Double donor-thiophene dendron-perylene monoimide: Efficient light-harvesting metal-free chromophore for solid-state dye-sensitized solar cells

Article abstract of DOI:10.1002/asia.201000895

Efficient dye-namics: A highly efficient perylene sensitizer (ν=3.8%) for solid-state dye-sensitized solar cells was obtained. Key factors for this excellent photovoltaic performance are 1) careful balance of the orbital energies and color tuning, in orde

Full text of DOI:10.1002/asia.201000895

COMMUNICATION
DOI: 10.1002/asia.201000895
Double Donor-Thiophene Dendron-Perylene Monoimide: Efficient Light-
Harvesting Metal-Free Chromophore for Solid-State Dye-Sensitized Solar
Cells
Henrike Wonneberger,[a] Neil Pschirer,[c] Ingmar Bruder,[c] Jan Schçneboom,[c]
Chang-Qi Ma,[b] Peter Erk,[c] Chen Li,*[a] Peter Bꢀuerle,[b] and Klaus Mꢁllen*[a]
On the occasion of the 150th anniversary of the Department of Chemistry, The University of Tokyo
Dye-sensitized solar cells (DSCs) based on a stable large-
band nanostructured semiconductor, such as titanium diox-
ide, are low cost and easily processable alternatives to con-
ventional silicon wafers, and as such have lately drawn much
attention.[1] In particular, the solid-state DSCs show great
potential owing to their increased stability compared to
liquid DSCs. The reason for this stability is the exchange of
the liquid electrolyte, which often bears the problem of
leakage and electrode corrosion, for a solid hole-conducting
material, mainly 2,2’,7,7’-tetrakis(N,N-para-dimethoxyphenyl-
amino)-9,9’-spirobifluorene (spiro-MeOTAD).[2] However,
compared to the parent liquid DSCs, solid-state DSCs have
shown much lower efficiencies. Whilst ruthenium-based sen-
sitizers, and now a first porphyrin sensitizer,[3] have shown
efficiencies up to approximately 11% in liquid cells, solid-
state DSCs only reach values of up to 6%.[4] Most of the
more efficient sensitizers are ruthenium-based, which have
drawbacks such as cost, sustainability, and limited ease of
band-gap manipulation. One very stable and metal-free al-
ternative are sensitizers based on perylene monoimides,
which are known for their excellent chemical, photochemi-
cal, and thermal stability as well as high absorptivity and ac-
ceptor ability. Another outstanding class of chromophores
are thiophenes, in particular oligothiophenes, for their
highly variable optical properties according to their architec-
ture, extraordinary charge transport properties, and extinc-
tion.[2,5] Both perylenes and thiophenes have found wide ap-
plication in optoelectronic devices.
Herein, we present a donor–acceptor perylene monoimide
with a branched terthiophene spacer group[6,7] and a triphe-
nylamine donor moiety[8] (1a, Scheme 1) as well as a naph-
thalene analogue (1b, Scheme 1). As reported by Thomas
et al. and Fischer et al., the combination of a triphenylamine
donor and a branched oligothiophene spacer in combination
with a 2-cyanoacrylate acceptor gave good efficiencies of up
to 6.15%[6] and 6.8%[7] in liquid DSCs and up to 2.6%[7] in
a solid-state DSC. Furthermore, three moieties—triphenyla-
mine, oligothiophene, and perylene monoimide—have re-
cently caused a stir in a p-DSC (NiO), showing a sevenfold
increase in energy conversion efficiencies compared to pre-
ceding sensitizers.[9] However, we have designed our sensitiz-
ers for n-DSCs (TiO2), in which the perylene sensitizer 1a in
particular shows an outstanding efficiency of 3.8% under
1.5 AM illumination (1 sun). To the best of our knowledge,
this is an unprecedented performance for a perylene sensi-
tizer, the best perylene sensitizer for solid-state DSCs so far
being ID176 with an efficiency of 3.2%.[10]
[a] H. Wonneberger, Dr. C. Li, Prof. Dr. K. Mꢀllen
Max-Planck Institute for Polymer Research
Ackermannweg 10, 55128 Mainz (Germany)
Fax : (+49)6131-379-100
Both compounds were prepared by the initial introduction
of the terthiophene spacer group by Suzuki coupling with
the brominated perylene (or naphthalene) imide, successive
Suzuki coupling with the triphenylamine donor, and finally
saponification and imidization with glycine to yield the final
product (Scheme 1).
