Page 134 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 134
Chapter 5
the optical characteristics.
The morphological and structural properties of the films were determined by a FEI
Quanta 600 Scanning Electron Microscope (SEM) combined with Energy Dispersion
X-ray (EDX) EDAX Genesis 4000 and a Philips X’pert 5068 powder diffractometer,
o
o
equipped with a Cu Kα source (λ=0.154 nm) in the 5 to 95 2θ range with a step size of
0.02 . No internal standard was used. Raman spectra were taken with a Renishaw Raman
o
spectrophotometer using a 514 nm laser.
Time-Of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS) depth profiling was
performed in the negative mode using an Ion-Tof TOF-SIMS IV instrument, with a
beam of 50x50 μm of 2 keV Cs .
2
+
The X-ray photoemission measurements were performed on a Kratos AXIS Ultra
spectrometer using a monochromatic Al K X-ray source at 150 W for core levels and
α
2
valence band. The analyzed area was 700×300 μm . The energy scale was calibrated
using Au 4f at 83.97 eV and Cu 2p peaks at 923.63 eV measured from sputter-
7/2
3/2
cleaned gold and copper films. The overall energy resolution, as determined from the
Fermi edge of a silver reference, was 0.47±0.03 eV at 20 eV pass energy. The Kratos
charge neutralizer system was used during all the experiments. The wide spectra
and detailed spectra (core levels and valence band) were obtained using a 160 eV
and 40 eV pass energy respectively. All high-resolution spectra were analysed and
fitted with CasaXPS (N. Fairley, Copyright 2005 Casa Software Ltd). The photoemission
contributions were fitted using a Shirley function for the background. For the Mo 3d-
Se 3s signal, the spectrum of the non-degraded Mo2 in Figure 5.27 has been used to
create a Line Shape (LS), which has been used for further fitting. Then pseudo-Voigt
5+
functions for the Mo 3d doublets relative to Mo and Mo have been used with a 3.14
6+
eV spin-orbit-splitting and an area ratio for the 3d5/2 / 3d3/2 peaks of 1.5.
Unless mentioned otherwise, the results of degradation in this article are shown after
the damp heat treatment.
5.3 Results
In these studies, two important deposition parameters for molybdenum in CIGS
modules have been taken into account, which are the sputter pressure and the
presence of selenium.
Sputter pressure
When molybdenum is deposited by DC magnetron sputtering, a correlation is observed
between the sputter gas pressure and the stress in the film [10,14]. Under low argon
pressure, molybdenum films become tightly packed and have a relatively low resistivity.
132
the optical characteristics.
The morphological and structural properties of the films were determined by a FEI
Quanta 600 Scanning Electron Microscope (SEM) combined with Energy Dispersion
X-ray (EDX) EDAX Genesis 4000 and a Philips X’pert 5068 powder diffractometer,
o
o
equipped with a Cu Kα source (λ=0.154 nm) in the 5 to 95 2θ range with a step size of
0.02 . No internal standard was used. Raman spectra were taken with a Renishaw Raman
o
spectrophotometer using a 514 nm laser.
Time-Of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS) depth profiling was
performed in the negative mode using an Ion-Tof TOF-SIMS IV instrument, with a
beam of 50x50 μm of 2 keV Cs .
2
+
The X-ray photoemission measurements were performed on a Kratos AXIS Ultra
spectrometer using a monochromatic Al K X-ray source at 150 W for core levels and
α
2
valence band. The analyzed area was 700×300 μm . The energy scale was calibrated
using Au 4f at 83.97 eV and Cu 2p peaks at 923.63 eV measured from sputter-
7/2
3/2
cleaned gold and copper films. The overall energy resolution, as determined from the
Fermi edge of a silver reference, was 0.47±0.03 eV at 20 eV pass energy. The Kratos
charge neutralizer system was used during all the experiments. The wide spectra
and detailed spectra (core levels and valence band) were obtained using a 160 eV
and 40 eV pass energy respectively. All high-resolution spectra were analysed and
fitted with CasaXPS (N. Fairley, Copyright 2005 Casa Software Ltd). The photoemission
contributions were fitted using a Shirley function for the background. For the Mo 3d-
Se 3s signal, the spectrum of the non-degraded Mo2 in Figure 5.27 has been used to
create a Line Shape (LS), which has been used for further fitting. Then pseudo-Voigt
5+
functions for the Mo 3d doublets relative to Mo and Mo have been used with a 3.14
6+
eV spin-orbit-splitting and an area ratio for the 3d5/2 / 3d3/2 peaks of 1.5.
Unless mentioned otherwise, the results of degradation in this article are shown after
the damp heat treatment.
5.3 Results
In these studies, two important deposition parameters for molybdenum in CIGS
modules have been taken into account, which are the sputter pressure and the
presence of selenium.
Sputter pressure
When molybdenum is deposited by DC magnetron sputtering, a correlation is observed
between the sputter gas pressure and the stress in the film [10,14]. Under low argon
pressure, molybdenum films become tightly packed and have a relatively low resistivity.
132
