Resumen de Development of efficient mn-based redox materials for thermochemical heat storage in concentrated solar power plants
Alfonso Juan Carrillo del Teso
Concentrating solar power (CSP) is meant to play a major role in achieving the targeted objectives of renewable energy generation in the next decades. This is partly due to the possibility of implementing thermal energy storage (TES) systems allowing increasing the annual capacity factor and enhancing the energy generation dispatchability. The excess of heat collected by the solar field can be stored by such systems via three different mechanisms: sensible, latent and thermochemical heat storage. To date, only sensible heat storage using molten salts has been successfully implemented in commercial plants. However, new advances in tower plants working with volumetric air receivers will allow producing air at temperatures above 1000 ºC, which is higher than the stability range of commercial salts. Relying on this fact, thermochemical heat storage (TCS) based on redox reactions of metal oxides, has emerged as a potentially more efficient option for storing excess of heat at high temperatures. In particular, the Mn2O3/Mn3O4 redox couple is considered as a promising candidate due to its low cost and toxicity and favorable thermodynamics at 1000 ºC. However, in order to guarantee the feasibility of the process, selected materials should present long term durability, which requires the evaluation of the redox couple cyclability for extended periods. Therefore, the main objectives of the present work have been to evaluate experimentally the viability of the Mn2O3/Mn3O4 redox couple for thermochemical heat storage and to develop more efficient Mn-based oxides for such application.
First, commercial Mn oxides were studied as to corroborate the redox reversibility of such system. One of the tested samples showed redox cyclability, evaluated in thermobalance, confirming the feasibility of this system for TCS applications. However, it was found that after several cycles the material suffered deactivation mainly caused by particle sintering, as observed through SEM analyses. Furthermore, it was observed that reduction was faster than oxidation. This indicated that efforts should be focused on improving the oxidation kinetics at the same time that cycle stability is guaranteed.
First approach performed for that purpose was based on morphological modifications. Several synthesis methods (precipitation, Pechini) were evaluated in order to modify structural and textural properties of Mn oxides and study how these variations affected to the overall cycling behavior. Through varying synthesis parameters of the precipitation method it was possible to tune the particle size and specific surface area of Mn2O3 samples. Materials formed of smaller particles (<88 nm) showed deactivation after some cycles, owing to a higher degree of densification, caused by an increased sintering effect. On the other hand, materials with larger particles (>127 nm) showed 30 redox-cycle stability, as measured in thermobalance. However, it was observed that sintering also affected to these samples, causing a cycle-to-cycle decay on the oxidation rate, mainly caused by hindrances to oxygen diffusion due to particle coarsening. Conversely, reduction rates were stable over the whole cycling test. Pechini method was also utilized to prepare Mn2O3 materials, owing to the high homogeneity of the final product that can be achieved with this route, resulting in samples with high cycling stability, although suffered from the same oxidation rate decay. Mn-oxides with induced macroporosity were also synthesized as to evaluate the effect of additional porosity on enhancing the oxidation rate. It was observed that porosity generated with the aid of porogens (starch, methyl cellulose or ammonium carbonate) was not beneficial for redox cycling, whereas through an additive-free method, Mn2O3 with higher macroporosity was observed to present faster oxidation rates. However, rates of such reaction again suffered a cycle-to-cycle decay caused by sintering.
In the second approach metal doping was used to modify the structure of both Mn2O3 and Mn3O4 oxides. Several Mn oxides doped with different metal cations (Al3+, Cr3+, Cu2+, Fe3+) were prepared through Pechini method. In particular, it was observed that Fe incorporation, which is also an abundant element, improved the redox performance of this system, while Al-doping totally hampered the redox reversibility, and Cr and Cu-doping did not show remarkable improvements. Namely, a material with (Mn0.8Fe0.2)2O3 stoichiometry exhibited the fastest oxidation reactions and presented stable redox reversibility over 75 cycles without loss of reactivity. Furthermore, the heat storage density was increased and the redox thermal hysteresis (temperature difference between reduction and oxidation) narrowed. Due to its promising characteristics, kinetics of reduction and oxidation reactions of such sample were evaluated in more detail. Rate laws were determined for both reactions. Special attention was put on a better understanding of the oxidation reaction. It was suggested that such process takes place through a nucleation and growth mechanism. In addition, incorporation of Fe was found to alter the Mn-O bonding, which in turn facilitated bond rearrangement, enhancing the oxidation rate.
In summary, the feasibility of the Mn oxide-based TCS system has been demonstrated in the present work. Fe-Mn oxides were developed and found to be promising candidates for up-scaling in TCS reactors coupled to CSP plants due to its favorable thermodynamics and outstanding kinetic and reversibility behavior over prolonged cycling.
