Selective triplet-state formation during charge recombination in a fullerene/Bodipy molecular dyad (Bodipy = borondipyrromethene)

Article abstract of DOI:10.1002/chem.200900440

A conformationally restricted molecular dyad has been synthesized and subjected to detailed photophysical examination. The dyad comprises a borondipyrromethene (Bodipy) dye covalently linked to a buckminsterfullerene Q₆₀ residue, and is equippe

Full text of DOI:10.1002/chem.200900440

DOI: 10.1002/chem.200900440
Selective Triplet-State Formation during Charge Recombination in a
Fullerene/Bodipy Molecular Dyad (Bodipy=Borondipyrromethene)
Raymond Ziessel,*[a] Ben D. Allen,[b] Dorota B. Rewinska,[b] and Anthony Harriman*[b]
Dedicated to the memory of Josꢀe Charbonneau
Abstract: A conformationally restrict-
ed molecular dyad has been synthe-
sized and subjected to detailed photo-
physical examination. The dyad com-
and could compete with intramolecular
energy transfer in the same direction.
The driving force for light-induced
electron abstraction from Bodipy by
the singlet excited state of C60 depends
critically on the solvent polarity. Thus,
in non-polar solvents, light-induced
electron transfer is thermodynamically
uphill, but fast excitation energy trans-
fer occurs from Bodipy to C60 and is
followed by intersystem crossing and
subsequent equilibration of the two
triplet excited states. Moving to a polar
solvent switches on light-induced elec-
tron transfer. Now, in benzonitrile, the
charge-transfer state (CTS) is posi-
tioned slightly below the triplet levels,
such that charge recombination re-
stores the ground state. However, in
CH2Cl2 or methyltetrahydrofuran, the
CTS is slightly higher in energy than
the triplet levels, and decays, in part, to
form the triplet state localized on the
C60 residue. This step is highly specific
and does not result in direct formation
of the triplet excited state localized on
the Bodipy unit. Subsequent equilibra-
tion of the two triplets takes place on a
relatively slow timescale.
prises
a
borondipyrromethene
(Bodipy) dye covalently linked to a
buckminsterfullerene C60 residue, and
is equipped with hexadecyne units at
the boron centre in order to assist solu-
bility. The linkage consists of a diphe-
nyltolane, attached at the meso posi-
tion of the Bodipy core and through an
N-methylpyrrolidine ring at the C60 sur-
face. Triplet states localised on the two
terminals are essentially isoenergetic.
Cyclic voltammetry indicates that light-
induced electron transfer from Bodipy
to C60 is thermodynamically favourable
Keywords: cyclic voltammetry
·
dyes/pigments · electron transfer ·
fullerenes
cules, the triplet manifold is rarely accessed by direct excita-
tion, unless pronounced spin-orbit coupling is induced by
heavy-atom perturbation,[2] but is readily populated through
indirect routes. Prominent among such secondary processes
are intersystem crossing from the first singlet excited state
and charge recombination between relevant radical ions.
This latter reaction might involve diffusive contact between
isolated radical ions,[3] charge migration along conjugated
polymers[4] or intramolecular charge recombination[5] be-
tween closely spaced radical ions. Narrowing the field even
further, we note that intramolecular charge recombination
might involve orbital contact between species linked
through a flexible bridge[6] or through-bond electron transfer
between reagents connected by a rigid spacer.[7] It is this
latter type of system that concerns us here. Such processes
are extremely frustrating when related to systems intended
to produce an unusually long-lived, charge-separated state.[8]
For example, recent claims regarding the production of ex-
ceptionally long-lived, charge-separated states[9] have been
Introduction
The triplet excited state has played a major role throughout
the history of molecular photophysics[1] and continues to be
a dominant force within the field. For simple organic mole-
[a] Dr. R. Ziessel
Laboratoire de Chimie Organique et
Spectroscopies Avancꢀes (LCOSA)
Ecole Chimie Polymꢁres, Matꢀriaux (ECPM)
25 rue Becquerel 67087 Strasbourg Cedex (France)
Fax : (+33)90-24-27-61
:
E-mail: iessel@chimie.u-strasbg.fr
[b] B. D. Allen, D. B. Rewinska, Prof. Dr. A. Harriman
Molecular Photonics Laboratory
School of Chemistry, University of Newcastle
Newcastle upon Tyne, NE1 7RU (UK)
Fax : (+44)191-222-8660
E-mail: Anthony.harriman@ncl.ac.uk
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200900440.
7382
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7382 – 7393

FULL PAPER
re-interpreted by other authors in terms of fast charge re-
combination to form the triplet state.[10] At a more funda-
mental level, triplet formation by way of through-bond elec-
tron transfer offers some interesting mechanistic challeng-
es.[11]
5 K in solid neon[20] doped with xenon as a heavy-atom per-
turber. Related phosphorescence studies[21] made in rare gas
matrices place ET at 12925 cmÀ1. Using the external heavy-
atom effect, ET for C60 in a methylcyclohexane glass at 77 K
has been located[22] as being 12690 cmÀ1, which can be com-
pared to values of 12600 and 13100 cmÀ1 determined[23] by
opto-acoustic spectroscopy. Calculated values for ET are
19560, 18070, and 13245 cmÀ1.[24] The actual ET value for a
C60 residue comprising a molecular dyad is likely to change
slightly according to the nature of the linkage. Indeed, sensi-
tized phosphorescence has been reported[25] for some C60-
based dyads in a low-temperature glass for which ET corre-
sponds to 12105 cmÀ1, and the emission is surprisingly long-
lived, having a lifetime of 4.8 ms at 77 K. A key feature of
C60, and a factor that is keenly exploited in this work, is that
intersystem crossing to the triplet manifold is almost quanti-
tative.[26]
Early highlights in this field involved the observation of
very fast generation of the triplet state localized on the
pyrene moiety during charge recombination in a series of
pyrene–amine heteroexciplexes.[12] It was noted that the rate
of intersystem crossing showed a marked dependence on the
mutual configuration of donor and acceptor groups and also
on the solvent polarity. Triplet formation was likewise found
to be very fast in some naphthalene–imine molecular
dyads,[13] in which the donor and acceptor functions were
held in a rigid conformation. Again, the efficiency of triplet
formation was dependent on the polarity of the surrounding
solvent. Several 9-aryl-10-methylacridinium cations have
been shown to favour triplet formation during charge re-
combination.[10,14] Here, the redox-active subunits are held
in an orthogonal geometry because of steric constraints, and
triplet formation is highly efficient. Closely comparable be-
haviour has been reported for an orthogonal molecular dyad
In marked contrast to C60, intersystem crossing in boron-
dipyrromethene (Bodipy) dyes is extremely inefficient and
the detection of phosphorescence from such molecules has
been highly elusive.[27] The first such report[28] described the
sensitized emission from a Bodipy-based molecular dyad
and located the triplet energy at 12920 cmÀ1. The sensitizer
based[15] on
a borondipyrromethene chromophore. A
common feature of all these systems is the close positioning
of donor and acceptor units. Recently, Wasielewski et al.[16]
used time-resolved EPR spectroscopy to monitor intersys-
tem crossing within intramolecular charge-transfer states of
several anthracene–julolidine dyads, and confirmed that the
efficiency is at a maximum when the reactants are held in a
perpendicular geometry. These authors have probed[16] the
mechanism in some detail, and conclude that the overall
effect is likely to be applicable to a wide variety of electron
donor-acceptor systems, provided the reactants are nearby.
was a ruthenium(II)–polyACTHGNUTERNNU(G pyridine) complex linked directly
to the Bodipy dye through a short, rigid tether. The lifetime
of this phosphorescence was 50 ms in a butyronitrile glass at
77 K.[28] Subsequently, phosphorescence was observed for
two Bodipy derivatives in a low temperature glass using the
external heavy-atom effect.[29] The ET values were deter-
mined as being 12740 and 13330 cmÀ1 and both phosphores-
cence lifetimes were found to be 1.2 ms under these condi-
tions. Again, we might expect some modest variation in ET
according to the appendages attached to the Bodipy nucleus.
It must also be recognized that some kind of mutual pertur-
bation that would cause minor modification of the triplet
energy levels might be expected for a Bodipy–C60 dyad.
Even so, it will not have escaped attention that the ET
values for C60 and Bodipy derivatives are expected to be re-
markably similar. This situation provides the opportunity to
examine equilibration between the two triplets, especially
since the isolated triplets are relatively long-lived in deoxy-
genated solution. Furthermore, C60 has high electron affini-
ty[30] and should enter into electron-transfer reactions with
the excited-singlet state of Bodipy. By linking together these
units through a short spacer and building in some degree of
orthogonality at the linkage, we might anticipate[12] that
charge recombination within the initially-formed radical ion
pair would favour triplet formation. If so, we could explore
the possibility that one triplet would be formed
preferentially.
We now describe the results of an investigation into the
photophysical properties of a closely spaced molecular dyad
comprising terminals formed from an organic dye and a full-
erene (see above). There have been numerous such systems
studied in recent years,[17] mostly using the fullerene as an
electron acceptor. Alternative systems have used the fuller-
ene as a triplet reservoir[18] due to its triplet level being situ-
ated at low energy. In fact, the triplet energy of C60 is quite
well defined despite its extremely poor phosphorescence
yield in a low temperature glass. A luminescence study[19] of
C60 in glassy toluene at 77 K has located the triplet energy
(ET) as being 12614 cmÀ1, whereas a value of 12773 cmÀ1
has been derived from phosphorescence spectra recorded at
Results and Discussion
Synthesis and electrochemistry: The target dyad, Bodipy–
C60, has the two redox-active terminals linked by a diphenyl-
AHCTUNGTREGaNNNU cetylene spacer. The connection at the meso-site (posi-
Chem. Eur. J. 2009, 15, 7382 – 7393
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7383

