Page 233 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 233
The impact of alkali elements
A negative effect of sodium migration on CIGS solar cells was observed by Fjällström
o
et al. [3]. After the exposure of a 50V bias combined with 85C on CIGS solar cells,
sodium migrated from the glass substrate to the pn-junction, which resulted in a
rapid decrease of the conversion efficiency. In this case, the relatively high voltage bias
was apparently able to drive sodium from the glass, leading to a potential induced
degradation-(PID-)like process, which was also reported by Colli [14].In our work,
it is not yet possible to distinguish between the impact of sodium and potassium
migration. Because of the limited potassium migration and the results of Fjällström
[3], who observed that sodium migration led to degradation, it will be assumed that
sodium migration is the most important factor.
7.4.1 Degradation model
In our case the exposure conditions are different since during the damp heat
treatment, no voltage bias is applied. The majority of sodium atoms in CIGS are
located at the grain boundaries [15], most probably bonded to the adjacent grains
+
dangling bonds. The presence of water likely enables the release of Na ions through
the hydrolysis of the bonds between CIGS and sodium. The soda lime glass can also
serve as a source of the sodium, but it should be noted that, as far as can be judged
from the SIMS spectra, the sodium content in the complete stack does not increase
greatly. Therefore, it is likely that the majority of the mobile Na ions originate from
+
the active layer. These Na ions are then free to migrate along the grain boundaries;
+
however, this phenomenon alone can hardly explain the accumulation of sodium
at the pn-junction because simple diffusion should rather result in a homogeneous
distribution of sodium throughout the layers. The migration of released Na ions is
+
therefore driven by an additional force. In the following, we argue that the driving
mechanism is related to illumination.
We consider the energy band diagram in the region of grain boundaries in the
CIGS absorber. This band diagram with band bending due to trapped positive
+
charges is shown in Figure 7.16. In the dark, the positive Na should drift from the
grain boundaries to the grains due to the internal electric field in the region of
+
the grain boundaries. It is proposed that migration of Na does not occur because
of limited solubility of Na in CIGSe compared to that of Cu, while it has the same
+
+
driving force for migration. This hypothesis is supported by the fact that although
copper vacancies are certainly present within the grains, allowing the formation of
+
p-type CIGSe, sodium accumulates at the grain boundaries. Therefore, the Na ions
are pinned in the grain boundaries. However, the band bending in the region of the
grain boundaries is dependent on the density of the photogenerated carriers. Under
231
A negative effect of sodium migration on CIGS solar cells was observed by Fjällström
o
et al. [3]. After the exposure of a 50V bias combined with 85C on CIGS solar cells,
sodium migrated from the glass substrate to the pn-junction, which resulted in a
rapid decrease of the conversion efficiency. In this case, the relatively high voltage bias
was apparently able to drive sodium from the glass, leading to a potential induced
degradation-(PID-)like process, which was also reported by Colli [14].In our work,
it is not yet possible to distinguish between the impact of sodium and potassium
migration. Because of the limited potassium migration and the results of Fjällström
[3], who observed that sodium migration led to degradation, it will be assumed that
sodium migration is the most important factor.
7.4.1 Degradation model
In our case the exposure conditions are different since during the damp heat
treatment, no voltage bias is applied. The majority of sodium atoms in CIGS are
located at the grain boundaries [15], most probably bonded to the adjacent grains
+
dangling bonds. The presence of water likely enables the release of Na ions through
the hydrolysis of the bonds between CIGS and sodium. The soda lime glass can also
serve as a source of the sodium, but it should be noted that, as far as can be judged
from the SIMS spectra, the sodium content in the complete stack does not increase
greatly. Therefore, it is likely that the majority of the mobile Na ions originate from
+
the active layer. These Na ions are then free to migrate along the grain boundaries;
+
however, this phenomenon alone can hardly explain the accumulation of sodium
at the pn-junction because simple diffusion should rather result in a homogeneous
distribution of sodium throughout the layers. The migration of released Na ions is
+
therefore driven by an additional force. In the following, we argue that the driving
mechanism is related to illumination.
We consider the energy band diagram in the region of grain boundaries in the
CIGS absorber. This band diagram with band bending due to trapped positive
+
charges is shown in Figure 7.16. In the dark, the positive Na should drift from the
grain boundaries to the grains due to the internal electric field in the region of
+
the grain boundaries. It is proposed that migration of Na does not occur because
of limited solubility of Na in CIGSe compared to that of Cu, while it has the same
+
+
driving force for migration. This hypothesis is supported by the fact that although
copper vacancies are certainly present within the grains, allowing the formation of
+
p-type CIGSe, sodium accumulates at the grain boundaries. Therefore, the Na ions
are pinned in the grain boundaries. However, the band bending in the region of the
grain boundaries is dependent on the density of the photogenerated carriers. Under
231
