Page 232 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 232
Chapter 7
the ZnO:Al layer. This effect is stronger within the spots (dotted purple line after 165
hours) than for regions without spots. The sodium migration continued in the same
way after 365 hours (not depicted). Migration of potassium also occurred, but was
very limited and mainly led to an increase in the top of the ZnO:Al layer.
For the alkali-poor sample, sodium migrated from the Mo/CIGS interface to the pn-
junction region. Furthermore, the potassium and sodium profiles were relatively
stable.
7.3.7 Reproducibility
Experiments similar to this one have been executed multiple times before on similar
samples, as is for example shown in reference [4]. These experiments have shown
that the decrease of the solar cell parameters is much slower for samples with a low
alkali content. Since several of the earlier samples had spots with a larger size, the
SIMS measurements could be executed more precisely. In earlier SIMS profiles, the
differences between positions on and next to the spots was therefore even stronger.
7.4 Discussion
The in-situ IV measurements indicated that alkali-poor samples were stable under
damp heat-illumination, while the performance of the alkali-rich samples decreased
quickly, especially during the first 100 hours. This degradation was accompanied
by the appearance of surface spots. SIMS and SEM-EDX measurements in alkali-rich
samples indicated that especially sodium and to a lesser account potassium migrated
from the CIGS/Mo interface and the CIGS bulk towards the pn-junction and ZnO:Al
layer. The sodium migration is stronger on the spots than on spot-free areas. These
measurements also indicated that the alkali-rich samples showed a complete change
of the sample structureand composition after 778 hours.
The first factor playing a role in the degradation process of the CIGS solar cells is an
increase in hydroxide content, which is mainly observed in the ZnO:Al layers in both
the alkali-rich and poor samples. This hydroxide increase is likely caused by water and
possible CO ingression, which leads to a slow but steady increase of the series resistance
2
for all samples, for example by the formation of hydrozincite in the grain boundaries (see
chapter 6) [13]. However, this increase of the series resistance only has a small impact on
the performance change and does not explain the differences in degradation behaviour
between the samples. We propose that the migration of sodium and possibly potassium
has the largest impact on the performance change and leads to the formation of surface
spots as well as the decrease of the solar cell parameters.
230
the ZnO:Al layer. This effect is stronger within the spots (dotted purple line after 165
hours) than for regions without spots. The sodium migration continued in the same
way after 365 hours (not depicted). Migration of potassium also occurred, but was
very limited and mainly led to an increase in the top of the ZnO:Al layer.
For the alkali-poor sample, sodium migrated from the Mo/CIGS interface to the pn-
junction region. Furthermore, the potassium and sodium profiles were relatively
stable.
7.3.7 Reproducibility
Experiments similar to this one have been executed multiple times before on similar
samples, as is for example shown in reference [4]. These experiments have shown
that the decrease of the solar cell parameters is much slower for samples with a low
alkali content. Since several of the earlier samples had spots with a larger size, the
SIMS measurements could be executed more precisely. In earlier SIMS profiles, the
differences between positions on and next to the spots was therefore even stronger.
7.4 Discussion
The in-situ IV measurements indicated that alkali-poor samples were stable under
damp heat-illumination, while the performance of the alkali-rich samples decreased
quickly, especially during the first 100 hours. This degradation was accompanied
by the appearance of surface spots. SIMS and SEM-EDX measurements in alkali-rich
samples indicated that especially sodium and to a lesser account potassium migrated
from the CIGS/Mo interface and the CIGS bulk towards the pn-junction and ZnO:Al
layer. The sodium migration is stronger on the spots than on spot-free areas. These
measurements also indicated that the alkali-rich samples showed a complete change
of the sample structureand composition after 778 hours.
The first factor playing a role in the degradation process of the CIGS solar cells is an
increase in hydroxide content, which is mainly observed in the ZnO:Al layers in both
the alkali-rich and poor samples. This hydroxide increase is likely caused by water and
possible CO ingression, which leads to a slow but steady increase of the series resistance
2
for all samples, for example by the formation of hydrozincite in the grain boundaries (see
chapter 6) [13]. However, this increase of the series resistance only has a small impact on
the performance change and does not explain the differences in degradation behaviour
between the samples. We propose that the migration of sodium and possibly potassium
has the largest impact on the performance change and leads to the formation of surface
spots as well as the decrease of the solar cell parameters.
230