A. Harriman, R. Ziessel et al.
tion 8) and the presence of the two methyl groups (posi-
tions 1 and 7) on the Bodipy unit forces the phenylene ring
to be orthogonal with the Bodipy nucleus, and quantum
chemical calculations made at the Hartree–Foch level with
the 6–31G** basis set indicate that the dihedral angle for
this connection is 888, this being imposed by steric interac-
tions. Attachment to the C60 residue is through a pyrrolidine
linkage. The closest contact between the terminals was esti-
mated to be 13.5 ꢃ by computational studies, but it is
known that this type of bridge is somewhat flexible and fa-
cilitates more intimate approach of the terminals.[31] We
report in Scheme 1 the synthetic protocol used (starting
to the diastereotopic methylene protons. Remarkably, the
Bodipy unit exhibits a singlet at 6.00 ppm corresponding to
the two pyrrole protons and two singlets at 2.76 and
1.40 ppm for the two sets of pyrrole methyl protons. The
MS-FAB at 1604.5 [M+H]+ is in keeping with the proposed
molecular structure of Bodipy–C60.
Cyclic voltammograms were recorded for Bodipy–C60 in
CH2Cl2 containing background electrolyte and referenced to
1
ferrocene (E = =0.38 V vs. SCE) as an internal standard
2
(Figure 1). Peak assignment was made by adding an equimo-
Scheme 1. i) C14H29CꢀCMgBr (2.5 equiv), THF, 708C, 12 h, 79%; ii) [Pd-
ACHTUNGTRENNUNG(PPh3)4] (3 mol%), C6H6, iPr2NH, 708C, 24 h, 75%.
Figure 1. Cyclic voltammograms of the dyad 4 (top) and of a stoichiomet-
ric mixture of 4 + 2 (bottom), in CH2Cl2 containing 0.1m tetra-n-buty-
lammonium hexafluorophosphate at 208C; scan rate 200 mVsÀ1. All
1
peaks were calibrated against ferrocene (Fc) at E = =0.38 V vs. SCE.
Peaks marked with an asterisk correspond to compound 2.
2
from the boron precursor 1) for preparation of Bodipy–C60
and appropriate reference compounds. Note, in particular,
the presence of the two solubilising groups attached to the
boron centre.[32]
Cross-linking the phenyliodo-substituted derivative 2 with
N-methyl-2-(4’-ethynyl)phenyl-3,4-fulleropyrrolidine (3)[33]
was carried out in benzene and diisopropylamine using [Pd-
lar amount of the Bodipy compound lacking the C60 append-
age, 2. On oxidative scans, two peaks are observed. The first
1
peak corresponds to a half-wave potential, E = , of 1.03 V vs.
2
A
SCE and is quasi-reversible. This process can be ascribed to
usual B F groups were replaced with B (1-hexadecynyl)
groups, using the desired Grignard reagent under conditions
that have been described previously.[34] Furthermore, substi-
tution at the boron site greatly improves the stability of
these dyes, as concluded previously for somewhat related
compounds.[32] Purification of dyes 2 and 4 and of the C60
precursor 3 was carried out by flash column chromatogra-
phy. All new compounds were characterized by FAB-MS,
1H and 13C NMR spectroscopy, and elemental analysis. The
most diagnostic NMR signals are those that relate to the N-
methylpyrrolidine fragment, with one singlet at 2.86 ppm
due to the methyl substituent on the nitrogen atom, one sin-
glet at 5.00 ppm due to the proton in the a-position of the
pyrrolidine ring, and two doublets at 5.03 and 4.31 ppm due
formation of the Bodipy p-radical cation.[35,36] The second
peak seen under these conditions is due to oxidation of the
C60 unit (possibly the tertiary amine of the pyrrolidine), and
is electrochemically irreversible, with the anodic peak po-
tential being 1.40 V vs. SCE.[37] A series of peaks are seen
on reductive scans, but these are easily assigned by compari-
son with the cyclic voltammogram recorded for the equimo-
lar mixture (Figure 1).
Thus, three successive reduction peaks can be resolved for
1
the C60 unit, with E = values of À0.69, À1.11 and À1.60 V vs.
2
SCE, respectively, and each can be ascribed to the addition
of one electron.[38] Each of these reduction steps is electro-
chemically reversible. One-electron reduction of the Bodipy
nucleus occurs with an E = value of À1.40 V vs. SCE.[39]
1
2
7384
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7382 – 7393

Charge Recombination in a Fullerene/Bodipy Molecular Dyad
FULL PAPER
From these cyclic voltammograms it can be concluded that
two electrons can be added to the C60 residue before reduc-
tion of the Bodipy nucleus takes place. Furthermore, it is
clear from comparison of cyclic voltammograms recorded
for Bodipy–C60 and the mixture that prior reduction of the
C60 unit[40] does not affect the reduction potential of the
Bodipy unit. As such, electrostatic interactions between the
subunits, and also their mutual electronic coupling, are mini-
mal.
Several intramolecular electron-transfer reactions are pos-
sible for Bodipy–C60, and the corresponding thermodynamic
driving forces can be estimated on the basis of the cyclic vol-
tammetry and using excited-state energies derived from
emission spectroscopy (see below). In so doing, we neglect
any redox reactions that might involve oxidation of the C60
residue and, for the moment, restrict our attention to
CH2Cl2 (note that driving forces in other solvents were cal-
culated using the so-called Weller approach).[41] Now, the
singlet excited state of Bodipy is able to transfer an electron
to the nearby C60 residue (DG8=À0.84 eV), but the corre-
sponding triplet excited state is less likely to do so (DG8=
+0.05 eV). Similarly, the triplet excited state localized on
the C60 unit is not expected to oxidize the appended Bodipy
unit (DG8=+0.04 eV). The singlet excited state associated
with the C60 fragment is not a particularly strong oxidant
and barely possesses the thermodynamic capability to oxi-
dize the Bodipy unit (DG8=À0.16 eV). In making the vari-
ous estimates for the driving force, a minor correction has
been made for changes in the electrostatic interaction be-
tween the reactants.[42] The two thermodynamically favoura-
ble electron-transfer reactions give rise to a common radical
ion pair, for which charge recombination to reform the
ground state (DG8=À1.60 eV) is highly favourable on ther-
modynamic grounds. It is also recognized that the driving
force for electron transfer from Bodipy to the singlet excited
state of C60, this being critically balanced in CH2Cl2, will
depend on the polarity of the surrounding medium.
Figure 2. Absorption spectrum recorded for Bodipy–C60 in toluene and
the fluorescence spectrum recorded for the Bodipy component following
excitation at 490 nm. Also shown are an expansion of the far-red region
and an overlay of the fluorescence from the C60 residue following excita-
tion at 570 nm.
The Bodipy-based reference compound 2 shows strong
fluorescence centred at 516 nm in toluene, for which the
emission quantum yield (FF) is 0.50 and the excited-state
lifetime (tS) is 3.4 ns. This information can be used to locate
the singlet excitation energy (ES) as being 19670 cmÀ1. Like-
wise, the C60 reference compound, 3, shows very weak fluo-
rescence (FF =0.0003) centred at 720 nm in toluene, with
tS =1.15 ns. On this basis, we estimate ES for the C60 unit as
being 14160 cmÀ1. Comparison to C60 shows that the linkage
has no evident effect on the excitation energy. Using these
data, it is concluded that the C60 unit present in Bodipy–C60
can be selectively excited over much of the visible region,
and the Bodipy-based chromophore can be excited with
high specificity at wavelengths around 500 nm. Fluorescence
from each subunit is easily distinguished by virtue of the dis-
parate spectral distributions.
Initial experiments were carried out in toluene at 208C in
order to assess the significance of electronic energy transfer
in the dyad. Excitation (l=615 nm) into the C60-based unit
in Bodipy–C60 gave rise to weak fluorescence centred at
around 720 nm. This emission decayed with mono-exponen-
tial kinetics, for which the lifetime was 1.20 ns. Since this is
very similar to the value found for the reference compound
3, for which tS =1.26 ns, it can be argued that the singlet ex-
cited state of the C60 unit is not quenched by the appended
Bodipy residue under these conditions. It should be noted
that light-induced electron transfer (DG8=+0.50 eV) is
highly unfavourable in toluene. Decay of the singlet excited
state is primarily through intersystem crossing to form the
first excited triplet state (T1) localized on C60. This latter
species, which shows a pronounced absorption band[44] cen-
tred at around 730 nm (Figure 3), enters into energy-transfer
reactions with the corresponding Bodipy-based T1 state, as
will be discussed below. In contrast, excitation into the S1
state localized on the Bodipy unit shows that fluorescence
from this species is heavily quenched. Indeed, tS recorded in
Light-induced processes in non-polar media: The absorption
spectrum recorded for Bodipy–C60 in toluene is given in
Figure 2. The S0–S1 absorption transition associated with the
Bodipy unit is easily recognized by its pronounced maxi-
mum at 501 nm and by comparison to spectra recorded for
2. This wavelength is typical of a B-ethyne Bodipy deriva-
tive[43] and testifies to the absence of pronounced electronic
coupling between the redox-active units at the ground-state
level. Absorption transitions localized on the C60 residue
cover a wide wavelength range, with the S0–S1 transition
being evident at 693 nm (Figure 2). There are more pro-
nounced absorption maxima centred at ꢁ310 and 335 nm,
which can be assigned to the dipole-allowed transition to a
much higher lying excited state. These latter bands obscure
the S0–S2 transition localized on the Bodipy unit. Limited
solubility of the dyad prevented a systematic study of how
these absorption transitions depend on the nature of the sol-
vent but, in any case, such effects are expected to be
modest.
Chem. Eur. J. 2009, 15, 7382 – 7393
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7385

