Cyclic Tetramers of Zinc Chlorophylls as a Coupled Light-Harvesting Antenna-Charge-Separation System

Article abstract of DOI:10.1002/chem.201503789

A coupled light-harvesting antenna-charge-separation system, consisting of self-assembled zinc chlorophyll derivatives that incorporate an electron-accepting unit, is reported. The cyclic tetramers that incorporated an electron acceptor were constructed by the co-assembly of a pyridine-appended zinc chlorophyll derivative, ZnPy, and a zinc chlorophyll derivative further decorated with a fullerene unit, ZnPyC₆₀. Comprehensive steady-state and time-resolved spectroscopic studies were conducted for the individual tetramers of ZnPy and ZnPyC₆₀ as well as their co-tetramers. Intra-assembly singlet energy transfer was confirmed by singlet-singlet annihilation in the ZnPy tetramer. Electron transfer from the singlet chlorin unit to the fullerene unit was clearly demonstrated by the transient absorption of the fullerene radical anion in the ZnPyC₆₀ tetramer. Finally, with the co-tetramer, a coupled light-harvesting and charge-separation system with practically 100 % quantum efficiency was demonstrated.

Full text of DOI:10.1002/chem.201503789

DOI: 10.1002/chem.201503789
Full Paper
&
Self-Assembly
Cyclic Tetramers of Zinc Chlorophylls as a Coupled Light-
Harvesting Antenna–Charge-Separation System
Yoshinao Shinozaki,[a] Kei Ohkubo,[b, d] Shunichi Fukuzumi,*[b, c, d] Kosuke Sugawa,[a] and
Joe Otsuki*[a]
Abstract: A coupled light-harvesting antenna–charge-sepa-
ration system, consisting of self-assembled zinc chlorophyll
derivatives that incorporate an electron-accepting unit, is re-
ported. The cyclic tetramers that incorporated an electron
acceptor were constructed by the co-assembly of a pyridine-
appended zinc chlorophyll derivative, ZnPy, and a zinc chlor-
ophyll derivative further decorated with a fullerene unit,
ZnPyC60. Comprehensive steady-state and time-resolved
spectroscopic studies were conducted for the individual tet-
ramers of ZnPy and ZnPyC60 as well as their co-tetramers.
Intra-assembly singlet energy transfer was confirmed by sin-
glet–singlet annihilation in the ZnPy tetramer. Electron
transfer from the singlet chlorin unit to the fullerene unit
was clearly demonstrated by the transient absorption of the
fullerene radical anion in the ZnPyC60 tetramer. Finally, with
the co-tetramer, a coupled light-harvesting and charge-sepa-
ration system with practically 100% quantum efficiency was
demonstrated.
lished synthetic methodology.[4] However, compared with
models of either one of light-harvesting antenna or charge-
separation systems, coupled antenna–charge-separation sys-
tems are still rare.[5] For example, Kobuke et al. reported cyclic
bis(zinc porphyrin) arrays as an antenna that could incorporate
a tripodal fullerene derivative as an electron acceptor.[6]
Introduction
The high efficiency of natural photosynthesis relies on the
presence of light-harvesting antennas acting in concert with
the charge-separating reaction center. The antenna system re-
sides in the vicinity of the reaction center to capture and
supply energy from sunlight to the reaction center through ex-
cited energy transfer among (bacterio)chlorophyll molecules.[1]
The reaction center converts the excitation energy from the
antenna system to potential energy in the form of charge sep-
aration by consecutive electron-transfer processes.[2]
Semisynthetic chlorophylls are also an attractive candidate
for artificial photosynthetic systems due to strong light absorp-
tion in the visible region and favorable electrochemical proper-
ties of chlorophyll derivatives, although the synthetic chemistry
of chlorophylls is more demanding than that of porphyrins.[7]
Although a variety of chlorophyll-based antenna models[8] and
chlorophyll-based charge-separation models[7a,9] have been re-
ported, chlorophyll-based coupled systems that integrate
a light-harvesting antenna and charge separation is very limit-
ed. We are aware of J-type aggregates of zinc chlorophyll de-
rivatives that incorporate a bacteriochlorin–fullerene dyad as
the only example of such a coupled system.[10] However, no
clear evidence for the charge-separated state was reported.
We and other groups previously reported discrete and well-
defined self-assembled cyclic oligomers of zinc porphyrin[11]
and chlorophyll derivatives[8c,d,12] carrying a pyridine unit. In
this motif, the pyridine unit introduced onto the periphery of
the chlorophyll core coordinates to the zinc center in another
molecule in a vertical fashion, which leads to the formation of
cyclic structures in solution. Such coordination-directed cyclic
arrays are considered a promising motif for light-harvesting ar-
chitecture. Indeed, energy transfer within the cyclic arrays has
been demonstrated.[8c,11b,12a] Also, charge-separation functional-
ities have been introduced in one of such cyclic arrays.[12b]
However, no coupled system of light harvesting and charge
separation has been achieved within this motif. We have pre-
pared a chlorophyll–fullerene dyad (ZnPyC60) by introducing
To understand and artificially reproduce natural photosyn-
thetic systems, photophysical events that occur in model pho-
tosynthetic systems have been intensely studied.[3] Porphyrins,
which are synthetic chlorophyll analogues, are most often se-
lected as an electron donor and/or a light-harvesting chromo-
phore owing to their intense light absorption and well-estab-
[a] Dr. Y. Shinozaki, Dr. K. Sugawa, Prof. J. Otsuki
College of Science and Technology, Nihon University
1-8-14 Kanda Surugadai, Chiyoda-ku, Tokyo 101-8308 (Japan)
E-mail: otsuki.joe@nihon-u.ac.jp
[b] Prof. K. Ohkubo, Prof. S. Fukuzumi
Department of Material and Life Science
Graduate School of Engineering, ALCA and SENTAN, JST
Osaka University, 2-1 Yamada-oka, Suita, Osaka 565-0871 (Japan)
E-mail: fukuzumi@chem.eng.osaka-u.ac.jp
[c] Prof. S. Fukuzumi
Department of Bioinspired Science, Ewha Womans University
Seoul 120-750 (Korea)
[d] Prof. K. Ohkubo, Prof. S. Fukuzumi
Faculty of Science and Engineering, Meijo University
ALCA and SENTAN, JST, Nagoya, Aichi 468-0073 (Japan)
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under http://dx.doi.org/10.1002/chem.201503789.
