Buckminsterfullerene derivatives bearing a fluoroalkyl group for use in organic photovoltaic cells

Article abstract of DOI:10.1016/j.jfluchem.2012.09.009

Six novel fluoroalkylpyrrolidine-substituted [60]fullerene derivatives were synthesized and their ability to perform as new n-type organic photovoltaic materials was evaluated. The fullerene derivatives were soluble in common organic solvents, affording good processability properties for the fabrication of photovoltaic cells, and showed an absorption range and molar extinctions similar to those of [6,6]-phenyl C₆₁ butyric acid methyl ester. Bulk-heterojunction photovoltaic cells using poly(3-hexylthiophene):fullerene derivative blends as the photovoltaic active layers were fabricated and characterized. The performances of the photovoltaic cells were notably affected by the substituents on the fullerene derivatives. Short fluoroalkyl (C ₄F₉) chains on pyrrolidine-linked phenyl groups were suitable substituents for the photovoltaic materials in the current study. This fullerene derivative bearing a C₄F₉-phenyl group showed a moderate power conversion efficiency of 0.53% during simulated AM 1.5 G solar irradiation at 100 mW/cm². This is so far the first report of the use of fluoroalkyl fullerene derivatives as the active materials in organic photovoltaic cells.

Full text of DOI:10.1016/j.jfluchem.2012.09.009

Journal of Fluorine Chemistry 144 (2012) 51–58
Contents lists available at SciVerse ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Buckminsterfullerene derivatives bearing a fluoroalkyl group for use in organic
photovoltaic cells
Makoto Karakawa a, Takabumi Nagai b, Tomomi Irita b, Kenji Adachi b, Yutaka Ie a, Yoshio Aso a,
*
a Department of Soft Nanomaterials, Nanoscience and Nanotechnology Center, The Institute of Scientific and Industrial Research (ISIR), Osaka University, 8-1 Mihogaoka, Ibaraki,
Osaka 567-0047, Japan
b Fundamental Technology Group, Chemical R&D Center, Daikin Industries, Ltd., 1-1 Nishi Hitotsuya, Settsu, Osaka 566-8585, Japan
A R T I C L E I N F O
A B S T R A C T
Article history:
Six novel fluoroalkylpyrrolidine-substituted [60]fullerene derivatives were synthesized and their ability
to perform as new n-type organic photovoltaic materials was evaluated. The fullerene derivatives were
soluble in common organic solvents, affording good processability properties for the fabrication of
photovoltaic cells, and showed an absorption range and molar extinctions similar to those of [6,6]-
phenyl C61 butyric acid methyl ester. Bulk-heterojunction photovoltaic cells using poly(3-hexylthio-
phene):fullerene derivative blends as the photovoltaic active layers were fabricated and characterized.
The performances of the photovoltaic cells were notably affected by the substituents on the fullerene
derivatives. Short fluoroalkyl (C4F9) chains on pyrrolidine-linked phenyl groups were suitable
substituents for the photovoltaic materials in the current study. This fullerene derivative bearing a
C4F9-phenyl group showed a moderate power conversion efficiency of 0.53% during simulated AM 1.5 G
solar irradiation at 100 mW/cm2. This is so far the first report of the use of fluoroalkyl fullerene
derivatives as the active materials in organic photovoltaic cells.
Received 2 July 2012
Received in revised form 25 September 2012
Accepted 25 September 2012
Available online 3 October 2012
Keywords:
Fullerene derivative
Perfluoroalkyl
n-Type semiconductor
Photovoltaic cell
ß 2012 Elsevier B.V. All rights reserved.
1. Introduction
derivatives have been developed in recent years [15–21]. Cao et al.
[15] reported the synthesis of methanofullerenes containing
Organic photovoltaic (OPV) cells based on conjugated polymers
and soluble fullerene derivatives have attracted a great deal of
attention because of their solution processability and mechanical
flexibility properties and their potential low-cost large-module
fabrication [1–14]. The soluble buckminsterfullerene derivative,
[6,6]-phenyl C61 butyric acid methyl ester (PC61BM), is currently
the most promising acceptor material for OPV cells. The power
conversion efficiency (PCE) of poly(3-hexylthiophene) (P3HT)/
PC61BM bulk-heterojunction (BHJ) solar cells has been improved
up to 6% under specially optimized conditions [1–5].
To make further progress in OPV materials and physics,
developing a deeper understanding of the compatibility, miscibili-
ty, and dispersibility in OPV materials is necessary to allow for the
fabrication of an ideal p–n juncture between a donor and an
acceptor material in a blend film. We cannot, however, select an
alternative acceptor material to PC61BM for OPV fabrication at
present. This lack of alternative acceptor materials has led to an
overall lack of fundamental understanding and has severely
limited the development of any further improvement in OPV
performance. In light of these limitations, several new fullerene
several long-alkyl ester groups from PC61BM. They subsequently
concluded that the synthesis of an acceptor that has good
compatibility with donor polymer materials is important for
improving device performance. Blom et al. [16] reported that
fullerene bis-adducts with higher LUMO levels effectively en-
hanced the open-circuit voltage and PCE values to a greater extent
than that of the corresponding fullerene mono-adducts. In 2008,
Wudl et al. [17] developed functionalized methanofullerenes
based on the PC61BM with a branched alkoxy side chain on a
phenyl group. These compounds provided PCE values of 1.73% from
a standard photovoltaic cell. The efficiency was further improved
up to 2.64% when a TiOx layer was introduced as a spacer between
the active layer and the Al cathode. In a separate development in
2009, Wudl et al. [18] reported an isomeric iminofullerene and its
positive effect in enhancing short-circuit currents. In a more
detailed study, two other groups [19–21] investigated the
relationship between material structure and photovoltaic perfor-
mance, and the studies revealed that the solubility of the fullerene
derivatives and their intermolecular interactions had an impact on
the OPV cell parameters. These reports have provided us with some
important information about the relationship between the
chemical structures of the materials and the device characteristics.
The suitably modified chemical structures of the materials can
induce not only the function of high carrier mobility but also the
* Corresponding author. Tel.: +81 6 6879 8476; fax: +81 6 6879 8479.
E-mail address: aso@sanken.osaka-u.ac.jp (Y. Aso).
