon emissions, and promoting the integration of solar power into sustainable energy grids.

1. Introduction

To deal with global warming and high carbon emissions, we need to increase the proportion of clean and renewable energy in primary energy consumption. Among many renewable energy sources, solar energy has attracted the attention of many scholars because of its wide distribution, inexhaustible supply, and great development potential [1,2,3,4]. Due to the intermittent instability of the solar energy supply, solar power generation systems cannot be fully integrated into the grid, which has restricted the rapid development of solar power generation technology for a long time [5,6,7].

It was found that the combination of a thermal energy storage system and solar power generation system could effectively solve the problem of unstable power generation [8,9,10]. The thermal energy storage system mainly stores the temporarily unused or redundant solar energy in the medium through a certain heat storage method, then releases and uses it when the demand increases or solar energy is supplied intermittently [11,12,13].

Thermochemical energy storage (TCS) stores heat in the chemical bond of a substance through certain chemical reactions [14,15,16]. This kind of method can keep heat stable for a long time and adapt to a wide range of temperatures [17,18]. The thermochemical energy storage technology based on metal oxides mainly uses its reversible oxidation-reduction reaction to store and release heat [19,20]. The higher reaction (500~1200 °C) meets the requirements of the solar power generation system [21]. In addition, the redox reaction has the advantages of higher energy storage density, a simpler device, and low heat loss [22,23,24].

Among many metal oxides, Co₃O₄/CoO, CuO/Cu₂O, Fe₂O₃/Fe₃O₄, BaO₂/BaO, Mn₂O₃/Mn₃O₄, and so on were found to have the potential for thermochemical energy storage by some scholars’ research [25]. Among them, a Mn₂O₃/Mn₃O₄ reaction system has attracted the interest of some scholars because of its abundant reserves, low price, non-toxicity, and the fact that it is harmless [26]. The redox reaction equation of the reaction system is as follows:

6Mn₂O₃ ↔ 4Mn₃O₄ + O₂(g)  ∆H = 202 kJ/kg

(1)

Gokon et al. [27] compared the heat release kinetics of the Fe-substituted Mn₂O₃/Mn₃O₄ material in high-temperature thermochemical energy storage between long-term thermal cycling and initial preparation, revealing changes in different reaction models and kinetic parameters and their effects on the energy storage performance of the next generation of concentrated solar power generation systems. Huang et al. [28] modified manganese-based oxide with silicon doping, which significantly improved the thermochemical energy storage performance of the material, solved the problem of insufficient reoxidation capacity of pure manganese trioxide, and provided theoretical support and application guidance for the development of low-cost environmentally friendly large-scale thermochemical energy storage materials. Fetisov et al. [29] studied the redox characteristics of a Mn₂O₃/Mn₃O₄ reaction system with a thermogravimetric analysis. It was found that Mn₃O₄ could not be oxidized to Mn₂O₃ at the cooling stage. However, the research on Mn series oxides by Müller et al. [30] and Block et al. [31] showed that Mn₂O₃/Mn₃O₄ could realize the reversible redox reaction cycle. According to the analysis results of Zhao et al. [32], the chemical activity of Mn₂O₃ is related to its physical properties, which may be one of the reasons for the above contradiction. In the follow-up study on the physical nature of Mn₂O₃, it was found that the parameters in the synthesis of Mn₂O₃ had a great influence on its physical nature, especially the concentration of the precursor solution. In addition, Carillo et al. [23] emphatically studied the effect of the initial particle size of Mn₂O₃/Mn₃O₄ on redox reactions. They found that the reaction rate and conversion temperature of the oxidation reaction decreased gradually as the particle size decreased. Meanwhile, the decrease in particle size also brought about a sintering phenomenon, which led to the end of the cycle. Liu et al. [33] investigated the redox and melting properties of mangan-based ores to test their potential applications in thermochemical energy storage. The results show that manganese ore with a high oxygen capacity and deformation temperature has the potential to be used as a TCS material.

