Light-harvesting supramolecular porphyrin macrocycle accommodating a fullerene-tripodal ligand

Article abstract of DOI:10.1002/chem.200701720

Trisporphyrinatozinc(II) (1-Zn) with imidazolyl groups at both ends of the porphyrin self-assembles exclusively into a light-harvesting cyclic trimer (N-(1-Zn)₃) through complementary coordination of imidazolyl to zinc(II). Because only the two terminal porphyrins in 1-Zn are employed in ring formation, macrocycle N-(1-Zn)₃ leaves three uncoordinated porphyrinatozinc(II) groups as a scaffold that can accommodate ligands into the central pore. A pyridyl tripodal ligand with an appended fullerene connected through an amide linkage (C₆₀-Tripod) was synthesized by coupling tripodal ligand 3 with pyrrolidine-modified fullerene, and this ligand was incorporated into N-(1-Zn)₃. The binding constant for C ₆₀-Tripod in benzonitrile reached the order of 10⁸M ^-1. This value is ten times larger than those of pyridyl tetrapodal ligand 2 and tripodal ligand 3. This behavior suggests that the fullerene moiety contributes to enhance the binding of C₆₀-Tripod in N-(1-Zn) ₃. The fluorescence of N-(1-Zn)₃ was almost completely quenched (≈97%) by complexation with C₆₀-Tripod, without any indication of the formation of charge-separated species or a triplet excited state of either porphyrin or fullerene in the transient absorption spectra. These observations are explained by the idea that the fullerene moiety of C ₆₀-Tripod is in direct contact with the porphyrin planes of N-(1-Zn)₃ through fullerene-porphyrin π-π interactions. Thus, C₆₀-Tripod is accommodated in N-(1-Zn)₃ with a π-π interaction and two pyridyl coordinations. The cooperative interaction achieves a sufficiently high affinity for quantitative and specific introduction of one equivalent of tripodal guest into the antenna ring, even under dilute conditions (≈10^-7M) in polar solvents such as benzonitrile. Additionally, complete fluorescence quenching of N-(1-Zn)₃ when accommodating C₆₀-Tripod demonstrates that all of the excitation energy collected by the nine porphyrins migrates rapidly over the macrocycle and then converges efficiently on the fullerene moiety by electron transfer.

