Photoinduced processes in fullerenopyrrolidine and fullerenopyrazoline derivatives substituted with an oligophenylenevinylene moiety

Article abstract of DOI:10.1039/b200432a

Fullerene derivatives A-3PV and B-3PV in which an oligophenylenevinylene trimeric subunit (3PV) is attached to C₆₀ through, respectively, a pyrrolidine or a pyrazoline ring have been prepared. The electrochemical and excited-state properties of

Full text of DOI:10.1039/b200432a

View Article Online / Journal Homepage / Table of Contents for this issue
Photoinduced processes in fullerenopyrrolidine and
fullerenopyrazoline derivatives substituted with an
oligophenylenevinylene moiety{
Nicola Armaroli,*a Gianluca Accorsi,a Jean-Paul Gisselbrecht,b Maurice Gross,*b
Victor Krasnikov,c Dimitris Tsamouras,c Georges Hadziioannou,*c{ Mar´ıa J.
Go´mez-Escalonilla,d Fernando Langa,*d Jean-Franc¸ois Eckerte and
Jean-Franc¸ois Nierengarten*e
aIstituto per la Sintesi Organica e la Fotoreattivita`, Laboratorio di Fotochimica,
Consiglio Nazionale della Ricerche, via Gobetti 101, 40129 Bologna, Italy.
E-mail: armaroli@frae.bo.cnr.it; Fax: 139-051-6399844
bLaboratoire d’Electrochimie et de Chimie Physique du Corps Solide, Universite´ Louis Pasteur
et CNRS (UMR 7512), 4 rue Blaise Pascal, 67008 Strasbourg, France.
E-mail: gross@chimie.u-strasbg.fr
cDepartment for Polymer Chemistry and Materials Science Center, University of Groningen,
Nijenborgh 4, 9747 AG Groningen, The Netherlands. E-mail: hadzii@chem.rug.nl
dFacultad de Ciencias del Medio Ambiente, Universidad de Castilla-La Mancha,
45071 Toledo, Spain. E-mail: flanga@amb-to.uclm.es
eInstitut de Physique et Chimie des Mate´riaux de Strasbourg, Groupe des Mate´riaux
Organiques, Universite´ Louis Pasteur et CNRS (UMR 7504), 23 rue du Loess
67037 Strasbourg, France. E-mail: Jean-Francois.Nierengarten@ipcms.u-strasbg.fr
Received 2nd January 2002, Accepted 22nd March 2002
First published as an Advance Article on the web 30th April 2002
Fullerene derivatives A-3PV and B-3PV in which an oligophenylenevinylene trimeric subunit (3PV) is attached
to C60 through, respectively, a pyrrolidine or a pyrazoline ring have been prepared. The electrochemical and
excited-state properties of the multicomponent arrays A-3PV and B-3PV have been investigated in solution
using the related oligophenylenevinylene derivative 3PV, fullerenopyrrolidine A and fullerenopyrazoline B as
reference compounds. In A-3PV quantitative OPV A C60 photoinduced singlet–singlet energy transfer has been
observed. Population of the lowest fullerene singlet excited state is followed by nearly quantitative intersystem
crossing to the lowest fullerene triplet excited state in CH2Cl2 and toluene, whereas OPV A C60 electron
transfer is able to compete significantly in the more polar solvent benzonitrile. In the case of B-3PV, the
excited-state properties are more complex due to the electron donating ability of the pyrazoline ring. As
observed for A-3PV, quantitative OPV A C60 photoinduced singlet–singlet energy transfer occurs in B-3PV.
However, in this case, the population of the lowest fullerene singlet excited state is followed by an efficient
electron transfer from the pyrazoline ring in CH2Cl2 and benzonitrile. In B-3PV, studies of the dependence of
photoinduced processes on solvent polarity, addition of acid, and temperature also reveal that this compound
can be considered as a fullerene-based molecular switch, the switchable parameters being the photoinduced
processes. Finally, A-3PV and B-3PV have been tested as active materials in photovoltaic devices and the
differences of light to energy conversion efficiencies found for the two compounds have been rationalised on
the basis of their photophysical properties.
upon light irradiation can be controlled by pH and solvent
Introduction
polarity.8–13 It should also be pointed out that the fullerene
sphere is a particularly interesting electron acceptor in artificial
photosynthetic models because of its symmetrical shape, its
large size, and the properties of its p-electron system.14 The
characteristics of C60 successfully compare with those of
common acceptors with smaller size such as benzoquinone.15,16
In fact, accelerated charge separation and decelerated charge
In the light of their particular electronic properties, fullerene
derivatives appear to be interesting building blocks for the
construction of new photochemical molecular devices.1–3 Ful-
lerene derivatives covalently bound to a number of donors
have been prepared and these systems may exhibit photo-
induced intramolecular processes such as electron or energy
transfer.1,2,4–7 Among others, amines are an interesting type of
electron donor, since their ability to reduce the carbon sphere
recombination have been observed in
a fullerene-based
acceptor–donor system when compared to the equivalent
benzoquinone-based system.15,16 This has been rationalized by
the smaller reorganization energy (l) of C60 compared with
those of other acceptors: the smaller reorganization energy of
{Electronic supplementary information (ESI) available: synthetic
procedures and full characterization of all new compounds. See
http://www.rsc.org/suppdata/jm/b2/b200432a/
{Present address: Ecole Europe´enne Chimie Polyme`res Mate´riaux
(ECPM), Universite´ Louis Pasteur, 25 rue Becquerel, F-67087
Strasbourg Cedex 2, France.
C60 positions the photoinduced charge separation rate upward
along the normal region of the Marcus parabolic curve,
while forcing the charge recombination rate downward in the
inverted region.17 The efficient photogeneration of long lived
DOI: 10.1039/b200432a
J. Mater. Chem., 2002, 12, 2077–2087
This journal is # The Royal Society of Chemistry 2002
2077

View Article Online
charge-separated states by photoinduced electron transfer is
of particular interest for initiating photocatalytic reactions or
for solar energy conversion.1,18
Following the first preparation of photovoltaic devices from
fullerene–oligophenylenevinylene conjugates,19 a great deal of
attention has been devoted to C60 fullerene arrays substituted
with p-conjugated oligomers such as oligophenylenevinylenes
(OPV)19–23 and related systems24–27 When compared to photo-
voltaic devices based on interpenetrating blends of donor
(conjugated polymer) and acceptor (C60), chemically linked
OPV–fullerene derivatives offer the possibility of obtaining a
homogeneous molecular network of molecular p–n junctions.
Even if the efficiencies of the molecular photovoltaic systems
are still low, it is important to highlight that the behavior of
a unique molecule in a photovoltaic cell and the study of its
electronic properties allow us to obtain easily structure–activity
relationships for a better understanding of the photovoltaic
system. In particular, photophysical studies of OPV–C60
hybrids in solution have demonstrated that ultrafast OPV A
C60 singlet–singlet energy transfer competes strongly with the
desired electron transfer process, and this can be one of the
reasons for their low photovoltaic efficiency.21
In this paper, we describe the synthesis and the electro-
chemistry of a new fullerenopyrazoline molecule (B) and of a
new OPV–C60 molecular array (A-3PV). In dichloromethane
solution, photoinduced electron transfer (pyrazoline A C60) is
detected in the former, whereas in the latter an OPV A C60
energy transfer is preferred. A more sophisticated array
containing both the pyrazoline and the OPV moiety has been
prepared and characterized (B-3PV), with the aim of integrat-
ing both the energy donor (OPV) and the electron donor
(pyrazoline) subunit onto the C60 moiety. The photophysical
properties of B, A-3PV, and B-3PV have been also studied
in toluene (PhMe) and benzonitrile (PhCN) both at room
temperature and at 77 K, in order to test the dependence and
competition of photoinduced processes by solvent polarity and
temperature. Finally, A-3PV and B-3PV have been tested as
active materials in photovoltaic devices and the differences
of light to energy conversion efficiencies found for the two
compounds rationalised on the basis of their photophysical
properties.