This study innovatively investigates the Mn₂O₃/Mn₃O₄ redox system through integrated advanced characterization techniques and kinetic analysis to elucidate its thermochemical energy storage mechanisms. The redox reaction characteristics, microstructure properties, and phase transitions of Mn₃O₄ and synthetic Mn₂O₃ were analyzed using synchronous thermal analysis (STA), X-ray diffraction (XRD), and scanning electron microscopy (SEM). The key findings reveal that Mn₃O₄ undergoes an initial transformation to Mn₂O₃ prior to participating in reversible redox cycles, thereby clarifying the reaction pathway. Furthermore, the influence of a precursor solution concentration on Mn₂O₃’s physical properties and reactivity was systematically explored, emphasizing the critical role of synthesis conditions in material design. The oxidation-reduction kinetics of both Mn₃O₄ and Mn₂O₃ were rigorously analyzed, providing a theoretical foundation for optimizing reaction conditions to enhance energy conversion efficiency.

By integrating these approaches, this work advances the fundamental understanding of the Mn₂O₃/Mn₃O₄ system while offering practical guidance for its application in solar thermal energy storage systems, thereby facilitating the development of efficient and sustainable energy solutions.

2. Materials and Methods

2.1. Materials

High-purity Mn₃O₄ (99.97%) supplied by McLean (Shanghai, China) was used in the experiments. Mn₂O₃ was synthesized via the ammonia precipitation method using manganese nitrate (50% by mass, analytical reagent, AR) and ammonia solution (25–28% by mass, AR).

The sample preparation procedure was as follows: First, a series of precursor solutions with varying concentration gradients were prepared by mixing Mn(NO₃)₂ (50% by mass, AR) with distilled water. Subsequently, NH₃·H₂O (25–28% by mass, AR) was added dropwise to the Mn(NO₃)₂ solution. The mixture was allowed to stand for 20 h to facilitate solidification. After filtration and washing, the samples were dried overnight at 80 °C. The dried solid was ground into a fine powder. A brown coloration of the powder indicates that Mn₃O₄ is the dominant phase. The powder was then calcined in a muffle furnace at 700 °C for 4 h with a heating rate of 2 °C/min under static air to yield the final required sample. For clarity in subsequent descriptions, the samples were labeled Mn_x, where x represents the concentration of the manganese nitrate precursor solution in mol/L. For example, a sample prepared from a 0.025 mol/L manganese nitrate solution was denoted as Mn_0.025.

Samples were prepared using manganese nitrate solutions with initial concentrations of 0.025, 0.05, 0.1, 0.5, 1, 2, 3, 4, 5, 6, and 7.6 mol/L, and were named Mn_0.025, Mn_0.05, Mn_0.1, Mn_0.5, Mn₁, Mn₂, Mn₃, Mn₄, Mn₅, Mn₆, and Mn_7.6, respectively.

2.2. Thermal Analysis

The samples were analyzed using a synchronous thermal analysis (STA) instrument (STA 449F3, Netzsch, Germany). This STA system combines thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC), enabling simultaneous measurement of mass changes and heat flow during oxidation-reduction reactions. The experiments were performed in a reactive gas atmosphere containing 20% oxygen and 80% nitrogen. A precisely weighed sample was loaded into the STA instrument for thermal analysis, with the temperature ranging from ambient to 1000 °C at a constant heating and cooling rate of 2 °C/min. Furthermore, the cyclic performance of the sample was evaluated through a cycle experiment, which consisted of two parts: heating to 1000 °C at 500 °C and cooling back to 500 °C at a rate of 2 °C/min.

2.3. Kinetic Analysis

Kinetic analysis methods are primarily categorized into isothermal and non-isothermal approaches. The isothermal method involves heating the sample under constant temperature conditions to obtain the thermal analysis curve. The non-isothermal method can be further divided into single-heating rate and multi-heating rate methods, depending on whether the heating and cooling rates are fixed.

