We have numerically studied the thermal-adaptive response of the lattice-matched GaxIn1−xAsySb1−y/InAs0.91Sb0.09 dual-junction thermophotovoltaic (TPV) cell. It is shown that the general linear temperature coefficient (κ) is only applicable to model the cell temperature (TC)-induced variations of the optimal structure and resulting performance in the near room temperature regime. With increasing spectrum temperature (TBB), κ extracted from a near room temperature regime has a nonlinear evolution, yielding reasonably an exponential decay function, i.e., a2exp(−TBB/a1)+a0, for the grid-finger separation, while a cubic function, i.e., a3TBB3+a2TBB2+a1TBB+a0, for the rest structure and performance parameters with aj(j=0,1,2,3) being the extracted coefficients. For the general TBB of realistic TPV systems, i.e., 1000–2000 K, κ for the short-circuit current density, open-circuit voltage, and efficiency of studied dual-junction cells are, respectively, in the range of −0.01 to 0.01 A/cm2K, −2.2 to −1.8 mV/K, and −0.07 to −0.01%/K. Our work formulates an effective method to rapidly predict the required structure and performance of a dual-junction TPV cell for the operation condition beyond the room temperature, the common scenarios in the realistic TPV systems.
Lou et al. (2026) studied this question.