Scheme 2 Reagents and conditions: (i) C60, N-phenylglycine, toluene,
reflux, 16 h (10%).
azomethine ylide generated in situ from aldehyde 2 and
N-phenylglycine. The reaction of C60 with 2 in the presence
of an excess of N-phenylglycine in refluxing toluene afforded
the N-phenylfullerenopyrrolidine derivative A-3PV in 28%
yield. Thanks to the presence of the three dodecyloxy sub-
stituents, compound A-3PV is well soluble in common organic
solvents such as CH2Cl2, CHCl3, THF, or toluene, and
complete spectroscopic characterization was easily achieved. In
1
the H-NMR spectrum of A-3PV recorded in CDCl3, all the
anticipated signals are observed. In addition to the resonances
arising from the 3PV moiety and the phenyl ring, a singlet and
an AB quartet are seen at d 6.09 and 5.32 ppm, respectively,
corresponding to the signals of the protons of the pyrrolidine
ring in full agreement with the C1 symmetry of A-3PV. This
lack of symmetry resulting from the presence of the stereogenic
C atom in the pyrrolidine ring is also evidenced by the signal
dispersion observed in the 13C-NMR spectrum of A-3PV. The
structure of A-3PV is confirmed by FAB-mass spectrometry
with the molecular ion peak at m/z ~ 1672.8 ([M 1 H]1,
calculated for C126H98NO3: 1672.74).
Model fullerenopyrrolidine A was obtained by treatment of
C60 with benzaldehyde 330 and N-phenylglycine in refluxing
toluene (Scheme 2).
The preparation of B-3PV is depicted in Scheme 3.
Hydrazone 4 was obtained in 93% yield by treatment of
aldehyde 2 with p-nitrophenylhydrazine in refluxing ethanol in
the presence of acetic acid. Reaction of 4 with N-chlorosuc-
cinimide (NCS) and subsequent reaction of the resulting
nitrilimine (nitrile imide) intermediate with C60 under micro-
wave irradiation31,32 afforded B-3PV in 57% yield. The 1H- and
13C-NMR spectra of B-3PV show all the expected signals and
are in full agreement with its Cs symmetry. The structure of
B-3PV is also confirmed by MALDI-MS which shows the
expected molecular ion peak at m/z ~ 1717.6 ([M]1, calculated
for C125H93N3O5: 1717.13).
Model compound B was prepared from aldehyde 3 following
the same two step procedure (Scheme 4). Reaction of 3 with
p-nitrophenylhydrazine followed by treatment of the resulting
5 with NCS and subsequent reaction with C60 under microwave
irradiation gave B.
Results and discussion
Synthesis
The synthesis of 3PV and A-3PV is shown in Scheme 1.
Compound 1 was prepared as previously reported.28 Treat-
ment of 1 with CF3CO2H in CH2Cl2–H2O and subsequent
LiAlH4 reduction of the resulting aldehyde 2 yielded refer-
ence compound 3PV. The functionalization of C60 with the
OPV group was based on the 1,3-dipolar cycloaddition29 of the
Scheme 1 Reagents and conditions: (i) CF3CO2H, H2O, CH2Cl2, rt, 5 h (91%); (ii) C60, N-phenylglycine, toluene, reflux, 7 days (28%); (iii) LiAlH4,
THF, 0 uC, 1 h (86%).
2078
J. Mater. Chem., 2002, 12, 2077–2087

View Article Online
Table 1 Redox potential of the studied species observed in CH2Cl2 1
0.1 M Bu4NPF6 on a glassy carbon working electrode. All potentials
are given versus ferrocene as standard
Red3/Va Red2/Va Red1/Va
Ox1/Vb
Ox2/Vb
A 22.00
A-3PV 21.99
21.49
21.44
21.36
21.34
21.10
21.09
20.97
20.97
10.79 (Ep)
10.60 (Ep) 10.91 (Ep)
10.80 (Ep)
10.60 (Ep)
B
21.65c
B-3PV 21.64c
3PV
22.40 (Ep) 10.62 (Ep)
20.98
C60
21.81
21.37
aReversible reduction: Eo ~ (Epc 1 Epa)/2. bPeak potential at n ~
0.1 Vs21 for the irreversible oxidation step. cReduction of the nitro
group.
Scheme 3 Reagents and conditions: (i) p-nitrophenylhydrazine, AcOH,
EtOH, reflux, 4.5 h (93%); (ii) NCS, pyridine, CHCl3, 0 uC to rt, then
C60, Et3N, toluene, microwave irradiation, 40 min (57%).
Electrochemistry
The results of cyclic voltammetry experiments are reported
in Table 1. The reference compound A exhibits the typical
fullerenopyrrolidine-type three reversible one-electron reduc-
tion steps, at about 100 mV more negative potentials than
pristine C60.33,34 An irreversible one-electron oxidation step is
also detected in A which is not observed in C60 and is attributed
to the pyrrolidine nitrogen. Further oxidation also occurs, but
the process is not well defined due to electrode inhibition. 3PV
is oxidized at 10.62 V and reduced at a rather negative potential
(22.40 V), close to the electrolyte discharge. The three one-
electron reduction processes in A-3PV are easily attributed to
the fullerene unit, whereas the oxidation step occurs on the
OPV branch and not on the fullerenopyrrolidine nitrogen.
...
Fig. 1 Absorption spectra of A (——), B (
), and 3PV (– – –) in
Cl . Inset: fluorescence spectra, corrected for the detector response,
of A (——) and 3PV (– – –) in CH2Cl2 rigid matrix at 77 K.
CH2
2
77 K (lmax ~ 458 nm).37 The luminescence lifetime at room
temperature is monoexponential (1.3 ns), whereas at 77 K a
biexponential decay is observed (t1 ~ 2.8 ns, t2 ~ 12.3 ns),
probably as a consequence of the presence of different ‘‘frozen’’
rotameric forms in the solid matrix.38 Accordingly, the
fluorescence spectrum of 3PV at 77 K could be the result of
a superimposition of two slightly shifted spectral profiles, and
this would explain the lack of marked vibrational resolution
(Fig. 1). In general, the spectroscopic properties of 3PV are
in line with those of previously reported OPV’s in organic
solvents.39
Fullerenopyrrolidines are among the most widely investi-
gated C60 derivatives.29 The N-phenylfullerenopyrrolidine
derivative A exhibits photophysical properties very similar to
those of analogous N-methylfullerenopyrrolidines.20,29 For
the latter, ground and/or excited state electronic interactions
between the fullerene carbon sphere and the nitrogen atom
are not present as revealed, for instance, by fluorescence
experiments in solvents of different polarity.40 Therefore it is
reasonable that, for A, conjugation of the nitrogen atom with a
phenyl ring does not substantially affect the photophysical
properties of the distant fullerene chromophore. The room
temperature steady state electronic absorption spectrum and
77 K fluorescence band of A in CH2Cl2 are reported in Fig. 1.
The singlet excited state lifetimes and fluorescence quantum
yields of A in the same solvent are listed in Table 2. The
The fullerenopyrazolines
B and B-3PV exhibit three
reversible one-electron reduction steps; the first two processes
are attributable to the fullerene core. Fullerene derivatives
usually possess higher reduction potentials than pristine C60
but, due to the inductive effect of the pyrazoline nitrogen
directly attached to the carbon sphere, the potentials of pristine
C60 are restored in B and B-3PV.32,35 The third reduction at
21.65 V occurs on the nitrophenyl group as argued upon
comparison with nitrobenzene, which is reduced at 21.61 V
under these experimental conditions.36 B-3PV is oxidized at
10.60 V, most likely on the OPV branch, whereas the first
oxidation of B occurs at 10.80 V and can be localized on the
pyrazoline sp3 nitrogen.
In summary the electrochemical results suggest that in the
dyads A-3PV and B-3PV the first oxidation step is located on
the 3PV core and the two first reductions on the carbon cage.
Interestingly, the two moieties also behave as independent
redox centers in B-3PV where the OPV-type unit is conjugated
with the pyrazoline ring.
Photophysical properties
Reference compounds 3PV and A. The electronic absorption
spectrum of 3PV in CH2Cl2 solution is reported in Fig. 1. 3PV
exhibits intense fluorescence both in fluid CH2Cl2 solution at
298 K (lmax ~ 460 nm; Wfl ~ 0.77) and in a rigid matrix at
Scheme 4 Reagents and conditions: (i) p-nitrophenylhydrazine, AcOH, EtOH, reflux, 4 h (85%); (ii) NCS, pyridine, CHCl3, 0 uC to rt, then C60, Et3N,
toluene, microwave irradiation, 40 min (43%).
