Page 44 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 44
Chapter 2
open circuit voltages after damp heat may be caused by changes inherent to the poly -
crystalline CIGS absorber such as point defects, doping, elemental diffusion, oxida-
tion or grain boundaries. Or, they may be related to the properties of the pn-junction,
e.g. band bending, Fermi-level-pinning or interface defect states. Either way, these
effects are typically studied by electrical measurement techniques, e.g. admittance
spectroscopy, which rely on the existence of a pn-junction.
This chapter is limited to the stability issues measurable in the CIGS absorber (this
chapter), while complete solar cells and modules are described in chapter 2.4. In this
chapter, we will first discuss the limitations considering the study of the stability of the
CIGS absorber. Then, the intrinsic stability, which is the stability without environmental
exposure, is presented. Finally, the stability of the CIGS absorber with respect to chem -
ical changes and to sodium content under exposure to humidity and/or temperature
will be discussed.
2.3.2.1 Experimental aspects regarding the long-term stability of CIGS absorber layer
When investigating the properties of a CIGS absorber in a solar cell under ALT con-
ditions, one faces the challenge of revealing properties of one specific layer within a
stack of interacting layers. This is particularly the case since creating model samples,
i.e. an 'isolated' CIGS layer, of similar quality as in the actual device is nearly impossible.
For example, if one removes the TCO/buffer/CIGS stack from the back contact, then
the buffer and TCO, which are more conductive than the CIGS, are still underneath
the CIGS. Thus, electrical measurements on the CIGS absorber alone are difficult. If, on
the other hand, one stops the solar cell manufacturing process after absorber depo-
sition or chemically etches away layers from the top, creating a substrate/molybde-
num/CIGS sample, the more conductive molybdenum back contact causes the same
problems. In the case of etching, an additional challenge is to stop the etching process
right at the absorber surface.
Another option to allow testing of the CIGS absorber would be the growth of the ab-
sorber directly onto an alternative, non-conducting substrate. In that case, however,
the crystal growth would be different, the sodium and potassium contents would be
different, no MoSe could form and the layer properties would not be comparable to
2
a real solar cell anymore. Apart from the fact that the absorber properties are most
likely different from the CIGS layer in the solar cell, the aforementioned techniques are
destructive. This is also the case for many valuable analysis techniques for structure
and composition, like SIMS and XPS. Thus, the comparison of the electrical properties
of the as-prepared and aged solar cells as well as the individual layer properties at the
same time is impossible.
42
open circuit voltages after damp heat may be caused by changes inherent to the poly -
crystalline CIGS absorber such as point defects, doping, elemental diffusion, oxida-
tion or grain boundaries. Or, they may be related to the properties of the pn-junction,
e.g. band bending, Fermi-level-pinning or interface defect states. Either way, these
effects are typically studied by electrical measurement techniques, e.g. admittance
spectroscopy, which rely on the existence of a pn-junction.
This chapter is limited to the stability issues measurable in the CIGS absorber (this
chapter), while complete solar cells and modules are described in chapter 2.4. In this
chapter, we will first discuss the limitations considering the study of the stability of the
CIGS absorber. Then, the intrinsic stability, which is the stability without environmental
exposure, is presented. Finally, the stability of the CIGS absorber with respect to chem -
ical changes and to sodium content under exposure to humidity and/or temperature
will be discussed.
2.3.2.1 Experimental aspects regarding the long-term stability of CIGS absorber layer
When investigating the properties of a CIGS absorber in a solar cell under ALT con-
ditions, one faces the challenge of revealing properties of one specific layer within a
stack of interacting layers. This is particularly the case since creating model samples,
i.e. an 'isolated' CIGS layer, of similar quality as in the actual device is nearly impossible.
For example, if one removes the TCO/buffer/CIGS stack from the back contact, then
the buffer and TCO, which are more conductive than the CIGS, are still underneath
the CIGS. Thus, electrical measurements on the CIGS absorber alone are difficult. If, on
the other hand, one stops the solar cell manufacturing process after absorber depo-
sition or chemically etches away layers from the top, creating a substrate/molybde-
num/CIGS sample, the more conductive molybdenum back contact causes the same
problems. In the case of etching, an additional challenge is to stop the etching process
right at the absorber surface.
Another option to allow testing of the CIGS absorber would be the growth of the ab-
sorber directly onto an alternative, non-conducting substrate. In that case, however,
the crystal growth would be different, the sodium and potassium contents would be
different, no MoSe could form and the layer properties would not be comparable to
2
a real solar cell anymore. Apart from the fact that the absorber properties are most
likely different from the CIGS layer in the solar cell, the aforementioned techniques are
destructive. This is also the case for many valuable analysis techniques for structure
and composition, like SIMS and XPS. Thus, the comparison of the electrical properties
of the as-prepared and aged solar cells as well as the individual layer properties at the
same time is impossible.
42
