Critical issues in the design of polycrystalline, thin-film tandem solar cells

We use an empirical technique for modeling the efficiency of thin-film tandem solar cells and calculate an approximate upper limit on the range of performance of these hypothetical devices. This is shown to be approximately 28-2%, without losses due to inactive layers at the front of the device, or other parasitic sources. Reduction of the value of the reverse saturation current density by a factor of ten, increases the lossless efficiency by approximately 4% absolute. This change also greatly broadens the range of top and bottom cell bandgaps that would lead to efficiencies greater than 25%, the project goal. These observations emphasize the critical importance of focusing future research on gaining a thorough understanding of recombination losses. We then calculate daily energy density outputs for various direct spectra, computed from meteorological data, and show that the optimum bandgap pairs are relatively insensitive to the detail of the spectral irradiance. We also show that the use of daily energy density output may be a more useful criterion than efficiency in designing tandem thin-film solar cells. We compute contours of equal daily energy density output and show that the range of potentially suitable bandgap pairs is much larger than simple maximization of efficiency implies. The simple parametric approach enables us to investigate the effect of partial loss of photons with energies less than that of the bandgap of the top cell, but greater than that of the bottom cell. These photons are essential to the project goal of 25% efficiency, which emphasizes the need to evaluate the optical properties in this wavelength range very carefully. We also discuss the reduction of the thickness, or the area, of the top cell. When the top subcell generates a greater current than the bottom subcell, either of these parameters may be reduced to enable current-matching, and increased efficiencies, to be achieved. Again, this approach greatly extends the range of bandgaps that could lead to a 25% tandem thin-film cell. Next, we consider the case of concentrated sunlight and show that the optimum bandgap pairs decrease with concentration ratio. This is due to the atmospheric absorption bands. The efficiency increases by approximately 4% absolute per decade increase in concentration ratio. Finally, we comment on some of the practical difficulties that can already be anticipated in constructing these devices. Published in 2003 by John Wiley Sons, Ltd.