J. Mater. Chem., 2002, 12, 2077–2087
2079

View Article Online
Table 2 Fullerene luminescence data in CH2Cl2, unless otherwise
noted
298 K
max/nma Wem 6 104 t/nsb
77 K
max/nma,c t/nsb
l
l
A
714
5.5
5.5d
1.0
1.1
722
718
1.5
1.6
A-3PV
B
710d
706
f,g
v 0.1e
1.3
698
1.5
B 1 TFAh
B-3PV
690
0.76
f
700
1.1i 704i
0.75
f
f
708d
776i
v 0.1d
f
3.5i
0.16
1.2i
B-3PV 1 TFAh 694d
f
f
aFrom emission spectra corrected for the photomultiplier response.
blexc ~ 337 nm, single photon counting spectrometer (see Experi-
mental section). cSpectra obtained upon subtraction of the back-
ground signal due to the non-transparent matrix. dExcitation either
on the OPV-type (365 nm) or fullerene-type (450 nm) moiety.
Fig. 3 Absorption spectra (——) of A-3PV (top) and B-3PV (bottom)
in CH2Cl2 compared to those obtained (– – –) by summing the spectra
of the corresponding reference compounds A 1 3PV and B 1 3PV.
f
g
ey0.35 in toluene. Not measured due to signal weakness. v0.5 ns
in toluene. h30% (v/v) addition of trifluoroacetic acid (TFA). iIn
toluene solution.
Excitation of A-3PV at 365 nm in CH2Cl2 results in a 1000-
fold decrease of fluorescence intensity relative to the reference
compound 3PV under the same conditions. In parallel,
sensitization of the fullerene lowest singlet excited state is
observed since solutions of A and A-3PV displaying the same
absorbance at 365 nm exhibit the same fullerene fluorescence
intensity (Fig. 4). These results suggest that intramolecular
singlet–singlet OPV A C60 energy transfer takes place.20,21
This process is expected to be extremely fast (within about
1 ps following light excitation) considering the lifetime of
unquenched 3PV (1.3 ns) and the dramatic decrease (about
1000 times) of the 3PV fluorescence. Model calculations on
dipole–dipole Fo¨rster-type OPV A C60 energy transfer,
performed as described earlier,21 give an energy transfer rate
constant (ken) in the range 1 6 1012 to 1 6 1011 s21 for donor–
triplet–triplet transient absorption spectrum of A in toluene
is shown in Fig. 2; the triplet lifetime is 282 ns and 45 ms in
air-equilibrated and air purged solution, respectively.
Fullerenopyrrolidine array A-3PV. The absorption spectrum
of A-3PV in CH2Cl2 is reported in Fig. 3, along with the profile
obtained by summing the spectra of its component units A
and 3PV; some differences are evident. In OPV–C60 arrays
investigated earlier, where the through bond and through
space distances between the chromophores are identical to
the present case, matching between the experimental and the
model spectra was also not obtained.21 This was attributed to
different rotameric conformations of the OPV subunit unit
when attached (or not) to the cumbersome C60, rather than
to strong ground state electronic interactions between the
chromophores, that were not evident by electrochemistry.21
Even in the present case the electrochemical results do not
point to sizeable interactions; therefore a similar rationale to
explain the differences found in the spectral profiles of Fig. 3
(top panel) can be proposed. Interestingly, when OPV and C60
subunits are located further apart, the absorption spectra of the
arrays are identical to the sum of the component subunits.20
It is important to point out that the absorption features of
the A and 3PV chromophores (Fig. 1) allow excitation
selectivity in the dyad A-3PV to a large extent. In particular
in the range 230–300 nm a nearly selective excitation of the
C60 unit is possible, which becomes complete above 420 nm.
On the other hand in the 350–390 nm range a good degree of
selectivity for the OPV subunit is obtained; at 365 nm about
75% of the light is absorbed by such units.
˚
acceptor distances between 10 and 15 A and a critical transfer
radius Rc (i.e. the distance between the partners at which ken
equalizes the rate of the intrinsic deactivation of the donor)
˚
of 34 A. Thus energy transfer is expected to be extremely fast
as suggested earlier21 and recently confirmed experimentally for
OPV–C60 arrays.41
The occurrence of photoinduced energy or electron transfer
processes in C60-containing multicomponent arrays can be
controlled by solvent polarity. Usually, energy transfer occurs
in apolar media (e.g. toluene), whereas in polar solvents such as
Fig. 4 Top: fluorescence spectra of A (——) and A-3PV (– – –) in
toluene (PhMe), dichloromethane (CH2Cl2), and benzonitrile (PhCN);
lexc ~ 365 nm, A ~ 0.200 for all samples. Bottom: sensitized singlet
oxygen luminescence of A (#) and A-3PV ($) in the same solvents and
under identical experimental conditions as above. The same spectral
intensity ratios are obtained upon selective excitation of the fullerene
moiety at 430 nm for both type of experiments.
Fig. 2 Transient absorption spectrum of A at 298 K in toluene air-
equilibrated solution upon laser excitation at 355 nm (energy ~ 3 mJ
per pulse). The spectra were recorded at delays of 100, 200, 400, 600,
and 1000 ns following excitation. The inset shows the time profile of DA
(700 nm) from which the spectral kinetic data were obtained; the fitting
is monoexponential and gives a lifetime of 282 ns.
2080
J. Mater. Chem., 2002, 12, 2077–2087

View Article Online
benzonitrile electron transfer tends to prevail. These trends
have been observed in a number of cases42–44 including dyads
containing organic conjugated oligomers.22,26,45
We have tested the luminescence behaviour of A-3PV in
toluene (PhMe, static relative permittivity e ~ 2.4) and
benzonitrile (PhCN, e ~ 25.2), and compared it to that in
dichloromethane (CH2Cl2, e ~ 8.9). Excitation of the 3PV
moiety in PhMe and PhCN (and CH2Cl2, see above) leads to
the same results: a 1000-fold decrease of the OPV fluorescence
band intensity is observed both at 298 and 77 K. Importantly,
upon excitation of the OPV moiety (365 nm), quantitative
sensitization of the fullerene fluorescence is observed in PhMe,
as in CH2Cl2; on the contrary, in PhCN, the recovery of this
band is only 40% (Fig. 4). An identical trend in the relative
intensities of the fullerene fluorescence bands, as a function
of solvent polarity, is observed upon exclusive excitation
of the fullerene moiety at l ¢ 430 nm. These results indicate
that for A-3PV in benzonitrile, irrespective of the excitation
wavelength, there is a 60% loss of fullerene singlet excited
states, not observed in less polar media. This suggests that
an additional deactivation pathway, quite competitive with
the extensive intersystem crossing typical of C60 fullerenes,46 is
made available.
Fig. 5 Energy-level diagram describing excited state intercomponent
processes following light excitation of the OPV subunit in A-3PV.
Charge separated states are indicated as dashed lines, whereas full lines
represent localized singlet and triplet electronic excited levels. EnT and
ElT stand for energy-transfer and electron-transfer, respectively. For
details on how the energy of the charge-separated and electronic excited
states were estimated, see text.
The above experimental data can be rationalized with the
aid of the diagram shown in Fig. 5. The energies of the lowest
3PV (A-13PV*) and fullerene (1A*-3PV) singlet excited states
have been estimated from the highest energy feature of the
In order to get further insight into the photophysical
processes following light excitation of A-3PV, the yield of
formation of the lowest triplet excited state in the three solvents
was measured. This was accomplished by taking advantage of
the singlet oxygen sensitization process brought about by the
lowest triplet state of fullerenes according to the following
reaction scheme:46,47
3
77 K fluorescence spectra. A value of 1.50 eV for A*-3PV is
derived from the phosphorescence spectra of similar full-
erenopyrrolidines.54 The energy of the charge separated state in
dichloromethane is obtained from the first oxidation and
reduction potentials in this solvent (Table 1). The energy of the
charge separated states in toluene and benzonitrile, for which
electrochemical data are not available, are estimated by using
the Weller equation:55
hn
?
isc
?
