628
22 Light-to-Electricity Conversion
(a)
40
30
20
10
0
E (eV)
g
2
3
0
1
GaAs
C=1000
CdTe
Cu O
2
e
G
i
S
C=1
CdS
CuInS 2
CuInSe 2
CuGaS 2
CuGaSe 2
(b)
FF
j sc
projected
V oc
10
0
10
2
10
1
10
3
0.80
0.75
0.70
24
22
20
18
16
14
0.8
0.7
0.6
V (V)
10
1
10
-2
2
I (A/cm )
sc
2
Fig. 22.58 a Ideal quantum efficiency of solar cells (single junction) as a function of band gap and light concentration C.
The band gaps of some important semiconductors are denoted by arrows. Adapted from [1877]. b Properties of a silicon
solar cell (water cooled) as a function of light concentration. Solid line is theoretically projected efficiency, dashed lines
are guides to the eye. Adapted from [574]
(2) represents the value of E m . The efficiency is the ratio of E m n ph and the area under curve (1). The
efficiency as a function of the band gap is shown in Fig. 22.58a. It has a fairly broad maximum such
that many semiconductors can be used for solar cells, in principle. The maximum theoretical efficiency
for a single junction is 31% for nonconcentrated sunlight (AM1.5). This limit corresponds to the
classic Shockley-Queisser limit [1878–1880], assuming radiative recombination as the only chargecarrier recombination mechanism. In [1881], the limit for a single material is found to be 43% for an
optimally tailored band structure that allows carrier multiplication by optically excited hot carriers.
The solar cell as a heat engine is dicussed in [1882].
When the sunlight is concentrated, e.g. by a lens, the efficiency increases (Fig. 22.58b). The shortcircuit current increases linearly. The effect is mostly due to the increase of the open-circuit voltage.
For C = 1000, the maximum theoretical efficiency for a single-junction solar cell is 38%.
A further increase of efficiency can be achieved with multiple junctions using various materials for
absorption. In a tandem cell (two junctions), the upper layer absorbs the higher-energy photons in a
wide band gap material. The material with the lower band gap makes use of the low-energy photons.
Thus, the cell works with two different values of E m (Fig. 22.57b). With band gaps of 1.56 eV and
0.84 eV, an efficiency of 50% can be reached theoretically. With three materials 56%, and for a large
number of materials 72% is the limit. Between the junctions, tunneling diodes (Sect. 21.5.9) must be
used to allow carrier transport through the entire structure. It is a nontrivial task to fabricate multiple
heterojunctions due to incompatibilities of the lattice constants. Besides heteroepitaxy, wafer bonding
can also be used for fabrication. A lattice-matched InGaP/GaAs/InGaAsN cell seems a viable solution
for high-efficiency solar cells.
22 Light-to-Electricity Conversion
(a)
40
30
20
10
0
E (eV)
g
2
3
0
1
GaAs
C=1000
CdTe
Cu O
2
e
G
i
S
C=1
CdS
CuInS 2
CuInSe 2
CuGaS 2
CuGaSe 2
(b)
FF
j sc
projected
V oc
10
0
10
2
10
1
10
3
0.80
0.75
0.70
24
22
20
18
16
14
0.8
0.7
0.6
V (V)
10
1
10
-2
2
I (A/cm )
sc
2
Fig. 22.58 a Ideal quantum efficiency of solar cells (single junction) as a function of band gap and light concentration C.
The band gaps of some important semiconductors are denoted by arrows. Adapted from [1877]. b Properties of a silicon
solar cell (water cooled) as a function of light concentration. Solid line is theoretically projected efficiency, dashed lines
are guides to the eye. Adapted from [574]
(2) represents the value of E m . The efficiency is the ratio of E m n ph and the area under curve (1). The
efficiency as a function of the band gap is shown in Fig. 22.58a. It has a fairly broad maximum such
that many semiconductors can be used for solar cells, in principle. The maximum theoretical efficiency
for a single junction is 31% for nonconcentrated sunlight (AM1.5). This limit corresponds to the
classic Shockley-Queisser limit [1878–1880], assuming radiative recombination as the only chargecarrier recombination mechanism. In [1881], the limit for a single material is found to be 43% for an
optimally tailored band structure that allows carrier multiplication by optically excited hot carriers.
The solar cell as a heat engine is dicussed in [1882].
When the sunlight is concentrated, e.g. by a lens, the efficiency increases (Fig. 22.58b). The shortcircuit current increases linearly. The effect is mostly due to the increase of the open-circuit voltage.
For C = 1000, the maximum theoretical efficiency for a single-junction solar cell is 38%.
A further increase of efficiency can be achieved with multiple junctions using various materials for
absorption. In a tandem cell (two junctions), the upper layer absorbs the higher-energy photons in a
wide band gap material. The material with the lower band gap makes use of the low-energy photons.
Thus, the cell works with two different values of E m (Fig. 22.57b). With band gaps of 1.56 eV and
0.84 eV, an efficiency of 50% can be reached theoretically. With three materials 56%, and for a large
number of materials 72% is the limit. Between the junctions, tunneling diodes (Sect. 21.5.9) must be
used to allow carrier transport through the entire structure. It is a nontrivial task to fabricate multiple
heterojunctions due to incompatibilities of the lattice constants. Besides heteroepitaxy, wafer bonding
can also be used for fabrication. A lattice-matched InGaP/GaAs/InGaAsN cell seems a viable solution
for high-efficiency solar cells.