A. Harriman, R. Ziessel et al.
competes poorly with energy transfer under these condi-
tions.
In deoxygenated toluene at 208C, the T1 state centred on
the C60 residue is the only species detected[46] by transient
absorption spectroscopy on timescales of around 10 ns, re-
gardless of which subunit is irradiated. On longer timescales,
the strong absorption band associated with the Bodipy unit
is partially bleached due to population of the Bodipy T1
state (Figure 5). This bleaching process is the consequence
Figure 3. Transient differential absorption spectra recorded after excita-
tion into the C60 unit present in Bodipy–C60 in deoxygenated toluene and
showing the course of intersystem crossing to the C60 triplet state. Delay
times: 0.1, 0.3, 0.6, 1.0, 1.5, 2, 3, 4, and 6 ns.
toluene is reduced to 170 ps, compared to 3.4 ns as found for
the reference compound 2. Quenching, for which the de-
rived rate constant is 5.6ꢄ109 sÀ1, is attributed to intramolec-
ular electronic energy transfer to the nearby C60 unit, al-
though there is a modest driving force (DG8=À0.18 eV) for
light-induced electron transfer in toluene; such behaviour
has been seen previously for certain C60 derivatives.[45] Inter-
estingly, fluorescence from the Bodipy unit is only slightly
recovered by cooling the sample to 77 K. We can conclude,
therefore, that the activation energy for the quenching step
is small under these conditions.
Figure 5. Transient differential absorption spectra recorded after excita-
tion into the C60 unit present in Bodipy–C60 in deoxygenated toluene and
showing equilibration between the two triplet states. Delay times: 0.25, 1,
2, 4, 8, 15, 30, 60, 100, and 150 ms.
of T1–T1 energy transfer from the C60 triplet state, although
this step is slightly endergonic (DETT ꢁ 400 cmÀ1). Triplet
energy transfer is incomplete and results in the formation of
an equilibrated mixture of triplet states that decays with a
lifetime of ꢁ70 ms in deoxygenated toluene. This triplet life-
time is somewhat shorter than those measured for the isolat-
ed reference compounds, with triplet lifetimes being 140 ms
for both 2 and 3. The equilibrium is established on a rather
slow timescale, for which the averaged rate constant is only
8ꢄ105 sÀ1. We presume that the slowness of this step reflects
the uphill energy balance. In the presence of 20% v/v iodo-
ethane, added as an external heavy-atom perturber, the
equilibrium mixture is formed on a slightly faster timescale,
the rate constant increasing to ꢁ1ꢄ106 sÀ1, but decays much
faster, with a lifetime of ꢁ6 ms. The overall photophysical
processes delineated in toluene are summarized in
Scheme 2.
This intramolecular S1–S1 energy-transfer process was con-
firmed by laser flash photolysis studies following excitation
into the Bodipy unit at 490 nm (Figure 4). Here, the initially
Figure 4. Transient differential absorption spectra recorded after excita-
tion into the Bodipy unit present in Bodipy–C60 in deoxygenated toluene
and showing the course of energy transfer to the C60 unit. The inset
shows a decay profile recorded at 525 nm. Delay times: 30, 70, 120, 180,
250, 350, 500, 650, and 800 ps.
formed S1 state localized on the Bodipy unit evolves first to
the corresponding S1 associated with the C60 residue, this
species absorbing primarily at 960 nm,[44] and subsequently
to the T1 state localized on C60. The energy gap for S1–S1
electronic energy transfer is 5510 cmÀ1 while the measured
efficiency is approximately 95%. There is no indication of
transient formation of the C60 p-radical anion, which is
known from spectro-electrochemical studies to exhibit a
characteristic absorption transition at around 1000 nm in tol-
uene. As such, it can be concluded that electron transfer
Scheme 2. Outline of the various processes observed following pulsed ex-
citation of either subunit for Bodipy–C60 in toluene. Key: EnT=electron-
ic energy transfer, ISC=intersystem crossing, NR=nonradiative decay.
Energies: BOD*S =2.44 eV, C60*S =1.76 eV, BOD*T =1.61 eV, C60*T =
1.56 eV (BOD=Bodipy).
7386
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7382 – 7393