Chem. Eur. J. 2016, 22, 1165 – 1176
1165
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
a fullerene unit into the pyridine-appended chlorophyll mole-
cule (ZnPy) and investigated the photodynamics of their co-as-
sembly, as well as the assemblies of individual components in
solution (Scheme 1). We herein demonstrate the first successful
construction of a coupled antenna–charge-separation system
by using chlorophyll derivatives in which clear evidence is ob-
tained for both energy transfer and charge-separation process-
es. Furthermore, the energy- and electron-transfer processes
are both much faster than those of the excited-state lifetime,
which leads to practically 100% quantum efficiency. We used
noncoordinating, nonpolar benzene as the solvent for this
study, which not only stabilized the assemblies, but also mim-
icked the low dielectric environment surrounding natural pho-
tosynthetic pigments.[13] It is reported that an appropriate
value of the dielectric constant representing the protein envi-
ronment of a reaction center of photosynthesis is approximate-
ly two.[13a]
Results and Discussion
Syntheses
The synthetic procedures for preparing the zinc chlorophyll de-
rivative appended by pyridine, ZnPy, and its fullerene dyad,
ZnPyC60, are shown in Schemes 2 and 3. The fullerene moiety
8 in ZnPyC60 was synthesized in seven steps from 1,3,5-trihy-
droxybenzene (1). The hydroxy groups in 1 were etherified by
1-bromododecane to give 1,3,5-tridodecanoxybenzene (2). The
Friedel–Crafts acylation of the alkoxybenzene with glutaric an-
hydride gave 4-benzoylbutylic acid 3.[14] Methyl ester 4 was ob-
tained through Fischer esterification of acid 3 in methanol in
the presence of a catalytic amount of sulfuric acid. Substitution
of the carbonyl group of 4 with tosyl hydrazide in methanol
did not proceed, probably due to steric repulsion from the
alkoxy groups.[15] Instead, tosyl hydrazone 5 was almost quanti-
tatively obtained under the general conditions used for the
imine synthesis, with p-TsOH as an acid catalyst. The 1,3-dipolar
Scheme 2. Synthetic procedures for preparing the fullerene moiety. a) 1-Bro-
mododecane, K2CO3, DMF, reflux, 24 h; b) glutaric anhydride, AlCl3, CH2Cl2,
08C, 10 min, RT, overnight; c) H2SO4, MeOH, reflux, overnight; d) p-TsNHNH2
(Ts=tosyl), p-TsOH·H2O, toluene, reflux, 2.5 h; e) i) NaOMe, pyridine, 1 h;
ii) C60, o-dichlorobenzene (o-DCB), reflux, overnight; f) o-DCB, reflux, 5 days;
g) diisobutyl aluminum hydride (DIBAL-H), heptane, toluene, overnight.
addition of the diazo compound generated in situ by the
base-induced decomposition of hydrazone 5 to form fullerene
gave a mixture of [5,6] (6) and [6,6] (7) isomers.[15] The com-
plete isomerization of the [5,6] isomer to the [6,6] isomer was
accomplished by heating the mixture in o-DCB for five days.[15]
Scheme 1. Co-assembly of the zinc chlorophyll, ZnPy, and its fullerene dyad, ZnPyC60, (3:1) into (ZnPy)3(ZnPyC60), which is the most abundant species in the
statistical mixture.
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1166
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ZnPyC60 almost quantitatively. The zinc chlorophyll ZnPy with-
out a fullerene unit was prepared by a similar procedure with
5-phenylpentanol derivative 10 (Scheme S1 in the Supporting
Information), which was obtained in four steps from 1, instead
of alcoholic fullerene 8.
Self-assembly
1
The H NMR spectra of the zinc complexes, ZnPy and ZnPyC60,
in coordinatable [D8]THF and non-coordinatable C6D6 (10 mm,
298 K) are shown in Figure 1. Whereas the signals of the pyrid-
yl protons of the zinc complexes appeared in the aromatic
region in [D8]THF (Pya: d=8.36 ppm for ZnPy, d=8.28 ppm
for ZnPyC60; Pyb: d=7.66 ppm for ZnPy, d=7.65 ppm for
ZnPyC60), remarkable upfield shifts were observed in C6D6
(Pya: d=3.45 ppm for ZnPy, d=3.52 ppm for ZnPyC60; Pyb:
d=5.42 ppm for ZnPy, d=5.46 ppm for ZnPyC60). The sharp
and well-structured signals in the spectra in [D8]THF indicate
that the zinc complexes exist as THF-coordinated monomeric
species.[8d] On the other hand, the remarkable upfield shifts ob-
served in the spectra of ZnPy and ZnPyC60 in C6D6 relative to
Scheme 3. Synthetic procedure for preparing the chlorophyll derivatives.
a) OsO4, NaIO4, AcOH, H2O, THF, overnight; b) Bestmann–Ohira reagent 17,
CsCO3, THF, MeOH, 3 h; c) 4-iodopyridine, [Pd2(dba)3] (dba=dibenzylidenea-
cetone), P(o-tolyl)3, Et3N, toluene, reflux, overnight; d) bis(dibutylchloroti-
n(IV)) oxide, toluene, reflux, overnight; e) Zn(OAc)2·2H2O, MeOH, CHCl3, 4 h.
The methyl ester group in 7 was reduced with DIBAL-H to
afford alcoholic fullerene 8.[16]
Pyridine-appended chlorophyll derivative 19 was synthe-
sized in three steps from methyl pyropheophorbide-a (11;
Scheme 3), which was obtained by following a previously re-
ported procedure.[17] The vinyl group in 11 was oxidized to the
formyl group by Lemieux–Johnson oxidation with OsO4 and
NaIO4 to give methyl pyropheophorbide-d (12).[18] After the
formyl group was converted into the ethynyl group by Sey-
ferth–Gilbert homologation with Bestmann–Ohira reagent
17,[19] the copper-free Sonogashira coupling of 18 with 4-iodo-
pyridine gave the pyridine-appended chlorophyll derivative
19.[20] Transesterification of 19 with bis(dibutylchlorotin(IV))
oxide as a catalyst successfully afforded the free-base chloro-
phyll FbPyC60.[21] Zinc insertion into the free-base chlorophyll
derivative with Zn(OAc)2·2H2O afforded the zinc complex
Figure 1. 1H NMR spectra of the zinc complexes a) ZnPy and b) ZnPyC60 in
[D8]THF (upper) and C6D6 (lower; 10 mm, 298 K). Asterisks signify the residual
chloroform.
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1167
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
those in [D8]THF indicate the formation of the intermolecular
NPyÀZn bond.[8d,22] The values of these chemical shifts for ZnPy
and ZnPyC60 were almost insensitive to the wide range of tem-
perature of 298–343 K (Figure S1 in the Supporting Informa-
tion), which was indicative of a high stability of axial coordina-
tion. The exceptional stability of axial coordination is a result
of the formation of cyclic oligomers.[8d,23] Considering the coor-
dination geometry of the zincÀpyridine axial coordination, in
which a pyridine moiety stands upright with respect to the
planar chlorin macrocycle, it is natural to assume that a pre-
dominant species is a cyclic tetramer. Indeed, we have previ-
ously shown that another zinc chlorophyll derivative, which
was structurally very similar, formed cyclic tetramers on the
basis of molecular-weight determination by means of vapor
pressure osmometry.[8d] In pioneering works, Wasielewski and
co-workers demonstrated that structurally related chlorin–pyri-
dine conjugates formed cyclic tetramers on the basis of small-
angle X-ray scattering experiments.[8c,12b] However, the possibil-
ity of the formation of other cyclic oligomers, such as tri- and
pentamers, as minor species in the present case cannot be
ruled out.