0022-1139/$ – see front matter ß 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jfluchem.2012.09.009

52
M. Karakawa et al. / Journal of Fluorine Chemistry 144 (2012) 51–58
compatibility in the blend film, which are important character-
istics of the materials.
synthesized according to the literature [25–30]. Compound 10 was
purchased from Aldrich.
Chikamatsu et al. developed long-chain perfluoroalkyl-substi-
tuted C60 derivatives and investigated their field-effect transistor
(FET) properties. They demonstrated a field-effect mobility of the
order of 10ꢀ1 cm2/V s as well as the device-operation air stability of
the compounds [22,23]. We believed that high performance
materials for n-channel FETs such as perfluoroalkyl substituted
C60 derivatives could be potential candidates for acceptor
materials in the next generation of OPV cells. The characteristic
nature of the perfluoroalkyl group, however, must be considered.
For instance, Hashimoto et al. reported the attempted use of a
fullerene derivative with a fluoroalkyl chain (F-PC61BM) in an OPV
cell. The F-PC61BM did not work as the active material of the BHJ
type device, but was used as an additive to form a self-organized
buffer layer, improving the fill factor of the device as a consequence
of its strong phase separation ability [24]. As stated above, there
are no reported examples of the use of fullerene derivatives
containing fluoroalkyl groups as the active acceptor materials in
OPV cell.
In the current paper, we report the synthesis of several new
fullerene derivatives containing fluoroalkyl groups of different
lengths, together with their photophysical, electrochemical, and
photovoltaic properties. Our study of the BHJ OPV cell using
P3HT:fullerene-derivative blend film as the active layer provided a
PCE value up to 0.53% under simulated AM 1.5 G solar irradiation at
100 mW/cm2. This represents so far the first demonstration of the
successful use of a fluoroalkyl fullerene derivative as the active
material in organic photovoltaic cells.
2.2. Apparatus
UV–vis absorption spectra were recorded on Shimadzu UV-
3100PC. All spectra were obtained in spectrograde solvents. 1H and
19F NMR were recorded with JEOL JNM-ECS400 (400 MHz) and 13
C
NMR spectra were recorded with a Bruker BioSpin AVANCE III 700
(700 MHz) spectrometer in chloroform-d [chemical shifts in parts
per million (ppm) downfield from tetramethylsilane as an internal
standard for 1H and 13C and upfield from fluorotrichloromethane
as an internal standard for 19F]. Cyclic voltammetry was carried out
on a BAS CV-50W voltammetric analyzer using a Pt electrode in a
0.1 mol/L Bu4NPF6 in chlorobenzene/CH3CN (5:1, v/v) solution. The
synthesized fullerene derivatives were purified by a recycling
mode of preparative gel-permeation liquid chromatography
(GPLC) with the use of Jasco LC-9201S equipped with a refractive
index detector using JAIGEL 1H/2H columns and Japan Analytical
Industry LC-908 equipped with JAIGEL 1H/2H columns (eluent:
CHCl3). The HPLC measurements were carried out by gel
permeation chromatography in chloroform at 40 8C. Hitachi UV/
vis detector (L-2400), Hitachi pump (L-2130), TOSOH GPC system
(GPC-8200 model II ver. 6.0), and Shodex column (GPC K-803L and
GPC K-G Guard Column) were used. The flow rate was 1.0 mL/min.
Column chromatography was performed on silica gel, KANTO
Chemical silica gel 60N (40–50
mm). Elemental analyses were
performed on Perkin Elmer LS-50B by the Elemental Analysis
Section of Comprehensive Analysis Center (CAC) of ISIR, Osaka
University. The thicknesses of the thin films were measured using a
HORIBA Jobin Yvon UVISEL LT NIR-NNG spectroscopic ellips-
ometer. The surface morphology of the deposited organic film was
observed using an atomic force microscope (Shimadzu, SPM9600).
X-ray diffraction measurements of the organic films were carried
2. Experimental
2.1. Materials
out on Rigaku Ultima IV using Cu K
a curved graphite monochromator. The diffractions were mea-
sured from 38 to 308 in the 2 scan mode with a 0.018 step.
a radiation (40 kV, 40 mA) with
Reagents were purchased from Wako Pure Chemical Industries,
Tokyo Kasei Chemical Industries, Merck Ltd., and Aldrich, and used
without further purification. [60]Fullerene was purchased from
Honjo Chemical Corporation. Compounds 1, 8, 9 (Scheme 1), amino
acid derivatives [2-(methylamino)decanoic acid, N-dodecylglycine,
and 2-(dodecylamine)hexadecanoic acid], and perfluroalkylbenzal-
dehyde derivatives (4-heptadecafluorooctylbenzaldehyde, 4-tride-
cafluorohexylbenzaldehyde, and 4-nonafluorobenzaldehyde) were
u–u
2.3. Device fabrication and measurements
Glass slides patterned with ITO (Sanyo Vacuum Industries Co.,
Ltd.) were treated in an O2 plasma oven for 5 min. Thereafter, the
Scheme 1. Synthetic route to the fullerene derivatives, 1–9, and the structure of compound 10.

M. Karakawa et al. / Journal of Fluorine Chemistry 144 (2012) 51–58
53
ITO-coated grass slides were spin-coated (500 rpm for 5 s and
3000 rpm for 60 s) with a filtered (0.45 m pore size PTFE
142.59, 142.63, 142.64, 142.75, 143.05, 143.20, 144.40, 144.46,
144.61, 144.76, 145.23, 145.27, 145.32, 145.39, 145.49, 145.54,
145.57, 145.59, 145.68, 145.71, 145.99, 146.00, 146.16, 146.20,
146.23, 146.26, 146.32, 146.37, 146.38, 146.47, 147.38, 152.67,
m
membrane filter) aqueous suspension of poly(ethylene dioxythio-
phene) doped with poly(styrenesulfonic acid) (PEDOT-PSS) (Cle-
vios P VP Al 4083), and then baked at 135 8C for 10 min in air. P3HT
(Merck Ltd.) and a fullerene derivative (1:1 or 2:1, wt/wt) were
dissolved in chloroform at room temperature and then spun-cast
on the top of the PEDOT-PSS layer at 1500 rpm for 60 s. The
thicknesses of the resulting P3HT/fullerene-derivative films were
found to be ca. 180 nm, which were confirmed by ellipsometry
measurements. The device fabrication was completed by the
vacuum deposition of Al cathode (ca. 800 nm). The active area of
each solar cell device was 3 mm ꢁ 3 mm. The solar cells were
subsequently tested under simulated air mass (AM) 1.5 G solar
irradiation (100 mW/cm2, SAN-EI ELECTRIC CO., LTD. XES-301S).