Charge Recombination in a Fullerene/Bodipy Molecular Dyad
FULL PAPER
Light-induced electron transfer in benzonitrile: On account
of solvation effects, strongly polar solvents serve to in-
crease[41,42] the thermodynamic driving force for light-in-
duced electron transfer within the dyad. In particular, the
solvated CTS in benzonitrile (PhCN) is now positioned at
an energy somewhat lower than that of the S1 state localized
on the C60 residue (Scheme 3). The net effect is that fluores-
Scheme 3. Outline of the various processes observed following pulsed ex-
citation of either subunit for Bodipy–C60 in benzonitrile. Key: CR=
charge recombination, ElT=electron transfer. Energies: BOD*S =
2.44 eV, C60*S =1.76 eV, BOD*T =1.61 eV, C60*T =1.56 eV, CTS=1.46 eV
(BOD=Bodipy).
cence from the C60 unit can no longer be resolved from the
baseline in PhCN at room temperature, and it was not possi-
ble to estimate tS for this state by time-resolved fluores-
cence spectroscopy because of the weak signal. Transient ab-
sorption spectroscopy made following laser excitation into
the C60 unit at 320 nm (at which the Bodipy unit absorbs
ꢁ5% of incident photons) confirmed that an additional
quenching process takes place in PhCN and allowed deter-
mination of tS as being 125 ps. Under these conditions, the
S1 state decays by first-order kinetics to form the CTS
(DG8=À0.32 eV), which can be recognized by the charac-
teristic differential absorption spectrum of the C60 p-radical
anion and also from the pronounced bleaching of the
Bodipy chromophore (Figure 6). Decay curves recorded for
the CTS are reasonably mono-exponential and correspond
to a lifetime of 440 ps. Deactivation involves reformation of
the ground state, by way of charge recombination, probably
without intermediate formation of the triplet manifold.
Indeed, in PhCN, the CTS lies some 11770 cmÀ1 above the
ground state, but about 800 cmÀ1 below the T1 state of the
Figure 6. a) Transient differential absorption spectra recorded after
pulsed excitation of the C60 unit present in Bodipy–C60 in deoxygenated
benzonitrile. Delay times: 50, 100, 150, 250, 400, 600, 900, 1200, and
1600 ps. b) Kinetic traces recorded at 525 and 1080 nm for the experi-
ment mentioned above.
formed, but this probably arises from competitive intersys-
tem crossing from the corresponding S1 state. In PhCN, the
triplet lifetime is reduced to ꢁ20 ns. It is notable that in this
polar solvent, and in marked contrast to the case found for
toluene, there is no indication of intermediate formation of
the T1 state on the Bodipy unit. This is because charge re-
combination with the CTS does not populate the triplet
manifold in PhCN and energy transfer from the C60 T1 state
is too slow for this step to compete with electron transfer to
form the CTS.
Electron transfer in methyltetrahydrofuran: In searching for
intermediate behaviour, further studies were undertaken in
methyltetrahydrofuran (MTHF), the dielectric constant
(eS =7.0) of which is between those of toluene (eS =2.4) and
PhCN (eS =26.0). At room temperature, time-resolved emis-
sion decay profiles recorded for fluorescence from the C60
fragment were bi-exponential. Averaged values for the fit-
ting parameters correspond to fluorescence lifetimes of 0.81
(73) and 14.7 ns (27%). Under these conditions, the energy
levels for the S1 state localized on the C60 unit (ES =
14160 cmÀ1) and the expected charge-transfer state (CTS)
are likely to be closely comparable. As such, the dual-expo-
nential decay curves can be interpreted[46] in terms of rever-
sible population of the S1 state from the CTS in accordance
with Scheme 4. Analysis of the experimental data in terms
of this scheme allows calculation[47] of the rate constants for
forward (kF =7.4ꢄ108 sÀ1) and reverse (kR =8.3ꢄ108 sÀ1)
C
60 residue.
Direct excitation into the Bodipy chromophore in PhCN
results in immediate population of the corresponding S1
state, which decays rapidly with first-order kinetics (tS =
36 ps) to form the CTS. There is a large thermodynamic
driving force[41] for this process, DG8=À1.00 eV, but the de-
rived rate constant is not much higher than that found for
electron transfer to the S1 state of the C60 unit (Scheme 3).
Possibly, the former reaction lies in the Marcus inverted
region while the latter falls within the normal region. Again,
decay of the CTS occurs through charge recombination to
restore the ground state with
a lifetime of 420 ps
(Scheme 3). Monitoring at 740 nm shows that a small
amount of the T1 state associated with the C60 unit is
Chem. Eur. J. 2009, 15, 7382 – 7393
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7387

A. Harriman, R. Ziessel et al.
Scheme 4. Outline of the various processes observed following pulsed ex-
citation of either subunit for Bodipy–C60 in MTHF. Energies: BOD*S =
2.44 eV, C60*S =1.76 eV, BOD*T =1.61 eV, C60*T =1.56 eV, CTS=1.77 eV
(BOD=Bodipy).
Figure 7. Transient decay curves recorded after laser excitation into the
Bodipy unit present in Bodipy–C60 in deoxygenated MTHF and recorded
*
*
at 525 ( ) or 1080 nm ( ).
electron-transfer steps and also for non-radiative decay
(kNR =2ꢄ108 sÀ1) of the CTS (note that this analytical
method is not reliable for determination of kNR). This last
process involves a combination of internal conversion to
reform the ground state and triplet population. From the
ratio of kF and kR, the energy gap can be estimated as being
ꢁ20 cmÀ1 (i.e., 0.002 eV), with the CTS being the higher
energy species, and is therefore in respectable accord with
expectations based on cyclic voltammetry (DG8=À0.09 eV).
Intersystem crossing to the triplet manifold is expected to
compete effectively with formation of the CTS, and transi-
ent absorption spectroscopy confirms fast formation of the
T1 state localized on the C60 residue (see below). A summa-
ry of the main findings observed in MTHF is given in
Scheme 4.
Excitation into the Bodipy-based chromophore present in
Bodipy–C60 dissolved in MTHF at ambient temperature re-
sults in the appearance of fluorescence characteristic of that
unit. However, FF is reduced to a value that is only 0.6% of
that of 2, and there is a corresponding reduction in tS from
3.4 ns to 65 ps. The decay profile is mono-exponential such
that the rate constant for intramolecular fluorescence
quenching can be estimated by comparison to the lifetime of
2 as being 1.5ꢄ1010 sÀ1. Fluorescence quenching might be
expected to involve a combination of both electron transfer
(DG8=À0.77 eV) and energy transfer (DESS =0.68 eV) and,
on the basis that the rate of energy transfer is solvent inde-
pendent, the rate constant for electron transfer can be esti-
mated to be ꢁ1ꢄ1010 sÀ1. Evidence for light-induced elec-
tron transfer was sought from laser flash photolysis of the
dyad in MTHF with excitation at 490 nm. Thus, the S1 state
localized on the Bodipy unit is present immediately after
the laser pulse (Figure 7). This species decays with a lifetime
of ꢁ70 ps to form a mixture of products. The corresponding
S1 state associated with the C60 residue is clearly recognized
by its broad absorption profile centred around 960 nm, but
there is also the absorption band characteristic of the C60 p-
radical anion centred at around 1050 nm. This latter species
is formed with a time constant of 75 ps and decays by first-
order kinetics with a lifetime of 170 ps (Figure 7). The
Bodipy p-radical cation, which does not possess any charac-
teristic absorption bands but is easily monitored by ground-
state bleaching, also decays by first-order kinetics with a
lifetime of 185 ps. Deactivation of the CTS leads to partial
formation of the C60-based T1 state, as shown by the evolu-
tion of an absorption band centred at 740 nm. It should be
stressed that there is no indication of intermediate forma-
tion of the T1 state associated with the Bodipy unit under
these conditions.
A transient differential absorption spectrum recorded im-
mediately after laser excitation with a 4 ns pulse delivered
at either 490 or 532 nm shows that the dominant metastable
species is the T1 state localized on the C60 residue. This spe-
cies arises both from intersystem crossing from the S1 state
of the C60 residue, and by way of (partial) charge recombi-
nation within the CTS. Its deactivation establishes the ther-
mal mixture of triplet states described above. In deoxygenat-
ed MTHF, the equilibrium is formed with a lifetime of 2 ms
and decays to the ground state with a lifetime of 40 ms
(Figure 8). Addition of 20% v/v iodoethane has little effect
Figure 8. Kinetic traces recorded after laser excitation of Bodipy–C60 in
deoxygenated MTHF at 532 nm, with the signal being measured at
a) 760 nm and b) 525 nm.
on the rate of formation of the mixture, with the lifetime
being ꢁ1 ms, whereas the lifetime of the equilibrium mixture
of triplets falls to ꢁ10 ms. As found also in toluene, the equi-
librated mixture is formed on a timescale that is rather slow,
but perhaps fully reflective of the poor thermodynamics. It
is important to emphasize that the lifetime of the CTS in
MTHF (tCTS =170 ps) is significantly shorter than that found
7388
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7382 – 7393