E-mail: lichen@mpip-mainz.mpg.de
muellen@mpip-mainz.mpg.de
[b] Dr. C.-Q. Ma, Prof. Dr. P. Bꢁuerle
Institute of Organic Chemistry II and Advanced Materials
University of Ulm
Albert-Einstein-Allee 11, 89081 Ulm (Germany)
[c] Dr. N. Pschirer, Dr. I. Bruder, Dr. J. Schçneboom, Dr. P. Erk
BASF SE
As described above, both sensitizers consist of an acceptor
unit with a carboxylic acid anchor in the imide structure, a
67056 Ludwigshafen (Germany)
ꢀ
ꢀ
branched terthiophene (a a connection and a b connection
of the thiophene units) and a triphenylamine donor. In
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/asia.201000895.
1744
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 1744 – 1747

HOMO, on the other hand, can
be found mainly on the donor
moiety, which is fully conjugat-
ed through the thiophene
groups. Both the HOMO and
LUMO also extend to the ter-
thiophene spacer connecting
the donor and acceptor. Hence,
considerable decoupling can be
observed, whilst still allowing
communication between the
donor and acceptor.
The branched terthiophene
moiety was chosen as a spacer
because thiophenes exhibit ex-
cellent charge-transport proper-
ties and enhance absorptivi-
ty.[2,5] As our group has shown
before, variation of the archi-
tecture of
a
terthiophene
moiety can significantly alter
the absorption properties.[12]
Using a branched terthiophene
instead of a single thiophene or
linear oligothiophene followed
our intentions of adding more
than two thiophene groups, but
still keeping the distance be-
tween donor and acceptor for
efficient intramolecular charge-
transfer reasonably short, and
being able to introduce
a
second donor moiety. As one
ꢀ
thiophene is a b connected,
the triphenylamine donor con-
nected to this thiophene loses
some of its strength due to
weak conjugation, but still can
generate electron transfer to-
wards the perylene (or naphtha-
lene) core and will not only
affect the HOMO level but also
shift the first reduction poten-
tial to a more-negative poten-
Scheme 1. Synthesis of the donor–acceptor–perylene monoimides with thiophene dendron spacer groups (1a
and 1b): a) K2CO3, [Pd(PPh3)4], toluene, water, ethanol, 808C, overnight (yields: 4a 77%, 4b 92%); b) NBS,
(PPh3)4], toluene, water, ethanol, 808C, over-
night (yields: 7a 43%, 7b 19%); d) 2-methyl-2-butanol, KOH, reflux, overnight (yields: 90% for n=1, 92%
for n=0); e) glycine, imidazole, 1408C, overnight (yields: 1a 89%, 1b 91%). TMS=trimethylsilyl, NBS=N-
bromosuccinimide, THF=tetrahydrofuran.
ACHTUNGTRENNUNG
THF, 08C!RT, 24 h (yields: 5a 82%, 5b 84%); c) K2CO3, [Pd
ACHTUNGTRENNUNG
order to achieve a stronger push–pull effect in the naphtha-
lene sensitizer, the outer benzene rings of the triphenyl-
amine donor bear para-methoxy groups.
tial. This elevation of the LUMO level energy enhances the
electron injection from the dye to the conduction band of
the TiO2. Moreover, this second spacer–donor branch in-
creases the bulkiness of the molecule and presumably pre-
vents aggregation, and hence unwanted recombination be-
tween the dye molecules.[13] Furthermore, bulkiness can
function as a barrier between the TiO2 and the hole-trans-
porting layer.[14] Obviously, this effect is always a trade-off
between the aforementioned positive effects and the unde-
sired consequence of lower dye concentrations on the TiO2,
thereby leading to less light harvesting.[15]
A strong push-pull effect is desirable for broad absorp-
tion, and this fundamental concept has been employed in
many donor–acceptor sensitizers. A strong vectorial excita-
tion from the donor to the acceptor supports efficient
charge separation. The insertion of a p-spacer leads to stron-
ger orbital partitioning (OP), thus impeding the recombina-
tion of electrons in TiO2 with holes located on the dye mole-
cules.[11] In fact, quantum-mechanical calculations show that
for compounds 1a and 1b the LUMO is mainly located on
the naphthalene and perylene cores, respectively. The
Concerning the absorption characteristics, 1a appears
dark purple, and 1b as brownish-orange to the human eye.