O2
ꢀ
ꢀ
ꢀ
C60
1C60
3C60 ? C60z1O2
(1)
where isc denotes intersystem crossing, and 1O2* stands for
O2(1Dg), commonly named ‘‘singlet oxygen’’. The excited 1O2*
state deactivates to the ground state giving rise to
a
q2
characteristic IR emission band centred at l ~ 1268 nm.48
The quantum yield of singlet oxygen sensitization (WD) and of
fullerene triplet formation (WT) are found to be identical for
C60, for its closed cage49 or open cage50 derivatives as well as
DGCS~q(Eox(D){Ered(A)){E00{
4pe0esRDA
ꢁ
(2)
ꢀ
ꢁꢀ
q2
1
1
1
1
{
z
{
er es
8pe0 rz r{
for higher fullerenes like C70 and C76.52 Therefore the
51
where q is elementary charge, Eox(D) and Ered(A), are the
oxidation and reduction potentials of the donor and the
acceptor in the solvent with relative permittivity er for which
electrochemical data are available; e0 is the vacuum permitti-
vity; E00 is the spectroscopic energy of the excited state from
which the charge separation is promoted; RDA is the center-
to-center distance of the positive and negative charges in the
charge-separated state; r1 and r2 are the radii of the oxidized
determination of WD (derived by recording the sensitized IR
emission of 1O2*) can be taken as an indirect evaluation of WT
for all fullerenes.53 In Fig. 4 (bottom panel) are reported the
sensitized singlet oxygen luminescence bands of A-3PV in
the three solvents, with A used as internal reference in all cases.
The relative 1O2* luminescence intensity ratios (A : A-3PV) are
identical by exciting both at 365 nm (OPV moiety) and at
430 nm (C60 subunit).
and reduced species; es is the relative permittivity of the studied
solvent. For our calculations we employed r2 ~ 5.6 A and
The spectra of Fig. 4 show that the relative amount of singlet
that is formed (monitored by C60 fluorescence) is identical to
that of triplet (monitored by sensitized 1O2 luminescence) in
any solvent. The singlet (Table 2) and triplet (Table 3) lifetimes
of A and A-3PV are identical both in CH2Cl2 and in PhMe. In
PhCN, instead, the singlet lifetime is reduced from 2.4 ns (A) to
1.1 ns (A-3PV), whereas the triplet lifetime is unchanged for the
two compounds both in aerated and in deaerated solution
(Table 3).
56
˚
r1 ~ 4.6 A (roughly estimated from the van der Waals volume
˚
of 3PV).22 The center-to-center RDA distance was estimated to
˚
be 12.5 A from molecular modelling; the redox potentials are
in Table 1; finally we used E00 ~ 1.72 eV corresponding to the
energy of the lowest fullerene singlet excited state.
In toluene the pattern of photoinduced processes is
straightforward and identical to that in dichloromethane.
Table 3 Relative intensities of sensitized singlet oxygen luminescence and triplet lifetimes obtained by transient absorption spectroscopy
PhMe
CH2Cl2
PhCN
IRel
t
aer/nsa
t
dea/msb
IRel
t
aer/nsa
t
dea/msb
IRel
t
aer/nsa
t
dea/msb
A
100
100
v10
100
282
288
45
45
100
100
v10
v10
577
541
28
29
100
40
v10
v10
408
403
36
35
A-3PV
B
B-3PV
c
c
c
c
c
c
c
c
c
c
307
44
c
aAir-equilibrated solution. bAir-free solutions. The signals are too weak to obtain reliable results.
J. Mater. Chem., 2002, 12, 2077–2087
2081

View Article Online
The quenching of the OPV fluorescence, followed by
sensitization of the fullerene lowest singlet and triplet states,
suggests the occurrence of efficient intramolecular OPV A C60
energy transfer followed by fullerene intersystem crossing. In
PhMe electron transfer from 1A*-3PV is not observed since
it would be endergonic (DGCS ~ 0.34 eV). Importantly, such
a process would be slightly exergonic in CH2Cl2 (DGCS ~
20.03 eV), but it is disfavoured by a sizeable activation energy
barrier (see below).
In PhCN, the fullerene singlet is quenched with a rate
constant k ~ 4.9 6 108 s21, as derived from luminescence
lifetimes measurements. This process is ascribable to exergonic
electron transfer (DGCS ~ 20.29 eV), see Fig. 5. The identical
relative yields (compared to the reference compound A) of
singlet and triplet formation and the unquenched triplet
lifetimes show that electron transfer from the triplet state
does not occur in this solvent. This implies that the observed
triplet states are generated via intersystem crossing from the
upper singlet level (Fig. 5, dashed arrow).
(DGCS w 2l)17 and the reaction exergonicity is higher than
the activation barrier. This allows sizeable competition with
fullerene intrinsic deactivation, unlike the unfavourable cases
of CH2Cl2 and PhMe. Attempts to directly detect the charge
separated state via picosecond transient absorption spectro-
scopy in PhCN were unsuccessful since A-3PV decomposes
under long-term high energy laser irradiation. In particular,
UV–Vis ground state absorption spectra obtained after these
experiments show a dramatic decrease of the OPV-type band
centred at 365 nm. However, similar OPV–C60 dyads exhibit
intramolecular charge separated states living about 50 ps in
o-dichlorobenzene solution.41
Eqn. (6) describes the dependence of the rate for charge
separation (kCS) on the reorganization energy l, the electron
coupling between donor and acceptor V, and the activation
energy DG#CS, in the so-called non-adiabatic regime:56
"
#
DGC#S
2p3=2
pffiffiffiffiffiffiffiffiffiffiffi
lkBT
2
kCS
~
V exp {
(6)
kBT
h
Using the experimental kCS in benzonitrile (4.9 6 108 s21), we
can estimate a value of 26 cm21 for V, when electron transfer is
promoted from the lowest fullerene singlet excited state. Such a
weak donor–acceptor coupling41 is consistent with the electro-
chemical results and can support the interpretation of the
ground state absorption spectrum (Fig. 3, top panel).
Competition between energy and electron transfer. In the
investigated solvents the experimental results do not suggest
the occurrence of electron transfer from the lowest singlet
excited state of the 3PV moiety, though thermodynamically
allowed (Fig. 5). According to the Marcus theory the maximum
rate for charge separation is obtained for DGCS ~ 2l,17
where l is the reorganization energy and is the result of two
contributions:
Fullerenopyrazoline B. The absorption spectrum of B in
CH2Cl2 is reported in Fig. 1. An intense shoulder extending up
to 500 nm is recorded, not present in the fullerenopyrrolidine
A; a very weak band at about 680 nm (e # 200 M21 cm21) is
attributable to the S0 A S1 transition.
B exhibits a faint fluorescence band in CH2Cl2 at room
temperature (lmax ~ 706 nm), and the emission quantum
yield is about two orders of magnitude lower than that of A
(Table 2). The fluorescence intensity is dependent on solvent
polarity; the lower the polarity the stronger the emission
(Fig. 6). Fluorescence enhancing is obtained upon addition of
trifluoroacetic acid (TFA) in CH2Cl2. At about 30% addition
of TFA (v/v) an intensity plateau is reached (Fig. 6). The lumi-
nescence signal reverts to its initial intensity after extraction of
l~lizle
(3)
where li and le are the internal (nuclear rearrangements) and
external (of solvent origin) reorganization energies. For C60
dyads li is 0.3 eV,8,22 whereas le can be estimated from the
following equation:56,57
ꢀ
ꢁꢀ
ꢁ
q2
1
1
1
1
1
le~
z
{
{
(4)
4pe0 2rz 2r{ RDA
n2 es
where n is the solvent refractive index. le values of 0.05,
0.65, and 0.66 are calculated for PhMe, CH2Cl2 and PhCN,
respectively. Within the Marcus approach it is then possible
to calculate the energy of the activation barrier for charge
separation according to eqn. (5):17
(DGCSzl)2
DGC#S
~
(5)
4l
The data calculated from eqn. (2)–(5) and collected in
Table 4 help to rationalize the trend of photoinduced processes
displayed in Fig. 5. From these data the electron transfer from
A-13PV* is concluded to be in the Marcus inverted region
(DGCS v 2l)17 in all solvents and this disfavours successful
competition towards the ultrafast (subpicosecond)21,41 singlet–
singlet energy transfer. As for photoinduced processes arising
1
from the lowest fullerene singlet A*-3PV, it is seen (Table 4)
that the activation energy barrier in CH2Cl2 is too high
(0.22 eV) to allow the occurrence of charge separation, for
which DGCS ~ 20.03 eV; this explains the lack of quenching
processes in this solvent. In PhCN the charge separation
process is located in the normal region of the Marcus parabola
Fig. 6 (a) Fluorescence spectra of B in various solvents, under the same
experimental conditions; lexc ~ 430 nm, A ~ 0.400. (b) Fluorescence
spectrum of B in CH2Cl2 upon addition of increasing amount of acid
up to 30% (v/v). lexc ~ 424 nm, isosbestic point.