Current–voltage (J–V) characteristics were recorded using a PC-
controlled Keithley 2400 source meter.
152.94, 154.05, 156.26; UV–vis (CHCl3): lmax
(e) = 256 nm
(114,000); Anal. calcd for C88H32NF17: C 74.11, H 2.26, N 0.98;
Found: C 74.04, H 2.35, N 1.00.
2.4.4. C60-fused N-dodecyl-5-tetradecyl-2-(4-
heptadecafluorooctyl)phenylpyrrolidine (4)
4 was synthesized according to the typical procedure; Yield
56.9%; 1H NMR (400 MHz, CDCl3,
d): 0.80–1.00 (m, 6H), 1.00–2.05
(m, 44H), 2.41–2.60 (m, 1H), 2.60–2.75 (m, 1H), 2.90–3.05 (m, 1H),
3.05–3.20 (m, 1H), 4.96 (d, J = 8.0 Hz, 1H), 5.63 (s, 1H), 7.60 (d,
J = 8.0 Hz, 2H), 7.88 (d, J = 8.0 Hz, 2H); 19F NMR (372 MHz, CDCl3,
d
): ꢀ80.59 (t, J = 10.0 Hz, 3F), ꢀ110.53 (t, J = 12.2 Hz, 2F), ꢀ120.86
to ꢀ121.20 (m, 2F), ꢀ121.46 to ꢀ122.10 (m, 6F), ꢀ122.42 to
ꢀ122.72 (m, 2F), ꢀ125.85 to ꢀ126.06 (m, 2F); 13C NMR (175 MHz,
2.4. Synthesis
CDCl3, d): 14.22, 14.24, 22.82, 25.00, 27.53, 28.29, 28.92, 29.45,
29.51, 29.62, 29.79, 29.82, 29.84, 29.88, 30.02, 32.06, 46.32, 70.65,
73.59, 76.55, 76.71, 106–120 (m, C8F17), 126.83 (bs), 128.78 (t, JC–C–
F = 24.3 Hz), 130.14 (bs), 135.93, 136.48, 136.90, 136.96, 139.30,
139.60, 140.06, 140.27, 141.56, 141.64, 141.88, 142.01, 142.03,
142.20, 142.25, 142.33, 142.35, 142.63, 142.66, 142.71, 142.88,
143.16, 143.26, 144.44, 144.59, 144.70, 144.74, 145.23, 145.25,
145.31, 145.34, 145.45, 145.48, 145.55, 145.61, 145.73, 145.76,
146.06, 146.08, 146.22, 146.24, 146.35, 146.37, 146.38, 146.40,
146.63, 146.99, 147.41, 147.43, 152.89, 153.64, 153.91, 157.46;
2.4.1. Typical procedure for the synthesis of fulleropyrrolidine
derivative
A mixture of amino acid (4.0 mmol), benzaldehyde (1.0 mmol),
and C60 (2.0 mmol) was refluxed in toluene (300 mL) under a N2
atmosphere for 20–64 h. After vacuum evaporation of the solvent,
the crude product was purified by preparative GPLC using
chloroform as eluent to give the desired product.
2.4.2. C60-fused N-methyl-5-octyl-2-(4-
heptadecafluorooctyl)phenylpyrrolidine (2)
UV–vis (CHCl3): lmax
102H60NF17: C 75.50, H 3.73, N 0.86; Found: C 75.33, H 3.74, N
0.95.
(e) = 256 nm (106,000); Anal. calcd for
C
2 was synthesized according to the typical procedure; Yield
24.0%; 1H NMR (400 MHz, CDCl3): 0.87 (t, J = 8.0 Hz, 3H), 1.18–1.67
(m, 10H), 1.67–1.93 (m, 2H), 2.50–2.64 (m, 1H), 2.50–2.64 (m, 1H),
2.86 (s, 3H), 4.91 (d-d, J = 8.1, 2.4 Hz, 1H), 5.58 (s, 1H), 7.62 (d,
J = 8.0 Hz, 2H), 7.90 (d, J = 8.0 Hz, 2H); 19F NMR (372 MHz, CDCl3,
2.4.5. C60-fused N-dodecyl-2-(4-
tridecafluorohexyl)phenylpyrrolidine (5)
5 was synthesized according to the typical procedure; Yield
16.8%; 1H NMR (400 MHz, CDCl3,
d): 0.89 (t, J = 6.8 Hz, 3H), 1.20–
1.60 (17 H, m), 1.60–1.75 (m, 1H), 1.82–2.05 (m, 2H), 2.50–2.62 (m,
1H), 3.12–3.23 (m, 1H), 4.15 (d, J = 9.6 Hz, 1H), 5.12 (d, J = 9.6 Hz,
1H), 5.13 (s, 1H), 7.63 (d, J = 8.7 Hz, 2H), 7.97 (bs, 2H); 19F NMR
d
): ꢀ80.65 (t, J = 9.8 Hz, 3F), ꢀ110.58 (t, J = 13.1 Hz, 2F), ꢀ120.90 to
ꢀ121.20 (m, 2F), ꢀ121.55 to ꢀ121.95 (m, 4F), ꢀ121.95 to ꢀ122.15
(m, 2F), ꢀ122.44 to ꢀ122.74 (m, 2F), ꢀ125.88 to ꢀ126.10 (m, 2F);
13C NMR (175 MHz, CDCl3,
d): 14.14, 22.71, 26.25, 29.00, 29.22,
29.38, 29.94, 31.86, 35.29, 73.78, 75.38, 106–120.5 (m, C8F17),
126.84 (bs), 128.85 (t, JC–C–F = 23.5 Hz), 129.99 (bs), 135.79, 136.47,
136.75, 139.33, 139.58, 141.51, 141.59, 141.78, 141.82, 141.94,
141.97, 142.08, 142.12, 142.15, 142.17, 142.28, 142.47, 142.57,
