Page 211 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 211
Degradation mechanisms of the aluminium doped zinc oxide front contact
Furthermore, white spots appeared, containing elements that migrated from the
glass, like silicon and calcium, which reacted with elements from the environment,
including oxygen, carbon and chlorine.
Treatment with atmospheric species helped the identification of the species
most detrimental to ZnO:Al. We have shown that the driving force behind ZnO:Al
degradation is the combined presence of H O and CO . Individually, gaseous CO does
2
2
2
not impact the degradation at all during the tested period, while the individual impact
of H O is minor: the latter leads to slow diffusion of water down the grain boundaries,
2
where it reacts, possibly resulting in the formation of Zn(OH) . This leads to a decrease
2
in the electrical mobility. However, in the presence of CO, the electrical and optical
2
properties change very quickly. Depth profiling showed that the concentration of
hydroxide is a factor 20 higher in the bulk, and even higher at the air/ZnO:Al and the
ZnO:Al/glass interface, while carbon, hydrogen, chlorine, sulphur were also observed.
Exposure to HO and CO also led to local dissolution of the ZnO:Al at the ZnO:Al/
2
2
glass interface. Additionally, it was observed that the roles of oxygen and nitrogen in
ZnO:Al degradation are very small.
Together, these experiments shown that ZnO:Al degradation mostly led to a decrease
in mobility, likely caused by an increased potential barrier at the grain boundary. This
-
effect on the mobility is limited when only HO/OH diffuses down, which can lead
2
to the formation of Zn(OH) or adsorption of OH. In the presence of CO, the effect
-
2
2
on the mobility is larger, probably due to the formation of Zn(OH) (CO ) or a similar
5
6
3 2
molecule in the grain boundaries. Furthermore, chlorine and sulphide were found in
the top layer of the degraded samples, which suggest that e.g. Zn(OH) Cl •H O and
8
2
2
5
Zn SO (OH) •nH O can also be present.
6
2
4
4
6.6 Acknowledgements
I would like to thank Henk Steijvers, Linda van de Peppel, Harmen Rooms, Arthur Eijk,
Hans van der Veer, Emile van Veldhoven and Ester de Vrees (TNO) and Bertil Okkerse
(Philips Innovation Services) for their assistance by the measurements, analysis and
the building of the setup. I would also like to thank Arjan Hovestad and Paul Poodt
(TNO) for the fruitful discussions.
6.7 References
[1] J. Wennerberg, J. Kessler, L. Stolt, Proc. 16 EU- Klenk, Thin Solid Films 520 (2011) 1285–1290
th
PVSEC (2000) 309–312 [3] E. Ando, M. Miyazaki, Thin Solid Films 516 (2008)
[2] D. Greiner, S. Gledhill, C. Köble, J. Krammer, R. 4574–4577
209
Furthermore, white spots appeared, containing elements that migrated from the
glass, like silicon and calcium, which reacted with elements from the environment,
including oxygen, carbon and chlorine.
Treatment with atmospheric species helped the identification of the species
most detrimental to ZnO:Al. We have shown that the driving force behind ZnO:Al
degradation is the combined presence of H O and CO . Individually, gaseous CO does
2
2
2
not impact the degradation at all during the tested period, while the individual impact
of H O is minor: the latter leads to slow diffusion of water down the grain boundaries,
2
where it reacts, possibly resulting in the formation of Zn(OH) . This leads to a decrease
2
in the electrical mobility. However, in the presence of CO, the electrical and optical
2
properties change very quickly. Depth profiling showed that the concentration of
hydroxide is a factor 20 higher in the bulk, and even higher at the air/ZnO:Al and the
ZnO:Al/glass interface, while carbon, hydrogen, chlorine, sulphur were also observed.
Exposure to HO and CO also led to local dissolution of the ZnO:Al at the ZnO:Al/
2
2
glass interface. Additionally, it was observed that the roles of oxygen and nitrogen in
ZnO:Al degradation are very small.
Together, these experiments shown that ZnO:Al degradation mostly led to a decrease
in mobility, likely caused by an increased potential barrier at the grain boundary. This
-
effect on the mobility is limited when only HO/OH diffuses down, which can lead
2
to the formation of Zn(OH) or adsorption of OH. In the presence of CO, the effect
-
2
2
on the mobility is larger, probably due to the formation of Zn(OH) (CO ) or a similar
5
6
3 2
molecule in the grain boundaries. Furthermore, chlorine and sulphide were found in
the top layer of the degraded samples, which suggest that e.g. Zn(OH) Cl •H O and
8
2
2
5
Zn SO (OH) •nH O can also be present.
6
2
4
4
6.6 Acknowledgements
I would like to thank Henk Steijvers, Linda van de Peppel, Harmen Rooms, Arthur Eijk,
Hans van der Veer, Emile van Veldhoven and Ester de Vrees (TNO) and Bertil Okkerse
(Philips Innovation Services) for their assistance by the measurements, analysis and
the building of the setup. I would also like to thank Arjan Hovestad and Paul Poodt
(TNO) for the fruitful discussions.
6.7 References
[1] J. Wennerberg, J. Kessler, L. Stolt, Proc. 16 EU- Klenk, Thin Solid Films 520 (2011) 1285–1290
th
PVSEC (2000) 309–312 [3] E. Ando, M. Miyazaki, Thin Solid Films 516 (2008)
[2] D. Greiner, S. Gledhill, C. Köble, J. Krammer, R. 4574–4577
209