The diffusion-ordered spectroscopy (DOSY) measurements
were also consistent with tetramer formation of ZnPy and
ZnPyC60. The diffusion coefficients (D) obtained by DOSY
measurements in C6D6 (10 mm, 298 K) were significantly small-
er for the zinc complexes, ZnPy and ZnPyC60 ((2.81Æ0.01)”
10À10 and (2.28Æ0.01)”10À10 m2 sÀ1, respectively), than those
for the free-base compounds, FbPy and FbPyC60 ((4.23Æ
0.02)”10À10 and (3.58Æ0.01)”10À10 m2 sÀ1, respectively). The
hydrodynamic radii (r) were estimated on the basis of these
diffusion constants by use of the Stokes–Einstein equation
[Eq. (1)]:
Figure 2. Molecular models of monomeric free-base derivatives and tetra-
meric zinc complexes optimized by MM+. a) FbPy and ZnPy. b) FbPyC60
and ZnPyC60. The circles were drawn by use of the calculated hydrodynamic
radiuses obtained from DOSY measurements according to equation 1.
kBT
ð1Þ
r ¼
6phD
in which M, Mn and Cn represent the concentrations of mono-
mer, linear n-mers, and cyclic n-mers, respectively. Ki is the in-
termolecular association constant for polymerization, which is
assumed to be identical for all n, and Kc,4 is the equilibrium
constant for the cyclization of the linear tetramer.[25] Only the
cyclic tetramer is taken into account for simplicity. The value of
Ki was estimated to be 1.1”104 mÀ1 by NMR spectroscopy titra-
tion in C6D6 with two model compounds (Scheme 4): a zinc
chlorophyll derivative M-1 and 4-phenylpyridine M-2 (Figure S3
in the Supporting Information).[26] The total concentration, MT,
of the monomer unit is given by Equation (3):
in which kB, T, and h are the Boltzmann constant, absolute tem-
perature, and the viscosity of the solvent, respectively.[24] The
radii obtained for FbPy, FbPyC60, ZnPy, and ZnPyC60 were 8.5,
10.1, 12.9, and 15.9 Š, respectively. Although a sphere with the
obtained radius (8.5 Š) for the free-base FbPy fits the chlorin
core of the molecule, as shown in Figure 2, a sphere for
FbPyC60 (10.1 Š) is about 20% larger in radius due to the pres-
ence of a fullerene unit. A sphere (12.0 Š) for ZnPy fits closely
the chlorin part of the proposed cyclic tetramer. A sphere
(15.9 Š) for ZnPyC60 is again about 20% larger in radius.
We estimated the stability of the cyclic tetramers by use of
the concentration-dependent absorption spectra, which
showed redshifts in the Soret and Q bands (Figure S2 in the
Supporting Information). The equilibrium for this system can
be represented by Equation (2):
1
x
Kc;4
Ki
ð3Þ
MT ¼
þ 4
x4
Ki 1 À x2
in which x is defined as x=KiM. The value of Kc,4 was then esti-
mated by curve fitting to the absorption data by assuming
that the uncoordinated and coordinated chlorophyll molecules
afforded distinct spectra. The stabilities of the cyclic tetramers
were very similar for ZnPy and ZnPyC60 with values of Kc,4 of
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1168
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 3. Steady-state absorption spectra of ZnPy, ZnPyC60, and fullerene
derivative 7 (ꢀ2 mm) in a) neat benzene and b) benzene containing 1% pyri-
dine.
Scheme 4. Structures of the model compounds used in this study.
1.1”103 and 1.0”103, respectively. By using the values of Ki
and Kc,4 obtained herein, we simulated the fraction of self-as-
sembled species, including monomer, dimer, linear oligomers
and polymers, and cyclic tetramer in solution as a function of
the total zinc chlorophyll concentration (Figure S4 in the Sup-
porting Information). Although an appreciable amount of mon-
omeric species exists at less than the total concentration of
0.1 mm, the cyclic tetramer is almost exclusively present at
more than 1.0 mm. Therefore, we chose the total concentration
of 1.0 mm for the photophysical analyses because we were pri-
marily interested in photodynamics that occurred in the cyclic
tetramers.
in the region longer than lꢀ400 nm are due to absorption by
the chlorophyll moiety, whereas the band at lꢀ330 nm is that
of the fullerene moiety. A slight redshift and band broadening
at both the Soret and Q bands were observed in the spectra of
ZnPyC60, relative to those of ZnPy in either solution, which
suggested that the chlorin and fullerene moieties in ZnPyC60
weakly interacted intramolecularly.
Fluorescence spectra of the zinc complexes ZnPy and
ZnPyC60 in benzene (ꢀ2 mm), photoexcited at l=630 nm, at
which the zinc chlorin moiety is exclusively excited, are shown
in Figure 4. Intense fluorescence was observed for ZnPy,
whereas that of ZnPyC60 was significantly quenched both in
neat benzene (f =0.31 and 5.3”10À3 for ZnPy and ZnPyC60,
f
respectively) and in benzene containing 1% pyridine. Slight
broadening of the fluorescence spectra is noted in pure ben-
zene (Figure 4a), which may be due to partial assembly of the
molecules, which is expected, even in such a dilute solution
(see Figure S4 in the Supporting Information), and will be com-
pletely disrupted by the addition of 1% pyridine (Figure 4b).
Fluorescence quenching in ZnPyC60 is ascribed to intramolecu-
lar electron transfer from the photoexcited zinc chlorin moiety
to the fullerene moiety, as proven by the characteristic band of
the fullerene anion radical in transient absorption spectra (see
below). The normalized fluorescence spectra of ZnPyC60 are in
good agreement with those of ZnPy, which indicates that the
weak fluorescence of ZnPyC60 originates from the zinc chlorin
Steady-state spectroscopy
The absorption spectra of the zinc complexes ZnPy and
ZnPyC60 in benzene (ꢀ2 mm) with and without 1% pyridine
are shown in Figure 3. We expected that the chlorophyll mole-
cules would exist mainly as monomeric species at this low con-
centration (see Figure S4 in the Supporting Information); this
was further ensured by the addition of pyridine.[8c,12b] Complex
ZnPy had spectra typical of a zinc chlorophyll with five bands:
the Soret band at lꢀ430 nm and four Q bands over the range
of l=530–680 nm, of which the Qy(0,0) band was the most in-
tense. The spectra of ZnPyC60 was composed of absorptions
by the zinc chlorophyll and fullerene moieties. The five bands
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1169
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 4. Fluorescence spectra of ZnPy and ZnPyC60 (ꢀ2 mm) in a) neat ben-
zene and b) benzene containing 1% pyridine. Normalized spectra are shown
as insets.