Full text of DOI:10.1002/chem.200701720

FULL PAPER
DOI: 10.1002/chem.200701720
Light-Harvesting Supramolecular Porphyrin Macrocycle Accommodating a
Fullerene–Tripodal Ligand
Yusuke Kuramochi,[a] Akiharu Satake,[a] Mitsunari Itou,[b] Kazuya Ogawa,[a]
Yasuyuki Araki,[b] Osamu Ito,[b] and Yoshiaki Kobuke*[a, c]
Abstract: Trisporphyrinatozinc(II) (1–
Zn) with imidazolyl groups at both
ends of the porphyrin self-assembles
exclusively into a light-harvesting cyclic
trimer (N-(1–Zn)3) through comple-
mentary coordination of imidazolyl to
zinc(II). Because only the two terminal
porphyrins in 1–Zn are employed in
ring formation, macrocycle N-(1–Zn)3
leaves three uncoordinated porphyrina-
tozinc(II) groups as a scaffold that can
accommodate ligands into the central
pore. A pyridyl tripodal ligand with an
appended fullerene connected through
an amide linkage (C60–Tripod) was syn-
thesized by coupling tripodal ligand 3
with pyrrolidine-modified fullerene,
and this ligand was incorporated into
N-(1–Zn)3. The binding constant for
C60–Tripod in benzonitrile reached the
order of 108 mꢀ1. This value is ten times
larger than those of pyridyl tetrapodal
ligand 2 and tripodal ligand 3. This be-
havior suggests that the fullerene
moiety contributes to enhance the
binding of C60–Tripod in N-(1–Zn)3.
The fluorescence of N-(1–Zn)3 was
almost completely quenched (ꢁ97%)
by complexation with C60–Tripod, with-
out any indication of the formation of
charge-separated species or a triplet
excited state of either porphyrin or
fullerene in the transient absorption
spectra. These observations are ex-
plained by the idea that the fullerene
moiety of C60–Tripod is in direct con-
tact with the porphyrin planes of N-(1–
Zn)3 through fullerene–porphyrin p–p
interactions. Thus, C60–Tripod is accom-
modated in N-(1–Zn)3 with a p–p inter-
action and two pyridyl coordinations.
The cooperative interaction achieves a
sufficiently high affinity for quantita-
tive and specific introduction of one
equivalent of tripodal guest into the
antenna ring, even under dilute condi-
tions (ꢁ10ꢀ7 m) in polar solvents such
as benzonitrile. Additionally, complete
fluorescence quenching of N-(1–Zn)3
when accommodating C60–Tripod dem-
onstrates that all of the excitation
energy collected by the nine porphyrins
migrates rapidly over the macrocycle
and then converges efficiently on the
fullerene moiety by electron transfer.
Keywords: fullerenes · photochem-
istry · photosynthesis · porphyri-
noids · self-assembly
Introduction
tion center (RC).[1] LH systems in purple bacteria, in partic-
ular, have been extensively studied and fully characterized.
Purple bacteria have two types of cyclic antenna, LH1[2] and
LH2.[3] They operate in biological membranes to transport
the energy captured by LH2 to the RC located in LH1with
Photosynthetic organisms have developed characteristic
light-harvesting antenna (LH) systems to capture weak
energy from sunlight and efficiently transfer it to the reac-
[a] Dr. Y. Kuramochi, Prof. A. Satake, Prof. K. Ogawa, Prof. Y. Kobuke
Graduate School of Materials Science
Nara Institute of Science and Technology
Takayama 8916-5, Ikoma
Uji, Kyoto 611-0011 (Japan)
Fax : (+81)774-38-3508
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author. It contains
MALDI-TOF mass spectral and GPC analyses of C-(1–Zn)3, GPC
analyses of 9 and 10; an HMQC spectrum of N-(1–Zn)3, Q band curve
fitting of N-(1–Zn)3 by N-(5–Zn)2 and 6–Zn, UV/Vis titration spectra
of 2 in benzonitrile, time-resolved spectra (fluorescence decay and
transient absorption) of the complex of N-(1–Zn)3 and C60-Tripod, and
1H and 13C NMR spectra of C60-Tripod.
Nara 630-0101 (Japan)
E-mail: kobuke@iae.kyoto-u.ac.jp
[b] Dr. M. Itou, Prof. Y. Araki, Prof. O. Ito
Institute of Multidisciplinary Research for Advanced Materials
Tohoku University, Katahira
Sendai, 980-8577 (Japan)
[c] Prof. Y. Kobuke
Present address: Institute of Advanced Energy
Kyoto University, Gokasho
Chem. Eur. J. 2008, 14, 2827 – 2841ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2827

high efficiency.[4] Excitation of the special pair initiates
charge separation followed by multistep electron-transfer re-
actions in the RC to give a transmembrane proton gradient
that is coupled with adenosine triphosphate (ATP) produc-
tion through an ATP-synthase enzyme.[1] This is the primary
energy source of all of the living organisms on earth, and ap-
plication of the photosynthetic-energy-conversion system
has great potential to resolve energy and environmental
problems in the world.
A number of researchers have attempted to construct arti-
ficial light-harvesting systems by covalent[5,6] or noncova-
lent[5,7] approaches. Both approaches provide important con-
tributions not only to understanding the natural system, but
also in creating efficient energy-harvesting systems. The
noncovalent self-assembly approach has the merit of using
simple construction units with minimum synthetic effort. It
also allows us to mimic the sophisticated construction princi-
ples of nature.
We previously reported that bis(imidazolylporphyrinato-
zinc) linked through an m-phenylene or m-bis(ethynylene)-
phenylene moiety afforded a macrocyclic hexamer and pen-
tamer by complementary coordination of imidazolyl to
zinc(II) (Figure 1).[11,12,13] These macrocycles mimicked the
structure of LH2 and achieved rapid energy transfer in por-
phyrinic pigments without losing their fluorescence proper-
ties.[12,14] In a previous communication,[15] this supramolec-
ular methodology was applied to 1–Zn, which had imidazol-
yl groups at both porphyrin ends and exclusively gave mac-
rocyclic trimer N-(1–Zn)3 (Figure 1). In this macrocycle, the
two terminal porphyrins in 1–Zn participated in ring-form-
ing complementary coordinations, and the uncoordinated
central porphyrinatozinc(II) provided the scaffold to specifi-
cally accommodate a ligand, such as tetrakis{4-[(4-pyridyl)-
ethynyl]phenyl}methane (2), in the pore of N-(1–Zn)3 by
three-point coordination from pyridyl to the zinc moieties.
In composite 2/N-(1–Zn)3, ligand 2 possesses a fourth arm
Mimicking the RC is also im-
portant for further conversion
of the captured energy into
electrochemical potential. Full-
erene has been widely used for
various synthetic porphyrin
models as an electron acceptor
because of its small reorganiza-
tion energy (l) in photoinduced
electron-transfer reactions.[8,9]
In these studies, the porphyrin
and the fullerene units were
also connected through either
covalent[9] or noncovalent link-
ers.[10] Polar solvents, such as
benzonitrile, are often used in
the study of photoinduced elec-
tron transfer of covalently
linked systems because they
stabilize the charge-separated
species and can maintain
a
long-lived charge-separated
state. However, the use of polar
solvents produces difficulties
for noncovalent linker systems
because coordination and hy-
drogen-bonding interactions are
weakened, which leads to cleav-
age of the conjugates in polar
solvents. To overcome this diffi-
culty, multipoint binding be-
tween the porphyrin and the
fullerene units was used,[10f] but
addition of excess amounts of
ligand was necessary to obtain
significant amounts of the por-
phyrin–fullerene conjugate in
benzonitrile.
Figure 1. Structures of the porphyrin macrocycles. Arrows indicate the uncoordinated porphyrinatozinc sites.
2828
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2827 – 2841

