In brief summary, the ideal single-junction cell at best captures only about 30% of the light energy. Going beyond the single-junction cell, a well-developed technology is to stack two or three junctions in series (tandem), engineered so that light not converted in the first junction may be converted in the second or third junction. This approach brings the measured efficiency for a (concentrated) tandem cell up to the vicinity of 41.8%. Tandem solar cells, typically based on GaAs, are commercially available but are expensive. On the other hand, in conjunction with mirrors or lenses to concentrate light, as mentioned in connection with Figure 7.7, this approach may possibly be competitive, even with electricity from coal-powered plants. Light dispersing optics has also been demonstrated to separate the spectrum and to steer different spectral regions to locations of appropriate solar cells.
The second, less developed, idea, is to use dye molecules in wide-area receiving plates, to funnel reemitted light to a dedicated array of solar cells. A large inventory of dye molecules has long been available, and this particular use of dyes is based on the most fundamental role of a dye, to convert “blue” or “green” photons to “red” photons. This proposal could turn out to be important.
The third approach, in principle, to improve efficiency is to multiply the number of charges produced per photon. This could make use of the blue portion of the solar spectrum, not fully utilized in the single-junction cell, on the conventional assumption of one charge per photon, with excitons. This effect is well known in the physics of the “Li-drifted germanium detector,” a device for measuring the energy ofa cosmic ray or other energetic particle by counting how many charges it generates by the exciton multiplication process in a pure sample of germanium. We will return to this topic in Chapter 8, and to the further possibility of an “intermediate band solar cell” (see the second curve in Figure 7.1).