Communications
DOI: 10.1002/anie.201003357
Solar Cells
Enhanced Photovoltaic Performance of Low-Bandgap Polymers with
Deep LUMO Levels**
Huaxing Zhou, Liqiang Yang, Samuel C. Price, Kelly Jane Knight, and Wei You*
As a potential low-cost alternative to mainstream silicon solar
cells, bulk heterojunction (BHJ) polymer solar cells have
attracted a significant amount of attention in the research
community.[1] Fullerene derivatives (such as [6,6]-phenyl-C61-
butyric acid methyl ester, PC61BM) have been extensively
used as the n-type semiconductor in BHJ solar cells because
of their superior electron-accepting and transport behavior.
However, these fullerene derivatives are usually poor light
absorbers, thereby leaving the task of light absorbing to the
conjugated polymers. Moreover, fullerene derivatives usually
have fixed energy levels (e.g., a lowest unoccupied molecular
orbital (LUMO) of 4.3 eV), which dictate that the proposed
“ideal” conjugated polymer should exhibit a low highest
occupied molecular orbital (HOMO) energy level of À5.4 eV
and a small bandgap of 1.5 eV.[2] Therefore, a significant
amount of effort has been devoted to engineering the
bandgap and energy levels of conjugated polymers. As a
result, a few highly efficient polymers have been reported
with the record high efficiency surpassing 7%.[3]
To simultaneously lower the HOMO energy level and the
bandgap as required by the ideal polymer, a “weak donor–
strong acceptor” strategy was proposed.[2c] A few such
materials, by incorporating weak donor moieties based on
fused aromatic systems and a strong acceptor based on 4,7-
dithien-2-yl-2,1,3-benzothiadiazole (DTBT), have been suc-
cessfully demonstrated with high efficiency in typical BHJ
devices.[4] In these conjugated polymers, close to ideal HOMO
energy levels were achieved (e.g., À5.33 eV), which led to an
observed open circuit voltage (Voc) as high as 0.83 V.[4a]
However, the bandgaps of these materials were still larger
than the proposed 1.5 eV of ideal polymers, which explains
why mediocre short-circuit currents (Jsc) were obtained.
Logically, to further improve the efficiency, a smaller bandgap
is needed to achieve a higher short-circuit current (Jsc), while
the low HOMO energy level should still be maintained.
Fortunately, our previous study indicated that the LUMO of
donor–acceptor copolymers largely resides on the acceptor
moiety.[5] Therefore, we envisioned that incorporating a more
electron deficient acceptor to lower the LUMO would lead to
a smaller bandgap and maintain the low HOMO energy level
in the newly designed materials.
Compared with benzene, pyridine is p-electron deficient.
Therefore, if we replaced the benzene in the 2,1,3-benzothia-
diazole (BT) unit with pyridine, the new acceptor,
thiadiazolo[3,4-c]pyridine (PyT), would be one such stronger
acceptor. A similar strategy has been demonstrated recently
by Leclerc et al.[6] The copolymer of a carbazole unit with a
thienyl-flanked PyT unit (PCDTPT) did show a much lower
LUMO level compared with that of the copolymer with a BT
unit. However, a low efficiency was obtained, presumably
because of the low molecular weight and low solubility of
PCDTPT. To solve these issues, we employed the strategy of a
“soluble” acceptor[4a,5a] by flanking the PyT moiety with two
alkylated thienyl units, which converted the PyT into the new,
soluble, stronger acceptor DTPyT. As demonstrated in our
previous study,[5a] anchoring of alkyl chains to the 4-position
of the thienyl units of DTPyT would only significantly
improve the molecular weight and solubility of the resulting
polymers without introducing much steric hindrance.
Herein, we report the synthesis of a series of weak donor–
strong acceptor polymers, PNDT–DTPyT, PQDT–DTPyT,
and PBnDT–DTPyT, by copolymerizing various donor moi-
eties, namely naphtho[2,1-b:3,4-b’]dithiophene (NDT),
dithieno[3,2-f:2’,3’-h]quinoxaline (QDT), and benzo[1,2-
b:4,5-b’]dithiophene (BnDT), with the newly conceived
soluble DTPyT acceptor moiety (Scheme 1). Our preliminary
investigation on the photovoltaic properties of these polymers
in typical BHJ devices using PC61BM as the electron acceptor
showed highly respectable power conversion efficiencies
(PCEs) of over 5.5% for PQDT–DTPyT, and over 6% for
PBnDT–DTPyT and PNDT–DTPyT.
The synthesis of the alkylated DTPyT is modified from
the reported procedure[6] (see the Supporting Information for
experimental details). The other comonomers—alkylated
NDT, QDT, and BnDT—were prepared by established
literature procedures.[4a,7] Three polymers, PNDT–DTPyT,
PQDT–DTPyT, and PBnDT–DTPyT, were synthesized by
the microwave-assisted Stille polycondensation[1e] between
alkylated dibrominated DTPyT and the corresponding dis-
tannylated monomers. Crude polymers were purified by
[*] H. Zhou, S. C. Price, K. J. Knight, Prof. Dr. W. You
Department of Chemistry
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599-3290 (USA)
Fax: (+1)919-962-2388
E-mail: wyou@email.unc.edu
index.html
L. Yang, Prof. Dr. W. You
Curriculum in Applied Sciences and Engineering
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599-3287 (USA)
[**] The authors would like to thank the University of North Carolina at
Chapel Hill and the ONR (Grant No. N000140911016) for financial
support, and are grateful for a DuPont Young Professor Award and a
NSF CAREER Award (DMR-0954280). We also thank Prof. Richard
Jordan and Zhongliang Shen of the University of Chicago for GPC
characterization. LUMO=lowest unoccupied molecular orbital.
Supporting information for this article is available on the WWW
7992
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7992 –7995