SOLAR CELL

The photovoltaic effect, the generation of voltage when a device is exposed to light, was first discovered by Becquerel back in 1839, in a junction formed between an electrode and an electrolyte. Similar effects on selenium were observed by Adams and Day in 1876 and by Lange in 1930 and on cuprous oxide by Schottky in 1930 and by Grondahl in 1933. Photovoltaic effect on Ge was reported by Benzer in 1946 and by Pantchechnikoff in 1952. It was not until 1954 that the solar cell received increased interest, initiated by the works of Chapin et al. on single-crystal silicon cells, and of Reynolds et al. on cadmium sulfide cells. For other materials, Gremmelmaier presented results on GaAs in 1955 and Carlson and Wronski generated much interest in amorphous silicon in 1976. Analytical studies had been performed by Cummerow, Rittner, Prince and Loferski.

A solar cell can be made of a p-n junction or a Schottky barrier. The p-n junction version (p on n or n on p) is more common because it has better reliability and higher open-circuit voltage. The metallurgical junction is usually shallow, typically 0.5-1 nm deep. Too shallow a junction will increase the sheet resistance of the top layer, and also increase the dark current. A deep junction is not efficient in collecting carriers excited near the surface, especially from light of shorter wavelength which has high absorption coefficient. Another possible structure is the heterojunction solar cell. It has a material of higher energy gap at the surface so little light is absorbed in that layer. When this flayer is heavily doped, it has the combination of low sheet resistance, low dark current and good short-wavelength response. A similar structure is the heteroface solar cell where an isotype heterojunction is incorporated. The Schottky-barrier version offers certain advantages because no high-temperature processing is needed. This, for example, avoids enhanced diffusion along grain boundaries in poly-crystal materials. Another advantage is that the junction is at the surface for better short-wavelength response. An inherent disadvantage of the Schottky-barrier cell is a larger dark current. This can be improved somewhat by introducing a thin tunneling oxide layer of 10-20 A between the metal and the semiconductor, resulting in an MIS solar cell. The right choice of oxide thickness can decrease the dark current without affecting the light-generated current. The benefit of a low dark current will be shown later.

Since the solar cell is used as a power-generating source, series resistance is a critical factor, unlike the case of photodetectors. A metal grid structure is used to form ohmic contact. To design the shape of the grid, a compromise is struck between series resistance and area lost to exposure to light. Transparent conducting films such as InSn oxide (ITO) have been explored. Another film of antireflection coating is usually deposited to minimize overall reflection. Cells with texturized surface have also been made so that some of the reflected light can be reabsorbed.

More common materials for single-crystal cells are Si, GaAs, InP, and CdTe. Deposited thin-film solar cells offer low cost and large-area capability, but they are limited to low efficiency. These materials include polycrystal and amorphous Si, GaAs, and CdS (Cu2S-CdS heterostructure). The thickness of the solar-cell body needs to be only a few absorption lengths in order to absorb most of the power since the penetration of light decreases with distance.