7 Crystalline Silicon Solar Cells: Heterojunction Cells
189
that constitute recombination centres are formed. Such free bonds can be passivated
with hydrogen (hydrogenation).
The deposition of the a-Si layers is usually carried out by PE-CVD (Plasma
Enhanced Chemical Vapour Deposition, see also Chap. 6). In a mixture of silane
(SiH 4 ) and hydrogen a plasma is ignited. Within the plasma, the molecules decompose and the atoms are directed via an electric field onto the substrate, e.g. onto the
wafer. There they are deposited as an amorphous layer. In this way, the intrinsic,
undoped passivation layer a-Si(i) is formed without the addition of further gases.
By adding phosphine (PH 3 ) or diborane (B 2 H 6 ), n- or p-doping of the amorphous
layer can be achieved. The a-Si:H(p) deposition causes more complications regarding impurities and thermal stability. This is the reason why the p-layer is deposited
after the n-layer.
During the process, it must be ensured that no unwanted contamination by the
doping materials boron (diborane) and phosphorus (phosphine) occurs, since such
“cross-contamination” can lead to severe efficiency losses. The cell can be protected
from this, by using a separate transport box (transport carrier) for each layer and
the PECVD deposition equipment is equipped with separate plasma chambers. With
parallel plate plasma reactors and a box-in-box vacuum chamber system, large areas
can be deposited homogeneously with a thickness tolerance of less than 5%, at layer
thicknesses of a few nm. The deposition takes place at 13.56 MHz, at 10
−3 mbar and
at temperatures of 150–250 °C. It is absolutely necessary to avoid the amorphous
layers becoming too hot. Otherwise a restructuring of the originally amorphous
layers up to crystallization occurs, whereby the layer loses its excellent passivation
properties. The deposition of the four amorphous layers takes only approximately
60 s; despite this fast deposition, very high lifetime values of more than 4 ms for
the minority carriers are achieved, with industrially produced silicon (measured at
the reference excess charge carrier concentration of 10
15 cm
−3 ). With high-quality
floatzone (FZ) silicon, lifetime values of up to 10 ms can be achieved. Admittedly, FZ
silicon is very expensive and is not suitable for mass production in the photovoltaic
industry—however, this result indicates the basic potential for further improvement.
7.4.4 Coating of the TCO Layer
After the four amorphous silicon layers have been deposited, the front and back of
the cell are each coated with TCO. With the ITO (indium tin oxide) generally used
for HJT cells, the coating is usually done with sputtering. On the front side, the TCO
layer is responsible for light trapping (e.g. it forms the ARC—the Anti Reflection
Coating); furthermore, it has to provide an Ohmic contact to the amorphous silicon
and to the metal contact grid. On the back, where the p-n-junction is preferentially
located, the TCO layer can be slightly thicker (see Sect. 7.2.3) to ensure high electrical
conductivity. In the case of bifacial HJT cells, the back TCO layer additionally acts as
an ARC layer—just like the front one. Indeed, bifacial cells receive on their back side
the scattered light of the environment and the light reflection of the ground, in order
189
that constitute recombination centres are formed. Such free bonds can be passivated
with hydrogen (hydrogenation).
The deposition of the a-Si layers is usually carried out by PE-CVD (Plasma
Enhanced Chemical Vapour Deposition, see also Chap. 6). In a mixture of silane
(SiH 4 ) and hydrogen a plasma is ignited. Within the plasma, the molecules decompose and the atoms are directed via an electric field onto the substrate, e.g. onto the
wafer. There they are deposited as an amorphous layer. In this way, the intrinsic,
undoped passivation layer a-Si(i) is formed without the addition of further gases.
By adding phosphine (PH 3 ) or diborane (B 2 H 6 ), n- or p-doping of the amorphous
layer can be achieved. The a-Si:H(p) deposition causes more complications regarding impurities and thermal stability. This is the reason why the p-layer is deposited
after the n-layer.
During the process, it must be ensured that no unwanted contamination by the
doping materials boron (diborane) and phosphorus (phosphine) occurs, since such
“cross-contamination” can lead to severe efficiency losses. The cell can be protected
from this, by using a separate transport box (transport carrier) for each layer and
the PECVD deposition equipment is equipped with separate plasma chambers. With
parallel plate plasma reactors and a box-in-box vacuum chamber system, large areas
can be deposited homogeneously with a thickness tolerance of less than 5%, at layer
thicknesses of a few nm. The deposition takes place at 13.56 MHz, at 10
−3 mbar and
at temperatures of 150–250 °C. It is absolutely necessary to avoid the amorphous
layers becoming too hot. Otherwise a restructuring of the originally amorphous
layers up to crystallization occurs, whereby the layer loses its excellent passivation
properties. The deposition of the four amorphous layers takes only approximately
60 s; despite this fast deposition, very high lifetime values of more than 4 ms for
the minority carriers are achieved, with industrially produced silicon (measured at
the reference excess charge carrier concentration of 10
15 cm
−3 ). With high-quality
floatzone (FZ) silicon, lifetime values of up to 10 ms can be achieved. Admittedly, FZ
silicon is very expensive and is not suitable for mass production in the photovoltaic
industry—however, this result indicates the basic potential for further improvement.
7.4.4 Coating of the TCO Layer
After the four amorphous silicon layers have been deposited, the front and back of
the cell are each coated with TCO. With the ITO (indium tin oxide) generally used
for HJT cells, the coating is usually done with sputtering. On the front side, the TCO
layer is responsible for light trapping (e.g. it forms the ARC—the Anti Reflection
Coating); furthermore, it has to provide an Ohmic contact to the amorphous silicon
and to the metal contact grid. On the back, where the p-n-junction is preferentially
located, the TCO layer can be slightly thicker (see Sect. 7.2.3) to ensure high electrical
conductivity. In the case of bifacial HJT cells, the back TCO layer additionally acts as
an ARC layer—just like the front one. Indeed, bifacial cells receive on their back side
the scattered light of the environment and the light reflection of the ground, in order