Charge Recombination in a Fullerene/Bodipy Molecular Dyad
FULL PAPER
in PhCN (tCTS =430 ps). This is attributed to competing trip-
let formation.
bleaching of the Bodipy chromophore at around 500 nm.
Whereas the absorption band at ꢁ740 nm evolves during
decay of the CTS, there is complete recovery of the Bodipy
ground-state absorption band. We conclude, therefore, that
intersystem crossing within the CTS leads to selective, or at
least preferential, population of the C60 T1 state. It has to be
stressed, however, that this route is in competition with
charge recombination to reform the ground state.
Now, direct excitation into the C60-based chromophore in
CH2Cl2 gives rise to the corresponding S1 state, which
decays by first-order kinetics to form the CTS (Figures 10
and 11). The lifetime of the S1 state under these conditions
is 310 ps, but the driving force for light-induced electron
transfer is only À0.16 eV.[41] If we allow for the MTHF re-
sults, which indicate that cyclic voltammetry overestimates
the size of the driving force by ꢁ0.09 eV, it appears that in
CH2Cl2 the CTS is at an energy almost midway between the
S1 and T1 states localized on the C60 fragment; the energy
gap between the CTS and the T1 state is ꢁ890 cmÀ1 (i.e.,
0.11 eV). In this solvent, the C60 S1 state decays by mono-ex-
Electron transfer in dichloromethane: The events that
follow from selective excitation of the Bodipy unit in
CH2Cl2 (eS =8.9) are similar to those described above for
MTHF; the electrochemical properties were described earli-
er. Fluorescence is difficult to resolve from the baseline and
transient absorption spectroscopy indicates that the CTS is
formed rapidly. The S1 state has a lifetime of 42 ps under
these conditions, for which there is a high driving force
(DG8=À0.84 eV) for light-induced electron transfer. In
CH2Cl2, it is predicted[41] that the energy of the CTS will be
intermediate between the S1 and T1 levels localized on the
C60 residue. Indeed, the lifetime of the CTS is only 160 ps,
compared to 430 ps in PhCN, and deactivation leads to par-
tial formation of the T1 state resident on the C60 fragment
(Scheme 5, Figure 9). It is quite clear from these spectral re-
cords that decay of the CTS does not lead to population of
the T1 state localised on the Bodipy unit. This latter species
has a fairly characteristic signature, aided by pronounced
Scheme 5. Outline of the various processes observed following pulsed ex-
citation of either subunit for Bodipy–C60 in CH2Cl2. Energies: BOD*S =
2.44 eV, C60*S =1.76 eV, BOD*T =1.61 eV, C60*T =1.56 eV, CTS=1.67 eV
(BOD=Bodipy).
Figure 10. Transient differential absorption spectra recorded after pulsed
excitation into the C60 unit present in Bodipy–C60 in deoxygenated
CH2Cl2. Delay times: 20, 40, 60, 100, 150, 200, 300, 400, 600, 1000, and
1500 ps.
Figure 9. a) Transient differential absorption spectra recorded after laser
excitation into the Bodipy unit present in Bodipy–C60 in deoxygenated
CH2Cl2. Delay times: 20, 40, 60, 100, 150, 200, 300, 400, 600, 1000, and
*
*
*
)
Figure 11. Decay profiles recorded at 525 ( ), 760 ( ) and 1085 nm (
for the experiment described in Figure 10.
*
*
*
1500 ps. b) Decay profiles recorded at 525 ( ), 760 ( ) and 1085 nm ( ).
Chem. Eur. J. 2009, 15, 7382 – 7393
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7389

A. Harriman, R. Ziessel et al.
ponential kinetics, although the signal is fairly weak, such
that formation of the CTS is essentially irreversible. In turn,
the CTS has a lifetime of 190 ps; this is shorter than the pre-
cursor S1 state and is derived by fitting the transient bleach-
ing of the Bodipy absorption band. As above, decay of the
CTS leads to partial restoration of the ground state and to
population of the C60 T1 state. There is no accompanying
charge recombination to form the Bodipy T1 state. Instead,
slow energy transfer occurs from the C60 T1 state to establish
the equilibrium mixture of triplets. Triplet decay occurs by
first-order kinetics, with a lifetime of 35 ms in deoxygenated
CH2Cl2; this decreases to 5 ms in the presence of 15% (v/v)
iodoethane. As in MTHF and toluene, the rate of energy
transfer between triplet states is slow and occurs over a few
ms, regardless of the presence of iodoethane (Scheme 5).
or below the corresponding S1 state. Fine-tuning of these rel-
ative energy levels allows the CTS to be positioned below
the C60 T1 state or between the S1 and T1 states. There is a
large driving force for electron transfer from the Bodipy S1
state in all solvents, but this process has to compete with
energy transfer to form the C60 S1 state. Light-induced elec-
tron transfer is dominant in all cases, except for non-polar
solvents like toluene. Finally, the two isolated triplet states
lie close in energy, but with the C60 T1 state being slightly
below that of the Bodipy unit. Solvent polarity has no effect
on these latter energy levels. It is also worth highlighting
that intersystem crossing is extremely effective in C60 but in-
efficient in Bodipy. We can comment now on how changes
in solvent polarity affect these intramolecular processes by
reference to Schemes 2–5. A summary of the key rate con-
stants is given in Table 1.
Table 1. Compilation of the various rate constants derived following se-
lective excitation into either chromophore for Bodipy–C60 in solution. All
Conclusions
values given in nsÀ1, except kTT and kD, which are given in msÀ1
.
For the target compound in a non-polar solvent, rapid S1–S1
electronic energy transfer from Bodipy to C60 competes ef-
fectively with light-induced electron transfer.[48] This step is
followed by fast intersystem crossing to populate the T1
state localized on the C60 unit in quantitative yield. Because
this latter metastable state survives for a reasonable period,
T1–T1 energy transfer occurs despite the uphill thermody-
namic position. The net result is a somewhat enhanced trip-
let yield for the Bodipy unit, although this is kept modest by
the poor energetics. A similar strategy has been proposed
for increasing the triplet yield in oligophenylene-C60 blends
and termed the “ping-pong” mechanism.[49] For Bodipy–C60,
the equilibrated mixture of triplet states is dominated by the
C60 triplet. Thus, from the energy gap, the percentage
Bodipy triplet in the mixture is calculated to be 15% at
258C, whereas spectral records indicate that the mixture
contains ꢁ25% of the Bodipy triplet. This level of agree-
ment is judged to be satisfactory and adds confirmation to
the proposed energy levels. Since the triplet yield in isolated
Bodipy is less than a few percent, this represents a substan-
tial increase in triplet population. Coincidently, both isolated
terminals share a common triplet lifetime of ꢁ140 ms,
whereas the equilibrated mixture has a lifetime of 70 ms in
deoxygenated toluene. The difference might be due to mild
electronic interactions as indicated by the quantum chemical
calculations. This lifetime is shortened in the presence of an
external heavy atom, because of increased spin-orbital cou-
pling.[1,2] Under these conditions, the rate of formation of
the Bodipy-based T1 state increases slightly, but the yield is
unaffected. Since intersystem crossing is quantitative in the
C60 unit, the latter situation is to be expected. The apparent
heavy-atom effect on the rate of formation of the Bodipy-
based T1 state, which corresponds to a factor of approxi-
mately two, is less obvious, and might simply reflect the
changed environment.
Process
Toluene
PhCN
MTHF
CH2Cl2
[a]
kSS
5.6
5.6
22.2
7.2
2.4
NA
NA
5.6
9.8
0.74
5.9
0.5
5.6
17.2
2.4
5.7
0.6
k
k
kCR
kTT
kD
G
NA[g]
NA
NA
0.8
(C60)[c]
0.014
0.025
0.028
[a] S1–S1 energy transfer from Bodipy to C60. [b] Electron transfer from
S1 on Bodipy. [c] Electron transfer to S1 on C60. [d] Decay of the CTS.
[e] T1–T1 energy transfer. [f] Decay of equilibrium mixture of triplets.
[g] NA=not applicable.
There is no indication of a Marcus inverted effect,[50] in-
stead the rate of electron transfer increases progressively
with increasing driving force. This finding is based on the as-
sumption that the rate of S1–S1 energy transfer is independ-
ent of solvent polarity. The rate of T1–T1 energy transfer is
unaffected by changes in solvent polarity, but the rate of
triplet decay (kD =1/tT) shows a small solvent effect. This
might reflect a minor contribution from a competing elec-
tron-transfer process and it is notable that electron transfer
does take place from the C60 T1 state in PhCN. The com-
piled rate constants indicate that decay of the CTS is about
two times faster in CH2Cl2 and MTHF than in PhCN. This
effect is attributed to intersystem crossing to the T1 state
competing with charge recombination to the ground state.
This, in turn, is due to the relative energy gaps and to the
mutual positioning of the various energy levels.
The key feature of this system is the realization that the
CTS undergoes (partial) charge recombination to form the
triplet state localized on the C60 residue. Despite the compa-
rable triplet energies for each of the two terminals, there is
no experimental evidence to support the idea of charge re-
combination forming the Bodipy-based triplet state, regard-
less of which chromophore absorbs the incident photon. We
might expect triplet formation to compete with direct
charge recombination to the ground state because of its
Solvent polarity affects the driving force for light-induced
electron transfer to the excited states of the C60 residue, and
can be used to position the CTS energy level either above
7390
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7382 – 7393

