Comparing the performance of thermophotovoltaic (TPV) devices is challenging due to a lack of standard operation conditions. Here, we propose a universal figure of merit (FOM) that can be used to evaluate the performance of TPV devices that operate in the far-field regime relative to their thermodynamic bounds. The introduced FOM alleviates temperature dependence and accounts for the fundamental trade-off between power density and efficiency. Based on this FOM, we present a classification of TPV performances reported in recent experiments.

Thermophotovoltaics (TPVs) differs from solar photovoltaics (PV) because pairwise efficiency and electrical power cannot be optimized simultaneously, as a consequence of spectral selectivity or photon recycling. A review of around thirty experiments conducted so far is carried out, and the achieved performances are compared with those obtained in the detailed balance limit. The link between optimal cell bandgap and emitter temperature is highlighted as a function of out-of-band radiation exchange between the emitter and the cell. The analysis reveals that almost all the experimental data reported are far from power-maximizing conditions and more focused on optimizing efficiency. At high temperature, thermal management is obviously an issue and optimizing efficiency is required to minimize heat generation. In general, it is argued that in addition to pairwise efficiency and electrical power density, heat power density is a third metric that should be considered in the design of TPV devices.