Figure 5. Fluorescence of mixtures of ZnPy and ZnPyC60 (1.0 mm) in ben-
zene measured in the reflection configuration: a) spectra, and b) intensities
against the fraction of ZnPyC60
.
moiety (Figure 4, inset) and excludes the possibility of exciplex
emission.[27]
we expected to disrupt coordination-based assemblies. If we
make reasonable assumptions that 1) the zinc chlorophylls ex-
clusively form tetramers and 2) incorporation of ZnPy and
ZnPyC60 into the tetramer is statistical, which is consistent with
the similar values of Kc,4, the fraction of tetramers consisting
only of ZnPy will be (1Àc)4. Very good agreements between
the fluorescence intensities and the (1Àc)4 curve was found
(Figure 5b). It is noted that some of the data points lie lower
than the fourth-order curve. Part of the reason may be dynam-
ic quenching, as discussed above, because the fluorescent in-
tensities were not recovered completely by the addition of 1%
pyridine, which we expect to disrupt the assembly (Figure S6
in the Supporting Information, plots in black), although the in-
volvement of pentamers cannot be ruled out, as mentioned
earlier. Nevertheless, the good agreement indicates that fluo-
rescence from the tetramers, including at least one molecule
of ZnPyC60, is nearly completely quenched and only assemblies
consisting only of ZnPy are emissive. This, in turn, strongly
suggests that near-quantitative energy transfer within the as-
sembly occurs followed by near-quantitative electron transfer
to the fullerene moiety in the assemblies containing at least
one ZnPyC60 molecule. More detailed analyses have been per-
formed with time-resolved fluorescence and absorption spec-
troscopy, the results of which are presented and discussed in
the next section.
Fluorescence spectra for mixtures of ZnPy and ZnPyC60 are
shown in Figure 5a. For these measurements, the fraction of
ZnPy and ZnPyC60 was varied, with total concentrations con-
stant at 1.0 mm. The fluorescence intensity is plotted against
the fraction of ZnPyC60 (c) in the binary mixture in Figure 5b.
The fluorescence decreased as the ratio of ZnPyC60 increased
because the fluorescence of ZnPyC60 was negligibly smaller
than that of ZnPy, as described above. If the molecules of
ZnPy and ZnPyC60 emitted fluorescence independently with-
out interaction, the fluorescence intensity from the mixture
would have been linearly dependent on c, as indicated by the
dotted straight line on the graph in Figure 5b. The observed
intensities were lower than the straight line and showed an
upward curvature. This means that the fluorescence from ZnPy
is quenched by the action of ZnPyC60. Dynamic quenching of
the fluorescence of ZnPy by the fullerene unit attached to
ZnPyC60 should be only a minor cause of quenching because
the fluorescence of a model zinc chlorophyll without a coordi-
nation site, M-1, is quenched by 1 mm of model fullerene com-
pound 7 by at most 8%, as shown in Figure S5 in the Support-
ing Information. Furthermore, the fluorescence intensities were
largely, but not completely, recovered upon the addition of
1% pyridine (Figure S6 in the Supporting Information), which
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1170
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Time-resolved spectroscopy
We performed measurements of nanosecond fluorescence
decay (a light-emitting diode, lex =630 nm, lem =680 nm), fem-
tosecond transient absorption (lex =393 or 441 nm), and nano-
second transient absorption (lex =441 nm). The internal con-
version from the second singlet excited state of the chlorin
unit, which is formed upon excitation at the Soret band in the
transient absorption measurements, to the first singlet excited
state is expected to be faster than the time resolution of our
instruments. Therefore, the initially observed species upon ex-
citation is the first singlet excited state of the chlorin unit of
the zinc chlorophyll compounds. We discuss the results of
time-resolved measurements for ZnPy, ZnPyC60, and a 3:1 mix-
ture of ZnPy and ZnPyC60 in this order. For these samples,
1.0 mm solutions in benzene were used because the tetrameric
species was expected to be predominant at this concentration,
as described earlier.
The fluorescence from ZnPy exhibited a single exponential
decay with a lifetime of t0 =4.4 ns (k0 =2.3”108 sÀ1), as shown
in Figure 6, which was consistent with the typical lifetime of
Figure 7. Transient absorption of ZnPy (1.0 mm) in benzene: a) spectra, and
b) decay profiles at l=850 nm with different laser intensities (8, 4, and
2 mJpulseÀ1).
curve approached a single exponential with a lifetime of
4.6 ns, which agreed with the lifetime obtained by fluores-
cence decay. The laser-power-dependent decay of the transient
absorption is a manifestation of singlet–singlet annihilation,
which provides evidence for the occurrence of energy transfer
within the cyclic tetramer of ZnPy.[8c]
We then set out to estimate Fçrster-type energy-transfer
rates for comparison with the experimental results. To do so,
we needed structural information on the tetramer. There are
six possible configurations of tetramers, depending on the face
of the chlorophyll plane to which the pyridyl group coordi-
nates, anti (a face) or syn (b face) to the propionate residue.
We estimated the likely structures of the tetramer in all config-
urations with DFT calculations (B3LYP/3-21G). The center-to-
center distances, r, were (13.1Æ0.2) (Æstandard deviation) and
(18.2Æ0.9) Š and the orientation factors, k2, were 0.05Æ0.04
and 0.4Æ0.3 for the nearest-neighbor pairs and the diagonal
pairs, respectively. The orientation factors, k2, were calculated
by approximating the transition dipole moment as lying in the
direction of the diagonal line connecting the nitrogen in pyrro-
Figure 6. Fluorescence decay profile of ZnPy in benzene (1.0 mm). The semi-
log plot is shown in the inset. IRF: instrument response function.
zinc chlorophyll derivatives.[28] The femtosecond transient ab-
sorption spectra for the same system are shown in Figure 7a.
A broad absorption was observed after the excitation peaking
at lꢀ860 nm, in addition to the bleaching at l=650–700 nm
due to the disappearance of the Qy(0,0) band, which we as-
signed as absorption by the singlet excited state of the zinc
chlorin unit. With 1 mm chlorophyll derivatives, which we used
to ensure the integrity of the assembly, the large absorption
prevented us from using the Qy(0,0) band region for analysis.
Thus, we used the near-IR region for analysis of the transient
absorption. We found that the absorption decay was biexpo-
nential with relative amplitudes that depended on the excita-
tion laser power, as shown in Figure 7b. The fraction of the
shorter lifetime component of about 2 ps decreased with de-
creasing excitation laser power (41 (1.9 ps), 22 (2.2 ps), and 0%
at powers of 8, 4, and 2 mJpulseÀ1, respectively) and the decay
le A, containing the pyridylethynyl group, and the nitrogen in
Fçrster
EnT
pyrrole C.[8c] The Fçrster-type energy-transfer rate (k
) is
given by Equation (4).