Table 4 Reorganization energy (l), free energy change (DGCS), and activation barrier (DG#CS) for charge separation within A-3PV in toluene
(PhMe), CH2Cl2 and benzonitrile (PhCN) upon excitation of the 3PV (A-13PV*) or fullerene (1A*-3PV) lowest singlet states
A-13PV*
1A*-3PV
l/eV
DGCS/eV
DG#CS/eV
DGCS/eV
DG#CS/eV
PhMe
CH2Cl2
PhCN
0.35
0.95
0.96
20.79
21.16
21.42
0.14
0.01
0.06
0.34
20.03
20.29
0.34
0.22
0.12
2082
J. Mater. Chem., 2002, 12, 2077–2087

View Article Online
a function of
the CH2Cl2–TFA solution with a saturated water solution of
NaOH.
Table 5 Occurrence of fullerene fluorescence as
temperature and solvent
All these results suggest that the lowest fullerene singlet
excited state is quenched by electron transfer from the lone
pair of the pyrazoline sp3 nitrogen, in analogy with similar
substituted fullerenes.8–13 Accordingly, we can describe the
fullerenopyrazoline B as a redox dyad in which the electron
donor and acceptor units are linked through an sp3 carbon
as a very simple spacer, as suggested by Sun et al. for
aminofullerene derivatives.9
298 K
PhMe
77 K
CH2Cl2
PhCN
PhMe
CH2Cl2
PhCN
A
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
OFF
OFF
A-3PV
B
B-3PV
ONa
OFF
OFF
OFFb
ON
OFFb
OFFb
aReduced by 60% relative to A. bSome weak luminescence signal is
detected, that displays solvent dependence (Fig. 6). Anyway here we
labeled it as OFF, in a relative comparison with B-3PV in TOL, for
which the ‘‘genuine’’ unquenched emission of the fullerenopyrazoline
chromophore is observable.
Fullerenopyrazoline array B-3PV. The absorption spectrum
of B-3PV in CH2Cl2 is reported in Fig. 3 (bottom panel),
along with the profile obtained by summing the spectra of
the component units B and 3PV. Substantial differences are
evident, definitely more pronounced than those found in the
analogous comparison made for A-3PV (Fig. 3, top panel).
This can be ascribed to the extended conjugation of the 3PV
moiety in B-3PV as a result of the presence of an additional
double bond; accordingly the absorption of the OPV subunit is
found to be red-shifted.
Similarly to B a very weak fullerene-type fluorescence band
is detected in CH2Cl2 for B-3PV (lexc ¢ 430 nm), therefore, the
same pyrazoline A C60 electron transfer quenching process
must be active in the multicomponent array. Although in
B-3PV the OPV moiety is easier to oxidize than the pyrazoline
subunit (Table 1), prompt photoinduced electron transfer from
the heteroaromatic ring is observed. This can be related to its
closer vicinity to the electron accepting C60 moiety.
B exhibits faint emission in PhMe and CH2Cl2 (Fig. 6,
Wfluo v 1025). Nonetheless, recovery of fluorescence intensity
is achieved at 77 K, and ‘‘regular’’ fullerene singlet lifetimes of
1.2–1.5 ns are measured; this further substantiates the electron
transfer quenching mechanism at 298 K, previously discussed.
In the highly polar solvent PhCN electron transfer is not
prevented even in the rigid medium since the fluorescence is
OFF also at 77 K; in this solvent an identical behaviour is
found for B-3PV.
The trend of B-3PV fullerene fluorescence as a function of
solvent, temperature, and acid concentration is quite intri-
guing. In particular we note the following differences, relative
to B: (i) a strong band in PhMe at 298 K, (Fig. 7, Tables 2 and
5); (ii) less pronounced intensity recovery upon addition of acid
(about ten times less, Table 2).
These findings suggest that the electron donor unit of B (the
sp3 nitrogen lone pair) is not comparably effective in B-3PV.
We tentatively attribute this surprising behaviour to deloca-
lization of such electronic pair over the conjugated OPV–
pyrazoline system in B-3PV, which implies weakening of its
reducing (and basic) character. This process is able to prevent
pyrazoline A C60 electron transfer in apolar PhMe at 298 K,
but not OPV A C60 energy transfer.
Excitation of B-3PV in CH2Cl2 at 365 nm, where a sub-
stantial part of the light is absorbed by the 3PV chromophore
(Figs. 1 and 3), results in a 1000-fold decrease of fluorescence
intensity relative to the reference compound 3PV under the
same conditions. Although sensitization of the dramatically
quenched fullerene fluorescence cannot be used as a probe, this
quenching is attributed to singlet–singlet OPV A C60 energy
transfer. For B-3PV, we have calculated that Fo¨rster-type
singlet–singlet energy transfer can be as much as twice faster
than for A-3PV (see above);21 on the other hand electron
transfer from B-13PV* is deeper in the inverted region (DGCS
~
21.28 eV) than in the case of the fullerenopyrrolidine dyad
(DGCS ~ 21.16 eV), assuming the same l value (0.95 eV,
eqn. (4)) for both arrays. This suggests that electron transfer is
most likely not able to compete with energy transfer in B-3PV,
in analogy with A-3PV. Unfortunately, also for such an array,
direct detection of the charge separated state is prevented due
to sample decomposition under prolonged laser irradiation.
Control over photoinduced processes in B-3PV. It has been
demonstrated that in CH2Cl2 solution photoinduced energy
transfer occurs in A-3PV, whereas electron transfer is evi-
denced in B. B-3PV can be considered a more sophisticated
hybrid where both an OPV and a pyrazoline moieties are
coupled with the carbon sphere subunit. Electrochemical
results suggest a certain degree of electronic insulation between
the OPV fragment and the rest of the molecule but, on the other
hand, the photophysical results (see absorption spectra and
fluorescence in toluene) would not lead to the same conclusion.
This could raise debate on whether or not B-3PV can be strictly
defined as a molecular OPV–pyrazoline–fullerene triad; this is
a questionable matter and we prefer to stick to experimental
facts. In CH2Cl2 quenching of both the OPV and of the
Solvent and temperature dependence of the fullerene fluor-
escence. The addition of the same amount of acid to solutions
of B and B-3PV in CH2Cl2, causes a much lower recovery of
fullerene fluorescence for the latter compound (Table 2); this
suggests differences in the electron transfer quenching process
from the pyrazoline unit. We thus investigated in detail the
dependence of such emission bands on solvent polarity and
temperature, two factors that also affect electron-transfer
processes. In Table 5 is reported the presence or the lack (ON/
OFF) of the C60-type fluorescence band in PhMe, CH2Cl2,
and PhCN at 298 and 77 K; some indication of the emission
relative intensities are specified in the footnotes. The data for
A and A-3PV are also listed.
For the latter dyad the data of Table 5 provide further
support for extensive electron transfer (about 60% yield, see
above) in PhCN at 298 K. In fact the C60-type fluorescence
intensities of optically matched solution of A and A-3PV are
identical at 77 K, unlike at 298 K. The lack of solvent
repolarization in the rigid matrix destabilizes the A2-3PV1
charge separated state to such an extent that electron transfer
is no longer possible, differently from the behaviour at room
temperature (Fig. 5).