142.63, 142.65, 142.81, 143.11, 143.20, 144.37, 144.50, 144.60,
144.67, 145.19, 145.20, 145.28, 145.38, 145.41, 145.44, 145.45,
145.49, 145.66, 145.69, 146.00, 146.03, 146.15, 146.17, 146.28,
146.31, 146.33, 146.50, 146.79, 147.35, 147.37, 152.76, 153.59,
(372 MHz, CDCl3,
d
): ꢀ80.66 (t, J = 9.9 Hz, 3F), ꢀ110.56 (t,
J = 13.7 Hz, 2F), ꢀ121.10 to ꢀ121.40 (m, 2F), ꢀ121.90 to
ꢀ122.10 (m, 2F), ꢀ122.60 to ꢀ122.80 (m, 2F), ꢀ125.80 to
ꢀ126.10 (m, 2F); 13C NMR (175 MHz, CDCl3,
d): 14.13, 22.70,
27.50, 28.33, 29.39, 29.67, 29.70, 29.74, 31.94, 53.27, 66.86, 68.92,
76.33, 81.85, 106–120.5 (m, C6F13), 127.06 (bs), 128.84 (t, JC–C–
F = 24.2 Hz), 129.62 (bs), 135.59, 135.98, 136.41, 136.92, 141.50,
141.67, 141.82, 141.95, 142.01, 142.03, 142.06, 142.07, 142.12,
142.14, 142.24, 142.30, 142.56, 142.58, 142.59, 142.70, 143.01,
143.15, 144.36, 144.42, 144.56, 144.71, 145.18, 145.22, 145, 27,
145.35, 145.45, 145.49, 145.54, 145.55, 145.63, 145.67, 145.94,
145.96, 146.11, 146.15, 146.19, 146.22, 146.28, 146.31, 146.34,
146.42, 152.62, 152.87, 153.99, 156.21; UV–vis (CHCl3): lmax
157.15; UV–vis (CHCl3): lmax (e) = 256 nm (105,000); Anal. calcd
for C85H26NF17: C 73.76, H 1.89, N 1.01; Found: C 73.49, H 2.11, N
1.06.
2.4.3. C60-fused N-dodecyl-2-(4-
heptadecafluorooctyl)phenylpyrrolidine (3)
3 was synthesized according to the typical procedure; Yield
45.6%; 1H NMR (400 MHz, CDCl3,
d): 0.89 (t, J = 8.0 Hz, 3H), 1.20–
(e) = 256 nm (93,000); Anal. calcd for C86H32NF13: C 77.89, H 2.43,
N 1.06; Found: C 77.69, H 2.65, N 1.05.
1.75 (m, 18H), 1.80–2.06 (m, 2H), 2.52–2.66 (m, 1H), 3.10–3.26 (m,
2.4.6. C60-fused N-methyl-2-(4-nonafluorobutyl)phenylpyrrolidine
1H), 4.16 (d, J = 9.0 Hz, 1H), 5.13 (d, J = 9.0 Hz, 1H), 5.14 (s, 1H), 7.64
(6)
(d, J = 8.0 Hz, 2H), 7.98 (bs, 2H); 19F NMR (372 MHz, CDCl3,
d
):
6 was synthesized according to the typical procedure; Yield
ꢀ80.57 (t, J = 9.8 Hz, 3F), ꢀ110.55 (t, J = 13.7 Hz, 2F), ꢀ120.90 to
ꢀ121.20 (m, 2F), ꢀ121.50 to ꢀ122.10 (m, 6F), ꢀ122.44 to ꢀ122.70
55.5%; 1H NMR (400 MHz, CDCl3,
d
): 2.82 (s, 3H), 4.30 (d, J = 9.6 Hz,
1H), 5.01 (d, J = 9.6 Hz, 1H), 5.01 (s, 1H), 7.66 (d, J = 8.2 Hz, 2H), 7.97
(bs, 2H); 19F NMR (372 MHz, CDCl3,
): ꢀ80.85 (3F, t-t, J = 9.3,
(m, 2F), ꢀ125.85 to ꢀ126.05 (m, 2F); 13C NMR (175 MHz, CDCl3,
d):
d
14.12, 22.71, 27.51, 28.33, 29.39, 29.68, 29.70, 29.74, 31.96, 53.28,
66.92, 68.98, 76.39, 81.93, 106–120.5 (m, C8F17), 127.08 (bs),
128.94 (t, JC–C–F = 24.1 Hz), 129.80 (bs), 135.61, 136.00, 136.50,
136.96, 139.44, 139.94, 140.24, 140.26, 141.53, 141.72, 141.86,
141.87, 141.98, 142.06, 142.11, 142.17, 142.18, 142.28, 142.34,
2.7 Hz), ꢀ110.78 (t, J = 13.1 Hz, 2F), ꢀ122.59 to ꢀ122.79 (m, 2F),
ꢀ125.36 to ꢀ125.53 (m, 2F); 13C NMR (175 MHz, CDCl3,
d): 40.03,
70.07, 82.92, 106–120 (m, C4F9), 127.18 (bs), 128.94 (t, JC–C–
F = 24.2 Hz), 129.78 (bs), 135.65, 136.04, 136.38, 136.98, 139.51,
139.97, 140.27, 141.58, 141.74, 141.88, 141.90, 142.01, 142.04,

54
M. Karakawa et al. / Journal of Fluorine Chemistry 144 (2012) 51–58
142.11, 142.20, 142.26, 142.28, 142.60, 142.64, 142.65, 142.76,
143.06, 143.20, 144.39, 144.75, 145.25, 145.33, 145.36, 145.41,
145.54, 145.60, 145.68, 145.72, 146.00, 146.02, 146.17, 146.21,
146.27, 146.35, 147.37, 147.38, 152.49, 152.80, 153.79, 155.94;
UV–vis (CHCl3): lmax
(e) = 256 nm (116,000); Anal. calcd for
C
73H10NF9: C 81.80, H 0.94, N 1.31; Found: C 81.58, H 1.22, N 1.31.
