Page 64 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 64
Chapter 2
also present in the encapsulated modules. Dry heat treatment, on the other hand, did
not lead to discolouration and had only minor impact on sheet resistance and module
efficiency. The observed fill factor loss and discolouration were therefore concluded
to be caused by a reaction between H O, Ethylene-Vinyl Acetate (EVA) and ZnO:Al.
2
A German consortium [81,90] studied the impact of damp heat on different types of
ZnO:Al layers on nominally identical CIGS/CdS stacks. The ZnO:Al sheet resistance
was measured before, during and after exposure to a damp heat test and increased
for all samples. A clear correlation between this increase in series resistance and the
decrease of the fill factor of the module test structures was observed. It was also ob-
served that conditions which favour the growth of dense films with a good step cov-
erage, i.e. the use of RF sputtering at a low pressure and high temperature as well as
the use of a moving substrate, led to better stability.
Klaer et al. [91] used a sample design which allowed to determine the sheet resistance
of zinc oxide on top of different types of CIGS. They noticed that the degradation of
ZnO:Al when exposed to damp heat conditions can be very different depending on
the underlying layers: while the zinc oxide sheet resistance was more or less stable on
co-evapourated Cu(In,Ga)Se absorbers, (increase R =10 mΩ/ /h) during 540 hours
2
of damp heat, the sheet resistance increased very fast on sequentially deposited Cu-
InS (increase R =148 mΩ/ /h). This difference is probably caused by the very high
2
roughness of the CuInS, compared to the smooth Cu(In,Ga)Se. This effect was de-
2
2
scribed in reference [62] and in chapter 2.3.4.1.1 .
Furthermore, the impact of the ZnO:Al series resistance was shown by the comparison
of the series resistances in similar CuInS cells with a different cell length (3 mm vs 8
2
mm). This difference in series resistance increase has a large influence on the efficien-
2
cy of the solar cells: a difference of 2% absolute between the cells of 0.15 cm and 0.40
cm was observed.
2
The TCO itself can also function as a water barrier. Thompson et al. [22] discovered that a
layer of i-ZnO on top of the CIGS and its surroundings can prevent water ingression. Ex -
periments showed that 50 nm nanocrystalline i-ZnO and 150 nm amorphous ITO layers
2
-3
-3
had a water vapour transmission rate (WVTR) of 1.3x10 and 0.7x10 g/day/m respec-
tively. Therefore, a thin large area i-ZnO layer can function as a water barrier, if it is not
disrupted by scribes. This knowledge is especially relevant for thin film PV manufactur -
ers using traditional modules based on ‘tabbed cells, who can greatly save on the costs
of the additional water barriers necessary in a CIGS module in this way.
A similar effect was obtained by Selin-Tosun et al. [60], who deposited SnO 2 on top of un -
encapsulated CIGS cells with a ZnO/ITO TCO structure. This greatly decreased the water
62
also present in the encapsulated modules. Dry heat treatment, on the other hand, did
not lead to discolouration and had only minor impact on sheet resistance and module
efficiency. The observed fill factor loss and discolouration were therefore concluded
to be caused by a reaction between H O, Ethylene-Vinyl Acetate (EVA) and ZnO:Al.
2
A German consortium [81,90] studied the impact of damp heat on different types of
ZnO:Al layers on nominally identical CIGS/CdS stacks. The ZnO:Al sheet resistance
was measured before, during and after exposure to a damp heat test and increased
for all samples. A clear correlation between this increase in series resistance and the
decrease of the fill factor of the module test structures was observed. It was also ob-
served that conditions which favour the growth of dense films with a good step cov-
erage, i.e. the use of RF sputtering at a low pressure and high temperature as well as
the use of a moving substrate, led to better stability.
Klaer et al. [91] used a sample design which allowed to determine the sheet resistance
of zinc oxide on top of different types of CIGS. They noticed that the degradation of
ZnO:Al when exposed to damp heat conditions can be very different depending on
the underlying layers: while the zinc oxide sheet resistance was more or less stable on
co-evapourated Cu(In,Ga)Se absorbers, (increase R =10 mΩ/ /h) during 540 hours
2
of damp heat, the sheet resistance increased very fast on sequentially deposited Cu-
InS (increase R =148 mΩ/ /h). This difference is probably caused by the very high
2
roughness of the CuInS, compared to the smooth Cu(In,Ga)Se. This effect was de-
2
2
scribed in reference [62] and in chapter 2.3.4.1.1 .
Furthermore, the impact of the ZnO:Al series resistance was shown by the comparison
of the series resistances in similar CuInS cells with a different cell length (3 mm vs 8
2
mm). This difference in series resistance increase has a large influence on the efficien-
2
cy of the solar cells: a difference of 2% absolute between the cells of 0.15 cm and 0.40
cm was observed.
2
The TCO itself can also function as a water barrier. Thompson et al. [22] discovered that a
layer of i-ZnO on top of the CIGS and its surroundings can prevent water ingression. Ex -
periments showed that 50 nm nanocrystalline i-ZnO and 150 nm amorphous ITO layers
2
-3
-3
had a water vapour transmission rate (WVTR) of 1.3x10 and 0.7x10 g/day/m respec-
tively. Therefore, a thin large area i-ZnO layer can function as a water barrier, if it is not
disrupted by scribes. This knowledge is especially relevant for thin film PV manufactur -
ers using traditional modules based on ‘tabbed cells, who can greatly save on the costs
of the additional water barriers necessary in a CIGS module in this way.
A similar effect was obtained by Selin-Tosun et al. [60], who deposited SnO 2 on top of un -
encapsulated CIGS cells with a ZnO/ITO TCO structure. This greatly decreased the water
62
