Page 250 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 250
Chapter 8
disappearance of the current was also observed by EQE measurements for the H O/air
2
and H O/CO /N samples.
2
2
2
The changes in light shunt resistance were remarkable: for the HO/air and HO/
2
2
CO /N samples, the shunt resistance dropped when the efficiency dropped, but
2
2
later increased again. This can imply the formation of shunt paths that are later not
shunting anymore, but can also result from the disappearance of contact points,
thereby decreasing the possibilities for shunt paths. This decrease was not observed
for the dark shunt resistance. Another observation for especially the light but also the
dark shunt resistances was the decrease for the H O/N samples, which also has been
2
2
observed in similar experiments. Furthermore, the open circuit voltage also changed,
but this impact was minor compared to the changes series resistance and current.
When the initial increase during the first 50 hours is considered, it seems that this is
mostly caused by enhancement of the short circuit current, while the open circuit
voltage also has a small impact. It should be noted that in earlier experiments, the
impact of V on this efficiency increase was larger (20-30 mV).
oc
When the reference and air samples were measured before and after 1560 hours
under argon and air respectively, it was observed that these solar cells showed a
slight increase in efficiency (+0.2 and +0.3 percent point respectively). This was mostly
driven by small changes in the shunt resistance (glovebox sample) and open circuit
voltage (air). However, it is not expected that these changes are significant.
8.3.3 Compositional changes
The elemental depth profiles of non-degraded and degraded samples was measured
by SIMS (Figure 8.8). This was operated in positive mode, and the elements zinc,
oxygen, aluminium, cadmium, sulphur, copper, indium, gallium, selenium, chlorine,
bromine, sodium, potassium as well as the compound hydroxide were measured.
It should be noted that no absolute concentrations can be determined with SIMS,
since the intensity is not calibrated. Since the measured number of counts is largely
influenced by the surrounding matrix, no absolute comparison of concentrations
between different matrices (e.g. ZnO:Al or CIGS) or different elements can be executed.
Figure 8.8a shows a depth profile of a typical CIGS cell, which has only been exposed
to argon. The gallium, indium and selenium profiles are not included, since they do
not change significantly during degradation. In these SIMS spectra, the boundaries
between the various layers can be distinguished, although the width of the cadmium
peak also indicates that the depth resolution is limited. One remarkable observation
is the high concentration of sodium, potassium and sulfur in the ZnO:Al layer, but
this can be explained by the matrix effects and difference in electronegativity. Similar
248
disappearance of the current was also observed by EQE measurements for the H O/air
2
and H O/CO /N samples.
2
2
2
The changes in light shunt resistance were remarkable: for the HO/air and HO/
2
2
CO /N samples, the shunt resistance dropped when the efficiency dropped, but
2
2
later increased again. This can imply the formation of shunt paths that are later not
shunting anymore, but can also result from the disappearance of contact points,
thereby decreasing the possibilities for shunt paths. This decrease was not observed
for the dark shunt resistance. Another observation for especially the light but also the
dark shunt resistances was the decrease for the H O/N samples, which also has been
2
2
observed in similar experiments. Furthermore, the open circuit voltage also changed,
but this impact was minor compared to the changes series resistance and current.
When the initial increase during the first 50 hours is considered, it seems that this is
mostly caused by enhancement of the short circuit current, while the open circuit
voltage also has a small impact. It should be noted that in earlier experiments, the
impact of V on this efficiency increase was larger (20-30 mV).
oc
When the reference and air samples were measured before and after 1560 hours
under argon and air respectively, it was observed that these solar cells showed a
slight increase in efficiency (+0.2 and +0.3 percent point respectively). This was mostly
driven by small changes in the shunt resistance (glovebox sample) and open circuit
voltage (air). However, it is not expected that these changes are significant.
8.3.3 Compositional changes
The elemental depth profiles of non-degraded and degraded samples was measured
by SIMS (Figure 8.8). This was operated in positive mode, and the elements zinc,
oxygen, aluminium, cadmium, sulphur, copper, indium, gallium, selenium, chlorine,
bromine, sodium, potassium as well as the compound hydroxide were measured.
It should be noted that no absolute concentrations can be determined with SIMS,
since the intensity is not calibrated. Since the measured number of counts is largely
influenced by the surrounding matrix, no absolute comparison of concentrations
between different matrices (e.g. ZnO:Al or CIGS) or different elements can be executed.
Figure 8.8a shows a depth profile of a typical CIGS cell, which has only been exposed
to argon. The gallium, indium and selenium profiles are not included, since they do
not change significantly during degradation. In these SIMS spectra, the boundaries
between the various layers can be distinguished, although the width of the cadmium
peak also indicates that the depth resolution is limited. One remarkable observation
is the high concentration of sodium, potassium and sulfur in the ZnO:Al layer, but
this can be explained by the matrix effects and difference in electronegativity. Similar
248