Chem. Asian J. 2011, 6, 1744 – 1747
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
1745

C. Li, K. Mꢀllen et al.
COMMUNICATION
The absorption spectra of both dyes are displayed in
Figure 1. 1a shows a very broad absorption over the whole
visible region with two strong absorption bands, one band
As for the application in DSCs, not only the absorption of
the dyes but also their energy levels play an important role
in determining their efficiency. The energy levels of the
HOMO and LUMO, that is, the ionization potential and the
electron affinity, were calculated using density functional
theory (see the Supporting Information) and afterwards de-
termined by cyclic voltammetry. Theoretical predictions for
1a gave an energy level of ꢀ5.0 eV for the HOMO and
ꢀ3.6 for the LUMO, whilst cyclic voltammetry measure-
ments in dry dichloromethane with 0.1m tetrabutylammoni-
um hexafluorophosphate as the supporting electrolyte gave
ꢀ5.0 eV and ꢀ3.5 eV, respectively. For 1b, the calculated
HOMO and LUMO energy levels were ꢀ4.7 eV and
ꢀ3.0 eV, respectively. Again, there is a good coherence with
the values determined by cyclic voltammetry: ꢀ4.9 eV
(HOMO), ꢀ3.1 eV (LUMO). This result corresponds to
0.6 V (HOMO) and ꢀ0.9 V (LUMO) for 1a and 0.5 V
(HOMO) and ꢀ1.3 V (LUMO) for 1b vs. NHE (normal hy-
drogen electrode). Comparing the reduction potentials of 1a
and ID176 in solution, the LUMO of 1a is considerably
higher than for ID176 (LUMO: ꢀ0.68 V vs. NHE; see the
Supporting Information). As the energy of the conduction
band of titaniumdioxide is approximately at ꢀ0.5 V versus
NHE, the driving force of electron injection is roughly
0.4 eV for the perylene derivative 1a and 0.8 eV for the
naphthalene derivative 1b. Thus, the driving force lies well
above the desired minimum driving force of 0.2 eV for effi-
cient electron injection.[16] The ionization potential of the
hole-transporting material spiro-MeOTAD ꢀ4.77 eV in
vacuum[17] is above the HOMO values of both dyes, thus al-
lowing efficient dye regeneration.
From the electrochemical properties, the naphthalene sen-
sitizer 1b, with its higher LUMO, has the potential for much
better electron injection. However, this sensitizer does not
have an efficiency as high as the perylene sensitizer 1a. One
reason for this observation could be the lower driving force
for the dye regeneration. Another cause could surely be
found in the additional light harvesting of 1a. Even though
the absorptivity for 1b in the region of 320–460 nm is
higher, 1a harvests and converts more sunlight through its
additional absorption band in the visible region which—
bearing in mind the solar spectrum—is of greater impor-
tance for photovoltaics than the UV region. This effect,
which could already be assumed from the absorption spectra
in dichloromethane or on TiO2, is affirmed by the EQE
spectrum and reflected in the ISC. Furthermore, the VOC of
1a and the fill factor (63% for 1a compared to 51% for 1b)
are higher resulting in a three times higher efficiency for 1a.
In summary, we have reported the synthesis, optical, elec-
tronic, and photovoltaic properties of a perylene-monoimide
sensitizer with a terthiophene dendron spacer and a triphe-
nylamine donor that shows an outstanding efficiency of
3.8% in a solid state DSC under 1.5 AM light illumination;
to date, this is the highest reported efficiency for perylene-
monoimide sensitizers as well as for similar sensitizer sys-
tems using a cyanoacrylate acceptor.[7] For comparison, we
prepared its naphthalene analogue, which gave a much
Figure 1. a) Absorption spectra (in CH2Cl2) of 1a and 1b; b) Absorbance
on TiO2 of 1a and 1b; c) IPCE of 1a and 1b; d) I–V curve of 1a and 1b.
around l=365 nm and a strong band resulting from the
ꢀ
p p* transition of the PMI at l=527 nm. On the other
hand, 1b shows one strong main band in the visible region
with an absorption maximum of lmax =378 nm. Moreover, a
weak charge transfer band is observed around l=500–
600 nm.