In the study of the thermal conversion performance of Mn₃O₄, the isothermal method was employed to analyze the reaction kinetics of the Mn₂O₃/Mn₃O₄ system. The oxidation reaction kinetics were investigated through isothermal experiments, where the sample was subjected to oxidation at different constant temperatures. During the heating stage, the sample was heated to 1000 °C at a rate of 2 °C/min and maintained in an air atmosphere for 1 h. In the cooling stage, the temperature was reduced to the desired isothermal temperature at a rate of 2 °C/min under a nitrogen atmosphere to prevent premature oxidation. The temperature was held at the target isotherm for 10 min, after which the atmosphere was switched to air to initiate the oxidation reaction. The isothermal stage lasted approximately 5.5 h until the reaction was completed, followed by cooling to room temperature.

For the thermal conversion study of Mn₂O₃, the non-isothermal multi-heating rate method was utilized to determine the reduction reaction kinetics. Using sample Mn_0.5 as an example, the thermal analysis data at 50% conversion (α) obtained from experiments at heating rates of 2 °C/min, 5 °C/min, and 10 °C/min in an air atmosphere served as the basis for kinetic analysis. The reaction kinetics were subsequently calculated and analyzed.

2.4. Material Characterization

To investigate the impact of reactant morphology on reaction properties, comprehensive characterization techniques were employed, including X-ray diffraction (XRD), field emission scanning electron microscopy (SEM), Brunauer–Emmett–Teller (BET) analysis, X-ray photoelectron spectroscopy (XPS), nanoparticle size analysis, and zeta potential measurements.

The diffraction pattern of the sample was obtained using XRD, and the crystal size of the sample was calculated according to the Debye–Scherrer formula. For qualitative analysis of metal oxides, the sample crystals were analyzed with an XRD meter (Xpert pro, Almelo, The Netherlands), employing characteristic Cu Kα radiation (40 Kv, 40 mA). SEM was used to observe the microstructure of the samples, which were made in Tokyo, Japan (Joel 7800f). Zetasizer ultra (ZSE) was used to measure the nanoparticle size distribution. The XPS was used to measure the surface elemental valence states of samples, calibrating all binding energies of Mn 2p with C1s at 284.8 eV as the standard. BET analysis enables the determination of specific surface area and pore structure parameters, which play a critical role in governing the gas–reactant interaction and, consequently, the reaction performance of the material. Nanoparticle size analysis, conducted through physical methods, is primarily employed to measure the particle size distribution of the material.

3. Results and Discussion

3.1. Thermal Conversion Performance Evaluation of Mn₃O₄

3.1.1. Thermochemical Reaction Assessment

The purchased Mn₃O₄ was directly loaded into the thermogravimetric analyzer for experimental investigation at a heating and cooling rate of 2 °C/min. As shown in Figure 1, the TG curve exhibits two distinct stages of weight loss during the heating process. In the first stage, when the temperature reaches approximately 550 °C, the TG curve shows a significant decrease to about 97% of the initial weight. In the second stage, when the temperature rises to around 950 °C, the TG curve further declines to approximately 94%. During the cooling process, as the temperature decreases to 700 °C, the TG curve recovers to 97%, stabilizing at a level similar to that observed at the end of the first weight loss stage. Based on the changes in the TG curve illustrated in Figure 1, redox reaction cycles are inferred to occur between the second weight loss stage and the subsequent weight recovery stage.

Concurrently, the DSC curve reveals two prominent endothermic peaks during the heating process, corresponding to the two stages of weight loss observed in the TG curve. The first endothermic peak at around 550 °C indicates the initial phase transformation or decomposition of Mn₃O₄, while the second endothermic peak near 950 °C suggests the onset of the redox reaction. During cooling, an exothermic peak appears at approximately 700 °C, aligning with the weight recovery in the TG curve.