2.4.7. C60-fused N-dodecyl-2-(4-nonafluorobutyl)phenylpyrrolidine
(7)
7 was synthesized according to the typical procedure; Yield
35.3%; 1H NMR (400 MHz, CDCl3,
d): 0.88 (t, J = 7.0 Hz, 3H), 1.10–
1.60 (m, 17H), 1.60–1.75 (m, 1H), 1.80–2.05 (m, 2H), 2.55–2.65 (m,
1H), 3.10–3.25 (m, 1H), 4.15 (d, J = 10.2 Hz, 1H), 5.12 (d, J = 10.2 Hz,
1H), 5.13 (s, 1H), 7.64 (d, J = 9.0 Hz, 2H), 7.97 (bs, 2H); 19F NMR
(372 MHz, CDCl3,
J = 12.6 Hz, 2F), ꢀ122.61 to ꢀ122.78 (m, 2F), ꢀ125.35 to ꢀ125.52
(m, 2F); 13C NMR (175 MHz, CDCl3,
): 14.13, 22.71, 27.51, 28.32,
d
): ꢀ80.85 (t-t, J = 9.3, 2.9 Hz, 3F), ꢀ110.76 (t,
d
Fig. 1. UV–vis absorption spectra of the fullerene derivatives, 1–10 and PC61BM in a
29.39, 29.68, 29.70, 29.74, 31.96, 53.27, 66.92, 68.98, 76.38, 81.95,
106–120 (m, C4F9), 127.08 (bs), 128.84 (t, JC–C–F = 24.5 Hz), 129.76
(bs), 135.69, 136.00, 136.44, 136.98, 139.44, 140.25, 140.27,
141.56, 141.72, 141.87, 142.01, 142.06, 142.10, 142.11, 142.17,
142.18, 142.29, 142.59, 142.63, 142.64, 142.75, 143.05, 143.20,
144.40, 144.46, 145.23, 145.27, 145.31, 145.33, 145.39, 145.40,
145.49, 145.54, 145.57, 145.59, 145.71, 145.99, 146.01, 146.20,
146.32, 146.37, 146.48, 147.38, 152.67, 152.94, 154.05, 156.26;
CHCl3 solution.
3.2. Electronic absorption properties
The UV–vis absorption spectra of the fullerene derivatives 1–10
together with the spectrum of PC61BM in chloroform are shown in
Fig. 1. All the compounds showed an absorption band in the range
of 250–350 nm and the peak maximum was observed at 260 nm.
Weak absorption peaks were also detected in all of the compounds
UV–vis (CHCl3): lmax
(e) = 256 nm (109,000); Anal. calcd for
C
84H32NF9: C 82.28, H 2.63, N 1.14; Found: C 82.09, H 2.91, N 1.20.
around 700 nm, which correspond to the
p–p* transition of the
3. Results and discussion
fullerene framework. It is clear from the spectra that the
substituent groups on the fullerenes do not affect the electronic
3.1. Synthesis of fluoroalkyl [60]fullerene derivatives
structure of the fullerene p system because the spectra were all
very similar in shape and showed similar molar extinction
coefficients.
The motivation for the design and synthesis of these fullerene
derivatives was to look for novel fullerene derivatives for
application as the acceptors in next generation OPV materials.
We were also keen to investigate the effect that incorporating
fluoroalkyl chains in fullerene derivatives have on the photovoltaic
properties. Our strategy for developing novel fullerene derivatives
3.3. Electrochemical properties
To investigate the electrochemical properties of the fullerene
derivatives and estimate their LUMO energy levels, cyclic
voltammetry (CV) measurements were conducted in a 5:1 (v/v)
mixture of chlorobenzene:CH3CN. The LUMO energy levels
(ELUMOs) of the compounds were estimated using the following
empirical equation: ELUMO ¼ E11=2 þ 4:8 eV, where E11=2 is the value
of first reduction potential relative to that of ferrocene/ferroce-
nium (Fc/Fc+) [31]. Representative examples of the resulting CV
curves are shown in Fig. 2, and the numerical data of all the
compounds are summarized in Table 1 (the remaining CV curves
are shown in the supplementary material). For the sake of
comparison, the CV curve and associated data from PC61BM are
also shown in Fig. 2 and Table 1, respectively. All the fullerene
derivatives underwent reversible reduction processes and the CV
curves were unchanged following successive multiple potential
scans, indicating the high stability of the materials for electron
injection. It is clear from the CV curves and associated data that all
of the fullerene derivatives exhibited reduction potential values
similar to that of PC61BM. Thus the LUMO energy levels of all of the
derivatives are in the range of ꢀ3.6 to ꢀ3.7 eV. Although the
compounds contain the strong-electron-withdrawing perfluor-
oalkyl chains, these substituents do not affect the frontier-orbital
energy levels of the compounds.
with high carrier mobility involved the introduction of
fluoroalkyl group onto the fullerene framework through
a
a
pyrrolidine ring. It was envisaged that the fluoroalkyl group
would enhance the aggregation properties of the fullerene
derivatives and provide better intermolecular interactions. With
all of this in mind, a series of fulleropyrrolidine derivatives with a
p-perfluoroalkylphenyl group 1–7 were designed and synthesized.
The synthetic methodology is depicted in Scheme 1. Alkylated
amino acids [25–27] and p-perfluoroalkylbenzaldehyde [28]
derivatives are two key starting materials, which were prepared
according to standard synthetic procedures from the correspond-
ing amino acid and p-iodobenzaldehyde derivatives, respectively.
[60]Fullerene was treated under the Prato reaction conditions
[29,30] with the N-alkylated amino acids and the corresponding p-
perfluoroalkylbenzaldehyde derivatives. The reactions proceeded
smoothly and afforded moderate yields of the corresponding
fulleropyrrolidene derivatives, which were purified by preparative
GPLC using chloroform as eluent to remove any byproducts. Non-
fluorinated alkyl derivatives 8 and 9 were similarly prepared as
reference compounds. The desired products were obtained as black
or brown powders and were fully characterized with 1H, 19F, and
13C NMR spectroscopy, elemental analysis and HPLC measure-
ments (the HPLC chart are shown in the supplementary material).
N-Methylfulleropyrrolidine (10) was purchased from Aldrich. All of
the fulleropyrrolidine derivatives except compound 10 dissolve
well in chlorinated organic solvents, such as chloroform, chloro-
benzene, and o-dichlorobenzene, providing good processability
properties for OPV cell fabrication.
