Page 55 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 55
Stability of Cu(In,Ga)Se 2 Solar Cells
Steinhauser et al. [79] concluded the dominance of grain boundary degradation in
Low Pressure Chemical Vapour Deposition (LPCVD) ZnO:B based on measurements.
Calculations based on these experiments also allowed the estimation of the barrier
width and height. The change of the current path due to degradation of grain bound -
aries in ZnO:B in shown in Figure 2.9. They described that for non-degraded ZnO:B
samples, both tunneling and thermionic emission can be considered as the path of
the current through the grain boundary potential barrier. When samples were ex-
N
posed to a damp heat test, this leads to an increase of the trap state density at
t
the grain boundary and thus to a higher and wider potential barrier. The latter are
depicted by a larger barrier width W and barrier height V . Therefore, the current path
b
through the barrier by tunneling is decreased, so thermionic emission becomes the
dominant way of transport. When the sample is exposed to a damp heat test, a shift
toward a more thermionic-emission governed conductivity is therefore made. It can
be expected that a similar trend occurs for sputtered ZnO:Al.
Greiner et al. [62,63] showed very strong degradation of ZnO:Al conductivity when
ZnO:Al films are grown on rough substrates, due to the formation of ‘extended grain
eV b eV b
W
a) Initial b) Degraded
Thermionic emission E v eV b Barrier height
W
Trap state N t E fermi Barrier width
Tunneling
Figure 2.9
Schematic drawing of the current path through the barrier at a grain boundary based on [79] of a (a) as grown and (b)
damp heat degraded LPCVD ZnO:B sample. In the degraded state the barrier width W and height V b are larger, while more
traps states are present.
53
Steinhauser et al. [79] concluded the dominance of grain boundary degradation in
Low Pressure Chemical Vapour Deposition (LPCVD) ZnO:B based on measurements.
Calculations based on these experiments also allowed the estimation of the barrier
width and height. The change of the current path due to degradation of grain bound -
aries in ZnO:B in shown in Figure 2.9. They described that for non-degraded ZnO:B
samples, both tunneling and thermionic emission can be considered as the path of
the current through the grain boundary potential barrier. When samples were ex-
N
posed to a damp heat test, this leads to an increase of the trap state density at
t
the grain boundary and thus to a higher and wider potential barrier. The latter are
depicted by a larger barrier width W and barrier height V . Therefore, the current path
b
through the barrier by tunneling is decreased, so thermionic emission becomes the
dominant way of transport. When the sample is exposed to a damp heat test, a shift
toward a more thermionic-emission governed conductivity is therefore made. It can
be expected that a similar trend occurs for sputtered ZnO:Al.
Greiner et al. [62,63] showed very strong degradation of ZnO:Al conductivity when
ZnO:Al films are grown on rough substrates, due to the formation of ‘extended grain
eV b eV b
W
a) Initial b) Degraded
Thermionic emission E v eV b Barrier height
W
Trap state N t E fermi Barrier width
Tunneling
Figure 2.9
Schematic drawing of the current path through the barrier at a grain boundary based on [79] of a (a) as grown and (b)
damp heat degraded LPCVD ZnO:B sample. In the degraded state the barrier width W and height V b are larger, while more
traps states are present.
53