Charge Recombination in a Fullerene/Bodipy Molecular Dyad
FULL PAPER
14.5 ppm; FAB-MS: m/z (%): 1605.5 (100) [M+H]+, 1407.4 (25)
[MÀC16H29]; elemental analysis calcd (%) for C122H86N3B (Mr=1604.82):
C 91.31, H 5.40, N 2.62; found: C 91.18, H 5.24, N 2.39.
much smaller energy gap. Such intersystem crossing requires
mixing with one of the local excited triplet states, and it is
clear that the dominant process involves coupling with the
Physical measurements: Absorption spectra were recorded with a Hitachi
U3310 spectrophotometer and luminescence spectra were recorded with
a fully-corrected Jobin–Yvon Fluorolog tau-3 spectrophotometer. Emis-
sion spectra were recorded for optically dilute solutions after purging
with N2. Luminescence quantum yields were determined relative to C60
in toluene,[55] Rhodamine 6G in water[56] and in ethanol,[57] and fluores-
cein in 0.1m NaOH.[56] Emission lifetimes were measured at room tem-
perature with the tau-3 spectrophotometer. For lower temperatures, the
luminescence lifetimes were measured by time-correlated, single-photon
counting using a picosecond laser diode (fwhm=60 ps) as excitation
source. Emission was isolated from scattered laser light with a narrow
bandpass filter and detected with a microchannel plate photocell. After
deconvolution of the instrument response function, the time resolution of
this setup was ꢁ40 ps. Temperature-dependence studies were made with
an Oxford Instruments Optistat DN. All samples were purged thoroughly
with N2 before being sealed into optical cuvettes.
C
60 triplet. Mixing with the Bodipy triplet might be rendered
unfavourable by the orthogonal meso-phenylene linker.[51] It
is also notable that direct intersystem crossing is considera-
bly faster in C60 (kISC ꢁ7ꢄ108 sÀ1) than in Bodipy (kISC ꢁ2ꢄ
106 sÀ1). Furthermore, intersystem crossing to form the C60
triplet involves hole transfer from the HOMO, whereas for-
mation of the Bodipy T1 state demands electron transfer to
the LUMO. Our results are fully consistent with hole trans-
fer having the larger electronic coupling element. Precisely
the same situation has been reported recently for other
donor–bridge–acceptor systems and explained in terms of
frontier molecular orbital levels.[11] For Bodipy–C60, the
HOMOs on the terminals are close in energy but the corre-
sponding LUMOs differ markedly in energy. According to
simple super-exchange theory,[52] the electronic coupling
matrix element for charge recombination depends inversely
on the energy gap between orbitals on the donor and on the
bridge. These energy gaps clearly favour hole transfer and,
as a result, lead to preferential formation of the C60-based
triplet.
Laser flash photolysis studies were performed with an Applied Photophy-
sics Ltd LKS50 instrument. The sample was dissolved in deoxygenated
solvent and exposed to 4 ns pulses delivered from a frequency-doubled,
Q-switched Nd-YAG laser. A pulsed, high-intensity Xe arc lamp was
used as monitoring beam and transient differential absorption spectra
were compiled point by point. At least five individual records were aver-
aged at each wavelength. Kinetic measurements were made at fixed
wavelength, with about 100 individual traces being averaged. Improved
time resolution was achieved with a mode-locked, frequency-tripled, Nd-
YAG laser (fwhm=20 ps) fitted with an optical parametric amplifier.
The monitoring pulse was a white light continuum, delayed with respect
to the excitation pulse with a computer-controlled, optical delay line. The
two pulses were directed almost collinearly through the sample cell. The
monitoring pulse was dispersed with a Princeton Instruments spectro-
graph and detected with a dual-diode array spectrometer. Approximately
150 individual laser shots were averaged at each delay time.
Similar reasoning can be used to explain why T1–T1
energy transfer is so slow in Bodipy–C60. Although this pro-
cess is thermodynamically uphill by about 0.05 eV (i.e.,
400 cmÀ1), the rate is surprisingly slow (Table 1). The reor-
ganization energy for T1–T1 energy transfer is small (l=
0.12 eV)[53] and the coupling element (VDA) can be calculat-
ed as being ꢁ0.01 cmÀ1. In this case, VDA can be considered
as the product of individual coupling elements for electron
transfer through the bridge LUMO and hole transfer
through the HOMO of the bridge.[54] At least one of these
coupling elements must be close to zero and, on energetic
grounds, this is most likely to be that for electron transfer.
Fast transient spectroscopy was performed by pump–probe techniques
using femtosecond pulses delivered from a Ti:sapphire generator ampli-
fied with a multi-pass amplifier pumped through the second harmonic of
a Q-switched Nd:YAG laser. The amplified pulse energies varied from
0.3 to 0.5 mJ and the repetition rate was kept at 10 Hz. Part of the beam
(approximately 20%) was focused onto a second harmonic generator in
order to produce the excitation pulse. The residual output was directed
onto a 4 mm sapphire plate so as to create a white light continuum for
detection purposes. The continuum was collimated and split into two
equal beams. The first beam was used as reference and the second beam
was combined with the excitation pulse and used as the diagnostic beam.
The two beams were directed to different parts of the entrance slit of a
cooled CCD detector and used to calculate differential absorbance
values. The CCD shutter was kept open for 1 s and the accumulated spec-
tra were averaged. This procedure was repeated until about 100 individu-
al spectra had been averaged. Time-resolved spectra were recorded with
a delay line stepped in increments of 100 fs for short time measurements.
This step was increased for longer time measurements. The decay profiles
were fitted globally as the sum of exponentials and deconvoluted with a
Gaussian excitation pulse. The group velocity dispersion across one spec-
trum (ꢁ220 nm) was of the order of 1 ps and the overall temporal reso-
lution of this setup was about 0.8 ps. The sample, possessing an absorb-
ance of approximately 1 at 430 nm, was flowed through a quartz cuvette
(optical pathlength=1 mm) and maintained under an atmosphere of N2.
Experimental Section
Preparation of dyad 4: A stirred mixture of 3 (45 mg, 0.045 mmol) and 2
(45 mg, 0.052 mmol) in benzene (20 mL) and diisopropylamine (5 mL)
was purged with argon for 30 min. Then [PdACHTNURTGNE(UNG PPh3)4] (6 mol%) was
added rapidly under argon and the mixture heated and kept at 708C for
24 h. The cooled solution was evaporated to dryness and treated with
water, and the organic compounds were extracted with dichloromethane.
The organic phase was dried over cotton wool and concentrated to
ꢁ5 mL. Purification was ensured by flash column chromatography using
silica as the solid phase and a mixture of dichloromethane/petroleum
ether as the mobile phase (2:3, v/v). The final material was recrystallised
by slow diffusion of pentane into a solution of the dyad in dichlorome-
thane/cyclohexane . The dark brown precipitate was recovered by centri-
fugation and dried under high vacuum, affording 62 mg of 4 (75% isolat-
1
Electrochemical studies employed cyclic voltammetry with a convention-
al three-electrode system using a BAS CV-50W voltammetric analyser
equipped with a Pt microdisk (2 mm2) working electrode and a silver
wire counter-electrode. Ferrocene was used as an internal standard and
was calibrated against a saturated calomel reference electrode (SCE)
separated from the electrolysis cell by a glass frit pre-soaked with electro-
lyte solution. Solutions contained the electroactive substrate in deoxygen-
ated and anhydrous dimethylformamide containing tetra-n-butylammoni-
ed yield). H NMR (CDCl3, 400.1 MHz): d=7.85–7.80 (m, 2H), 7.64–7.61
3
2
(m, 4H), 7.30 (dd, J=8.2 Hz, 2H), 6.00 (s, 2H), 5.03 (d, J=9.2 Hz, 1H),
5.00 (s, 1H), 4.31 (d, 2J=9.2 Hz, 1H), 2.86 (s, 3H), 2.70 (s, 6H), 2.16 (t,
3J=7.3 Hz, 4H), 1.56 (s, 18H), 1.49 (s, 6H), 1.40 (s, 30H), 0.88 ppm (t,
3J=6.6 Hz, 6H); 13C{1H} NMR (CDCl3, 50.3 MHz): d=155.7, 147.7,
146.7, 146.6, 146.3, 146.0, 145.7, 145.0, 144.8, 143.0, 142.64, 142.57, 142.5,
142.3, 142.1, 142.0, 141.9, 140.7, 132.5, 129.6, 128.9, 123.9, 121.9, 98.2,
96.5, 83.7, 40.4, 32.3, 30.1–29.5 (8 singlets), 23.1, 20.3, 16.4, 15.1,
Chem. Eur. J. 2009, 15, 7382 – 7393
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7391