3.4. Photovoltaic properties
The OPV properties of the fluoroalkyl as well as non-fluoroalkyl
fullerene derivatives were investigated using BHJ OPV cells with a
device structure of ITO/PEDOT-PSS/P3HT:fullerene derivative/Al,
where the polymer P3HT was used as the electron donor and the

M. Karakawa et al. / Journal of Fluorine Chemistry 144 (2012) 51–58
55
20
10
0
20
(b)
(a)
10
0
-10
-20
-30
-40
-10
-20
-1.5
-1.0
-0.5
0.0
-2.0
-1.5
-1.0
-0.5
0.0
0.5
Potential (V vs Fc/Fc+)
Potential (V vs Fc/Fc+)
20
(d)
(c)
20
0
10
0
-10
-20
-30
-20
-40
-2.0
-1.5
-1.0
-0.5
0.0
0.5
-1.5
-1.0
-0.5
0.0
Potential (V vs Fc/Fc+)
Potential (V vs Fc/Fc+)
Fig. 2. Cyclic voltammograms of the fullerene derivatives: (a) compound 3, (b) compound 5, (c) compound 7, and (d) PC61BM, in a chlorobenzene/CH3CN (5:1, v/v) solution.
fullerene derivatives were used as the electron acceptors. The
active layer was fabricated by spin-cast from a chloroform solution
data of the OPV cells, including open-circuit voltage (Voc), short-
circuit current density (Jsc), fill factor (FF), and PCE values are
summarized in Table 2 for a clear comparison.
as opposed to
a chlorobenzene solution because the slow
evaporation of the chlorobenzene solvent induced the unfavorable
aggregation of the fluoroalkyl fullerene derivatives owing to the
strong intermolecular interactions between the fluoroalkyl groups.
For the sake of comparison on the same conditions, OPV cells using
fullerene derivatives without fluoroalkyl group were also fabricat-
ed according to the same procedures.
The current density–voltage (J–V) characteristics of the OPV
cells fabricated from compounds 3, 5, 6, and 7 in the absence of
light and under simulated AM 1.5 G solar irradiation at 100 mW/
cm2 are shown in Fig. 3. The OPV cell based on P3HT:PC61BM was
also fabricated and measured as a reference cell (the J–V curve of
the cell is shown in the supplementary material). The performance
The OPV cells containing long fluoroalkyl (C8F17) fullerene
derivatives as the acceptors showed significantly lower PCE values
of less than 0.14%. Furthermore, in spite of their good performance
from the perspective of the n-type field-effect transistor properties
of the compounds [32], the cells experienced a high series
resistance leading to lower FF and Jsc values. Relative to these
long fluoroalkyl fullerene derivatives, the PCE values of compounds
2 and 4 were half those of compounds 1 and 3. These differences
could be attributed to the chemical structures of compounds 2 and
4, which contain a long alkyl group at the C-5 position of the
pyrrolidine ring, whereas compounds 1 and 3 possess no alkyl
group at the same position. This led us to the formulation of a
tentative theory, which suggested that steric hindrance derived
from the long alkyl group at the C-5 position affected the
compatibility, leading to the observed lower PV cell performance.
Furthermore, N-substituents regardless of long or short alkyl
groups did not adversely affect the PV cell performance. Our
tentative theory was supported by the fact that the performance of
the OPV cell fabricated from compound 8 was improved to a PCE
value of 0.22% when no long alkyl groups were attached to the
pyrrolidine C-5 position and a 3-dodecylphenyl group was present
at the C-2 position. Unfortunately, compounds 9 and 10, which
contain different alkyl groups at the N-1 position of their
pyrrolidine rings, showed poor device performances. These poor
performance levels are attributed to their poor solubility in
chloroform and the other solvents usable for the fabrication at
room temperature, making it difficult to form a dense and uniform
film. It is worthy to note that compound 3 containing long alkyl
Table 1
Reduction potentials and LUMO energy levels of the fullerene derivatives.
Compound
E11=2 (V vs Fe/Fe+)
E21=2 (V vs Fe/Fe+)
LUMO (eV)a
1
ꢀ1.13
ꢀ1.14
ꢀ1.14
ꢀ1.17
ꢀ1.17
ꢀ1.11
ꢀ1.11
ꢀ1.13
ꢀ1.16
ꢀ1.14
ꢀ1.13
ꢀ1.52
ꢀ1.53
ꢀ1.52
ꢀ1.55
ꢀ1.54
ꢀ1.51
ꢀ1.51
ꢀ1.52
ꢀ1.55
ꢀ1.53
ꢀ1.51
ꢀ3.67
ꢀ3.66
ꢀ3.66
ꢀ3.63
ꢀ3.63
ꢀ3.69
ꢀ3.69
ꢀ3.67
ꢀ3.64
ꢀ3.66
ꢀ3.67
2
3
4
5
6
7
8
9
10
PC61BM
a
The values were estimated from the first reduction potentials of the
compounds.

56
M. Karakawa et al. / Journal of Fluorine Chemistry 144 (2012) 51–58
(a) 4.0
3.0
(b) 4.0
3.0
2.0
2.0
1.0
1.0
0.0
0.0
-1.0
-2.0
-3.0
-4.0
-1.0
-2.0
-3.0
-4.0
0.0
0.2
0.4
Voltage (V)
0.6
0.8
1.0
0.0
0.2
0.4
Voltage (V)
0.6
0.8
1.0
(c) 4.0
3.0
4.0
3.0
(d)
2.0
2.0
1.0
1.0
0.0
0.0
-1.0
-2.0
-3.0
-4.0
-1.0
-2.0
-3.0
-4.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Voltage (V)
Voltage (V)
Fig. 3. J–V curve of the OPV cells based on the blend of P3HT and the fullerene derivatives: (a) compound 3, (b) compound 5, (c) compound 6, and (d) compound 7 [dark current
density, dotted line; photo current density, solid line; triangle and square indicate the mixture ratios at 2:1 and 1:1 (D:A), respectively].
group at the N-1 position, achieved a higher level of compatibility
with P3HT than compound 1, which did not form a homogeneous
film because of the strong oil- and water-shedding properties of its
C8-perflluoroalkyl group. The coexistence of a long aliphatic alkyl
group with a long fluoroalkyl group appeared to effectively relieve
the oil- and water-shedding properties derived from the fluor-
oalkyl chain interactions.