For a clearer description of the reaction products at each phase, the initial sample and reaction products at each phase were numbered and analyzed. The initial sample was named No.1. The products obtained by two distinct decreases in the TG curve were named No.2 and No.3, respectively. The product after the TG curve rise was named No.4. In order to analyze the sample and products, XRD and XPS tests were conducted. The XPS diffraction results of the Mn 2p of the No.1 sample are shown in Figure 2a. The peak value of 641.3 eV was ascribed to Mn²⁺ and the value of 642.3 eV was put down to Mn³⁺. The XRD pattern of the No.1 sample shown in Figure 3a was in keeping with pure Mn₃O₄ (PDF#24-0734). Thus, it could be ensured that the sample purchased was Mn₃O₄. The XPS diffraction results of the Mn 2p of the No.2 product are shown in Figure 2b. The peak value of 641.7 eV was put down to Mn³⁺, showing that Mn was present as Mn³⁺. To further confirm this, an XRD test was carried out for the No.2 product. As shown in Figure 3b, the XRD pattern of the No.2 product was the same as that of pure Mn₂O₃ (PDF#78-0390). It could be determined that the No.2 product was Mn₂O₃. The XPS diffraction results of the Mn 2p of the No.3 product are shown in Figure 2c. The peak value of 641.1 eV was assigned to Mn²⁺, while the peak value of 642.0 eV was put down to Mn³⁺. It indicated that the No.3 product was Mn₃O₄, which was further validated by the XRD results shown in Figure 3c (Mn₃O₄, PDF#24-0734). The XPS diffraction results of the Mn 2p of the No.4 product are shown in Figure 2d, which is greatly similar to the spectra in Figure 2b. The peak value of 641.7 eV was ascribed to Mn³⁺, indicating that Mn existed in pure Mn³⁺. The XRD pattern of the No.4 product shown in Figure 3d was also consistent with pure Mn₂O₃ (PDF#78-0390). In summary, the thermodynamic analysis of Mn₃O₄ as the initial material showed that Mn₃O₄ first turned into Mn₂O₃, and then Mn₂O₃ was reduced to Mn₃O₄. Finally, Mn₃O₄ was oxidized to Mn₂O₃. The redox reaction equation is Equation (1). From the above, it can be seen that if Mn₃O₄ was used as the TCS material, it first reversibly changed to Mn₂O₃, and then Mn₂O₃/Mn₃O₄ actually acted as a circulating substance.

3.1.2. Cyclic Performance Assessment

Figure 4 shows the results of five cyclic experiments conducted on the Mn₂O₃/Mn₃O₄ system within a temperature range of 500–1000 °C, with heating and cooling rates of 2 °C/min. The results indicate that the redox reaction proceeds effectively, exhibiting a high conversion rate and a weight loss rate close to 3.3%. Figure 5 illustrates the variation in the reactant conversion with temperature over the five cycles. The reaction temperature range and reaction time remain relatively stable during the reduction reactions of cycles 2–5. However, during the oxidation reactions, the temperature range gradually shifts toward lower temperatures as the number of cycles increases. These results demonstrate that the redox reaction, with Mn₃O₄ as the initial reactant, exhibits excellent conversion efficiency and cycling performance, highlighting its potential for energy storage applications.

3.1.3. Reaction Kinetics Analysis

The experimental data obtained at different temperatures were processed to generate α–t curves at constant temperatures of 550, 600, 650, 700, 750, and 800 °C, as illustrated in Figure 6. As shown in Figure 6a, the reaction rate increases with the rising oxidation temperature of Mn₃O₄. Consequently, when T ≤ 700 °C, the oxidation reaction adheres to the mathematical description of Arrhenius’ Law. However, as depicted in Figure 6b, when the temperature exceeds 700 °C, the oxidation reaction rate begins to decline with further temperature increase. Thus, when T > 700 °C, the oxidation reaction no longer conforms to the Arrhenius’ Law. These results indicate that when the oxidation reaction temperature surpasses a certain threshold, the reaction rate decreases with increasing temperature, leading to a reduction in the conversion rate. Therefore, in practical applications, the reaction temperature should be carefully controlled to ensure a high conversion efficiency.

