Page 81 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 81
Stability of Cu(In,Ga)Se 2 Solar Cells
2.5.1.2 Degradation in P2
The Mo/TCO contact in the P2 scribe is necessary for the current transport between
the individual cells within the CIGS module. Loss of conductivity in this scribe could
thus have a large impact on the power output of the modules. However, the impact of
conductivity loss of ZnO:Al and Mo and the degradation in P2 are hard to distinguish,
since the measured series resistance includes the impact of all three elements. Several
references [27,91,104,111] have tried to distinguish between these effects, mostly by
the use of testing structures.
Wennerberg et al. [27] found that the degradation of P2 was the principle cause of fill
factor degradation due to damp heat exposure of unencapsulated interconnected
solar cells. It was observed that the contact resistance in the P2 scribe increased by
a factor of ten within ten hours of exposure. After one hundred hours, the resistance
had increased by a factor 50 Theelen et al. [108] observed that the series resistance
of two interconnected cells increased with 12±3 mΩ/h, while single cells increased
in series resistance with 4.7±0.3 mΩ/h. When corrected for the changed cell size, the
interconnected cells degraded faster than expected.. Therefore, it was suggested that
part of the increase was caused by increased resistivity in P2.
Similar observations were made by Westin et al. [104]. An unencapsulated test struc-
ture, that was exposed to damp heat allowed the separate determination of the in-
crease of the series resistance of zinc oxide/CdS layer (average 8 mΩ/ /h) and of the
total resistance (average 73 mΩ//h). Since it is expected that molybdenum degrada-
tion leads to a very sudden change in resistance, the difference (average 65 mΩ/ /h)
is considered to be an estimation of the impact of the degradation in P2. It is therefore
concluded that resistance of the P2 contact quickly became dominant under damp
heat exposure, while the increased resistance of the ZnO:Al front contact played a
minor role.
A similar approach was chosen by Klaer et al. [91], who used a test structure that al-
lowed the separation of sheet resistance of the zinc oxide and the sum of the contact
resistance of the Mo/ZnO plus the resistance of the Mo. They showed that the contact
resistance of a P2 scribe containing i-ZnO increased more rapidly (27 mΩ/h) than that
of a scribe that did not contain i-ZnO (2.8 mΩ/h).
Wennerberg et al. [27] proposed that the formation of an oxide layer between the
ZnO:Al and the molybdenum, or the depletion of free carriers at the ZnO:Al interface
could be the cause for the degradation in P2.
2.5.1.3 Corrosion in P3
As shown in chapter 2.3.1.2, the corrosion of molybdenum is a very rapid process,
leading to the formation of a non-conductive molybdenum oxide layer on top of a
79
2.5.1.2 Degradation in P2
The Mo/TCO contact in the P2 scribe is necessary for the current transport between
the individual cells within the CIGS module. Loss of conductivity in this scribe could
thus have a large impact on the power output of the modules. However, the impact of
conductivity loss of ZnO:Al and Mo and the degradation in P2 are hard to distinguish,
since the measured series resistance includes the impact of all three elements. Several
references [27,91,104,111] have tried to distinguish between these effects, mostly by
the use of testing structures.
Wennerberg et al. [27] found that the degradation of P2 was the principle cause of fill
factor degradation due to damp heat exposure of unencapsulated interconnected
solar cells. It was observed that the contact resistance in the P2 scribe increased by
a factor of ten within ten hours of exposure. After one hundred hours, the resistance
had increased by a factor 50 Theelen et al. [108] observed that the series resistance
of two interconnected cells increased with 12±3 mΩ/h, while single cells increased
in series resistance with 4.7±0.3 mΩ/h. When corrected for the changed cell size, the
interconnected cells degraded faster than expected.. Therefore, it was suggested that
part of the increase was caused by increased resistivity in P2.
Similar observations were made by Westin et al. [104]. An unencapsulated test struc-
ture, that was exposed to damp heat allowed the separate determination of the in-
crease of the series resistance of zinc oxide/CdS layer (average 8 mΩ/ /h) and of the
total resistance (average 73 mΩ//h). Since it is expected that molybdenum degrada-
tion leads to a very sudden change in resistance, the difference (average 65 mΩ/ /h)
is considered to be an estimation of the impact of the degradation in P2. It is therefore
concluded that resistance of the P2 contact quickly became dominant under damp
heat exposure, while the increased resistance of the ZnO:Al front contact played a
minor role.
A similar approach was chosen by Klaer et al. [91], who used a test structure that al-
lowed the separation of sheet resistance of the zinc oxide and the sum of the contact
resistance of the Mo/ZnO plus the resistance of the Mo. They showed that the contact
resistance of a P2 scribe containing i-ZnO increased more rapidly (27 mΩ/h) than that
of a scribe that did not contain i-ZnO (2.8 mΩ/h).
Wennerberg et al. [27] proposed that the formation of an oxide layer between the
ZnO:Al and the molybdenum, or the depletion of free carriers at the ZnO:Al interface
could be the cause for the degradation in P2.
2.5.1.3 Corrosion in P3
As shown in chapter 2.3.1.2, the corrosion of molybdenum is a very rapid process,
leading to the formation of a non-conductive molybdenum oxide layer on top of a
79