The fullerene derivatives, 5, 6, and 7, which contain short
fluoroalkyl (C6F13 and C4F9) groups, showed better performance
levels in a series of fluoroalkyl fullerene derivatives because of
their good compatibility with P3HT in the film blend. The highest
performance in this series was achieved by the OPV cell fabricated
from compound 7, which contains a C4F9 group, with a Voc of
0.49 V, Jsc of 3.18 mA/cm2, and FF of 0.34, resulting in a PCE value of
0.53%. Compound 6 containing a C4F9 group also provided similar
performance data, with a Voc of 0.38 V, Jsc of 3.40 mA/cm2, and FF of
0.41, resulting in a PCE level of 0.41%. These devices clearly
demonstrate the contribution of the short fluoroalkyl-substituted
fullerenes to the photocurrent generation. The effects of the
fluoroalkyl and non-fluoroalkyl chain structures of the fullerene
derivatives on the photovoltaic properties were consistent and
reproducible. We believe that the better performances observed
from the C4-perfluoroalkyl fullerene derivatives result from the
tuned nature of the aggregation of the fluoroalkyl groups, which
leads to better compatibility with the donor polymer and results in
a more effective BHJ structure in the blend film. On the basis of this
assertion and the general findings of the current study, there is
obviously a need for compromise between the intermolecular
interactions of the fluoroalkyl groups and their compatibility with
the donor polymer during the design of novel fluoroalkyl fullerene
derivatives for use in OPV cells.
3.5. Surface morphology and film crystallinity
The morphology and crystallinity of the photoactive-layer are
important factors in determining the overall performance of BHJ
OPV cells [1–3]. With this in mind, atomic force microscopy (AFM)
and X-ray diffraction measurements were conducted to assess
these properties in our own materials.
Table 2
Output parameters of the OPV cells based on blend films of P3HT and the fullerene
derivative.
Compound
P3HT:fullerene
(w/w)
Jsc (mA/cm2)
Voc (V)
FF
PCE (%)
A
comparison of the AFM images for the C8- and C4-
perfluoroalkyl fullerene derivatives is shown in Fig. 4, clearly
showing the smooth and flat surface images of the films, which
were confirmed by the fact that the surface root-mean-square
roughness of the P3HT blend films with compounds 3, 4, 6, and 7
were 0.72, 1.34, 1.26, and 0.93 nm, respectively. There was,
however, an obvious difference in the blend films. Fig. 4c and d
shows a nano-order grain-like topography of the 6 and 7 blends,
respectively, which indicates the potential aggregation of the C4-
perfluoroalkyl fullerene derivatives confined within the P3HT
matrix. The nano-grains observed here may have contributed to
1
2:1
1:1
1:1
1:1
2:1
2:1
2:1
2:1
1:1
2:1
1:0.7
1.17
0.90
1.82
0.88
2.27
3.40
3.18
2.86
1.23
0.40
7.40
0.46
0.41
0.42
0.42
0.33
0.38
0.49
0.31
0.27
0.20
0.59
0.26
0.16
0.19
0.22
0.31
0.41
0.34
0.24
0.25
0.24
0.50
0.14
0.06
0.14
0.08
0.23
0.41
0.53
0.22
0.08
0.02
2.18
2
3
4
5
6
7
8
9
10
PC61BM

M. Karakawa et al. / Journal of Fluorine Chemistry 144 (2012) 51–58
57
Fig. 4. AFM images of the blend films for OPV cells using the fullerene derivatives: (a) compound 3, (b) compound 4, (c) compound 6, and (d) compound 7.
6000
5000
4000
3000
2000
1000
0
6000
5000
4000
3000
2000
1000
0
(a)
(b)
2θ ≈ 22°
(100)
5
2θ ≈ 22°
(100)
10
15
20
25
30
5
10
15
20
25
30
2θ
(deg)
2θ
(deg)
6000
5000
4000
3000
2000
1000
0
6000
5000
4000
3000
2000
1000
0
(c)
(d)
(100)
(100)
2θ ≈ 22°
2θ ≈ 22°
(300)
(300)
(200)
(200)
5
10
15
20
25
30
5
10
15
20
25
30
2θ
(deg)
2θ (deg)
Fig. 5. X-ray diffractograms of the blend films for OPV cells using the fullerene derivatives: (a) compound 3, (b) compound 4, (c) compound 6, and (d) compound 7.

58
M. Karakawa et al. / Journal of Fluorine Chemistry 144 (2012) 51–58
Appendix A. Supplementary data
the better photovoltaic performance observed in the C4-perfluor-
oalkyl fullerene derivatives relative to the C8-perfluoroalkyl
fullerene derivatives 3 and 4 (Fig. 4a and b) and the other fluoro
and non-fluoroalkyl fullerene derivatives (AFM images not shown).
The X-ray diffraction results revealed clear differences in the
crystallinity of the blend films (Fig. 5), with the results resembling
the trend observed in the corresponding AFM images (Fig. 4). In
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jfluchem.2012.09.
009.
References
Fig. 5a and b, the peak at 2
graphic direction along the alkyl side chain (1 0 0) of P3HT [33].
The peak at 2
d, both of the high order ((2 0 0) and (3 0 0)) and intense (1 0 0)
peaks were observed, suggesting that the structures of the films
fabricated from the C4-perfluoroalkyl fullerene derivatives were
highly crystalline. Differences in the morphology and crystallinity
of the films have a decisive influence on the formation of an
efficient charge generation interface in the OPV cells [8].
u
ꢂ 58 corresponds to the crystallo-
[1] W. Ma, C. Yang, X. Gong, K. Lee, A.J. Heeger, Advanced Functional Materials 15
(2005) 1617–1622.
[2] G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nature Materials
4 (2005) 864–868.
[3] K. Kim, J. Liu, M.M.A.G. Namboothiry, D.L. Carroll, Applied Physics Letters 90
(2007) 163511.
[4] H. Tang, G. Lu, L. Li, J. Li, Y. Wang, X. Yang, Journal of Materials Chemistry 20 (2010)
683–688.
u
ꢂ 228 arises from the ITO electrode [1]. In Fig. 5c and
[5] M.T. Dang, L. Hirsch, G. Wantz, Advanced Materials 23 (2011) 3597–3602.
[6] C.N. Hoth, S.A. Choulis, P. Schilinsky, C.J. Brabec, Advanced Materials 19 (2007)
3973–3978.
[7] S.-S. Kim, S.-I. Na, J. Jo, G. Tae, D.-Y. Kim, Advanced Materials 19 (2007) 4410–
4415.
