138
H. Ohkita
with time but can be resolved to two components of polymer singlet excitons and
polarons as described in Ref. 36. Figure 6.11c and f show the time evolution of
polymer singlet excitons, polarons, and the total of them. As shown in the figure,
polymer polarons are generated at 0 ps more efficiently in PNTz4T/PC 71 BM than
in PNOz4T/PC 71 BM blend films, suggesting that PNTz4T/PC 71 BM blends have
a larger fraction of mixed regions and or small domains. Subsequently, polymer
singlet excitons decay and polymer polarons rise with the same time constant: 5.5 ps
for PNTz4T/PC 71 BM and 100 ps for PNOz4T/PC 71 BM blend films. These time
constants are consistent with the photoluminescence quenching efficiency mentioned
above because the singlet exciton lifetime is as long as a few hundreds of picoseconds. For PNTz4T/PC 71 BM, the polaron signals decay after 100 ps, which is ascribed
to the geminate recombination because the decay is independent of the excitation
intensities. For PNOz4T/PC 71 BM, on the other hand, the polaron signals do not
decay at all on this time scale, suggesting efficient charge dissociation. In summary,
polymer polarons are generated from singlet excitons with >95% but ~30% of them
recombine geminately in PNTz4T/PC 71 BM blend films. In other words, the exciton
diffusion efficiency η ED is as high as >95% but the charge dissociation efficient η CD
is limited to ~70% in PNTz4T/PC 71 BM blend films. On the other hand, ~35% of
singlet excitons radiatively deactivate to the ground state and hence only ~65% of
them are converted into polymer polarons with a dissociation efficiency of ~100%
in PNOz4T/PC 71 BM blend films. In other words, the exciton diffusion efficiency
η ED is as low as ~65% but the charge dissociation efficient η CD is as high as ~100%.
The difference in η ED is due to the difference in crystalline domain size, which is
revealed by TEM and AFM images of the blend films. As described in [36], the
LUMO–LUMO energy offset is as large as 0.31 eV for PNTz4T/PC 71 BM and as
small as 0.12 eV for PNOz4T/PC 71 BM, which are estimated by CV measurements.
The small energy offset in PNOz4T/PC 71 BM is consistent with an effective bandgap
energy evaluated from temperature dependence of V OC and electroluminescence
being similar to the photoluminescence of PNOz4T pristine films. These results
show that the charge dissociation efficiency η CD is as high as ~100% even for such a
small energy offset of ~0.1 eV. As a result, photon energy loss, which is defined by the
difference between the optical bandgap E g and the open-circuit voltage energy eV OC ,
is as small as 0.56 eV for PNOz4T/PC 71 BM solar cells. As described in Sect. 6.2, it
has been believed that the energy offset should be large enough to break the Coulomb
binding energy of electron–hole pairs in excitons, which has been considered to be
more than 0.3 eV for efficient devices [40], although there was a study reporting
that the minimum driving force is as small as 0.1 eV for the charge transfer from
singlet excitons to the CT state [41]. In other words, our finding clearly contradicts
the conventional photovoltaic conversion mechanism. After this study, similar small
photon energy losses have been reported for many systems [42]. Thus, further studies
are needed for revising the conventional photovoltaic conversion mechanism.
H. Ohkita
with time but can be resolved to two components of polymer singlet excitons and
polarons as described in Ref. 36. Figure 6.11c and f show the time evolution of
polymer singlet excitons, polarons, and the total of them. As shown in the figure,
polymer polarons are generated at 0 ps more efficiently in PNTz4T/PC 71 BM than
in PNOz4T/PC 71 BM blend films, suggesting that PNTz4T/PC 71 BM blends have
a larger fraction of mixed regions and or small domains. Subsequently, polymer
singlet excitons decay and polymer polarons rise with the same time constant: 5.5 ps
for PNTz4T/PC 71 BM and 100 ps for PNOz4T/PC 71 BM blend films. These time
constants are consistent with the photoluminescence quenching efficiency mentioned
above because the singlet exciton lifetime is as long as a few hundreds of picoseconds. For PNTz4T/PC 71 BM, the polaron signals decay after 100 ps, which is ascribed
to the geminate recombination because the decay is independent of the excitation
intensities. For PNOz4T/PC 71 BM, on the other hand, the polaron signals do not
decay at all on this time scale, suggesting efficient charge dissociation. In summary,
polymer polarons are generated from singlet excitons with >95% but ~30% of them
recombine geminately in PNTz4T/PC 71 BM blend films. In other words, the exciton
diffusion efficiency η ED is as high as >95% but the charge dissociation efficient η CD
is limited to ~70% in PNTz4T/PC 71 BM blend films. On the other hand, ~35% of
singlet excitons radiatively deactivate to the ground state and hence only ~65% of
them are converted into polymer polarons with a dissociation efficiency of ~100%
in PNOz4T/PC 71 BM blend films. In other words, the exciton diffusion efficiency
η ED is as low as ~65% but the charge dissociation efficient η CD is as high as ~100%.
The difference in η ED is due to the difference in crystalline domain size, which is
revealed by TEM and AFM images of the blend films. As described in [36], the
LUMO–LUMO energy offset is as large as 0.31 eV for PNTz4T/PC 71 BM and as
small as 0.12 eV for PNOz4T/PC 71 BM, which are estimated by CV measurements.
The small energy offset in PNOz4T/PC 71 BM is consistent with an effective bandgap
energy evaluated from temperature dependence of V OC and electroluminescence
being similar to the photoluminescence of PNOz4T pristine films. These results
show that the charge dissociation efficiency η CD is as high as ~100% even for such a
small energy offset of ~0.1 eV. As a result, photon energy loss, which is defined by the
difference between the optical bandgap E g and the open-circuit voltage energy eV OC ,
is as small as 0.56 eV for PNOz4T/PC 71 BM solar cells. As described in Sect. 6.2, it
has been believed that the energy offset should be large enough to break the Coulomb
binding energy of electron–hole pairs in excitons, which has been considered to be
more than 0.3 eV for efficient devices [40], although there was a study reporting
that the minimum driving force is as small as 0.1 eV for the charge transfer from
singlet excitons to the CT state [41]. In other words, our finding clearly contradicts
the conventional photovoltaic conversion mechanism. After this study, similar small
photon energy losses have been reported for many systems [42]. Thus, further studies
are needed for revising the conventional photovoltaic conversion mechanism.