hot carrier
temperature
T
hc
=
T
a
/(
1
−
β)
.
If the interaction with phonons is introduced, through the reaction
e
1
−
+
phonon
↔
e
2
−
, where
e
1
(
2
)
−
represents an electron with energy
ˆ
ε
1
(
2
)
, the fact that the phonon chem-
ical potential is zero leads to
ε
F
(
ˆ
ε
1
)
=
ε
F
(
ˆ
ε
2
)
. However,
ˆ
ε
1
= ˆ
ε
2
but
ˆ
ε
2
= ˆ
ε
1
+
ε
, where
ε
is the phonon energy and, therefore,
ε
F
(
ˆ
ε
1
)
=
ε
F
(
ˆ
ε
2
)
is only fulfilled if
β
=
0 and
consequently all the electrons have the same electrochemical potential
ε
F
0
and are at the
lattice temperature. Phonon interaction is not considered to occur in what follows.
If the interaction with photons is considered now,
e
1
−
+
photon
↔
e
2
−
, the equi-
librium of the reaction is represented by
ε
F
(
ˆ
ε
1
)
+
µ
ph
=
ε
F
(
ˆ
ε
2
)
where
µ
ph
is the pho-
ton chemical potential. Taking into account equation (4.73), it is obtained that
µ
ph
=
β(
ˆ
ε
1
− ˆ
ε
2
)
=
βε
where
ε
is the energy of the photon involved. The result is that we have
an energy-dependent photon chemical potential
µ
ph
. The Bose function that describes the
occupation probability of the photon energy level
ε
becomes
1
exp
ε
−
βε
kT
a
−
1
=
1
exp
ε
kT
a
/(
1
−
β)
−
1
(
4
.
75
)
VERY HIGH EFFICIENCY CONCEPTS
143
which can be seen as free radiation (zero chemical potential) at the hot carrier
temperature.
W¨urfel also points out that, provided we can find a material in which a weak
phonon coupling is produced in order to manufacture a solar cell, we still should be
careful with the contacts because they do not allow for the maintenance of hot carriers
inside the absorbing material. These special contacts must be able to cool the electrons
from the hot temperature
T
hc
to the contact temperature
T
a
reversibly by changing their
electrochemical potential from
ε
F
0
to the electron electrochemical potential at the contacts
(Fermi level at the metals)
ε
F
+
and
ε
F
−
. In Wurfel’s Reference [47], these special contacts
are devised as selective membranes (Figure 4.12) that only allow electrons with energy
centred around
ε
e
(left contact) and
ε
h
(right contact) to pass through. The reversible
change in temperature and electrochemical potential of the electrons at the membranes is
obtained by setting
ˆ
ε
e
−
µ
hc
kT
hc
=
ˆ
ε
e
−
ε
F
−
kT
a
⇔
ε
F
−
= ˆ
ε
e
1
−
T
a
T
hc
+
µ
hc
T
a
T
hc
ˆ
ε
h
−
µ
hc
kT
hc
=
ˆ
ε
h
−
ε
F
+
kT
a
⇔
ε
F
+
= ˆ
ε
h
1
−
T
a
T
hc
+
µ
hc
T
a
T
hc
(
4
.
76
)
where
T
hc
=
T
a
/(
1
−
β)
. The cell voltage will therefore be given by
qV
=
ε
F
−
−
ε
F
+
=
(
ˆ
ε
e
− ˆ
ε
h
)
1
−
T
a
T
hc
(
4
.
77
)
The current extracted from the cell is determined by the rate at which electron–hole pairs
of energy
ˆ
ε
e
− ˆ
ε
h
can be withdrawn from the cell. Since no energy is lost as heat, the
energy balance equation (first principle) leads to
I (
ˆ
ε
e
− ˆ
ε
h
)/q
= ˙
E(T
s
,
0
, ε
g
,
∞
, H
s
)
− ˙
E(T
hc
,
0
, ε
g
,
∞
, H
r
)
(
4
.
78
)
Metal
Metal
∆
∆
Electron
membrane
Absorber
Hole
membrane
+
eV
Figure 4.12
Band structure of a hot electron solar cell showing contacting scheme by means
of selective membranes. (Reprinted from Solar Energy Materials and Solar Cells V. 46, N. 1,
W¨urfel P.,
Solar Energy Conversion with Hot Electrons from Impact Ionisation
, 43 – 52,
1997
with permission from Elsevier Science)
144
THEORETICAL LIMITS OF PHOTOVOLTAIC CONVERSION
and the power extracted from the cell can be finally computed as
˙
W
=
IV
=
[
˙
E(T
s
,
0
, ε
g
,
∞
, H
s
)
− ˙
E(T
hc
,
0
, ε
g
,
∞
, H
r
)
]
1
−
T
a
T
hc
(
4
.
