5 Crystalline Silicon Solar Cells: Homojunction Cells
123
continues to approach the edge of the valence band. For the solar cell, this means that
a near-surface electric field is formed on the back side, which pushes the minority
charge carriers (in the case of p-doping, the electrons) back into the crystal bulk and,
thus, prevents recombination at the back. Expressed numerically, this reduces the
back surface recombination velocity S rear from 10
5 to 10
6 cm s
−1 to the range of
500–1000 cm s
−1 . Such values of S rear were sufficient for a long time, since recombination in the highly doped emitter (pn-junction) and at the interface gave rise to even
higher recombination velocities on the front side, and, thus, at that time, the back
side did not represent the limiting zone for the efficiency of the solar cell. As a result
of a worldwide focus on the optimization of the front side of the solar cell, based
on the development of new silver pastes for contacting high-impedance pn-junctions
(emitters) the back side of the solar cells became (around 2010) the focus of attention
as a limiting factor in cell efficiency.
5.4.4 The Structure of the PERC Cell
In particular, the PERC solar cell concept (Passivated Emitter Rear Cell) has shown
a pronounced potential for increasing performance with only moderate changes in
the cell production process. The advantage of this cell concept is that it has an
“evolutionary” character, whereby the existing process plants for standard solar cells
are preserved, and only small expenses for additional equipment are necessary. In the
following, therefore, the concept of the PERC cell will be introduced; the differences
between the standard cell (Al-BSF cell) and the PERC cell will be discussed in
more detail. Figure 5.13 illustrates the fabrication process sequences for the two cell
concepts. It should be noted that the three coating processes in the PERC cell can be
realized by a single machine.
Figure 5.14 shows the PERC cell in cross-section. In order to reduce the surface
recombination velocity, the entire back side is first treated with a 1–2 nm thick,
high-quality silicon oxide layer.
28 This oxide layer (SiO 2 ) has a similar crystal structure as silicon, thus, very effectively saturating the free surface bonds of the silicon
and forming an excellent interface with the silicon crystal. Thereafter, the cell is
passivated with a non-conductive dielectric layer. Aluminium oxide is particularly
suitable for this, because aluminium oxide is an excellent insulator. In contrast to
SiO 2 , Al 2 O 3 has a rhombohedral crystal structure. This leads to a mismatch with
SiO 2 , whereby the Al atoms of the Al 2 O 3 compound have, at the boundary to SiO 2 ,
a free bond that tends to trap electrons, forming thereby static, negative charges. The
result is a so-called “field effect”
29 that additionally protects the minority charge carriers (the electrons) in silicon from recombination at the SiO 2 interface [10]. In this
way, surface recombination velocities of 10–20 cm s
−1 can be achieved industrially.
28 This is optional and can be done (as usual) by oxidation (not deposition).
29 By “field effect” we wish to indicate that there is an electric field, which is built up like in the
Al-BSF cell, so that the electrons on the back side are reflected back into the cell, before recombining.
123
continues to approach the edge of the valence band. For the solar cell, this means that
a near-surface electric field is formed on the back side, which pushes the minority
charge carriers (in the case of p-doping, the electrons) back into the crystal bulk and,
thus, prevents recombination at the back. Expressed numerically, this reduces the
back surface recombination velocity S rear from 10
5 to 10
6 cm s
−1 to the range of
500–1000 cm s
−1 . Such values of S rear were sufficient for a long time, since recombination in the highly doped emitter (pn-junction) and at the interface gave rise to even
higher recombination velocities on the front side, and, thus, at that time, the back
side did not represent the limiting zone for the efficiency of the solar cell. As a result
of a worldwide focus on the optimization of the front side of the solar cell, based
on the development of new silver pastes for contacting high-impedance pn-junctions
(emitters) the back side of the solar cells became (around 2010) the focus of attention
as a limiting factor in cell efficiency.
5.4.4 The Structure of the PERC Cell
In particular, the PERC solar cell concept (Passivated Emitter Rear Cell) has shown
a pronounced potential for increasing performance with only moderate changes in
the cell production process. The advantage of this cell concept is that it has an
“evolutionary” character, whereby the existing process plants for standard solar cells
are preserved, and only small expenses for additional equipment are necessary. In the
following, therefore, the concept of the PERC cell will be introduced; the differences
between the standard cell (Al-BSF cell) and the PERC cell will be discussed in
more detail. Figure 5.13 illustrates the fabrication process sequences for the two cell
concepts. It should be noted that the three coating processes in the PERC cell can be
realized by a single machine.
Figure 5.14 shows the PERC cell in cross-section. In order to reduce the surface
recombination velocity, the entire back side is first treated with a 1–2 nm thick,
high-quality silicon oxide layer.
28 This oxide layer (SiO 2 ) has a similar crystal structure as silicon, thus, very effectively saturating the free surface bonds of the silicon
and forming an excellent interface with the silicon crystal. Thereafter, the cell is
passivated with a non-conductive dielectric layer. Aluminium oxide is particularly
suitable for this, because aluminium oxide is an excellent insulator. In contrast to
SiO 2 , Al 2 O 3 has a rhombohedral crystal structure. This leads to a mismatch with
SiO 2 , whereby the Al atoms of the Al 2 O 3 compound have, at the boundary to SiO 2 ,
a free bond that tends to trap electrons, forming thereby static, negative charges. The
result is a so-called “field effect”
29 that additionally protects the minority charge carriers (the electrons) in silicon from recombination at the SiO 2 interface [10]. In this
way, surface recombination velocities of 10–20 cm s
−1 can be achieved industrially.
28 This is optional and can be done (as usual) by oxidation (not deposition).
29 By “field effect” we wish to indicate that there is an electric field, which is built up like in the
Al-BSF cell, so that the electrons on the back side are reflected back into the cell, before recombining.