Furthermore, the kinetic mechanism function was determined. Based on the aforementioned conclusions, experimental data at isothermal temperatures below 700 °C were selected. Specifically, the experimental data obtained from the isothermal oxidation reaction at 600 °C were used to plot the t/t_0.5–α curve, with the experimental curve labeled “exp” in the diagram as shown in Figure 7, following the isothermal method for mechanism function determination. This curve was employed to identify the most probable mechanism function for the oxidation reaction. By comparing the experimental curve with standard mechanism functions, it was found that the experimental curve overlaps with mechanism function number 25. Therefore, the mechanism function for the oxidation reaction in this study is determined to be the phase boundary reaction (one-dimensional) R1, with n = 1. The corresponding mechanism function expressions are G(α) = α and f(α) = 1.

The experimental curve is labeled “exp” in the diagram.

It is known from the above that when the constant temperature is greater than 700 °C, the oxidation reaction no longer obeys Arrhenius’ law. Therefore, the linear curve shown in Figure 8 is obtained by using the experimental data of an oxidation reaction with a constant temperature of less than 700 °C and a conversion rate greater than 0.1. Using the isothermal method for calculating the activation energy and pre-exponential factor, the slope and intercept of the linear fit in Figure 8 were determined. The slope was calculated to be −2462.563, and the intercept was −1.315. Based on these values, the activation energy (E) for the oxidation reaction was found to be 20.47 kJ/mol, and the pre-exponential factor (A) was 0.268/s.

3.2. Thermal Conversion Performance Evaluation of Mn₂O₃

3.2.1. Thermochemical Reaction Assessment

Figure 9 shows the thermal analysis curves of the Mn₂O₃ samples. As illustrated in the figure, a reversible redox reaction can be achieved for the Mn₂O₃ powders prepared from precursor manganese nitrate solutions of different concentrations. However, due to the presence of a small amount of impurities, the samples exhibited a slight weight loss during the initial heating stage, which is attributed to impurity decomposition rather than the reduction reaction. The weight loss rate during the reduction reaction stage was approximately 3.3%. However, the weight gain rate during the oxidation reaction varied among samples prepared with different concentrations of manganese nitrate. For instance, the oxidation reaction weight gain rates for the Mn_0.025 and Mn_0.1 samples were only about 2%, indicating incomplete conversion. The remaining samples exhibited comparable weight gain rates during oxidation to their weight loss rates during reduction, achieving nearly complete conversion in the redox reactions.

In this study, the Mn_0.5 sample was selected as a representative example to analyze the reaction performance in detail. During heating, the TG curve showed a significant drop, indicating the occurrence of a reduction reaction, during which Mn₂O₃ transformed into Mn₃O₄. The reduction reaction began at 941 °C and ceased after approximately 47 min when the temperature reached 963 °C, with a weight loss rate of 3.4%, close to the ideal value. During the cooling stage, the TG curve exhibited a clear increase, indicating the oxidation of Mn₃O₄ back to Mn₂O₃. The oxidation reaction started at 803 °C and ended after about 53 min when the temperature dropped to 698 °C, with a weight gain rate of 3.3%, achieving nearly complete conversion. Throughout the entire reaction process, the reduction reaction began at a temperature 138 °C higher than that of the oxidation reaction, and its duration was 6 min shorter, suggesting that the reduction reaction proceeded faster than the oxidation reaction.