A. Harriman, R. Ziessel et al.
um hexafluorophosphate (0.1m) as supporting electrolyte. The quoted
half-wave potentials were reproducible within ꢁ10 mV.
lewski, D. R. Minsek, M. P. Niemczyk, W. A. Svec, N. C. Yang, J.
Am. Chem. Soc. 1990, 112, 2823–2824.
[14] a) H. van Willigen, G. Jones II, M. S. Farahat, J. Phys. Chem. 1996,
100, 3312–3316; b) M. L. Horng, K. Dahl, G. Jones II, M. Maroncel-
li, Chem. Phys. Lett. 1999, 315, 363–370.
[15] A. Harriman, L. J. Mallon, G. Ulrich, R. Ziessel, ChemPhysChem
2007, 8, 1207–1214.
Acknowledgements
[16] Z. E. X. Dance, S. M. Mickley, T. M. Wilson, A. B. Ricks, A. M.
Scott, M. A. Ratner, M. R. Wasielewski, J. Phys. Chem. A 2008, 112,
4194–4201.
[17] a) D. M. Guldi, Chem. Soc. Rev. 2002, 31, 22–36; b) D. Gust, T. A.
Moore, A. L. Moore, Acc. Chem. Res. 2001, 34, 40–48; c) Fullerenes:
From Synthesis to Opto-electronic Applications (Eds.: D. M. Guldi,
N. Armaroli), Kluwer Academic, Dordrecht, 2002.
[18] a) J. Baffreau, S. Leroy-Lhez, V. A. Nguyen, R. M. Williams, P. Hud-
homme, Chem. Eur. J. 2008, 14, 4974–4992; b) S. Karlsson, J.
Modin, H. C. Becker, L. Hammarstrçm, H. Grennberg, Inorg.
Chem. 2008, 47, 7286–7294; c) R. Berera, G. F. Moore, I. H. Van
Stokkum, G. Kodis, P. A. Liddell, M. Gervaldo, R. Van Grondelle,
J. T. Kennis, D. Gust, T. A. Moore, A. L. Moore, Photochem. Photo-
biol. Sci. 2006, 5, 1142–1149; d) S. A. Vail, D. I. Schuster, D. M.
Guldi, M. Isosomppi, N. Tkachenko, H. Lemmetyinen, A. Palker, L.
Echegoyen, X. Chen, J. Z. Zhang, J. Phys. Chem. B 2006, 110,
14155–14166.
[19] S. P. Sibley, S. M. Argentine, A. H. Francis, Chem. Phys. Lett. 1992,
188, 187–193.
[20] W.-C. Hung, C.-D. Ho, C.-P. Liu, Y.-P. Lee, J. Phys. Chem. 1996, 100,
3927–3932.
[21] A. Sassara, G. Zerza, M. Chergui, Chem. Phys. Lett. 1996, 261, 213–
220.
[22] Y. Zeng, L. Biczok, H. Linschitz, J. Phys. Chem. 1992, 96, 5237–
5239.
We thank the CNRS, EPSRC (EP/D032946/1), Universitꢀ Louis Pasteur
de Strasbourg and the University of Newcastle for financial support of
this work. Dr. Laure Bonardi and Julie Batcha Seneclauze are acknowl-
edged for providing samples of compounds 2 and 3, respectively.
[1] S. P. McGlynn, T. Azumi, M. Kinoshita, Molecular Spectroscopy of
the Triplet State, Englewood Cliffs, New Jersey, 1969.
[2] P. van Haver, M. van der Auweraer, L. Viaene, F. C. De Schryver,
J. W. Verhoeven, H. J. Van Ramesdonk, Chem. Phys. Lett. 1992, 198,
361–366.
[3] a) W. M. Bartczak, A. Hummel, Chem. Phys. Lett. 1993, 208, 232–
236; b) M. Murakami, K. Ohkubo, P. Mandal, T. Ganguly, S. Fuku-
zumi, J. Phys. Chem. A 2008, 112, 635–642; c) N. L. Lavrik, J.
Chem. Phys. 2001, 114, 9492–9495; d) S. Tero-Kubota, Pure Appl.
Chem. 2001, 73, 519–523.
[4] a) H. Ohkita, S. Cook, Y. Astuti, W. Duffy, M. Heeney, S. Tierney, I.
McCulloch, D. D. C. Bradley, J. R. Durrant, Chem. Commun. 2006,
3939–3941; b) C. Rothe, S. King, A. Monkman, Nat. Mater. 2006, 5,
463–466; c) C. Rothe, L. O. Palsson, A. Monkman, Chem. Phys.
2002, 285, 95–101; d) Z. Shuai, D. Beljonne, R. J. Silbey, J. L.
Brꢀdas, Phys. Rev. Lett. 2000, 84, 131–134.
[5] U. Hofstra, T. J. Schaafsma, G. M. Sanders, M. Van Dijk, H. C. Van
der Plas, D. G. Johnson, M. R. Wasielewski, Chem. Phys. Lett. 1988,
151, 169–172.
[6] a) T. Nishimura, N. Nakashima, N. Mataga, Chem. Phys. Lett. 1977,
46, 334–338; b) N. Mataga, M. Migita, T. Nishimura, J. Mol. Struct.
1978, 47, 199–210; c) N. Mataga, Radiat. Phys. Chem. 1983, 21, 83–
89; d) A. M. Swinnen, M. van der Auwerear, F. C. De Schryver, K.
Nakatani, T. Okada, N. Mataga, J. Am. Chem. Soc. 1987, 109, 321–
330; e) K. Nakatani, T. Okada, N. Mataga, F. C. De Schryver, Chem.
Phys. 1988, 121, 87–92.
[23] a) R. R. Hung, J. J. Grabowski, J. Phys. Chem. 1991, 95, 6073–6075;
b) M. Terazima, N. Hirota, H. Shinohara, Y. Saito, J. Phys. Chem.
1991, 95, 9080–9085.
[24] a) I. Laszlo, L. Udvardi, J. Mol. Struct. 1989, 192, 271–278; b) F.
Negri, G. Orlandi, F. Zerbetto, Chem. Phys. Lett. 1988, 144, 31–37;
c) J. Feng, J. Li, Z. Wang, M. C. Zerner, Int. J. Quantum Chem.
1991, 39, 331–344.
[7] a) D. Kuciauskas, P. A. Liddell, A. L. Moore, T. A. Moore, D. Gust,
J. Am. Chem. Soc. 1998, 120, 10880–10886; b) M. Di Valentin, A.
Bisol, G. Agostini, A. L. Moore, T. A. Moore, D. Gust, R. E. Pala-
cios, S. L. Gould, D. Carbonera, Mol. Phys. 2006, 104, 1595–1607;
c) Y. Kobori, S. Yamauchi, K. Akiyama, S. Tero-Kubota, H. Imahori,
S. Fukuzumi, J. R. Norris, Proc. Natl. Acad. Sci. USA 2005, 102,
10017–10022.
[8] a) J. W. Verhoeven, J. Photochem. Photobiol. C 2006, 7, 40–60;
b) J. W. Verhoeven, H. J. Van Ramesdonk, H. Zhang, M. M. Groene-
veld, A. C. Benniston, A. Harriman, Int. J. Photoenergy 2005, 7,
103–108; c) J. W. Verhoeven, H. J. Van Ramesdonk, M. M. Groene-
veld, A. C. Benniston, A. Harriman, ChemPhysChem 2005, 6, 2251–
2260.
[9] a) K. Okamoto, T. Hasobe, N. V. Tkachenko, H. Lemmetyinen, P. V.
Kamat, S. Fukuzumi, J. Phys. Chem. A 2005, 109, 4662–4670; b) S.
Fukuzumi, H. Kotyani, K. Ohkubo, S. Ogo, N. V. Tkachenko, H.
Lemmetyinen, J. Am. Chem. Soc. 2004, 126, 1600–1601.
[10] a) H. J. Van Ramesdonk, B. H. Bakker, M. M. Groeneveld, J. W.
Verhoeven, B. D. Allen, J. P. Rostron, A. Harriman, J. Phys. Chem.
A 2006, 110, 13145–13150; b) A. C. Benniston, A. Harriman, P. Y.
Li, J. P. Rostron, H. J. van Ramesdonk, M. M. Groeneveld, H.
Zhang, J. W. Verhoeven, J. Am. Chem. Soc. 2005, 127, 16054–16064.
[11] Z. E. X. Dance, M. J. Ahrens, A. M. Vega, A. B. Ricks, D. W.
McCamant, M. A. Ratner, M. R. Wasielewski, J. Am. Chem. Soc.
2008, 130, 830–832.
[25] F. Nastasi, F. Puntoriero, S. Campagna, S. Schergna, M. Maggini, F.
Cardinali, B. Delavaux-Nicot, J.-F. Nierengarten, Chem. Commun.
2007, 3556–3558.
[26] a) J. W. Arbogast, A. P. Darmanyan, C. S. Foote, Y. Rubin, F. N. Die-
derich, M. M. Alvarez, R. B. Whetten, J. Phys. Chem. 1991, 95, 11–
12; b) L. Biczok, H. Linschitz, R. I. Walter, Chem. Phys. Lett. 1992,
195, 339–346.
[27] a) R. Ziessel, G. Ulrich, A. Harriman, New J. Chem. 2007, 31, 496–
501; b) G. Ulrich, R. Ziessel, A. Harriman, Angew. Chem. 2008, 120,
1202–1219; Angew. Chem. Int. Ed. 2008, 47, 1184–1201; c) R. Zies-
sel, C. R. Acad. Sciences 2007, 10, 622–629; d) A. Loudet, K. Bur-
gess, Chem. Rev. 2007, 107, 4891–4932.
[28] M. Galletta, S. Campagna, M. Quesada, G. Ulrich, R. Ziessel,
Chem. Commun. 2005, 4222–4224.
[29] A. Harriman, J. P. Rostron, M. Cesario, G. Ulrich, R. Ziessel, J.
Phys. Chem. A 2006, 110, 7994–8002.
[30] S. A. Olsen, A. M. Bond, R. G. Compton, G. Lazarev, P. J. Mahon, F.
Marken, C. L. Raston, V. Tadesco, R. D. Webster, J. Phys. Chem. A
1998, 102, 2641–2649.
[31] B. D. Allen, A. C. Benniston, A. Harriman, L. J. Mallon, C. Pariani,
Phys. Chem. Chem. Phys. 2006, 8, 4112–4118.
[32] C. Goze, G. Ulrich, R. Ziessel, J. Org. Chem. 2007, 72, 313–322.
[33] M. Maggini, G. Scorrano, J. Am. Chem. Soc. 1993, 115, 9798–9799.
[34] G. Ulrich, S. Goeb, A. De Nicola, P. Retailleau, R. Ziessel, Synlett
2007, 1517–1520.
[35] R. Ziessel, L. Bonardi, P. Retailleau, G. Ulrich, J. Org. Chem. 2006,
71, 3093–3102.
[12] T. Okada, I. Karaki, E. Matsuzawa, N. Mataga, Y. Sakata, S.
Misumi, J. Phys. Chem. 1981, 85, 3957–3960.
[13] a) A. M. Brun, A. Harriman, Y. Tsuboi, T. Okada, N. Mataga, J.
Chem. Soc. Faraday Trans. 1995, 91, 4047–4057; b) W. Rettig, R.
Haag, J. Wirz, Chem. Phys. Lett. 1991, 180, 216–221; c) M. R. Wasie-
[36] a) A. C. Benniston, G. Copley, K. J. Elliott, R. W. Harrington, W.
Clegg, Eur. J. Org. Chem. 2008, 2705–2713; b) S. Hisato, U. Yasu-
7392
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7382 – 7393

