Our team has explored redox cycles of doped CaMnO_3-δ between air and low O₂ partial pressures (~10^-4 bar) for high-temperature thermochemical energy storage (TCES) applications.  In this study, we have explored both A-site and B-site doped compositions using earth-abundant cations to identify perovskites for cost-effective TCES in concentrated solar and other high-temperature, thermal storage applications. Reduction of doped CaMnO_3-δ above 800 °C in the low P_O2 requires some amount of dopant to avoid irreversible decomposition of the perovskite structure observed for reduction of undoped CaMnO_3-δ [1]. This study shows that small amounts (5%) of A-site or B-site dopant can stabilize the CaMnO_3-δ perovskite structure during reduction at temperatures up to 1100 °C and P_O2 = 10^-4 bar.  For selected, stable doped CaMnO_3-δ compositions, total specific TCES (Δh_tot) was defined by the sum of specific sensible energy (Δh_sens) and chemical energy (Δh_chem) captured during heating and reduction from air at a cool temperature (T_C fixed at 500 °C in this study) to low P_O2 at varying high temperatures (T_H). B-site doping with 5% Cr and A-site doping with 5% Sr provided thermodynamic limits of Δh_tot over 720 kJ kg^-1 and 790 kJ kg^-1 respectively for T_H = 1000 °C. These materials were also tested kientically to assess the rates at which energy storage is captured during reduction and released during oxidation as relevant for concentrated solar energy storage applications. The results indicate that the doped CaMnO_3-δ and in particular A-site doped Ca_0.95Sr_0.05MnO_3-δ have promise as a TCES storage media with Δh_chem providing just under half of the total energy stored during the combined heating and reduction. 

For the perovskite redox cycles, the heat of oxide reduction -ΔH_O can vary with non-stoichometry δ [2]. To find the integrated Δh_chem, the functional dependence of ΔH_O with δ must be detemined. Two methods have been undertaken to explore this functional dependence: 1) combined TGA-DSC calorimetry measurements with incremental changes in δ, and 2) point-defect model fitting to equilibrium δ vs. P_O2 data from TGA measurements [3,4]. The DSC measurements showed that the magnitude of ΔH_O decreases to a near constant value for δ > 0.1 as illustrated for Ca_0.95Sr_0.05MnO_3-δ in Figure 1. Similar trends were observed for the other perovskite compositions. The alternative method for finding ΔH_O as a function of δ by fitting the equilibrium δ vs. P_O2 involved modeling the perovskite reduction with two reversible point-defect reactions -- oxide-ion vacancy formation and Mn cation disproportionation from 4+ to 3+ and 5+ states [4]. The model fits originally assumed that these two reactions had  enthalpies independent of δ but the contribution of the disproportionation reaction increased with δ thereby causing the combined ΔH_O to have a minor dependence on δ as also shown in Figure 1. The point defect modeling did not show the same trends for ΔH_O vs. δ but provided very similar values at high δ > 0.1. The integration of ΔH_O over a change in δ (i.e. Δδ as in Table 1) provides a basis for calculating the integrated Δh_chem as a function of T_H. Integrated Δh_chem values for  CaCr_0.05Mn_0.95O_3-δ, Ca_0.95Sr_0.05MnO_3-δ, and Ca_0.9Sr_0.1MnO_3-δ are shown in Table 1 along with the total energy stored which includes Δh_sens derived from integration of C_P with respect to T over the redox cycle heating.  The higher Δh_chem and associated Δh_tot for Ca_0.95Sr_0.05MnO_3-δ stems from its increase reducibility without a loss in heat of reaction and makes it a more promising material for TCES applications in these temperature ranges.

For thermal storage application, kinetic rates of TCES capture are important, particularly for concentrated solar applications where solar receiver residence times are limited. To explore kinetics, reduction and oxidation rates of porous perovskite particle beds were measured for fitting thermodynamically consistent kinetic models to the observed rates. Kinetic testing and model fitting for the Ca_0.95Sr_0.05MnO_3-δ indicates that it achieves approximately 80% of its Δh_chem thermodynamic limit in 60 s exposure of low P_O2 gas for reduction at T_H = 900 °C.  On the other hand, at the same temperature reoxidation occurs rapidly and releases over 90% of the Δh_chem limit in 30 s.  The thermodynamic consistent model involving surface kinetics and ionic bulk diffusion in the perovskite particles for Ca_0.95Sr_0.05MnO_3-δ and for CaCr_0.05Mn_0.95O_3-δ provide a basis for designing reactors for redox cycles to be used in TCES for concentrated solar and other potential thermal energy storage applications.

References

[1] E.I. Leonidova, I.A. Leonidov, M.V. Patrakeev, and V.L. Kozhevnikov, J. Solid State Electrochem., 15 (2011) 1071-1075.

[2] J. Mizusaki, M. Yoshihiro, S. Yamauchi, K. Fueki, J. Solid State Chem., 67 (1988) 1–8.

[3] E.I. Goldyreva, E.I., I.A. Leonidov, M.V. Patrakeev, A.V.Chukin, I.I. Leonidov, V.L. Kozhevnikov, J. Alloys Comp., 638 (2015) 44-49.

[4] K.J. Albrecht, G.S. Jackson, R.J. Braun, Appl. Energy, accepted for publication (2015).