79
)
which is independent of the carrier-extracting energies of the contacts. In other words,
a large separation of the extracting energies leads to high voltage and low current and
vice versa
. Note that
T
hc
is a parameter for equation (4.79) and for equation (4.77). By
elimination, we obtain the
W (V )
and from it the derived IV curve. The cell efficiency is
obtained from the maximum of
W (V )
or
W (T
hc
)
.
For
ε
g
→
0 the limit efficiency becomes
η
=
1
−
T
4
hc
T
4
s
1
−
T
a
T
hc
(
4
.
80
)
just as in the TPV converters, leading to a limiting efficiency of 85.4%.
The monoenergetic membrane for electron and hole transfer to the contacting
metals might perhaps be an insulator with an impurity band, but the nature of the phonon-
insulated absorber is totally unknown. Ross and Nozik [48, 49] who have developed
concepts close to those expressed here are investigating the poorer coupling of the elec-
trons with the photons in materials with quantum dots with the idea of favouring this
hot electron type of cell or the impact ionisation one that we have described in the
preceding section.
4.5.6 Intermediate Band Solar Cell
One of the causes of efficiency reduction in single-junction solar cells is the transparency
of the semiconductor to sub–band gap photons [50]. The inclusion of an intermediate
band (IB) may greatly increase the efficiency. We show in Figure 4.13 a band diagram
of the photon absorption and emission in this intermediate band material. Photons are
absorbed not only by pumping electrons from the VB to the CB as in a traditional solar
cell (photons with energy
hν
3
) but also by transitions from the VB to the IB (photons
with energy
hν
2
) and from the IB to the CB (photons with energy
hν
1
). In total, two low-
energy photons are used to pump an electron from the valence band to the conduction
band, passing through the intermediate band. This certainly increases the cell current.
The three absorption mechanisms detailed above are effective if the IB is a band
partially filled with electrons. In this way, there are empty states in it to accommodate
the electrons from the VB and there are electrons to sustain a strong pumping to the CB.
The detailed balance imposes photon emissions that are opposite to each one of the three
absorption mechanisms.
The cell is contacted as shown in Figure 4.14. The electrons are extracted from the
VB and returned to the CB using two layers of
n
and
p
ordinary semiconductors. A high
voltage is produced as the sum of the two semiconductor junction voltages occurring at
both sides of the IB material.
VERY HIGH EFFICIENCY CONCEPTS
145
Conduction band
Intermediate
band
Valence band
hv
2
hv
1
hv
3
e
g
e
h
e
l
Figure 4.13
The band diagram of an intermediate band solar cell
In this IB material we admit that there are three quasi-Fermi levels, one for each
band. The cell voltage is
q
times the splitting of the CB and VB quasi-Fermi levels.
In general, there is an energy threshold for each one of the absorption mechanisms
described earlier. However, the ideal structure is the one in which the upper energy (lintel)
of a photon that can be absorbed in certain mechanisms – involving, for example, transi-
tions from the VB to the IB – is the threshold of the next one – for example, transitions
from the IB to the CB. More specifically, calling
ε
g
the energy interval between the con-
duction and the valence band,
ε
l
the interval between the Fermi level in the intermediate
band and the valence band and
ε
h
=
ε
g
−
ε
l
, and assuming that
ε
l
< ε
h
, we consider that
all the photons in the interval (
ε
l
, ε
h
) are absorbed by transitions from the VB to the IB,
all the photons in the interval (
ε
h
, ε
g
) are absorbed by transitions from the IB to the CB
and all the photons in the interval (
ε
g
,
∞
) are absorbed by transitions from the VB to
the CB.
Under such conditions the equations that rule the current–voltage characteristic of
the cell are
I /q
= ˙
N (T
s
,
0
, ε
g
,
∞
, π )
− ˙
N (T , µ
CV
, ε
g
,
∞
, π )
+
[
˙
N (T
s
,
0
, ε
h
, ε
g
, π )
− ˙
N (T , µ
CI
, ε
h
, ε
g
, π )
]
(4.81)
˙
N (T
s
,
0
, ε
l
, ε
h
, π )
− ˙
N (T , µ
I V
, ε
l
, ε
h
, π )
= ˙
N (T
s
,
0
, ε
h
, ε
g
, π )
− ˙
N (T , µ
CI
, ε
h
, ε
g
, π )
(
4
.
82
)
with
µ
XY
being the photon chemical potentials equal to the separation of quasi-Fermi
levels between band
X
and band
Y
. Equation (4.81) states the balance of electrons in the
CB (addition of those that are pumped from the IB and the VB by means of the absorption
of the corresponding photon less those that radiatively recombine). Equation (4.82) states
a similar balance equation for the electrons in the IB and takes into account that no current
is extracted from the IB. In addition,
qV
=
µ
CV
=
µ
CI
+
µ
I V
(
4
.
83
)
Do'stlaringiz bilan baham: |