Both dyes were incorporated into s-DSCs following the
same procedure (see the Experimental Section). The ab-
sorptivity on TiO2 as well as the IPCEs show a clear superi-
ority of the perylene sensitizer in the visible region that is
relevant for the cell performance. Whereas 1b reaches a top
IPCE value of 42% at 420 nm and then drops steadily with
longer wavelength, 1a has IPCE values of around 50% (top
IPCE of 52% at 470 nm) from 460 to 580 nm and values of
over 40% from 400–630 nm. Moreover, the IPCE curve of
1a extends out to 800 nm, roughly 100 nm more bathochro-
mic than the curve of 1b (Table 1). Integration of the IPCE
spectra results in current densities of 8.6 mAcmꢀ2 for 1a
and 3.7 mAcmꢀ2 for 1b, in good agreement with the I/V
measurements.
Table 1. Optical and electrochemical properties of 1a and 1b.
lmax [nm]/
e [cmꢀ1 mꢀ1
LUMO[a]
/
/
ISC
VOC
[mV]
FF
[%]
h [%]
]
HOMO[a]
bandgap
[eV]
[mAcmꢀ2
]
ACHTUNGTRENNUNG
1a
1b
365/41.033
527/34.881
ꢀ3.5/
ꢀ5.0/
1.5
ꢀ3.1/
ꢀ4.9/
1.8
ꢀ8.7
ꢀ4.2
680
580
63
51
3.8
1.2
378/47.967
[a] determined by cyclic voltammetry.
1746
www.chemasianj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 1744 – 1747

Solid-State Dye-Sensitized Solar Cells
lower efficiency of 1.2%. Careful balance of the orbital en-
ergies and color tuning, in order to achieve a broad spec-
trum with high absorptivity as well as a sound degree of or-
bital partitioning and a considerate steric architecture of the
molecule, have been key factors for the excellent photovol-
taic performance of 1a.
[1] M. Grꢁtzel, Nature 2001, 414, 338–344.
[2] A. Mishra, M. K. Fischer, P. Bꢁuerle, Angew. Chem. 2009, 121,
2510–2536; Angew. Chem. Int. Ed. 2009, 48, 2474–2499.
[3] T. Bessho, S. M. Zakeeruddin, C. Y. Yeh, E. W. Diau, M. Grꢁtzel,
Angew. Chem. 2010, 122, 6796–6799; Angew. Chem. Int. Ed. 2010,
49, 6646–6649.
[4] X. Liu, W. Zhang, S. Uchida, L. Cai, B. Liu, S. Ramakrishna, Adv.
Mater. 2010, 22, E150–E155.
[5] C. Li, M. Liu, N. G. Pschirer, M. Baumgarten, K. Mꢀllen, Chem.
Rev. 2010, 110, 6817–6855.
Experimental Section
Synthesis procedures and product characterization (1H NMR, 13C NMR,
IR, UV/Vis absorption, high resolution mass spectroscopy) can be found
in the Supporting Information.
[6] K. R. J. Thomas, Y. C. Hsu, J. T. Lin, K. M. Lee, K. C. Ho, C. H. Lai,
Y. M. Cheng, P. T. Chou, Chem. Mater. 2008, 20, 1830–1840.
[7] M. K. R. Fischer, S. Wenger, M. K. Wang, A. Mishra, S. M. Zakeer-
uddin, M. Grꢁtzel, P. Bꢁuerle, Chem. Mater. 2010, 22, 1836–1845.
[8] Z. J. Ning, H. Tian, Chem. Commun. 2009, 5483–5495.
[9] A. Nattestad, A. J. Mozer, M. K. Fischer, Y. B. Cheng, A. Mishra, P.
Bꢁuerle, U. Bach, Nat. Mater. 2010, 9, 31–35.
