In the early 1990s, the technology used for photo-voltaic space solar cells diverged from the silicon technology used for terrestrial panels. Instead, the spacecraft application shifted to gallium arsenide-based III-V semiconductor material compositions. These in turn evolved to the modern III-V multijunction photovoltaic cell used on spacecraft with architectures built of four or more junctions. Silicon solar cells begin life as single crystal silicon with implanted p- and n junctions that generate current when illuminated with light of greater energy than the bandgap of the material.
The physics of photo-voltaic (P-V) cells is based on the generation of current by the separation of mobile charges, electrons and holes, in semiconductor materials. A generic silicon cell is depicted in Figure 1. Doping with a small percentage of the appropriate material creates either an excess or a deficiency of electrons (hole), depending on the particular dopant atoms. When the two doped materials are joined, a P-N junction is formed. An electric field develops across the P-N junction by the diffusion of electrons and holes in opposite directions. When the energy of the light incident on a semiconductor P-N junction exceeds the energy with which the outer electrons in the valence band are bound, electrons-hole pairs are created and mobilized by the electric field. In silicon, that energy, known as the bandgap, corresponds to wavelengths shorter than ~1000 nm. Electrons diffuse to the N-type layer, and holes to the P-type layer. The mobile charges are collected by the top and bottom electrodes, and the external circuit returns the electrons and holes to be recombined, thus generating an external current that produces power.
Figure 1. Generic Silicon solar cell construction illustrating the separation of charges at the P-N junction.
The upper electrode is a fine gold wire grid produced by photolithography or by screen printing silver that is then fired. Contact is made to the grid, and to construct a panel many cells are joined in parallel and serial sets to produce the desired panel voltage and current (power). The small loss imposed by the metal grid is tolerated in the interest of economy for terrestrial cells. Developments are underway to replace it by a transparent conductor layer such as indium-tin-oxide (ITO) and other transparent conducting oxides (TCO) for higher performance cells such as those used in space power. The metal electrode at the bottom is often a reflective deposition that returns some photons for a second pass.
The practical physics of P-V cells is a bit more complicated than this simple model implies. Semi-conductor material properties and process issues team to inhibit the full current-generating capability. Defects in the bulk material and surface encourage recombination outside of the electrical circuit, thus reducing the current available for output power. Morphologies other than single crystal tend to have more internal recombination sites. Surface chemical conditions also affects recombination. Passivation layers are often required to minimize surface shunting.
The high refractive indices of semiconductor surfaces produce a loss of photons by reflection values between 30% and 20%, depending on the top layer semiconductor composition. Reflection loss is responsible for the reduction of external quantum efficiency. Increased power output is achieved with an anti-reflection (AR) surface. A layer of Si3N4 grown reactively on the silicon serves as an AR and passivation coating. It is as effective as an AR over a limited spectral range. The traditional AR coating for silicon cells is a two-layer TiO2 / Al2O3 design. Coverage is adequate for silicon cells responding to ~1000 nm, but not for the wider spectral band response of multi-junction designs. The AR design required for efficient transmission over ~350 nm to ~1800 nm has three or more layers using Ta2O5 and SiO2.
Click to read the full technical paper Evolution of Photo-Voltaic Solar Cell Technology.