4. Conclusion
[8] J.W. Jung, W.H. Jo, Advanced Functional Materials 20 (2010) 2355–2363.
[9] Y.-J. Cheng, S.-H. Yang, C.-S. Hsu, Chemical Reviews 109 (2009) 5868–5923.
[10] G. Dennler, M.C. Scharber, C.J. Brabec, Advanced Materials 21 (2009) 1323–1338.
[11] P.-L.T. Boudreault, A. Najari, M. Leclerc, Chemistry of Materials 23 (2011) 456–
469.
[12] H.J. Son, F. He, B. Carsten, L. Yu, Journal of Materials Chemistry 21 (2011) 18934–
18945.
[13] A. Facchetti, Chemistry of Materials 23 (2011) 733–758.
[14] H. Zhou, L. Yang, W. You, Macromolecules 45 (2012) 607–632.
[15] L. Zheng, Q. Zhou, X. Deng, M. Yuan, G. Yu, Y. Cao, Journal of Physical Chemistry B
108 (2004) 11921–11926.
[16] M. Lenes, G.-J.A.H. Wetzelaer, F.B. Kooistra, S.C. Veenstra, J.C. Hummelen, P.W.
Blom, Advanced Materials 20 (2008) 2116–2119.
[17] S.H. Park, C. Yang, S. Cowan, J.K. Lee, F. Wudl, K. Lee, A.J. Heeger, Journal of
Materials Chemistry 19 (2009) 5624–5628.
[18] C. Yang, J.Y. Kim, S. Cho, J.K. Lee, A.J. Heeger, F. Wudl, Journal of the American
Chemical Society 130 (2008) 6444–6450.
[19] P.A. Troshin, H. Hoppe, J. Renz, M. Egginger, J.Y. Mayorova, A.E. Goryachev, A.S.
Peregudov, R.N. Lyubovskaya, G. Gobsch, N.S. Sariciftci, V.F. Razumov, Advanced
Materials 19 (2009) 779–788.
[20] P.A. Troshin, E.A. Khakina, M. Egginger, A.E. Goryachev, S.I. Troyanov, A. Fuchs-
bauer, A.S. Peregudov, R.N. Lyubovskaya, V.F. Razumov, N.S. Sariciftci, Chem-
SusChem 3 (2010) 356–366.
[21] G. Zhao, Y. He, Z. Xu, J. Hou, M. Zhang, J. Min, H.-Y. Chen, M. Ye, Z. Hong, Y. Yang, Y.
Li, Advanced Functional Materials 20 (2010) 1480–1487.
[22] M. Chikamatsu, A. Itakura, Y. Yoshida, R. Azumi, K. Yase, Chemistry of Materials 20
(2008) 7365–7367.
[23] Y. Horii, K. Sakaguchi, M. Chikamatsu, R. Azumi, K. Yase, M. Kitagawa, H. Konishi,
Applied Physics Express 3 (2010) 101601.
[24] Q. Wei, T. Nishizawa, K. Tajima, K. Hashimoto, Advanced Materials 20 (2008)
2211–2216.
[25] J. Li, H. Grennberg, Chemistry: A European Journal 12 (2006) 3869–3875.
[26] E. Olmo, A. Plaza, A. Muro, A.R. Martinez-Fernandez, J.J. Nogal-Ruiz, J.L. Lopez-
Perez, A.S. Feliciano, Bioorganic and Medicinal Chemistry Letters 16 (2006) 6091–
6095.
[27] M.J. O’Donnell, W.A. Bruder, B.W. Daugherty, D. Liu, K. Wojciechowski, Tetrahe-
dron Letters 25 (1984) 3651–3654.
Novel [60]fullerene derivatives bearing a fluoroalkyl group
were synthesized, and their absorbance, electrochemical, and
photovoltaic properties were investigated. All of the fullerene
derivatives exhibited absorption ranges and reduction potential
values similar to those of PC61BM. The substituents on the
pyrrolidine ring did not affect the frontier-orbital energy levels of
the corresponding fullerene derivatives. The BHJ OPV cells were
fabricated with an ITO/PEDOT-PSS/P3HT:fullerene derivative/Al
configuration and examined under simulated AM 1.5 G solar
irradiation. The fullerene derivatives with a short fluoroalkyl (C4F9)
group gave a better level of performance than the other fluoroalkyl
fullerene derivatives, and clearly demonstrated the contribution of
the short fluoroalkyl fullerene derivative to the generation of the
photocurrent. Taken together, these results suggest the require-
ment for a compromise between the intermolecular interactions of
the fluoroalkyl groups and the compatibility of these with the
donor polymer during the design and synthesis of novel fluoroalkyl
fullerene derivatives for use in OPV cells.
These results represent so far the first report of the successful
use of fluoroalkyl fullerene derivatives as the acceptor material in a
polymer solar cell. As part of our ongoing research, we plan to focus
on the structural optimization of the fullerene derivatives to
further improve the morphology related to the PCE in the blend
film. Optimization work will involve changing the substituent
groups of the fullerene and replacing P3HT with other polymers to
effectively tune the compatibility of the polymer with the
fluoroalkyl fullerene derivatives.
[28] K.J.L. Paciorek, S.R. Masuda, J.G. Shih, J.H. Nakahara, Journal of Fluorine Chemistry
53 (1991) 233–248.
[29] P. Wang, B. Chen, R.M. Metzger, T. Da Ros, M. Prato, Journal of Materials Chemistry
7 (1997) 2397–2400.
Acknowledgements
[30] M. Maggini, G. Scorrano, M. Prato, Journal of the American Chemical Society 115
(1993) 9798–9799.
This work was supported through a cooperative research
program with Daikin Co., Ltd. We would like to extend our thanks
to the Comprehensive Analysis Center at the Institute for Scientific
and Industrial Research of Osaka University for their assistance with
the elemental analyses and NMR spectroscopic measurements.
[31] Y. Ie, K. Nishida, M. Karakawa, H. Tada, Y. Aso, Journal of Organic Chemistry 76
(2011) 6604–6610.
[32] Unpublished data, me ꢂ 10ꢀ2–10ꢀ3 cm2/V s.
[33] M. Karakawa, M. Chikamatsu, Y. Yoshida, M. Oishi, R. Azumi, K. Yase, Applied
Physics Express 1 (2008) 061802.