144
A. Shah
the hydrogen content is higher, then the “excess” hydrogen atoms are no more
linked to silicon atoms, but wander about freely within the amorphous layer,
creating havoc and ultimately leading to an increase in SWE.
(f) The inclusion of impurities like oxygen within the amorphous layer, leads to an
increased SWE.
3
3. Density of states within the bandgap of amorphous silicon layers
Because of the random or “chaotic” structure of amorphous silicon, amorphous
silicon does not have a real bandgap, like crystalline silicon. In comparison with
crystalline silicon, a-Si:H layers have:
(a) An “equivalent bandgap”, which is filled with gap states, acting (partly) as
recombination centres. This is shown schematically in Fig. 6.6.
(b) A higher “equivalent bandgap”, higher than the “real bandgap” of crystalline
silicon, i.e. approximately 1.75 eV instead of 1.12 eV
Figure 6.6 indicates the density of states N(E) within the “equivalent bandgap” (called
in the Figure “mobility gap”) of typical amorphous silicon layers. The “midgap
states” act as recombination centres. Their density is increased by the SWE. (Fig. 6.7).
4. Optical properties of amorphous silicon layers
The optical properties of a-Si:H layers are very different from those of c-Si. From a
practical point of view, the main differences are:
(1) Absorption starts at a higher value of photon energy for a-Si:H → i.e. at a shorter
wavelength λ of light: at λ ≈ 700 nm for a-Si:H as compared to λ ≈ 1100 nm
for c-Si
3 There is a long story behind this simple statement—when work started at IMT Neuchâtel, after
1985, on amorphous silicon layers and solar cells, there was the “justified hope” that by suppressing
(reducing) the oxygen content within the layers we were depositing, we would be able to totally
remove the SWE effect. We therefore invested heavily in so-called “gas purifiers”, so as to be able
to produce a-Si:H layers with very low oxygen content. To our misfortune it turned out that these
layers had just the same intensity of SWE as layers containing more oxygen. It was only above a
certain threshold that oxygen led to an enhancement of SWE. However, when we started to deposit
microcrystalline silicon layers, with our equipment, the gas purifiers were absolutely decisive: We
at IMT Neuchâtel became the first laboratory able to produce microcrystalline silicon solar cells
with conversion efficiencies well above 5%. Thanks to this discovery, the Author of this chapter
received in 2007 the Becquerel Prize. More importantly the Author of this Chapter would wish
to communicate to the Readers the following lessons he learnt:
1. Never underestimate instability effects. They generally turn out to be unavoidable. no matter
what tricks one tries to do. This should be remembered in the context of the present hype
regarding perovskite solar cells. These cells are at present unstable, and may remain so, no
matter what tricks researchers try to do.
2. Research often leads to results you are totally unable to predict—our layers with low oxygen
content were of no use for amorphous silicon solar cells, but they were a decisive asset, when
we started our work on microcrystalline solar cells. The present heavy research investment in
perovskite solar cells will probably also lead to some quite unexpected results.
A. Shah
the hydrogen content is higher, then the “excess” hydrogen atoms are no more
linked to silicon atoms, but wander about freely within the amorphous layer,
creating havoc and ultimately leading to an increase in SWE.
(f) The inclusion of impurities like oxygen within the amorphous layer, leads to an
increased SWE.
3
3. Density of states within the bandgap of amorphous silicon layers
Because of the random or “chaotic” structure of amorphous silicon, amorphous
silicon does not have a real bandgap, like crystalline silicon. In comparison with
crystalline silicon, a-Si:H layers have:
(a) An “equivalent bandgap”, which is filled with gap states, acting (partly) as
recombination centres. This is shown schematically in Fig. 6.6.
(b) A higher “equivalent bandgap”, higher than the “real bandgap” of crystalline
silicon, i.e. approximately 1.75 eV instead of 1.12 eV
Figure 6.6 indicates the density of states N(E) within the “equivalent bandgap” (called
in the Figure “mobility gap”) of typical amorphous silicon layers. The “midgap
states” act as recombination centres. Their density is increased by the SWE. (Fig. 6.7).
4. Optical properties of amorphous silicon layers
The optical properties of a-Si:H layers are very different from those of c-Si. From a
practical point of view, the main differences are:
(1) Absorption starts at a higher value of photon energy for a-Si:H → i.e. at a shorter
wavelength λ of light: at λ ≈ 700 nm for a-Si:H as compared to λ ≈ 1100 nm
for c-Si
3 There is a long story behind this simple statement—when work started at IMT Neuchâtel, after
1985, on amorphous silicon layers and solar cells, there was the “justified hope” that by suppressing
(reducing) the oxygen content within the layers we were depositing, we would be able to totally
remove the SWE effect. We therefore invested heavily in so-called “gas purifiers”, so as to be able
to produce a-Si:H layers with very low oxygen content. To our misfortune it turned out that these
layers had just the same intensity of SWE as layers containing more oxygen. It was only above a
certain threshold that oxygen led to an enhancement of SWE. However, when we started to deposit
microcrystalline silicon layers, with our equipment, the gas purifiers were absolutely decisive: We
at IMT Neuchâtel became the first laboratory able to produce microcrystalline silicon solar cells
with conversion efficiencies well above 5%. Thanks to this discovery, the Author of this chapter
received in 2007 the Becquerel Prize. More importantly the Author of this Chapter would wish
to communicate to the Readers the following lessons he learnt:
1. Never underestimate instability effects. They generally turn out to be unavoidable. no matter
what tricks one tries to do. This should be remembered in the context of the present hype
regarding perovskite solar cells. These cells are at present unstable, and may remain so, no
matter what tricks researchers try to do.
2. Research often leads to results you are totally unable to predict—our layers with low oxygen
content were of no use for amorphous silicon solar cells, but they were a decisive asset, when
we started our work on microcrystalline solar cells. The present heavy research investment in
perovskite solar cells will probably also lead to some quite unexpected results.