To further investigate the cycling stability of the samples, cyclic experiments were conducted on Mn_0.025, Mn_0.05, Mn_0.1, Mn_0.5, Mn₁, Mn₃, Mn₆, and Mn_7.6. The experiments were performed at a heating and cooling rate of 2 °C/min in a reaction atmosphere consisting of 80% nitrogen and 20% oxygen. As shown in Figure 10, the weight loss rate of the Mn_0.025 sample during the first reduction reaction was approximately 3.37%, while the weight gain rate was only about 1.64%, significantly lower than the ideal value of 3.37%. In the subsequent cycles, the oxidation capacity of the sample gradually decreased, resulting in a weight gain rate of only 0.3% by the fifth cycle.

Although the weight gain rate of the Mn_0.1 sample in the first cycle was lower than its weight loss rate, the sample maintained relatively stable weight loss and weight gain rates in subsequent cycles. This indicates that the Mn_0.1 sample exhibits good cycling stability and meets the requirements for energy storage systems. In addition, the Mn_0.1, Mn_0.5, Mn₁, Mn₃, and Mn₆ samples exhibited relatively stable cycling performance and good conversion efficiency over five cycles. However, after five cycles, their cycle stability and conversion efficiency deteriorated significantly. This decline was attributed to sintering-induced pore closure, which compromised their ability to maintain good cycling performance. In contrast, the Mn_0.05 and Mn_7.6 samples demonstrated consistently stable conversion performance throughout the cycling process. To further investigate their long-term durability in thermochemical energy storage applications, we extended their cycling to 10 cycles. After 10 cycles, their performance remained stable, suggesting that these samples likely retained favorable porosity and surface area characteristics.

3.2.2. Synthesis Parameters Assessment on Physical Properties

To confirm that all the prepared samples consist of Mn₂O₃ powder, an XRD analysis was performed on the synthesized samples. Figure 11 shows the XRD patterns of the homemade samples. From the results, it is evident that the Mn_0.5 and Mn_0.1 samples exhibit the highest characteristic peak intensities and the best crystallinity. In contrast, the Mn₁, Mn₂, Mn₃, and Mn₄ samples show lower peak intensities and reduced crystallinity. The Mn_0.025, Mn_0.05, Mn₅, Mn₆, and Mn_7.6 samples display the lowest peak intensities and the poorest crystallinity.

To further investigate the influence of the precursor solution concentration on physical properties such as particle size, specific surface area, and pore size, five samples (Mn_0.025, Mn_0.5, Mn₁, Mn₃, and Mn_7.6) were selected for detailed analysis. The calculated crystal particle sizes, as well as the results of nanoparticle size and zeta potential analysis, are presented in

Table 1. The crystal size, specific surface area, and particle size of the sample were determined.

Sample Name	[Mn(NO3)2]/M	Crystal Phase	d/nm	d50 */nm	SBET/(m2/g)	Pore Volume/(cm3/g)	Average Hole Diameter/nm
Mn0.025	0.025	Mn2O3	40.357	1099	14.063	0.047	17.580
Mn0.5	0.5	Mn2O3	26.964	833	4.952	0.018	16.227
Mn1	1	Mn2O3	42.513	591	5.724	0.018	14.755
Mn3	3	Mn2O3	46.255	675	6.724	0.032	21.640
Mn7.6	7.6	Mn2O3	36.268	807	6.574	0.018	8.573

* d₅₀ is the corresponding median aperture in the cumulative distribution curve of the aperture distribution.

and Figure 12. The particle size distribution of Mn₂O₃ ranges from approximately 600 to 1100 nm. According to the test results, as the precursor solution concentration increases, the particle size exhibits a non-monotonic dependence on precursor concentration, with a minimum observed at Mn₁. As shown in Figure 12, larger particle sizes are associated with more dispersed particle size distributions.

A nitrogen desorption test was performed on the above five Mn₂O₃ samples. The nitrogen desorption curve and the BJH desorption curve differential pore volume distribution were obtained. The nitrogen adsorption and desorption curves of the five samples were hysteresis curves, as shown in Figure 13.