Charge Recombination in a Fullerene/Bodipy Molecular Dyad
FULL PAPER
teru, K. Hirotatsu, N. Tetsuo, J. Am. Chem. Soc. 2007, 129, 5597–
5604.
[37] A. Lembo, P. Tagliatesta, D. M. Guldi, J. Phys. Chem. A 2006, 110,
11424–11434.
[38] a) D. Dubois, K. M. Kadish, S. Flanigan, R. E. Haufler, L. P. F. Chi-
banke, L. J. Wilson, J. Am. Chem. Soc. 1991, 113, 4364–4366; b) D.
Dubois, M. T. Jones, K. M. Kadish, J. Am. Chem. Soc. 1992, 114,
6446–6451.
[48] a) P. Van Haver, N. Helsen, S. Depaemelaere, M. Van der Auweraer,
F. C. De Schryver, J. Am. Chem. Soc. 1991, 113, 6849–6857; b) M.
Lor, J. Thielemans, L. Viaene, M. Cotlet, J. Hofkens, T. Weil, C.
Hampel, K. Mꢅllen, J. W. Verhoeven, M. Van der Auweraer, F. C.
De Schryver, J. Am. Chem. Soc. 2002, 124, 9918–9925; c) M. Cotlet,
S. Masuo, M. Lor, E. Fron, M. Van der Auweraer, K. Mꢅllen, J.
Hofkens, F. C. De Schryver, Angew. Chem. 2004, 116, 6242–6246;
Angew. Chem. Int. Ed. 2004, 43, 6116–6120.
[39] a) R. Ziessel, C. Goze, G. Ulrich, M. Cesario, P. Retailleau, A. Har-
riman, J. P. Rostron, Chem. Eur. J. 2005, 11, 7366–7378; b) Y. Gabe,
T. Ueno, Y. Urano, H. Kojima, T. Nagano, Anal. Bioanal. Chem.
2006, 386, 621–626.
[49] R. Schueppel, C. Uhrich, M. Pfeiffer, K. Leo, E. Brier, E. Reinhold,
P. Braeuerle, ChemPhysChem 2007, 8, 1497–1503.
[50] a) R. A. Marcus, J. Chem. Phys. 1956, 24, 966–978; b) R. A. Marcus,
J. Chem. Phys. 1956, 24, 979–989.
[40] a) H. Imahori, K. Tamaki, D. M. Guldi, C. Luo, M. Fujitsuka, O. Ito,
Y. Sakata, S. Fukuzumi, J. Am. Chem. Soc. 2001, 123, 2607–2617;
b) H. Imahori, D. M. Guldi, K. Tamaki, Y. Yoshida, C. Luo, Y.
Sakata, S. Fukuzumi, J. Am. Chem. Soc. 2001, 123, 6617–6628.
[41] A. Weller, Z. Phys. Chem. (Muenchen Ger.) 1982, 133, 93–98.
[42] J. Kroon, J. W. Verhoeven, M. N. Paddon-Row, A. M. Oliver,
Angew. Chem. 1991, 103, 1398–1401; Angew. Chem. Int. Ed. Engl.
1991, 30, 1358–1361.
[43] R. Ziessel, C. Goze, G. Ulrich, Synthesis 2007, 963–969.
[44] M. Gevaert, P. V. Kamat, J. Phys. Chem. 1992, 96, 9883–9888.
[45] Y. Araki, Y. Yasumura, O. Ito, J. Phys. Chem. B 2005, 109, 9843–
9848.
[51] a) J. Lappe, R. J. Cave, M. D. Newton, I. V. Rostov, J. Phys. Chem. B
2005, 109, 6610–6619; b) G. P. Wiederrecht, W. A. Svec, M. R. Wa-
sielewski, T. Galili, H. Levanon, J. Am. Chem. Soc. 2000, 122, 9715–
9722.
[52] H. M. McConnell, J. Chem. Phys. 1961, 35, 508–515.
[53] A. Harriman, A. Khatyr, R. Ziessel, A. C. Benniston, Angew. Chem.
2000, 112, 4457–4460; Angew. Chem. Int. Ed. 2000, 39, 4287–4290.
[54] G. L. Closs, P. Piotrowiak, J. M. MacInnis, G. R. Fleming, J. Am.
Chem. Soc. 1988, 110, 3751–3753.
[55] Y.-P. Sun, P. Wang, N. B. Hamilton, J. Am. Chem. Soc. 1993, 115,
6378–6381.
[56] D. Magde, G. E. Rojas, P. Seybold, Photochem. Photobiol. 1999, 70,
737–744.
[46] R. V. Bensasson, T. Hill, C. Lambert, E. J. Land, S. Leach, T. G.
Truscott, Chem. Phys. Lett. 1993, 201, 326–335.
[57] M. Fischer, J. Georges, Chem. Phys. Lett. 1996, 260, 115–118.
[47] a) M. R. Wasielewski, D. G. Johnson, W. A. Svec, K. M. Kersey,
D. W. Minsek, J. Am. Chem. Soc. 1988, 110, 7219–7221; b) H. Hei-
tele, P. Frinckle, S. Weeren, F. Pollinger, M. E. Michel-Beyerle, J.
Phys. Chem. 1989, 93, 5173–5179; c) A. Harriman, V. Heitz, M.
Ebersole, H. Van Willigen, J. Phys. Chem. 1994, 98, 4982–4989.
Received: February 17, 2009
Published online: June 23, 2009
Chem. Eur. J. 2009, 15, 7382 – 7393
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7393