Z
9ðln 19Þk2ꢀf
Fçrster
EnT
ð4Þ
k
¼
fðlÞeðlÞl4dl
128p5n4NAtr6
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1171
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
in which n, NA, t, f, f, e, and l are the refractive index of the
f
solvent, the Avogadro constant, the lifetime of ZnPy (4.4 ns),
the fluorescence quantum yield of the donor (ZnPy, 0.31), the
fluorescence intensity of ZnPy in terms of the number of pho-
tons emitted, the extinction coefficient of the acceptor (also
Fçrster
EnT
ZnPy), and the wavelength, respectively.[29] The values of k
obtained for the calculated tetramer structures were 0.9Æ0.7”
1011 and 0.8Æ0.6”1011 sÀ1 for the nearest-neighbor pairs and
the diagonal pairs, respectively. Comparable values for both
pathways are a coincidental result of compromise between the
orientation factors and the intermolecular distances; chlorin
units in a nearest-neighbor pair are closer in position, but in
less favorable mutual orientations, whereas chlorin units in a di-
agonal pair are farther apart, but in more favorable mutual ori-
entations. Although the calculated rates vary widely due to the
large variation in the orientation factor, depending on the
varied configurations, we may expect the average energy-
transfer rate to be in the order of 1011 sÀ1
.
To relate the observed rates in the femtosecond transient
absorption with the energy-transfer rates, we made a kinetic
model in which energy transfer occurred between any pair in
the cyclic assembly with the same rate, whether nearest-neigh-
bor or diagonal, for simplicity; see the Supporting Information.
According to the model, the observed larger and smaller rate
constants in the transient absorption correspond to 2kEnT and
k0, respectively. Thus, the rate of energy transfer, kEnT, of 2.6”
1011 sÀ1 (tEnT =3.8 ps) was obtained, which was of the same
order of magnitude as those calculated with Equation (4). This,
in turn, implies that the Dexter-type energy transfer does not
play a significant role, which is reasonable when considering
the small overlap of p systems on the chlorin units in the as-
sembly.
The nanosecond transient absorption for ZnPy microsec-
onds after excitation showed typical spectra for the triplet
chlorophyll derivatives (Figure S8a in the Supporting Informa-
tion), which decayed with a rate constant of 3.8”103 sÀ1
(260 ms; Figure S8c in the Supporting Information).
Figure 8. Transient absorption of ZnPyC60 (1.0 mm) in benzene: a) spectra,
b) decay profile at l=850 nm, and c) decay profile at l=1000 nm.
The femtosecond transient absorption measurements de-
scribed hereafter were conducted by excitation at about
2 mJpulseÀ1 to avoid the occurrence of the singlet–singlet anni-
hilation process. The femtosecond transient absorption spectra
for ZnPyC60 displayed in Figure 8a showed a clear band at l=
1000 nm as early as 1 ps after excitation, which was character-
transfer (charge recombination) from the charge-separated
state ZnPy +–C60 in ZnPyC60. The rate of charge separation,
kET, is so fast, relative to k0, that the charge-separation efficien-
cy is practically 100% in ZnPyC60.
C
CÀ
CÀ
istic of the radical anion of fullerene, C60 , accompanied by
The nanosecond transient absorption for ZnPyC60 2 ms after
excitation showed a typical spectrum for triplet chlorophyll de-
rivatives and no absorption due to triplet fullerene, which
would show a strong band at l=750 nm, was observed (Fig-
ure S8b in the Supporting Information). The triplet chlorin de-
cayed biexponentially with rate constants of 4.3”103 (233 ms)
and 4.3”104 sÀ1 (23 ms), as shown in Figure S8d in the Support-
ing Information. The faster triplet decay of ZnPyC60 compared
with that of ZnPy may be ascribed to the contribution of the
decay from the triplet excited state of the fullerene moiety,
which is slightly higher in energy, but has a shorter lifetime.[31]
Finally, we conducted time-resolved measurements on a 3:1
mixture of ZnPy and ZnPyC60. In this mixture, the major con-
stituents are the cyclic tetramers with compositions of
ground-state bleaching at lꢀ680 nm due to the chlorin unit
in ZnPyC60. These spectral features provide evidence of elec-
tron transfer from the singlet excited state of the chlorin unit
to the fullerene unit.[30] The decay at l=850 nm for this
system was biexponential and consisted of a very short com-
ponent of 2.6”1011 sÀ1 and a longer component of 1.9”109 sÀ1
(Figure 8b). The former component can be assigned to fast
charge separation from the singlet excited state of the chlorin
unit, which is consistent with the early formation of the fuller-
ene radical anion. After this fast process, the transient absorp-
tion spectrum decayed homogenously in the range of 800–
1100 nm, including the band at 1000 nm, with kBET =1.9”
109 sÀ1 (Figure 8c), which was attributed to back electron
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1172
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(ZnPy)3(ZnPyC60) and (ZnPy)4, which account for 42 and 32%,
respectively, assuming statistical assembly. The statistical com-
positions of the cyclic tetramers in the mixed systems are
shown in Scheme 1 and Figure S9 in the Supporting Informa-
tion. The fluorescence decay for this mixture is shown in
Figure 9. Apart from the minor, fast component (9%, 1.5”
Figure 9. Fluorescence decay profile of a 3:1 mixture of ZnPy and ZnPyC60
(total 1.0 mm). The semilog plot is shown in the inset.
109 sÀ1) close to the resolution limit, the major component
was a single exponential (3.1”108 sÀ1, 3.3 ns), which we as-
signed as k0 (or t0), that arose from (ZnPy)4 in the mixture.
The expected energy-transfer and charge-separation rates in
(ZnPy)3(ZnPyC60) were too fast for this analysis, and were ad-
dressed by the femtosecond transient absorption. Figure 10a
shows the femtosecond transient absorption for a 3:1 mixture
of ZnPy and ZnPyC60, which exhibits a band at l=1000 nm
that indicates the formation of C60·À. The decay at l=850 nm
shown in Figure 10b and c could be fitted with three exponen-
tials: 2.1”1011, 1.2”109, and 2.3”108 sÀ1. The last value was
fixed to be equal to k0 obtained above, which was guaranteed
by the presence of (ZnPy)4 in this mixture. The first component
may be assigned to coupled energy- and electron-transfer pro-
cesses (a composite of kEnT and kET), whereas the middle com-
ponent may be assigned to back electron transfer (kBET). See
the Supporting Information for the kinetic models on which
these assignments are based.