Figure 13a shows that all samples prepared with different manganese nitrate concentrations display Type IV nitrogen adsorption-desorption isotherms with distinct hysteresis loops. The generation of this kind of curve was due to the capillary condensation of the sample during the adsorption process, indicating that the pore size of the sample tested was distributed between 2 and 50 nm, which belonged to mesopores. As shown in Figure 13b, the pore volume change rate of Mn_0.025 in the pore diameter range of 80–90 nm and 2–3 nm was large, which indicated that the pore diameter of the sample was mainly distributed in these two ranges, and the average pore diameter showed that the number of pores in the pore diameter range of 2–3 nm was more than that in 80–90 nm. The pore volume of Mn_0.5, Mn₁, and Mn_7.6 was the largest in the range of 2–4 nm, so the pore diameter of these three samples was the majority in this range. The pore volume change rate of the Mn₃ sample was the largest in the range of a pore diameter of 60–80 nm, so the pore diameter distribution of the Mn₃ sample was more in this range. It can be seen from Table 1 that the specific surface area of Mn_0.025 is two to three times that of other samples, while the pore volume and pore diameter of Mn_0.025 are similar to those of other samples, indicating that the Mn_0.025 has fewer pores so that the gas could not fully make contact with the sample, which should be one of the reasons for the low weight gain rate of the re-oxidation reaction.

To further investigate the microstructure of the samples, a SEM analysis was conducted on the five selected samples (Mn_0.025, Mn_0.5, Mn₁, Mn₃, and Mn₄). Figure 14 shows the SEM images, with the left side displaying the initial morphology of the samples and the right side showing their morphology after cyclic reactions. Initially, the five Mn₂O₃ samples exhibited a structure composed of small, round primary particles with loose pores, which facilitated full contact between the sample and the gas, enabling complete reactions. In contrast, after cycling, the samples displayed a coral-like morphology with a smooth surface and no visible pores, likely due to sintering effects during the cyclic reactions.

3.2.3. Synthesis Parameters Assessment on Reaction Performance

Figure 15 presents the temperature dependence curves of the reduction and oxidation degrees for Mn_0.025, Mn_0.5, Mn₁, Mn₃, and Mn_7.6, respectively. A comparison of Figure 15 reveals that the initial particle size has a smaller influence on the reduction reaction compared to the oxidation reaction. The reduction reaction temperatures for Mn_0.025 and Mn_7.6, which have larger particle sizes, are higher than those of the other samples, while their oxidation reaction temperatures are lower. This suggests that larger initial particle sizes require more stringent temperature conditions for the reactions. Regardless of particle size, the temperature range for the reduction reaction of the five Mn₂O₃ samples is narrower than that for the oxidation reaction.

In this study, the maximum peak value of the differential curve of the TG curve with respect to time was used to determine the reaction rate. As shown in Figure 16a, the Mn_0.025 sample, influenced by its larger particle size, not only exhibited a significant decrease in the oxidation reaction weight gain rate but also showed a rapid decline in the reduction reaction rate as the number of cycles increased. In contrast, the reduction reaction rates of the other samples remained relatively stable over the five cycles. Additionally, samples with larger initial particle sizes demonstrated faster reduction reaction rates. Figure 16b indicates that the oxidation reaction rate remains relatively stable with increasing cycle numbers, showing no significant variation. Similar to the reduction reaction, the sample with larger initial particle sizes exhibited a higher oxidation reaction rate. However, for the Mn_0.025 sample, the oxidation reaction was influenced by factors such as particle size, specific surface area, and pore size distribution. As a result, the reaction kinetics slowed with successive cycles, preventing complete transformation and leading to a lower oxidation reaction rate.

3.2.4. Reaction Kinetics Analysis

The temperatures corresponding to the reduction reaction of Mn₂O₃ at different heating rates are summarized in

Table 2. The corresponding temperature of the Mn₂O₃ reduction reaction at different heating and conversion rates.