Fig. 7 Corrected fluorescence spectra of B-3PV (——) and B (– – –) in
toluene at 298 and (inset) 77 K. lexc ~ 430 nm, A ~ 0.400.
J. Mater. Chem., 2002, 12, 2077–2087
2083

View Article Online
Table 6 Summary of the photovoltaic characteristics for the devices prepared from A-3PV and B-3PV
ISC/A cm22
FFa
Tb (%)
S400/A W21
(%)
400
a
ge
Compound
VOC/V
A-3PV
B-3PV
a
0.38
0.39
3.4 6 1027
1.3 6 1027
0.29
0.26
31
3
3.3 6 1024
1.3 6 1024
3.8 6 1023
1.3 6 1023
The calculation of the overall energy conversion efficiency ge, has been done using the equation ge~ V
where VOC, ISC, FF and PINC
OCISCFF
PINC
are the open circuit potential, short circuit current, filling factor and incident light power, respectively. The filling factor is given by
MAXIMAX
FF~ V
where VMAX and IMAX are voltage and current respectively, at the point of maximum power output. Standard intensity of light
was 1 mW cm22 bT ~ transmission (corrected on glass, ITO and PEDOT-PSS).
VOCISC
.
fullerene fluorescence occur and this can be attributed to,
respectively, (OPV A C60) energy and (pyrazoline A C60)
electron transfer in analogy with the behaviour of B and
A-3PV. The occurrence of distinct photoinduced processes can
be confirmed by emission spectra in PhMe. In this solvent we
observe quenching of the OPV fluorescence moiety accom-
panied by quantitative sensitization of the fullerene fluor-
escence. The occurrence of this latter band signals the
suppression of photoinduced electron transfer in the apolar
medium, while the OPV A C60 energy transfer process is kept
active.
We can thus state that B-3PV is a multicomponent array
featuring both an energy (3PV) and an electron transfer
(pyrazoline) terminal, the fullerene moiety being able to act as
energy or electron acceptor. Depending on the excitation
wavelength one can control the nature of the photoinduced
process. By exciting the OPV unit (e.g. at 365 nm, y60%
selectivity) an energy transfer to the C60 moiety, followed by
pyrazoline A C60 electron transfer is obtained; by exciting the
C60 fragment (l w 500 nm, 100% selectivity) only electron
transfer is triggered. Also chemical input (protons) and
temperature (in CH2Cl2, Table 5) can affect the latter process,
with no effect on the OPV A C60 energy transfer process, as
already pointed out.
contribution of photoinduced electron transfer from the OPV
moiety to the fullerene subunit. Most likely energy transfer
is the main deactivation pathway also in the solid state, in
line with the solution behaviour. This is in good agreement
with previous studies on related systems.21 Surprisingly, the
performances of the devices obtained with B-3PV are not
improved even if efficient electron transfer has been evidenced
for this compound. Actually, it seems that the charge separa-
tion involving the fullerene moiety and the pyrazoline N atom
is not able to contribute to the photocurrent production since it
does not involve the hole conducting moiety (the 3PV unit). In
other words, only the photoinduced electron transfer from
the OPV unit to the fullerene moiety is efficient for the
photovoltaic effect. Since the latter is a minor deactivation
pathway, only a small part of the absorbed light is able to
contribute effectively to the photocurrent.
Incorporation in photovoltaic cells. Compounds A-3PV and
B-3PV have been tested as active materials in photovoltaic
devices, and the results are summarized in Table 6. Each
C60–OPV conjugate has been sandwiched between poly(3,4-
ethylenedioxythiophene)poly(styrenesulfonate) (PEDOT-PSS)-
covered indium tin oxide (ITO) and aluminium electrodes. The
organic layers were prepared by spin coating from chloroform
solutions and observations of the surface morphology by
atomic force microscopy (AFM) in tapping mode revealed
the obtaining of continuous and uniform films for all the
compounds.
The current–voltage (I–V) curves of both devices have been
recorded in the dark and under illumination and are depicted in
Fig. 8. In the dark, the current is found to be higher by at least
two orders of magnitude for the ITO/PEDOT-PSS/A-3PV/Al
device when compared to the ITO/PEDOT-PSS/B-3PV/Al one.
The latter observation is likely due to better charge transport
properties of the thin films prepared from compound A-3PV.
In the case of B-3PV, the lower conductivity may be ascribed to
the presence of the pyrazoline N atom which is able to act as a
trap due to its electron donating ability as evidenced by the
electrochemical and photophysical studies.
Fig. 8 Current–voltage (I–V) characteristics in the dark ($) and under
illumination (#) for ITO/PEDOT-PSS/A-3PV/Al (top) and ITO/
PEDOT-PSS/B-3PV/Al (bottom).
Under illumination both devices show a clear photovoltaic
behaviour although the rectifying ratio is not significantly
large. The cells prepared from A-3PV display better overall
performance characteristics although, by comparing their
respective transmittance values, it is obvious that B-3PV
exhibits enhanced absorption ability for the examined spectral
region (Fig. 9). The sensitivity spectrum of both devices also
follows their respective absorption spectrum as depicted in
Fig. 9.
Fig. 9 Absorbance (bottom) and sensitivity (top) as a function of the
illumination wavelength for the devices: ITO/PEDOT-PSS/A-3PV/Al
(——); ITO/PEDOT-PSS/B-3PV/Al: (– – –).
The monochromatic power conversion efficiency of the
devices prepared from A-3PV is rather poor due to the low
2084
J. Mater. Chem., 2002, 12, 2077–2087

View Article Online
Electrochemistry
The electrochemical experiments were carried out at 20 ¡ 2 uC
in CH2Cl2 containing 0.1 M Bu4NPF6 in a classical three
electrode cell. The working electrode was a glassy carbon disk
(3 mm in diameter) used motionless for cyclic voltammetry
(sweep rates from 10 mV s21 to 10 V s21). The auxiliary and
the pseudo reference electrodes were platinum wires. All
potentials are refereed to the ferrocenium/ferrocene (Fc1/Fc)
couple, which was used as internal standard. CH2Cl2 (Merck,
˚
spectroscopic grade) was dried over molecular sieves (4 A) and
used as such. Bu4NPF6 (Fluka, electrochemical grade) was
used without further purification. The accessible potential
window on the glassy carbon electrode in CH2Cl2 was 11.4
to 22.4 V vs. Fc1/Fc. The cell was connected to a computerized
multipurpose electrochemical device AUTOLAB (Eco Chemie
BV, Utrecht, The Netherlands) controlled by the GPSE
software running on a personal computer.
Fig. 10 Schematic representation summarizing the photoinduced
processes occurring in A-3PV, B, and B-3PV in CH2Cl2 solution.
Conclusions
Photophysics
A-3PV and B-3PV are new multicomponent arrays where
an OPV unit (3PV) has been linked to a fullerenopyrrolidine
(A) or a fullerenopyrazoline (B) moiety. The photophysical
properties of A, B, A-3PV, and B-3PV have been studied in
three different solvents of increasing polarity, namely toluene,
dichloromethane, and benzonitrile as well as at 298 and 77 K.
Pyrazoline A C60 photoinduced electron transfer is observed
at room temperature in B in all solvents and can be blocked
by addition of acid or by passing to a rigid matrix at 77 K
(PhMe and CH2Cl2 only). In A-3PV OPV A C60 energy trans-
fer is evidenced in all solvents, followed by charge separation
in PhCN to a significant extent.
B-3PV is arranged in such a way that the C60 unit can act
either as energy (for OPV) or electron (for pyrazoline) acceptor,
following light excitation. Excitation of the OPV moiety or of
the fullerene chromophore triggers distinct photoprocesses.
Further, besides the choice of excitation wavelength, the con-
trol over electron transfer is achieved by varying several
(even combined) parameters, i.e. solvent polarity, acidity,
and temperature. A schematic picture summarizing all these
features in CH2Cl2 solution is depicted in Fig. 10.
The photophysical investigations were carried out in CH2Cl2,
and toluene (Carlo Erba, spectrofluorimetric grade), and in
benzonitrile (Aldrich, HPLC grade). The samples were placed
in fluorimetric 1 cm path cuvettes and, when necessary, purged
from oxygen by at least four freeze–thaw–pump cycles.