Figure 10. Transient absorption of a 3:1 mixture of ZnPy and ZnPyC60 (total
1.0 mm): a) spectra, and b) decay profile at l=850 nm.
bly of a pyridyl-functionalized chlorophyll/porphyrin has been
used for a coupled antenna–charge-separation system. It is
also noteworthy that charge separation was observed in non-
polar benzene.[9a,13b]
We have thus demonstrated the successful implementation
of a coupled antenna–charge-separation system that employed
a self-assembly strategy. The rate of energy transfer, kEnT, be-
tween the chlorin units in the assembly and the rate of elec-
tron transfer, kET, from the chlorin unit to the fullerene unit are
both three orders of magnitude faster than the unimolecular
decay of the singlet excited state of the chlorin, k0, the overall
efficiency of charge separation, fCS, is essentially 100% in the
assembly (ZnPy)3(ZnPyC60), whether or not the initially excited
chlorin unit is ZnPy or ZnPyC60. This also explains the steady-
state, near-quantitative quenching results displayed in
Figure 5. To the best of our knowledge, this is the first chloro-
phyll-based coupled system in which clear evidence of charge
separation is obtained. Also, this is the first time that coassem-
Energetics of the photodriven processes
The Gibbs energy change (DGCS) for charge separation from
the singlet excited state of the chlorin unit to the charge-sepa-
rated state in ZnPyC60 was estimated by using the Rehm–
Weller equation [Eq. (5)]:
e2
Benzene
Ox
Benzene
Red
ð5Þ
DGes ¼ eðE
À E
Þ À E00 À
4pe0eBr enzeneRDA
Benzene
Ox
Benzene
Red
in which E
and E
are the first oxidation and reduc-
tion potentials of the chlorin unit and the fullerene unit in ben-
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1173 ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
zene, respectively; E00 is the 0–0 transition energy of the chlor-
in unit in the S0!S1 transition (1.82 eV); e0 is the electric per-
mittivity of vacuum; eBr enzene is the dielectric constants of ben-
zene (2.27); and RDA is the distance between the zinc atom in
the chlorin ring and the center of the fullerene sphere.[12b] The
transition energy E00 was determined by averaging the maxi-
mum wavelengths of the absorption and fluorescence spectra
(1.82 eV). Because the chlorin and fullerene units are linked by
a flexible spacer in ZnPyC60, the distance RDA may take various
values. Hence, stretched and closed conformations were each
optimized at the MM+ level (Figure 11). The obtained distan-
chloromethane; r+ and rÀ are the effective ionic radii of the
chlorin and fullerene units, respectively; and eCr H Cl is the die-
2
2
lectric constant of dichloromethane (8.93). The r+ value was
CH2Cl2
Ox
estimated by a DFT calculation for M-4 (6.6 Š), whereas E
CH2Cl2
and E
for ZnPyC60 were determined by differential pulse
Red
voltammetry (DPV) measurements for model components—
zinc complex M-4 and fullerene derivative M-3—to be 0.71
and À0.67 V versus a standard calomel electrode (SCE), respec-
tively (Figure S10 in the Supporting Information), and the
Benzene
Ox
Benzene
Red
value of e(E
ÀE
) was thus estimated to be À0.92 eV.
The values of DGCS were estimated to be in the range from
À0.37 to 0.09 eV. Given the fast electron-transfer rate (kET =
1.9”1011 À1), DGCS must be appreciably negative.[33] The nega-
s
tive value for DGCS is obtained if the distance RDA is less than
13.9 Š, which suggests that the molecule of ZnPyC60 is not
fully stretched. This is consistent with the slight redshift and
broadening of the absorption spectra described earlier, and
suggests the presence of a weak interaction between the
chlorin and fullerene units in ZnPyC60. The lower bound for
the energy of the charge-separated state of in ZnPyC60 is
1.45 eV (E00 +DGCS), corresponding to the closed conformation,
which is slightly higher than the energy of the triplet excited
chlorin unit (1.42 eV) estimated from the phosphorescence of
ZnPy (Figure S11 in the Supporting Information). Thus, the
final state before going back to the ground state is the triplet
state of the chlorin unit, as indicated by the triplet–triplet ab-
sorption of the chlorin unit in ZnPyC60, which was clearly ob-
served in the nanosecond transient absorption spectra (Fig-
ure S8b in the Supporting Information).
Conclusion
The relevant energy levels and photodriven processes are sum-
marized in Figure 12. In the 3:1 tetrameric assembly of ZnPy
and ZnPyC60, visible-light excitation leads to the initial forma-
tion of the singlet excited state of a chlorin unit. The excited
chlorin quickly transfers its energy to one of the other chlorin
units (kEnT). When the chlorin in ZnPyC60 is in the excited state,
whether by energy transfer or direct excitation, there are two
possibilities: one is further energy transfer to yet another chlor-
Figure 11. Optimized structures of ZnPyC60: a) stretched and b) closed struc-
tures. The dodecyl chains were omitted for clarity.
+
C
in unit and the other is charge separation to form ZnPy
CÀ
ÀC60 (kET). Because energy transfer can occur many times
almost without decaying back to the ground or the triplet
state (k0), practically all excited chlorin ultimately results in
charge separation. Then back electron transfer occurs (kBET), re-
sulting in the formation of the chlorin triplet, which decays
back to the ground state in tens to hundreds of microseconds.
We have thus achieved the first coupled light-harvesting
and charge-separation system with coordination-directed self-
assembled chlorophyll derivatives. The lower oxidation poten-
tial of the zinc chlorophyll derivative, compared with those of
porphyrin derivatives, stabilizes the charge-separated state,
which enables charge separation in nonpolar environments
that barely stabilize ionic species, although the exact determi-
nation of the energy of the charge-separated state is difficult
due to the flexibility of the linker. Each of the rates of elemen-
tary steps of photodriven processes (deactivation, energy
ces of 17.6 and 7.7 Š for the stretched and closed structures,
respectively, may represent the upper and lower limits for
a possible range of distances.[32] Because determination of the
redox potentials in nonpolar benzene was difficult, they were
estimated by using the values measured in dichloromethane
with Equation (6).
CH2Cl2
CH2Cl
Benzene
Ox
Benzene
Red
eðE
À E
Þ ¼ðEOx À ERed Þþ
ꢀ
ꢁꢀ
ꢁ
e2
1
1
1
1
ð6Þ
þ
À
8pe0 rþ rÀ
eBr enzene
eCr H Cl
2
2
CH2Cl2
Red
CH2Cl2
Ox
in which E
and E
are the first oxidation and reduction
potentials of the chlorin and fullerene units, respectively, in di-
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1174
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 12. Coupled antenna–charge-separation system. a) Energy diagram for the photodriven processes within the cyclic tetramers. b) Kinetic scheme for the
cyclic tetramer of (ZnPy)3(ZnPyC60).
