Page 254 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 254
Chapter 8
magnitudes in the zinc oxide layers in the H O/air and H O/CO /N samples. SIMS also
2
2
2
2
indicated that the sodium profiles did not change greatly, while it has been shown
to migrate under illumination or bias, leading to degradation of the solar cells [10,11].
Therefore, it is unlikely that similar reactions are occurring under these conditions.
Furthermore, these samples also showed a change in colour: initially, they were blue/
black, while after degradation, they showed a reddish colour.
When the dissolution of ZnO:Al is taken into account, the changes in series resistance
and short circuit current density are not surprising. When part of the zinc oxide has
disappeared, the grains are barely connected and the resistivity of the layer increases.
Therefore, the transport of the current through this layer becomes very hard, leading
to an increase in series resistance and a decrease in short circuit current density.
Additionally, the new layer might also play a role in these changes.
For the samples exposed to unpurged HO, H O/N and H O/O , some minor gaps in
2
2
2
2
2
the ZnO:Al were also observed. The gaps started at the ZnO:Al/air interface, while
some hydroxide was also present in the ZnO:Al layers. These effects seem to have
led to a small increase in series resistance over time, probably also by an increased
resistivity of the ZnO:Al layer. For the H O/N sample, a decrease in the light and dark
2
2
shunt resistance was also observed. No effects were observed for the sample exposed
to air.
Since the degradation experiments in aqueous solutions took place in darkness, it is
not expected that a voltage was formed across the solar cell. Therefore, the reason
for the degradation can likely be found in the diffusion of the atmospheric species,
followed by chemical or physical reactions with the materials from the solar cells.
8.4.1 Grain boundary dissolution
The impact of the various combinations of water and atmospheric gases was comparable
to the observations obtained for ZnO:Al layers on borosilicate glass [3]. In this reference,
it was shown that ZnO:Al layers rapidly degraded in the presence of a mixture of HO
2
and CO . Individually, CO did not impact the degradation at all during the tested period,
2
2
while the individual impact of HO was small. However, when CO was also present,
2
2
the concentration of OH increases greatly in the ZnO:Al bulk and even more at the air/
ZnO:Al and the ZnO:Al/glass interfaces. Carbon based species were then also present,
indicating that Zn (OH) (CO ) was formed at the grain boundaries.
5
3 2
6
Detailed cross-section pictures obtained by Helium Ion Microscope showed the
formation of voids at the ZnO:Al/glass interface, indicating the local dissolution of
ZnO:Al, probably due to a reaction of elements from the glass, like calcium or silicon,
with water and CO leading to a pH change. Similar to the current experiment, the
2
252
magnitudes in the zinc oxide layers in the H O/air and H O/CO /N samples. SIMS also
2
2
2
2
indicated that the sodium profiles did not change greatly, while it has been shown
to migrate under illumination or bias, leading to degradation of the solar cells [10,11].
Therefore, it is unlikely that similar reactions are occurring under these conditions.
Furthermore, these samples also showed a change in colour: initially, they were blue/
black, while after degradation, they showed a reddish colour.
When the dissolution of ZnO:Al is taken into account, the changes in series resistance
and short circuit current density are not surprising. When part of the zinc oxide has
disappeared, the grains are barely connected and the resistivity of the layer increases.
Therefore, the transport of the current through this layer becomes very hard, leading
to an increase in series resistance and a decrease in short circuit current density.
Additionally, the new layer might also play a role in these changes.
For the samples exposed to unpurged HO, H O/N and H O/O , some minor gaps in
2
2
2
2
2
the ZnO:Al were also observed. The gaps started at the ZnO:Al/air interface, while
some hydroxide was also present in the ZnO:Al layers. These effects seem to have
led to a small increase in series resistance over time, probably also by an increased
resistivity of the ZnO:Al layer. For the H O/N sample, a decrease in the light and dark
2
2
shunt resistance was also observed. No effects were observed for the sample exposed
to air.
Since the degradation experiments in aqueous solutions took place in darkness, it is
not expected that a voltage was formed across the solar cell. Therefore, the reason
for the degradation can likely be found in the diffusion of the atmospheric species,
followed by chemical or physical reactions with the materials from the solar cells.
8.4.1 Grain boundary dissolution
The impact of the various combinations of water and atmospheric gases was comparable
to the observations obtained for ZnO:Al layers on borosilicate glass [3]. In this reference,
it was shown that ZnO:Al layers rapidly degraded in the presence of a mixture of HO
2
and CO . Individually, CO did not impact the degradation at all during the tested period,
2
2
while the individual impact of HO was small. However, when CO was also present,
2
2
the concentration of OH increases greatly in the ZnO:Al bulk and even more at the air/
ZnO:Al and the ZnO:Al/glass interfaces. Carbon based species were then also present,
indicating that Zn (OH) (CO ) was formed at the grain boundaries.
5
3 2
6
Detailed cross-section pictures obtained by Helium Ion Microscope showed the
formation of voids at the ZnO:Al/glass interface, indicating the local dissolution of
ZnO:Al, probably due to a reaction of elements from the glass, like calcium or silicon,
with water and CO leading to a pH change. Similar to the current experiment, the
2
252
