Page 258 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 258
Chapter 8
materials which are porous for water. This could greatly change the demands on the
barrier material.
Additionally, knowledge about stability of ZnO:Al in low-pH environments can also
help in the choice of the encapsulation material in a CIGS module. One of the standard
encapsulation materials, Ethylene-Vinyl Acetate (EVA), is known to react with water.
This can result in the formation of acetic acid, which can buffer aqueous solutions to a
pH of 4.76 [6]. This would imply that a pH of below 6 can certainly be expected in CIGS
modules. An effect of EVA has been presented by Lee et al. [7], who reported chemical
reactions occurring in the presence of ZnO:Al, EVA and water, leading to an increase
in the ZnO:Al resistivity. This confirms the conclusion of Kempe et al. [6] that careful
consideration should be made when EVA is considered as encapsulant in CIGS mod-
ules. Similar questions should naturally be posed for other encapsulation materials.
8.5 Conclusions
In chapter 6, it has been shown that the combined impact of HO and CO leads to
2
2
the degradation of ZnO:Al layers in glass. In this chapter, similar observations have
been made for complete CIGS solar cells. These solar cells degraded very fast when
exposed to H O purged with CO and N as well as to H O purged with air, while they
2
2
2
2
degraded only slowly in unpurged H O and H O purged with N and O . The exposure
2
2
2
2
of samples to H O with large concentrations of CO led to the dissolution of part of the
2
2
ZnO:Al layers. This was especially visible due to the formation of gaps in the zinc oxide
near the CdS/i-ZnO region, alongside with disappearance of the zinc oxide in positi-
ons near the grain boundaries. Additionally, SEM pictures indicated the possible for-
mation of a new intermediate layer in the CdS/i-ZnO region. These material changes
led to an increased series resistance, likely related to an increase in resistivity of the
ZnO:Al layer. Furthermore, the short-circuit current of these samples decreased very
rapidly, leading to the complete loss of conversion efficiency. It is proposed that the
dissolution of the ZnO:Al was caused by a local change in pH, since overall pH levels
were outside the range at which zinc oxide normally dissolves in water.
Additionally, for all samples, a change in shunt resistance was observed. This effect
was especially strong for samples exposed to H O purged with N , possible due to the
2
2
formation of conductive phases in the presence of only these atmospheric species.
256
materials which are porous for water. This could greatly change the demands on the
barrier material.
Additionally, knowledge about stability of ZnO:Al in low-pH environments can also
help in the choice of the encapsulation material in a CIGS module. One of the standard
encapsulation materials, Ethylene-Vinyl Acetate (EVA), is known to react with water.
This can result in the formation of acetic acid, which can buffer aqueous solutions to a
pH of 4.76 [6]. This would imply that a pH of below 6 can certainly be expected in CIGS
modules. An effect of EVA has been presented by Lee et al. [7], who reported chemical
reactions occurring in the presence of ZnO:Al, EVA and water, leading to an increase
in the ZnO:Al resistivity. This confirms the conclusion of Kempe et al. [6] that careful
consideration should be made when EVA is considered as encapsulant in CIGS mod-
ules. Similar questions should naturally be posed for other encapsulation materials.
8.5 Conclusions
In chapter 6, it has been shown that the combined impact of HO and CO leads to
2
2
the degradation of ZnO:Al layers in glass. In this chapter, similar observations have
been made for complete CIGS solar cells. These solar cells degraded very fast when
exposed to H O purged with CO and N as well as to H O purged with air, while they
2
2
2
2
degraded only slowly in unpurged H O and H O purged with N and O . The exposure
2
2
2
2
of samples to H O with large concentrations of CO led to the dissolution of part of the
2
2
ZnO:Al layers. This was especially visible due to the formation of gaps in the zinc oxide
near the CdS/i-ZnO region, alongside with disappearance of the zinc oxide in positi-
ons near the grain boundaries. Additionally, SEM pictures indicated the possible for-
mation of a new intermediate layer in the CdS/i-ZnO region. These material changes
led to an increased series resistance, likely related to an increase in resistivity of the
ZnO:Al layer. Furthermore, the short-circuit current of these samples decreased very
rapidly, leading to the complete loss of conversion efficiency. It is proposed that the
dissolution of the ZnO:Al was caused by a local change in pH, since overall pH levels
were outside the range at which zinc oxide normally dissolves in water.
Additionally, for all samples, a change in shunt resistance was observed. This effect
was especially strong for samples exposed to H O purged with N , possible due to the
2
2
formation of conductive phases in the presence of only these atmospheric species.
256