[5] a) G. Kodis, P. A. Liddell, L. de La Garza, P. C. Clausen, J. S. Lindsey, A. L.
transfer, electron transfer, and back electron transfer in the sin-
Moore, T. A. Moore, D. Gust, J. Phys. Chem. A 2002, 106, 2036–2048;
glet manifold as well as triplet decay) has been derived with
b) G. Kodis, Y. Terazono, P. A. Liddell, J. AndrØasson, V. Garg, M. Ham-
consistency among three types of different assemblies (ZnPy,
bourger, T. A. Moore, A. L. Moore, D. Gust, J. Am. Chem. Soc. 2006, 128,
ZnPyC60, and the 3:1 mixture thereof). Because both energy-
and electron-transfer processes are three orders of magnitude
faster than unimolecular decay, a coupled light-harvesting and
charge-separation system with practically 100% efficiency has
been realized. Furthermore, back electron transfer is two
orders of magnitude slower than the charge-separation pro-
cess.
1818–1827; c) V. Garg, G. Kodis, P. A. Liddell, Y. Terazono, T. A. Moore,
A. L. Moore, D. Gust, J. Phys. Chem. B 2013, 117, 11299–11308; d) M. E.
El-Khouly, D. K. Ju, K.-Y. Kay, F. D’Souza, S. Fukuzumi, Chem. Eur. J. 2010,
16, 6193–6202; e) F. D’Souza, S. Gadde, D. M. S. Islam, C. A. Wijesinghe,
A. L. Schumacher, M. E. Zandler, Y. Araki, O. Ito, J. Phys. Chem. A 2007,
111, 8552–8560; f) J.-Y. Liu, M. E. El-Khouly, S. Fukuzumi, D. K. P. Ng,
Chem. Eur. J. 2011, 17, 1605–1613; g) E. Maligaspe, T. Kumpulainen,
N. K. Subbaiyan, M. E. Zandler, H. Lemmetyinen, N. V. Tkachenko, F.
D’Souza, Phys. Chem. Chem. Phys. 2010, 12, 7434–7444.
[6] Y. Kuramochi, A. Satake, M. Itou, K. Ogawa, Y. Araki, O. Ito, Y. Kobuke,
Chem. Eur. J. 2008, 14, 2827–2841.
[7] a) K. Ohkubo, S. Fukuzumi, Bull. Chem. Soc. Jpn. 2009, 82, 303–315;
b) K. Stranius, V. Iashin, T. Nikkonen, M. Muuronen, J. Helaja, N. Tkachen-
ko, J. Phys. Chem. A 2014, 118, 1420–1429; c) J. S. Kavakka, S. Heikkinen,
I. Kilpeläinen, N. V. Tkachenkob, J. Helaja, Chem. Commun. 2009, 758–
760.
Acknowledgements
This work was partly supported by the JSPS KAKENHI (grant
number 24550163 to J.O. and grant numbers 26620154 and
26288037 to K.O.) and the MEXT Strategic Research Foundation
at Private Universities. Y.S. thanks the JSPS for a Research Fel-
lowship for Young Scientists (PD).
[8] a) H. Tamiaki, A. R. Holzwarth, K. Schaffner, J. Photochem. Photobiol. B
1992, 15, 355–360; b) C. Rçger, Y. Miloslavina, D. Brunner, A. R. Holz-
warth, F. Würthner, J. Am. Chem. Soc. 2008, 130, 5929–5939; c) R. F.
Kelley, R. H. Goldsmith, M. R. Wasielewski, J. Am. Chem. Soc. 2007, 129,
6384–6385; d) Y. Shinozaki, G. Richards, K. Ogawa, A. Yamano, K. Ohara,
K. Yamaguchi, S.-i. Kawano, K. Tanaka, Y. Araki, T. Wada, J. Otsuki, J. Am.
Chem. Soc. 2013, 135, 5262–5265; e) V. I. Prokhorenko, A. R. Holzwarth,
M. G. Müller, K. Schaffner, T. Miyatake, H. Tamiaki, J. Phys. Chem. B 2002,
106, 5761–5768.
Keywords: chlorophylls · electron transfer · energy transfer ·
fullerenes · self-assembly
[9] a) N. V. Tkachenko, A. Y. Tauber, D. Grandell, P. H. Hynninen, H. Lemme-
tyinen, J. Phys. Chem. A 1999, 103, 3646–3656; b) N. V. Tkachenko, L.
Rantala, A. Y. Tauber, J. Helaja, P. H. Hynninen, H. Lemmetyinen, J. Am.
Chem. Soc. 1999, 121, 9378–9387; c) K. Ohkubo, H. Kotani, J. Shao, Z.
Ou, K. M. Kadish, G. Li, R. K. Pandey, M. Fujitsuka, O. Ito, H. Imahori, S.
Fukuzumi, Angew. Chem. Int. Ed. 2004, 43, 853–856; Angew. Chem.
2004, 116, 871–874; d) Y. Kawashima, K. Ohkubo, M. Kentaro, S. Fukuzu-
mi, J. Phys. Chem. C 2013, 117, 21166–21177; e) T. Nikkonen, M. M.
Oliva, A. Kahnt, M. Muuronen, J. Helaja, D. M. Guldi, Chem. Eur. J. 2015,
21, 590–600.
[1] a) G. D. Scholes, G. R. Fleming, A. Olaya-Castro, R. van Grondelle, Nat.
Chem. 2011, 3, 763–774; b) R. Croce, H. van Amerongen, Nat. Chem.
Biol. 2014, 10, 492–501.
[2] a) J. Dekker, R. Van Grondelle, Photosynth. Res. 2000, 63, 195–208; b) A.
Warshel, S. Creighton, W. W. Parson, J. Phys. Chem. 1988, 92, 2696–
2701.
[3] a) M. R. Wasielewski, Chem. Rev. 1992, 92, 435–461; b) R. Ziessel, A. Har-
riman, Chem. Commun. 2011, 47, 611 –631; c) F. Odobel, Y. Pellegrin, J.
Warnan, Energy Environ. Sci. 2013, 6, 2041–2052; d) P. D. Frischmann, K.
Mahata, F. Würthner, Chem. Soc. Rev. 2013, 42, 1847–1870.
[4] a) S. Fukuzumi, K. Ohkubo, J. Mater. Chem. 2012, 22, 4575–4587; b) N.
Aratani, D. Kim, A. Osuka, Acc. Chem. Res. 2009, 42, 1922–1934; c) D.
Gust, T. A. Moore, A. L. Moore, Acc. Chem. Res. 2009, 42, 1890–1898;
d) Y. Kobuke, Eur. J. Inorg. Chem. 2006, 2333–2351; e) H. Imahori, Org.
Biomol. Chem. 2004, 2, 1425–1433.
[10] M. Katterle, V. I. Prokhorenko, A. R. Holzwarth, A. Jesorka, Chem. Phys.
Lett. 2007, 447, 284–288.