[10] U. B. Cappel, M. H. Karlsson, N. G. Pschirer, F. Eickemeyer, J. Schç-
neboom, P. Erk, G. Boschloo, A. Hagfeldt, J. Phys. Chem. C 2009,
113, 14595–14597.
[11] a) J. Wiberg, L. J. Guo, K. Pettersson, D. Nilsson, T. Ljungdahl, J.
Martensson, B. Albinsson, J. Am. Chem. Soc. 2007, 129, 155–163;
b) C. Li, J. Schoneboom, Z. H. Liu, N. G. Pschirer, P. Erk, A. Herr-
mann, K. Mꢀllen, Chem. Eur. J. 2009, 15, 878–884.
[12] H. Wonneberger, C. Q. Ma, M. A. Gatys, C. Li, P. Bꢁuerle, K.
Mꢀllen, J. Phys. Chem. B 2010, 114, 14343–14347.
Preparation of Solid-State Dye-Sensitized Solar-Cell Layers
First, a TiO2 blocking layer was prepared on a fluorine-doped tin oxide
(FTO)-covered glass substrate using spray pyrolysis.[18] Next, a TiO2 paste
(Dyesol), diluted with terpineol, was applied by screen printing, resulting
in a film thickness of 1.8 mm. All films were then sintered for 1 h at
4508C, followed by treatment in a 40 mm aqueous solution of TiCl4 at
658C for 30 min, followed by another sintering step. The electrodes were
then dyed in 0.5 mm dye solution in CH2Cl2. The dye-coated TiO2 electro-
des with both dyes contain an organic coadditive (BASF-A1). Spiro-
MeOTAD was applied by spin-coating from a solution in chlorobenzene
also containing 20 mm LiACTHNUTRGNEUNG(CF3SO2)2N. Fabrication of the device was com-
pleted by evaporation of 200 nm of silver as the counter electrode. The
active area of the solid-state DSCs was defined by the size of these con-
tacts (0.13 cm2), and the cells were masked by an aperture of the same
area for measurements. Current–voltage characteristics (1000 Wmꢀ2, AM
1.5G) and the incident photon to current conversion efficiency (IPCE)
spectra were measured for all cells.
[13] H. Choi, C. Baik, S. O. Kang, J. Ko, M. S. Kang, M. K. Nazeeruddin,
M. Grꢁtzel, Angew. Chem. 2008, 120, 333–336; Angew. Chem. Int.
Ed. 2008, 47, 327–330.
[14] J. E. Kroeze, N. Hirata, S. Koops, M. K. Nazeeruddin, L. Schmidt-
Mende, M. Grꢁtzel, J. R. Durrant, J. Am. Chem. Soc. 2006, 128,
16376–16383.
[15] C. Li, Z. H. Liu, J. Schçneboom, F. Eickemeyer, N. G. Pschirer, P.
Erk, A. Herrmann, K. Mꢀllen, J. Mater. Chem. 2009, 19, 5405–5415.
[16] K. Hara, T. Sato, R. Katoh, A. Furube, Y. Ohga, A. Shinpo, S. Suga,
K. Sayama, H. Sugihara, H. Arakawa, J. Phys. Chem. B 2003, 107,
597–606.
[17] J. E. Kroeze, N. Hirata, L. Schmidt-Mende, C. Orizu, S. D. Ogier, K.
Carr, M. Grꢁtzel, J. R. Durrant, Adv. Funct. Mater. 2006, 16, 1832–
1838.
Acknowledgements
We gratefully acknowledge the financial support from the Deutsche For-
schungsgemeinschaft (DFG) (BA 880/3-1, ME 1246/15-1, MU 334/28-1),
the Bundesministerium fꢀr Bildung und Forschung (BMBF), and BASF
SE.
[18] B. Peng, G. Jungmann, C. Jager, D. Haarer, H. W. Schmidt, M. The-
lakkat, Coord. Chem. Rev. 2004, 248, 1479–1489.
Keywords: donor–acceptor systems · dye-sensitized solar
cells · organic photovoltaics · perylene · thiophene
Received: December 13, 2010
Published online: March 17, 2011
Chem. Asian J. 2011, 6, 1744 – 1747
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
1747