No abstract

Waste heat generated during daytime operation of a solar module will raise its temperature and reduce cell efficiency. In addition to thermalization and carrier recombination, one major source of excess heat in modules is the parasitic absorption of light with sub-bandgap energy. Parasitic absorption can be prevented if sub-bandgap radiation is reflected away from the module. We report on the design considerations and projected changes to module energy yield for photonic reflectors capable of reflecting a portion of sub-bandgap radiation while maintaining or improving transmission of light with energy greater than the semiconductor bandgap. Using a previously developed, self-consistent opto-electro-thermal finite-element simulation, we calculate the total additional energy generated by a module, including various photonic reflectors, and decompose these benefits into thermal and optical effects. We show that the greatest total energy yield improvement comes from photonic mirrors designed for the outside of the glass, but that mirrors placed between the glass and the encapsulant can have significant thermal benefit. We then show that optimal photonic mirror design requires consideration of all angles of incidence, despite unequal amounts of radiation arriving at each angle. We find that optimized photonic mirrors will be omnidirectional in the sense that they have beneficial performance, regardless of the angle of incidence of radiation. By fulfilling these criteria, photonic mirrors can be used at different geographic locations or different tilt angles than their original optimization conditions with only marginal changes in performance. We show designs that improve energy output in Golden, Colorado by 3.7% over a full year. This work demonstrates the importance of considering real-world irradiance and weather conditions when designing optical structures for solar applications.

Ca3Co4O9 thin films synthesized through solution processing are shown to be high-performing, p-type transparent conducting oxides (TCOs). The synthesis method is a cost-effective and scalable process that consists of sol-gel chemistry, spin coating, and heat treatments. The process parameters can be varied to produce TCO thin films with sheet resistance as low as 5.7 kΩ/sq (ρ ≈ 57 mΩ cm) or with average visible range transparency as high as 67%. The most conductive Ca3Co4O9 TCO thin film has near infrared region optical transmission as high as 85%. The figure of merit (FOM) for the top-performing Ca3Co4O9 thin film (151 MΩ−1) is higher than FOM values reported in the literature for all other solution processed, p-type TCO thin films and higher than most others prepared by physical vapor deposition and chemical vapor deposition. Transparent conductivity in misfit layered oxides presents new opportunities for TCO compositions.

Many existing commercially manufactured photovoltaic modules include a cover layer of glass, commonly coated with a single layer antireflection coating (ARC) to reduce reflection losses. As many common photovoltaic cells, including c-Si, CdTe, and CIGS, decrease in efficiency with increasing temperature, a more effective coating would increase reflection of sub-bandgap light while still acting as an antireflection coating for higher energy photons. The sub-bandgap reflection would reduce parasitic sub-bandgap absorption and therefore reduce operating temperature. This reduction under realistic outdoor conditions would lead to an increase in annual energy yield of a photovoltaic module beyond what is achieved by a single layer ARC. However, calculating the actual increase in energy yield provided by this approach is difficult without using time-consuming simulation. Here, we present a time-independent matrix model which can quickly determine the percentage change in annual energy yield of a module with a spectrally selective mirror by comparison to a baseline module with no mirror. The energy benefit is decomposed into a thermal component from temperature reduction and an optical component from increased transmission of light above the bandgap and therefore increased current generation. Time-independent matrix model calculations are based on real irradiance conditions that vary with geographic location and module tilt angle. The absolute predicted values of energy yield improvement from the model are within 0.1% of those obtained from combined ray-tracing and time-dependent finite-element simulations and compute 1000× faster. Uncertainty in the model result is primarily due to effects of wind speed on module temperature. Optimization of the model result produces a 13-layer and a 20-layer mirror, which increase annual module energy yield by up to 4.0% compared to a module without the mirror, varying depending on the module location and tilt angle. Finally, we analyze how spectrally selective mirrors affect the loss pathways of the photovoltaic module.

No abstract