Absorption spectra were recorded with
a Perkin-Elmer
l40 spectrophotometer. Uncorrected emission spectra were
obtained with a Spex Fluorolog II spectrofluorimeter (con-
tinuous 150 W Xe lamp), equipped with a Hamamatsu R-928
photomultiplier tube. The corrected spectra were obtained via a
calibration curve determined with a procedure described
earlier.5 Fluorescence quantum yields obtained from spectra
on an energy scale (cm21) were measured with the method
described by Demas and Crosby59 using as standards air
equilibrated solutions of quinine sulfate in 1 N H2SO4 (W ~
0.546)60 or [Os(phen)3]21 in acetonitrile (Wem ~ 0.005).61 To
record the 77 K luminescence spectra, the samples were put in
glass tubes (2 mm diameter) and inserted in a special quartz
dewar, filled up with liquid nitrogen. When necessary, spectra
of the glass matrix were recorded and then subtracted as
background signal, in order to eliminate the contribution from
light scattering.
Many cases of dyads made of a C60 unit and a conjugated
oligomer have been reported so far,19–22,24–27,45,58 and also
cases of C60 coupled with an amine type donor have been
described.8–13 The integration of both features onto a C60 unit
is reported here for the first time in B-3PV. This introduces
a very interesting pattern of the photophysical properties in
solution as shown schematically in Fig. 10, that make B-3PV a
fullerene-based molecular switch, the switchable parameters
being photoinduced processes. The incorporation of B-3PV in
photovoltaic devices results in a lower light to current efficiency
than in A-3PV, since charge separation involving the fullerene
moiety and the pyrazoline N atom is not able to contribute
to the photocurrent and, instead, the pyrazoline unit can act
as an electron trap. Nevertheless we believe that the design of
multicomponent arrays like B-3PV, featuring an antenna unit
(OPV) and a charge separation module (pyrazoline–C60) is very
appealing for the construction of devices for charge separation
and light energy conversion. Future work will be aimed at
designing and testing new systems in which exciton dissociation
is more efficiently achieved.
The steady-state IR luminescence spectra were obtained with
a constructed-in-house apparatus available at the Chemistry
Department of the University of Bologna (Italy) and described
in detail earlier.5 A continuous 450 W Xenon lamp was used
as light source, in order to be able to excite at any wavelength
in the range 300–600 nm. The determination of the relative
yields of singlet oxygen sensitization, according to Darmanyan
and Foote,62 was obtained by monitoring the singlet oxygen
luminescence at 1268 nm and by comparing them with the
suited C60 reference compound (A or B). Reference and dyad
solutions were set at the same absorbance in the same solvent,
making corrections due to these factors unnecessary.62
Emission lifetimes on the nanosecond time scale were
determined with an IBH single photon counting spectrometer
equipped with a thyratron gated nitrogen lamp working in
the range 2–40 kHz (lexc ~ 337 nm, 0.5 ns time resolution);
the detector was a red-sensitive (185–850 nm) Hamamatsu
R-3237-01 photomultiplier tube.
Transient absorption spectra in the nanosecond–microsecond
time domain were obtained by using the second or the third
harmonic (532 or 355 nm) of a Nd : YAG laser (JK Lasers)
with 20 ns pulse and 1–2 mJ of energy per pulse. Triplet
lifetimes were obtained by averaging five different decays
recorded around the maximum of the absorption peak (680–
720 nm) The details on the flash-photolysis system are reported
elsewhere.63
Experimental
Synthetic procedures and full characterization of all new
molecules are available as electronic supplementary informa-
tion.{
Experimental uncertainties are estimated to be ¡8% for
J. Mater. Chem., 2002, 12, 2077–2087
2085

View Article Online
16 H. Imahori and Y. Sakata, Eur. J. Org. Chem., 1999, 2445.
17 R. A. Marcus and N. Sutin, Biochim. Biophys. Acta, 1985, 811,
265.
18 L. C. Sun, L. Hammarstrom, B. Akermark and S. Styring, Chem.
Soc. Rev., 2001, 30, 36.
lifetime determinations, ¡20% for emission quantum yields,
¡5% for relative emission intensities in the NIR, ¡1 nm and
¡5 nm for absorption and emission peaks respectively.
Preparation and characterization of the photovoltaic devices
19 J. F. Nierengarten, J. F. Eckert, J. F. Nicoud, L. Ouali,
V. Krasnikov and G. Hadziioannou, Chem. Commun., 1999, 617.
20 N. Armaroli, F. Barigelletti, P. Ceroni, J. F. Eckert, J. F. Nicoud
and J. F. Nierengarten, Chem. Commun., 2000, 599.
21 J. F. Eckert, J. F. Nicoud, J. F. Nierengarten, S. G. Liu,
L. Echegoyen, F. Barigelletti, N. Armaroli, L. Ouali, V. Krasnikov
and G. Hadziioannou, J. Am. Chem. Soc., 2000, 122, 7467.
22 E. Peeters, P. A. van Hal, J. Knol, C. J. Brabec, N. S. Sariciftci,
J. C. Hummelen and R. A. J. Janssen, J. Phys. Chem. B, 2000, 104,
10174.
23 G. Accorsi, N. Armaroli, J. F. Eckert and J. F. Nierengarten,
Tetrahedron Lett., 2002, 43, 65.
24 S. Knorr, A. Grupp, M. Mehring, G. Grube and F. Effenberger,
J. Chem. Phys., 1999, 110, 3502.
25 T. Yamashiro, Y. Aso, T. Otsubo, H. Q. Tang, Y. Harima and
K. Yamashita, Chem. Lett., 1999, 443.
26 M. Fujitsuka, O. Ito, T. Yamashiro, Y. Aso and T. Otsubo,
J. Phys. Chem. A, 2000, 104, 4876.
27 P. A. van Hal, J. Knol, B. M. W. Langeveld-Voss, S. C. J. Meskers,
J. C. Hummelen and R. A. J. Janssen, J. Phys. Chem. A, 2000, 104,
5974.
28 J. F. Eckert, J. F. Nicoud, D. Guillon and J. F. Nierengarten,
Tetrahedron Lett., 2000, 41, 6411.
29 M. Prato and M. Maggini, Acc. Chem. Res., 1998, 31, 519.
30 G. Zerban and H. Meier, Z. Naturforsch., B: Chem. Sci., 1993, 48,
171.
Films of the active material were fabricated by spin-coating
from solution onto ITO–covered glass substrates which were
previously coated with a layer of poly (3,4-ethylenedioxythio-
phene)poly(styrenesulfonate) (PEDOT-PSS, BAYTRON P—
Bayer AG). The aluminium (Al) top electrode, which typically
had a thickness of about 100 nm, was vapour-deposited at
pressures of about 3 6 1027 torr onto the organic layer. All
other fabrication and measurement steps were carried out
under nitrogen atmosphere to minimise effects of water and
oxygen.
The devices were illuminated from the ITO side (400 nm,
1 mW cm22) using an Oriel 60100 xenon lamp in series with a
CVI Digikrom 120 monochromator, while the I–V curves were
measured with a Keithley 236 source measure unit. In forward
bias, the ITO electrode was wired as the anode. A standard
calibrated silicon photodiode was used to record the action
photovoltaic spectra. UV/Vis absorption spectra were recorded
on a Perkin Elmer 900 spectrophotometer and a Digital
Instruments Nanoscope IIIa atomic force microscope was used
for the imaging of the films.
31 P. de la Cruz, A. Diaz-Ortiz, J. J. Garcia, M. J. Gomez-
Escalonilla, A. de la Hoz and F. Langa, Tetrahedron Lett., 1999,
40, 1587.
32 F. Langa, P. de la Cruz, E. Espildora, A. de la Hoz,
J. L. Bourdelande, L. Sanchez and N. Martin, J. Org. Chem.,
2001, 66, 5033.
33 M. Maggini, A. Karlsson, G. Scorrano, G. Sandona, G. Farnia
and M. Prato, J. Chem. Soc., Chem. Commun., 1994, 589.
