Surface-barrier Solar Cells
Surface-barrier solar cells all use only one semiconductor, in one doping type. They all have an electrostatic-field barrier, which begins at the semiconductor surface, as the principal source of photovoltaic action. The built-in electrostatic field can extend across the whole semiconductor, analogous to a p-i-n structure, or, at the other extreme, it can be confined to the semiconductor's near-surface region. There are basically two types of surface-barrier cells; they can be distinguished by the materials system used to form the semiconductor electrostatic field barrier: one type is the all-solid-state cell and the other is the electrolyte-solid-state cell. The all-solid-state cell has two functional versions: a metal-semiconductor （M-S） configuration and a metal-intermediate layer-semiconductor （M-I-S） configuration. The electrolyte-solid-state cell uses a liquid-semiconductor configuration. The solid-state cells use the difference between the electrochemical potential of a metal and that of the semiconductor （typically measured for both materials by their workfunctions） to set up the surface electrostatic field in the semicondutor. The electrolyte-based cells use the
difference between the electrochemical potential of an electrolyte （typically measured for electrolytes by the redox potential） and that of the semiconductor to set up the surface electrostatic field. The M-S and M-I-S devices are often referred to as Schottky-barrier （SB） cells and the electrolyte-semiconductor devices are referred to as electrochemical photovoltaic cells （EPC）. The version of the metal-semiconductor configuration for which the field reaches across the whole semiconductor is sometimes referred to as an M-I-M cell. Here the I notation is again used but this time to denote a low doped （intrinsic） absorber. The two basic types of surface-barrier solar cells are shown in Figure 6.1. The electrolyte-semiconductor cell can be further broken down into two subtypes, as shown in Figure 6.2: （1） the regenerative cell, which has the oxidation （hole-capture from the semiconductor surface） of the redox couple completed by reduction at the anode without any net chemical change; and （2） the photo-cleaving （or photosynthesis） cell, which has the hole-capture （oxidation） at the semiconductor produce the evolution of one product, while the reduction at the anode produces the evolution of another product,as seen in the figure. The particular photosynthesis example given in Figure 6.2b shows photolysis, the decomposition of water to produce oxygen and hydrogen which was first demonstrated in 1972.
The lineage of surface-barrier solar cells can be traced back to the electrolyte-solid structures used by Becquerel2 in 1839 in the first reported studies of photovoltaic-type behavior. Investigation of the solid-state, surface-barrier cell began with the Cu-Cu2o structure, which was shown to be photosensitive by Hallwachs3 in 1904, and which was developed into a photovoltaic device by 1927.4 By the 1930s, Cu-Cu2o metal-semiconductor, surface-barrier devices were in production and were being used for applications such as photometry and light control.
During the early 1950s, surface-barrier photovoltaic structures were quickly eclipsed by the emerging p-n homojunction solar cell technology, just as Schottky-barrier diodes were quickly surpassed by p-n junction diodes during this time. By 1970-1972, Schottky-barrier solar cells had evolved to only p < 6% for single-crystal Si M-S devices and to only p < 9% for single-crystal GaAs M-S devices （terrestrial conditions）.5 The electrolyte-semiconductor （EPC） solar cells of the period were unstable devices with efficiencies that had yet to reach 1% under terrestrial sunlight.