A-site doped Ca1-xSrxMnO3-δ and B-site doped CaCryMn1-yO3-δ have been studied for high specific thermochemical energy storage (TCES) in concentrating solar power and other high-temperature applications. With low levels of doping (5 and 10%) for both compositions, these perovskites have demonstrated significant reduction in high temperatures at O2 partial pressures P O2 ≈10-4 bar with as much as 450 kJ/kg of chemical energy can be stored (for Ca0.95Sr0.05MnO3-δ) at 900 ºC on top of sensible energy in a TCES system [1]. The viability of such high specific TCES in these perovskites depends on their rates of energy accommodation during heating cycles and their phase and morphological stability during rapid reduction and reoxidation for energy capture and release respectively. In this study, Ca1-xSrxMnO3-δ and CaCryMn1-yO3-δ were studied for reduction at temperatures up to 1000 ºC and tested in extended redox cycling to explore their stability for energy storage and release at such high-temperatures where the materials undergo expansion/compression and phase transitions during these cycles. The results as highlighted below suggest that these materials have significant promise for high-temperature large-scale TCES. Ca1-xSrxMnO3-δ (x = 0.05 and 0.10) and CaCryMn1-yO3-δ (y = 0.05 and 0.10) were synthesized as particles with solid-state-reactive sintering [1] and cycled as particles (diameter range of 250-425 μm) between air P O2 (≈ 0.17 bar) and P O2 ≈10-4 bar both in isothermal experiments up to 1000 ºC and in long-term cycling experiments between 500 and 900 ºC to simulate TCES system operation. Over these conditions, the doped CaMnO3-δ perovskites undergo significant reduction (δ > 0.2 at 1000 ºC) at phase transitions (orthorhombic to cubic to tetragonal perovskite phase), reduction, and associated lattice expansion as observed in hot-stage XRD for similar composition [1]. The isothermal redox cycling tests in conjunction with thermodynamics derived from TGA/DSC measurements as reported elsewhere [2,3] provided a basis for fitting thermodynamically consistent surface reaction rates [4,5] and bulk ionic transport properties to measured transient evolution in oxygen non-stoichiometry δ which went as high as 0.3 for Ca0.95Sr0.05MnO3-δ at 1000 ºC. These experiments are illustrated with ten-cycle averaged plots in Figure 1 with data for Ca0.9Sr0.1MnO3-δ compared to packed bed models [4] used to fit surface kinetic and bulk transport rate expressions. Figure 1 also shows the chemical energy incorporated into the perovskite with increasing δ during reduction based on thermodynamics derived from fitting equilibrium δ over a range of T and P O2. The packed bed results and model fits show fast rates of reoxidation relative to reduction. The rapid reoxidation i.e., oxygen incorporation rates are limited by the inlet flow supply for a substantial fraction of the experiments. The high-degree of reduction and the rapid rates of reoxidation and associated phase transitions raise questions about the stability of these materials in non-isothermal redox cycling for high-temperature TCES system applications. The selected perovskite materials were thus cycled for over 1000 times under non-isothermal conditions to simulate TCES system operation. The long-term cycling for all four doped CaMnO3-δ perovskites occurred with heating from a fully oxidized state at 500 ºC to a reduced state at 900 ºC in P O2 ≈10-4 bar followed by subsequent reoxidation and cooling in air back to 500 ºC. Figure 2 shows the transient evolution of δ during reduction and heating, averaged over successive 100 cycles for all four materials. An initial induction period of increasing maximum δ is followed by a stable period of redox cycling in spite of the rapid reoxidation. Only CaCr0.05Mn0.95O3-δ shows evidence of a significant loss in maximum δ, and this composition showed the slowest to undergo phase transitions in DTA experiments. Ex situ X-ray diffraction shown in Figure 3 of the particles before and after the long-term non-isothermal redox testing show that the particles maintain their dominant perovskite phase with evidence of small fractions of a secondary spinel phase. The stability of these cyclic tests particular for the Ca1-xSrxMnO3-δ reveal the promise of these materials as a media for effective TCES in a range of high-temperature energy storage applications. Reference [1] E.I. Leonidova, I.A. Leonidov, M.V. Patrakeev, V.L. Kozhevnikov, Journal of Solid State Electrochemistry, 15 (2011) 1071-1075. [2] L. Imponenti, K.J. Albrecht, R.J. Braun, G.S. Jackson, ECS Transactions, 72 (2016) 11-22. [3] L. Imponenti, K.J. Albrecht, J.W. Wands, M.D. Sanders, G.S. Jackson, Solar Energy, 151 (2017) 1-13. [4] K.J. Albrecht, "Multiscale Modeling and Experimental Interpretation of Perovskite Oxide Materials in Thermochemical Energy Storage and Conversion in Applications for Concentrating Solar Power", Ph.D. Dissertation, Colorado School of Mines, 2016. [5] R. Merkle, J. Maier, Angewandte Chemie-International Edition, 47 (2008) 3874-3894. Figure 1
Read full abstract