[11] a) K. Funatsu, T. Imamura, A. Ichimura, Y. Sasaki, Inorg. Chem. 1998, 37,
1798–1804; b) R. A. Jensen, R. F. Kelley, S. J. Lee, M. R. Wasielewski, J. T.
Hupp, D. M. Tiede, Chem. Commun. 2008, 1886–1888; c) A. Tsuda, S. Sa-
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1175
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
kamoto, K. Yamaguchu, T. Aida, J. Am. Chem. Soc. 2003, 125, 15722–
15723.
[26] H. Tamiaki, S. Yagai, T. Miyatake, Bioorg. Med. Chem. 1998, 6, 2171–
2178.
[12] a) V. L. Gunderson, M. R. Wasielewski, Chem. Phys. Lett. 2013, 556, 303–
307; b) V. L. Gunderson, A. L. Smeigh, C. H. Kim, D. T. Co, M. R. Wasielew-
ski, J. Am. Chem. Soc. 2012, 134, 4363–4372.
[13] a) M. C. Hutter, J. M. Hughes, J. R. Reimers, N. S. Hush, J. Phys. Chem. B
1999, 103, 4906–4915; b) S. Fukuzumi, K. Ohkubo, W. E, Z. Ou, J. Shao,
K. M. Kadish, J. A. Hutchison, K. P. Ghiggino, P. J. Sintic, M. J. Crossley, J.
Am. Chem. Soc. 2003, 125, 14984–14985.
[14] C. Yang, J. Y. Kim, S. Cho, J. K. Lee, A. J. Heeger, F. Wudl, J. Am. Chem.
Soc. 2008, 130, 6444–6450.
[15] J. C. Hummelen, B. W. Knight, F. LePeq, F. Wudl, J. Org. Chem. 1995, 60,
532–538.
[27] a) V. Vehmanen, N. V. Tkachenko, A. Y. Tauber, P. H. Hynninen, H. Lemme-
tyinen, Chem. Phys. Lett. 2001, 345, 213–218; b) V. Vehmanen, N. V. Tka-
chenko, A. Efimov, P. Damlin, A. Ivaska, H. Lemmetyinen, J. Phys. Chem.
A 2002, 106, 8029–8038.
[28] A. R. Holzwarth, M. Katterle, M. G. Müller, Y.-Z. Ma, V. Prokhorenko, Pure
Appl. Chem. 2001, 73, 469–474.
[29] J. Otsuki, J. Porphyrins Phthalocyanines 2009, 13, 1069–1081.
[30] a) H. Imahori, D. M. Guldi, K. Tamaki, Y. Yoshida, C. Luo, Y. Sakata, S. Fu-
kuzumi, J. Am. Chem. Soc. 2001, 123, 6617–6628; b) M. Fujitsuka, N. Tsu-
boya, R. Hamasaki, M. Ito, S. Onodera, O. Ito, Y. Yamamoto, J. Phys.
Chem. A 2003, 107, 1452–1458.
[16] W.-B. Zhang, J. He, X. Dong, C.-L. Wang, H. Li, F. Teng, X. Li, C. Wesde-
miotis, R. P. Quirk, S. Z. D. Cheng, Polymer 2011, 52, 4221–4226.
[17] K. M. Smith, D. A. Goff, D. J. Simpson, J. Am. Chem. Soc. 1985, 107,
4946–4954.
[18] H. Tamiaki, S. Miyata, Y. Kureishi, R. Tanikag, Tetrahedron 1996, 52,
12421–12432.
[19] a) S.-i. Sasaki, K. Mizutani, M. Kunieda, H. Tamiaki, Tetrahedron Lett. 2008,
49, 4113–4115; b) S.-i. Sasaki, K. Mizutani, M. Kunieda, H. Tamiaki, Tetra-
hedron 2011, 67, 6065–6072.
[20] Y. Yamamoto, H. Tamiaki, Tetrahedron 2014, 70, 2731–2737.
[21] S.-i. Sasaki, H. Tamiaki, Tetrahedron Lett. 2006, 47, 4965–4968.
[22] a) A. Tsuda, T. Nakamura, S. Sakamoto, K. Yamaguchi, A. Osuka, Angew.
Chem. Int. Ed. 2002, 41, 2817–2821; Angew. Chem. 2002, 114, 2941–
2945; b) M. Vinodu, Z. Stein, I. Goldberg, Inorg. Chem. 2004, 43, 7582–
7584.
[31] S. Fukuzumi, K. Ohkubo, H. Imahori, J. Shao, Z. Ou, G. Zheng, Y. Chen,
R. K. Pandey, M. Fujitsuka, O. Ito, K. M. Kadish, J. Am. Chem. Soc. 2001,
123, 10676–10683.
[32] a) D. I. Schuster, P. Cheng, P. D. Jarowski, D. M. Guldi, C. Luo, L. Echegoy-
en, S. Pyo, A. R. Holzwarth, S. E. Braslavsky, R. M. Williams, G. Klihm, J.
Am. Chem. Soc. 2004, 126, 7257–7270; b) M. E. El-Khouly, Y. Araki, O. Ito,
S. Gadde, A. L. McCarty, P. A. Karr, M. E. Zandler, F. D’Souza, Phys. Chem.
Chem. Phys. 2005, 7, 3163–3171; c) H. Imahori, K. Hagiwara, M. Aoki, T.
Akiyama, S. Taniguchi, T. Okada, M. Shirakawa, Y. Sakata, J. Am. Chem.
Soc. 1996, 118, 11771–11782; d) T. D. M. Bell, K. P. Ghiggino, K. A. Jolliffe,
M. G. Ranasinghe, S. J. Langford, M. J. Shephard, M. N. Paddon-Row, J.
Phys. Chem. A 2002, 106, 10079–10088; e) F. D’Souza, R. Chitta, S.
Gadde, D. M. Shafiqul Islam, A. L. Schumacher, M. E. Zandler, Y. Araki, O.
Ito, J. Phys. Chem. B 2006, 110, 25240–25250.
[33] J.-H. Ha, H. S. Cho, D. Kim, J.-C. Lee, T.-Y. Kim, Y. K. Shim, ChemPhysChem
2003, 4, 951–958.
[23] Y. Shinozaki, I. Yoshikawa, K. Araki, K. Ohara, K. Yamaguchi, S.-i. Kawano,
K. Tanaka, Y. Araki, T. Wada, J. Otsuki, Chem. Lett. 2014, 43, 862–864.
[24] a) A. I. Oliva, K. Gómez, G. Gonzµleza, P. Ballester, New J. Chem. 2008, 32,
2159–2163; b) A. Pastor, E. Martínez-Viviente, Coord. Chem. Rev. 2008,
252, 2314–1345.
Received: September 21, 2015
[25] G. Ercolani, J. Phys. Chem. B 1998, 102, 5699–5703.
Published online on November 30, 2015
Chem. Eur. J. 2016, 22, 1165 – 1176
www.chemeurj.org
1176
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim