Photovoltaic Power Generation - Renewable Energy: Technology, Economics and Environment - page 251

scale reached efficiencies between 16 and 18 %; that is approximately 75 % of the
theoretical efficiency of the crystalline silicon technology. Both materials can
physically be deposited onto glass at a temperature of 600 °C. Since both materi-
als are direct semiconductors, photovoltaically active layers of a few µm thickness
are sufficient to absorb all photons of the solar spectrum with energy superior to
the energy gap E g of the respective absorber material. The energy gap of cadmium
telluride (CdTe) amounts to approximately 1.45 eV and that of copper indium di-
selenium (CuInSe 2 ) to about 1.04 eV. However, for the latest generation of chal-
copyrit solar cells instead of pure CuInSe 2 , Cu(In,Ga)Se 2 alloy with a gallium
share of 20 to 30 % in relation to the total indium (In) and gallium (Ga) content is
used. With an energy gap of 1.12 to 1.2 eV the alloy is closer to the theoretically
achievable optimum of the efficiency (see Fig. 6.8; see /6-21/, /6-22/).
Thin films of good electronic quality of cadmium telluride (CdTe) and copper
indium di-selenium (CuInSe 2 ) can only be manufactured p-doped. Thus, a second
n-doped material is required for solar cell manufacturing, which can be combined
with the first material to form a semiconductor p-n-hetero-junction. In both cases,
n-doped cadmium sulfide (CdS) is used. Such semiconductor hetero-structures
present the "slight" disadvantage that at the boundary surface between both mate-
rials, an increased recombination of photogenerated charge carriers occurs. How-
ever, this disadvantage is compensated by the advantage that the top layer may be
designed as a "window", thanks to the possibility to choose a semiconductor with
a high energy gap (such as CdS: E g = 2.4 eV); hence, the described top layer only
absorbs a very limited share of the solar spectrum, which is then lost for the
photocurrent. After the transmission of the remaining irradiation through this
window layer the major share of the incident radiation is absorbed very close to
the p-n-junction - thus at the point of the maximum electrical field strength. The
resulting separation of the photogenerated charge carriers is thus highly efficient.
Fig. 6.11, left, shows the layer sequence of a CdS/CdTe hetero-structure solar
cell. This cell technology is a superstrate structure; i.e. the transparent front elec-
trode exposed to the sunlight and made of indium stannous oxide (ITO) is usually
applied first by means of sputtering. Subsequently, cadmium sulfide (CdS) is de-
posited as window or buffer layer followed by the actual photovoltaically active
absorber layer consisting of cadmium telluride (CdTe). Usually, both layers (i.e.
the window layer with a thickness between 0.1 and 0.2 µm and the absorber layer
with a thickness of approximately 3 µm) are superimposed using the same tech-
nology (e.g. by the sublimation-condensation method or silk-screen process print-
ing). In order to obtain layers of a photovoltaically sufficient quality, an activation
step based on a temperature treatment in the presence of cadmium chloride
(CdCl 2 ) needs to be performed after deposition. Cell manufacturing is accom-
plished by depositing a metal rear electrode made of graphite, copper (Cu) or a
mixture of both. On a laboratory scale, cadmium telluride (CdTe) solar cells have
reached peak efficiencies of nearly 17 % when applied to small surfaces. Over the
last few years, several pilot production lines for large-surface CdTe modules of a