Recombination-limited near-field thermoradiative conversion of low-grade waste heat: from cell physics to DC-bus power, interconnection design and the conditions for junction stacking

Near-field thermoradiative cells can convert low-grade heat into electricity, but their performance has been assessed almost only for single, ideally contacted cells; how much power survives interconnection, module non-uniformity and power conversion, and whether stacking junctions helps, has remained open. Here, HgCdTe cells at 400-650 K and vacuum gaps of 10-50 nm are followed from the cell to a DC bus, combining fluctuational electrodynamics, Auger and Shockley-Read-Hall recombination, design optimization, module Monte Carlo analysis, and models of interconnect resistance and DC-DC conversion. Realistic recombination makes every cell a sub-millivolt, high-current source: the nonradiative conductance exceeds the radiative one about fiftyfold, fixing the voltage near 0.4 mV and the current near 4 A cm^−2. This mechanism explains three results. Parallel connection of a two-junction stack retains only 12-93% of the independently contacted power, whereas series connection retains at least 98% where the junction currents balance. An optimized single junction matches or exceeds a stack of equal total absorber thickness at every Auger level and in the radiative limit. Interconnect resistance, not optics, decides the system output: the single junction delivers more power to the bus below about 32 μΩ cm^2, the series stack above it. A 2-nm vacuum-gap spread costs 5% of the module power, and tenfold Auger suppression raises the power from 15 to about 130 W m^−2. The work establishes realistic performance expectations at system level and identifies vacuum-gap uniformity, interconnect resistance and Auger suppression as the quantities that limit usable output.