Page 38 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 38
Chapter 2
samples in water and dry heat did not change in thickness. The sample in the RT water
bath showed a slight increase in resistivity, while the resistivity of the other samples
remained largely unchanged.
o
Reference [18] showed that molybdenum also degraded under 60C/60% RH con-
ditions, but that the change in the properties of the molybdenum was much slower
o
than under 85 C/85% RH. A rough estimation indicated that the degradation occurs
approximately 20 times faster under the more severe conditions.
Pern and Noufi [30] showed that encapsulation also influenced the stability of molyb -
denum. When molybdenum was covered with a moisture permeable back sheet, no
colour change or significant resistance change was observed even after 1250 hours
damp heat exposure, although a number of small ‘dots’ appeared on the surface after
500 hours. This was surprising, since bare molybdenum on soda glass already degraded
after 50 to 100 hours. The authors concluded that the presence of a back sheet either
changed the reaction mechanism or that the reaction rate was largely reduced in the
encapsulated test structure, in which the moisture vapour had to diffuse through the
back sheet.
The relationship between long-time air exposure and short-term accelerated lifetime
testing conditions is hard to identify. In order to make this comparison, Schmid et al. [31]
compared the impact of several days of ambient air exposure and one hour 200 o C heat
treatment on molybdenum layers. The ambient air exposure combined with surface
cleaning (several minutes at 400 C in vacuum) led to the formation of MoO as well as
o
2
o
some higher oxides, like MoO . The 200 C heat treatment, however, resulted in the ex-
3
clusive formation of MoO , while sodium, which might have migrated from the substrate,
3
was also found in these layers. Therefore, it is suggested that the oxidation state of the
molybdenum might influence the sodium migration through the molybdenum layer.
In a later stage, CuInSe was deposited on the molybdenum film exposed to ambient
2
air, which led to a decrease in the oxygen content due to selenium incorporation.
This was due to the reduction of MoO to MoO and the formation of either MoSe or
3
2
2
Mo-O-Se. However, the oxygen was not completely replaced by selenium, even after
longer selenisation times.
By further studying of the CuInSe sample, it was observed that the MoO presence
2
2
led to the formation of a Schottky-type barrier of about 0.3 eV at the CIS-MoO/Mo
2
interface (Figure 2.3). It was suggested that this number is deviating from the 0.6 eV
which was found for an ideal single crystal CIS/Mo system, due to differences in the
interface chemistry, like the presence of oxygen.
36
samples in water and dry heat did not change in thickness. The sample in the RT water
bath showed a slight increase in resistivity, while the resistivity of the other samples
remained largely unchanged.
o
Reference [18] showed that molybdenum also degraded under 60C/60% RH con-
ditions, but that the change in the properties of the molybdenum was much slower
o
than under 85 C/85% RH. A rough estimation indicated that the degradation occurs
approximately 20 times faster under the more severe conditions.
Pern and Noufi [30] showed that encapsulation also influenced the stability of molyb -
denum. When molybdenum was covered with a moisture permeable back sheet, no
colour change or significant resistance change was observed even after 1250 hours
damp heat exposure, although a number of small ‘dots’ appeared on the surface after
500 hours. This was surprising, since bare molybdenum on soda glass already degraded
after 50 to 100 hours. The authors concluded that the presence of a back sheet either
changed the reaction mechanism or that the reaction rate was largely reduced in the
encapsulated test structure, in which the moisture vapour had to diffuse through the
back sheet.
The relationship between long-time air exposure and short-term accelerated lifetime
testing conditions is hard to identify. In order to make this comparison, Schmid et al. [31]
compared the impact of several days of ambient air exposure and one hour 200 o C heat
treatment on molybdenum layers. The ambient air exposure combined with surface
cleaning (several minutes at 400 C in vacuum) led to the formation of MoO as well as
o
2
o
some higher oxides, like MoO . The 200 C heat treatment, however, resulted in the ex-
3
clusive formation of MoO , while sodium, which might have migrated from the substrate,
3
was also found in these layers. Therefore, it is suggested that the oxidation state of the
molybdenum might influence the sodium migration through the molybdenum layer.
In a later stage, CuInSe was deposited on the molybdenum film exposed to ambient
2
air, which led to a decrease in the oxygen content due to selenium incorporation.
This was due to the reduction of MoO to MoO and the formation of either MoSe or
3
2
2
Mo-O-Se. However, the oxygen was not completely replaced by selenium, even after
longer selenisation times.
By further studying of the CuInSe sample, it was observed that the MoO presence
2
2
led to the formation of a Schottky-type barrier of about 0.3 eV at the CIS-MoO/Mo
2
interface (Figure 2.3). It was suggested that this number is deviating from the 0.6 eV
which was found for an ideal single crystal CIS/Mo system, due to differences in the
interface chemistry, like the presence of oxygen.
36