α	Heating Rate (°C/min)
	2	5	10
	T/K
0.1	954.80	957.72	960.72
0.2	959.80	962.00	964.57
0.3	962.12	965.14	967.69
0.4	965.03	967.64	970.42
0.5	966.66	970.19	973.06
0.6	969.41	972.83	975.98
0.7	970.19	975.82	979.05
0.8	976.93	979.16	982.63
0.9	981.80	984.17	987.82

. The corresponding temperature of the Mn₂O₃ reduction reactions is shown at different heating and conversion rates. As indicated in the table, at the same conversion rate, higher heating rates result in higher reaction temperatures. This is attributed to the fact that faster heating rates cause the sample to react more slowly as it reaches the required temperature.

To avoid errors in activation energy calculations arising from the determination of mechanism functions, the activation energy in this study was determined using an iterative method. The detailed calculation results are presented in

Table 3. The activation energy of the Mn₂O₃ reduction reaction was calculated using Ozawa and KAS methods.

α	Ozawa	KAS	$\mathbf{ln}\frac{\mathit{\beta }}{\mathit{H}}\cdots \frac{1}{\mathit{T}}$	$\mathbf{ln}\frac{\mathit{\beta }}{\mathit{h}{\mathit{T}}^2}\cdots \frac{1}{\mathit{T}}$
	Activation Energy E/(kJ/mol)
0.1	236.93	247.19	237.55	247.22
0.2	295.82	309.16	295.79	309.20
0.3	256.40	267.70	256.12	267.75
0.4	265.38	277.13	264.81	277.18
0.5	225.36	235.05	224.90	235.10
0.6	220.71	230.16	220.23	230.13
0.7	162.12	168.54	161.29	168.61
0.8	251.15	262.16	250.81	262.20
0.9	240.88	251.34	240.48	251.40
Ave.	239.42	249.82	239.11	249.87

. As shown in the table, the activation energies calculated by the two methods and their corresponding iterative approaches are 239.42 kJ/mol, 249.82 kJ/mol, 239.11 kJ/mol, and 249.87 kJ/mol, respectively. Based on the study by Dizaji et al. [11], it was determined that the activation energy calculated using the Kissinger–Akahira–Sunose (KAS) method is more accurate. Therefore, the activation energy for this reaction is 249.87 kJ/mol, and the KAS method is deemed more suitable for the kinetic analysis of this reaction.

4. Conclusions

Mn₃O₄, as the initial reactant, can undergo redox reactions with Mn₂O₃. In the initial stage, Mn₃O₄ first undergoes elemental restructuring to transform into Mn₂O₃, followed by subsequent redox reactions. In a five-cycle experiment with a heating and cooling rate of 2 °C/min, the reaction exhibited stable cyclic performance, with a conversion rate exceeding 95% and a weight loss of approximately 3.3%. As the cycling proceeded, the temperature range of the oxidation reaction shifted toward lower temperatures but the reaction rate and conversion rate remained largely unchanged, meeting the requirements for energy storage systems. A kinetic analysis of the oxidation reaction using the isothermal method revealed that the reaction follows Arrhenius’ Law at temperatures below 700 °C but deviates from it at temperatures above 700 °C. Using experimental data below 700 °C, the most probable mechanism function for the oxidation reaction was determined to be G(α) = α and f(α) = 1, with an activation energy (E) of 20.47 kJ/mol and a pre-exponential factor (A) of 0.268/s.

Using the ammonia precipitation method, Mn₂O₃ samples prepared from precursor solutions of varying concentrations, except for the 0.025 mol/L manganese nitrate solution, demonstrated excellent performance in cyclic reactions. The results indicate that Mn₂O₃ samples prepared using this method are suitable for thermochemical energy storage (TCS). However, the particle size of Mn₂O₃ samples varied with the precursor concentration. Particle size had a more pronounced effect on the oxidation reaction: larger particles required stricter temperature conditions for redox reactions. The activation energy for the reduction reaction of Mn₂O₃, calculated using the non-isothermal multi-heating rate method, was determined to be 249.87 kJ/mol.