8 Open-Circuit Voltage in Organic Solar Cells
205
To investigate the effect of doping near the D/A interface, trilayer devices with
an active layer consisting of H 2 Pc/doped-H 2 Pc/C 60 were fabricated. The device with
a 5 nm Cs 2 CO 3 -doped layer showed a V OC of 0.50 V, and the device with a 10 nm
Cs 2 CO 3 -doped layer showed an increased V OC of 0.52 V; these values are approximately the same as those of the bilayer device with a 50 nm Cs 2 CO 3 -doped layer
(0.52 V), as shown in Fig. 8.6c. These results indicate that 10-nm-thick doped layers
determined the magnitude of the increase in V OC , which is consistent with the bandmapping results. Notably, the device with a thinner Cs 2 CO 3 -doped layer showed
larger J SC and FF values. These results indicate that the addition of n-type dopants
only near the D/A interface could result in an increase in V OC while maintaining the
J SC and FF values.
8.3.3 Organic pn Homojunction Solar Cells
D/A heterojunctions are generally used in OSCs to split coulombically bound excitons formed in the semiconductor. As discussed in the Sect. 8.3.2, energy-level alignment creates an electric field at the interface, and this electric field can separate
excitons if they are formed at a single organic semiconductor interface. Doping was
applied to create a pn homojunction at an organic semiconductor interface, similar to
the pn junction in an inorganic SC [28]. A bilayer pn homojunction device was created
using diindenoperylene (DIP) as the host material, and MoO 3 and Cs 2 CO 3 as p-type
and n-type dopants, respectively (Fig. 8.7a, b). DIP is well known to have ambipolar
characteristics [29], which are necessary for the host material in pn homojunction
SCs, where electrons and holes are transported in a single material.
The J-V curves of the pn homojunction device with different dopant concentrations are displayed in Fig. 8.7c. The undoped device showed a very low J SC , as the
device could not split the coulombically bound excitons to generate a photocurrent.
In contrast, the pn homojunction device with higher dopant concentrations showed
a larger J SC . The J SC value became more than eight times larger with an increase
in dopant concentration from 0 to 5%. This is due to the large electric field (builtin-potential) near the pn homojunction interface formed by doping, as shown in
Fig. 8.7d. The internal quantum efficiency of the 5% doped device was found to be
30%, indicating that the pn homojunction created by doping can efficiently separate
excitons to form free charges.
However, V OC decreased with increasing dopant concentration, as shown in the JV curves. Analysis of the V OC variation helps to expose the critical difference between
organic and inorganic semiconductors with respect to how V OC is determined [30].
Plots of V OC of the pn homojunction devices as a function of temperature are shown
in Fig. 8.7e. The temperature dependence of V OC follows Eq. 8.6 in the Sect. 8.2.3,
and the magnitude of V OC is determined mainly by two parts: the energetic term (E
eff
g
is used here instead of E CT because the devices do not have a D/A interface), and the
recombination loss term, which is dependent on the temperature. The E
eff
g values of
the devices with 5% and 1% doped DIP, calculated from the intercepts of the plots,
205
To investigate the effect of doping near the D/A interface, trilayer devices with
an active layer consisting of H 2 Pc/doped-H 2 Pc/C 60 were fabricated. The device with
a 5 nm Cs 2 CO 3 -doped layer showed a V OC of 0.50 V, and the device with a 10 nm
Cs 2 CO 3 -doped layer showed an increased V OC of 0.52 V; these values are approximately the same as those of the bilayer device with a 50 nm Cs 2 CO 3 -doped layer
(0.52 V), as shown in Fig. 8.6c. These results indicate that 10-nm-thick doped layers
determined the magnitude of the increase in V OC , which is consistent with the bandmapping results. Notably, the device with a thinner Cs 2 CO 3 -doped layer showed
larger J SC and FF values. These results indicate that the addition of n-type dopants
only near the D/A interface could result in an increase in V OC while maintaining the
J SC and FF values.
8.3.3 Organic pn Homojunction Solar Cells
D/A heterojunctions are generally used in OSCs to split coulombically bound excitons formed in the semiconductor. As discussed in the Sect. 8.3.2, energy-level alignment creates an electric field at the interface, and this electric field can separate
excitons if they are formed at a single organic semiconductor interface. Doping was
applied to create a pn homojunction at an organic semiconductor interface, similar to
the pn junction in an inorganic SC [28]. A bilayer pn homojunction device was created
using diindenoperylene (DIP) as the host material, and MoO 3 and Cs 2 CO 3 as p-type
and n-type dopants, respectively (Fig. 8.7a, b). DIP is well known to have ambipolar
characteristics [29], which are necessary for the host material in pn homojunction
SCs, where electrons and holes are transported in a single material.
The J-V curves of the pn homojunction device with different dopant concentrations are displayed in Fig. 8.7c. The undoped device showed a very low J SC , as the
device could not split the coulombically bound excitons to generate a photocurrent.
In contrast, the pn homojunction device with higher dopant concentrations showed
a larger J SC . The J SC value became more than eight times larger with an increase
in dopant concentration from 0 to 5%. This is due to the large electric field (builtin-potential) near the pn homojunction interface formed by doping, as shown in
Fig. 8.7d. The internal quantum efficiency of the 5% doped device was found to be
30%, indicating that the pn homojunction created by doping can efficiently separate
excitons to form free charges.
However, V OC decreased with increasing dopant concentration, as shown in the JV curves. Analysis of the V OC variation helps to expose the critical difference between
organic and inorganic semiconductors with respect to how V OC is determined [30].
Plots of V OC of the pn homojunction devices as a function of temperature are shown
in Fig. 8.7e. The temperature dependence of V OC follows Eq. 8.6 in the Sect. 8.2.3,
and the magnitude of V OC is determined mainly by two parts: the energetic term (E
eff
g
is used here instead of E CT because the devices do not have a D/A interface), and the
recombination loss term, which is dependent on the temperature. The E
eff
g values of
the devices with 5% and 1% doped DIP, calculated from the intercepts of the plots,