Thermophotovoltaic efficiency

A hot block gives off infrared light. A photovoltaic cell converts some of that light into electricity, while a reflector sends unused light back to heat the block. Can the converter turn more than half of the radiation it absorbs into electricity? Here light crosses a gap much larger than its wavelength—the far-field regime.[1] [5]

The light source is a hot emitter, rather than sunlight. Photons below the cell’s bandgap lack the energy needed to generate its electrical charge carriers; reflecting them back to the emitter recovers energy that would otherwise be wasted. The stated converter metric is $\eta=P_{\mathrm{electric}}/(P_{\mathrm{incident}}-P_{\mathrm{reflected}})$. It treats reflected power as returned to the source. The 100 W incident, 60 W returned and 20 W electrical example gives 50% under this convention, rather than 50% of all incoming light. In hardware, mirror absorption and escaping radiation weaken recovery; generated charges can recombine before collection, and electrical resistance dissipates power. The far-field scope concerns an ordinary radiative gap much larger than the wavelength.[1] [5]

In a 2024 experiment, air-bridge cells reached up to 44% under 1,435°C blackbody illumination. Designs predicting more than 50% still need measured hardware that validates the prediction.[2] [4]

Photon reflection, carrier recombination, resistance and radiation geometry limit measured performance. The 2025 filter result uses simulation and detailed balance; converter efficiency omits storage charging and other plant losses.[1] [4]

Stored heat could supply electricity when it is needed. Better conversion would let a given supply of heat produce more electricity, but charging losses and heat escaping during storage also affect how much comes back.[1] [2]

A thermal battery must ultimately be judged by electricity returned relative to the energy used to charge it. Converter improvements address one part of that chain. Understanding which light is recovered and which is lost helps identify whether better cells, mirrors or geometry would improve useful output. It also prevents a favourable cell percentage from being mistaken for measured performance of a complete storage plant.[1] [2]