34 W. Kutner, K. Noworyta, G. R. Deviprasad and F. D’Souza,
J. Electrochem. Soc., 2000, 147, 2647.
35 F. Langa, M. J. Gomez-Escalonilla, E. Diez-Barra, J. C. Garcia-
Martinez, A. de la Hoz, J. Rodriguez-Lopez, A. Gonzalez-Cortes
and V. Lopez-Arza, Tetrahedron Lett., 2001, 42, 3435.
36 R. E. Martin, U. Gubler, C. Boudon, C. Bosshard,
J. P. Gisselbrecht, P. Gunter, M. Gross and F. Diederich,
Chem. Eur. J., 2000, 6, 4400.
Acknowledgement
This work was supported by the CNR-CNRS project
‘‘Supramolecular Fullerene Systems as Materials for Solar
Energy Conversion’’, ECODEV, French Ministry of Research,
Dutch Science Foundation and Spanish DGI (Project
BQU2001-1512). We thank Professor A. Juris (Department
of Chemistry, University of Bologna) for allowing the use of
the IR spectrofluorimeter and Dr Francesco Barigelletti for
calculations on energy transfer rates. G. A. thanks MIUR
(Progetto 5%), D.T. EU (Marie Curie Fellowship), and
J.-F. E. DGA for their fellowships. We further thank L.
Oswald and M. Minghetti for technical help and M. Schmitt
for NMR measurements.
37 N. Armaroli, J. F. Eckert and J. F. Nierengarten, Chem. Commun.,
2000, 2105.
38 U. Mazzucato and F. Momicchioli, Chem. Rev., 1991, 91, 1679.
39 H. Meier, Angew. Chem., Int. Ed. Engl., 1992, 31, 1399.
40 B. Ma, C. E. Bunker, R. Guduru, X. F. Zhang and Y. P. Sun,
J. Phys. Chem. A, 1997, 101, 5626.
References
1
D. Gust, T. A. Moore and A. L. Moore, Acc. Chem. Res., 2001, 34,
40.
2
3
D. M. Guldi and M. Prato, Acc. Chem. Res., 2000, 33, 695.
N. Martin, L. Sanchez, B. Illescas and I. Perez, Chem. Rev., 1998,
98, 2527.
41 P. A. van Hal, R. A. J. Janssen, G. Lanzani, G. Cerullo,
M. Zavelani-Rossi and S. De Silvestri, Phys. Rev. B, 2001, 6407,
5206.
4
5
6
7
H. Imahori, M. E. El-Khouly, M. Fujitsuka, O. Ito, Y. Sakata and
S. Fukuzumi, J. Phys. Chem. A, 2001, 105, 325.
N. Armaroli, G. Marconi, L. Echegoyen, J. P. Bourgeois and
F. Diederich, Chem. Eur. J., 2000, 6, 1629.
N. Martin, L. Sanchez, M. A. Herranz and D. M. Guldi, J. Phys.
Chem. A, 2000, 104, 4648.
N. Armaroli, F. Diederich, C. O. Dietrich-Buchecker, L. Flamigni,
G. Marconi, J. F. Nierengarten and J. P. Sauvage, Chem. Eur. J.,
1998, 4, 406.
R. M. Williams, J. M. Zwier and J. W. Verhoeven, J. Am. Chem.
Soc., 1995, 117, 4093.
Y. P. Sun, B. Ma and C. E. Bunker, J. Phys. Chem. A, 1998, 102,
7580.
42 D. Kuciauskas, S. Lin, G. R. Seely, A. L. Moore, T. A. Moore,
D. Gust, T. Drovetskaya, C. A. Reed and P. D. W. Boyd, J. Phys.
Chem., 1996, 100, 15926.
43 T. D. M. Bell, T. A. Smith, K. P. Ghiggino, M. G. Ranasinghe,
M. J. Shephard and M. N. Paddon-Row, Chem. Phys. Lett., 1997,
268, 223.
44 C. Luo, D. M. Guldi, H. Imahori, K. Tamaki and K. Sakata,
J. Am. Chem. Soc., 2000, 122, 6535.
45 J. L. Segura, R. Gomez, N. Martin, C. P. Luo, A. Swartz and
D. M. Guldi, Chem. Commun., 2001, 707.
8
9
46 R. R. Hung and J. J. Grabowski, J. Phys. Chem., 1991, 95, 6073.
47 J. W. Arbogast, A. P. Darmanyan, C. S. Foote, Y. Rubin,
F. N. Diederich, M. M. Alvarez, S. J. Anz and R. L. Whetten,
J. Phys. Chem., 1991, 95, 11.
10 K. G. Thomas, V. Biju, M. V. George, D. M. Guldi and
P. V. Kamat, J. Phys. Chem. A, 1998, 102, 5341.
11 G. E. Lawson, A. Kitaygorodskiy and Y. P. Sun, J. Org. Chem.,
1999, 64, 5913.
12 S. Komamine, M. Fujitsuka, O. Ito, K. Moriwaki, T. Miyata and
T. Ohno, J. Phys. Chem. A, 2000, 104, 11497.
13 F. Langa, P. de la Cruz, E. Espildora, A. Gonzalez-Cortes, A. de la
Hoz and V. Lopez-Arza, J. Org. Chem., 2000, 65, 8675.
14 D. M. Guldi, Chem. Commun., 2000, 321.
48 F. Wilkinson, W. P. Helman and A. B. Ross, J. Phys. Chem. Ref.
Data, 1995, 24, 663.
49 R. V. Bensasson, E. Bienvenue, C. Fabre, J. M. Janot, E. J. Land,
S. Leach, V. Leboulaire, A. Rassat, S. Roux and P. Seta, Chem.
Eur. J., 1998, 4, 270.
50 R. Stackow, G. Schick, T. Jarrosson, Y. Rubin and C. S. Foote,
J. Phys. Chem. B, 2000, 104, 7914.
51 R. V. Bensasson, M. Schwell, M. Fanti, N. K. Wachter,
J. O. Lopez, J.-M. Janot, P. R. Birkett, E. J. Land, S. Leach,
P. Seta, R. Taylor and F. Zerbetto, ChemPhysChem., 2001, 109.
15 H. Imahori, K. Hagiwara, T. Akiyama, M. Aoki, S. Taniguchi,
T. Okada, M. Shirakawa and Y. Sakata, Chem. Phys. Lett., 1996,
263, 545.
2086
J. Mater. Chem., 2002, 12, 2077–2087

View Article Online
52 R. V. Bensasson, E. Bienvenue, J. M. Janot, E. J. Land, S. Leach
and P. Seta, Chem. Phys. Lett., 1998, 283, 221.
57 J. Kroon, J. W. Verhoeven, M. N. Paddon-Row and A. M. Oliver,
Angew. Chem., Int. Ed. Engl., 1991, 30, 1358.
53 R. V. Bensasson, M. N. Berberan-Santos, M. Brettreich,
J. Frederiksen, H. Gottinger, A. Hirsch, E. J. Land, S. Leach,
D. J. McGarvey, H. Schonberger and C. Schroder, Phys. Chem.
Chem. Phys., 2001, 3, 4679.
54 D. M. Guldi and K. D. Asmus, J. Phys. Chem. A, 1997, 101, 1472.
55 A. Weller, Z. Phys. Chem. Neue Folge, 1982, 133, 93.
56 R. M. Williams, M. Koeberg, J. M. Lawson, Y. Z. An, Y. Rubin,
M. N. Paddon-Row and J. W. Verhoeven, J. Org. Chem., 1996, 61,
5055.
58 J. L. Segura, R. Gomez, N. Martin, C. P. Luo and D. M. Guldi,
Chem. Commun., 2000, 701.
59 J. N. Demas and G. A. Crosby, J. Phys. Chem., 1971, 75, 991.
60 S. R. Meech and D. J. Philips, J. Photochem., 1983, 23, 193.
61 E. M. Kober, J. V. Caspar, R. S. Lumpkin and T. J. Meyer,
J. Phys. Chem., 1986, 90, 3722.
62 A. P. Darmanyan and C. S. Foote, J. Phys. Chem., 1993, 97, 5032.
63 L. Flamigni, J. Phys. Chem., 1992, 96, 3331.
J. Mater. Chem., 2002, 12, 2077–2087
2087