Supramolecular Porphyrin Macrocycle
FULL PAPER
that does not participate in conjugate formation and can be
modified with energy or an electron-acceptor unit. Although
various types of light-harvesting and RC models that use
porphyrins and electron-acceptor molecules have been re-
ported independently, their combined architectures have
been constructed in only a limited number of cases to
date.[16] Herein, we report the synthesis of a fullerene–tripo-
dal ligand (C60–Tripod) by modifying one arm of the tetra-
podal ligand with a fullerene moiety (Figure 2) and incorpo-
rated it into N-(1–Zn)3 with three-point binding of the pyr-
idyl to the zinc moieties. The multipoint binding enabled us
to investigate the photochemical properties of the antenna/
electron acceptor composite, even in benzonitrile. Herein
we also describe detailed structural studies of the free-base
trisporphyrin (1), ring formation behavior of 1–Zn, and the
incorporation of trispyridine derivatives into N-(1–Zn)3.
(based on 4). Because the second condensation did not
afford porphyrin by-products except for tarry materials, the
purification of 1 was easily achieved by means of silica gel
column chromatography.
1
The H NMR spectrum of 1 in CDCl3 showed signals of a
slightly complex nature (Figure 3). The Im-4, Im-5, and N-
Me protons had four nonequivalent signals (at d=7.681,
7.677, 7.65, and 7.64 ppm for Im-4; 7.47, 7.45, 7.44, and
7.41ppm for Im-5; 3.40, 3.37, 3.36, and 3.30 ppm for N-Me,
respectively) with the same integration values. On the other
hand, the Ph-2 protons had six nonequivalent signals be-
tween d=8.97 and 9.13 ppm, in which the two signals indi-
*
cated by
had an integration value twice as large as the
*
four signals indicated by . Because each porphyrin is ori-
ented perpendicularly to the m-phenylene and N-methylimi-
dazolyl moieties to reduce steric hindrance, there is a possi-
bility that 1 has several atropisomers unless free rotation is
allowed. These nonequivalent signals for the imidazolyl and
phenyl moieties suggest the existence of their isomers.
Results
Synthesis of 1: Free base 1 was synthesized by two porphy-
rin condensation reactions based on Lindseyꢁs method
(Scheme 1).[17] The first condensation of 1-methylimidazole-
2-carboxaldehyde, 3-(4,4-dimethyl-2,6-dioxane-1-yl)benzal-
dehyde, and meso-(3-allyloxypropyl)dipyrromethane afford-
ed monomeric porphyrin 5 in a yield of 13%[11c] accompa-
nied by two porphyrin by-products, bis(3-(4,4-dimethyl-2,6-
dioxan-1-yl)phenyl)porphyrin 6 and bis(1-methylimidazol-2-
yl)porphyrin. Porphyrin by-product 6 was also isolated for
use as the reference material. The second condensation of 4,
which was obtained by deprotecting 5, and meso-methoxy-
carbonylethyldipyrromethane[18] gave 1 in a yield of 24%
Synthesis of C60–Tripod: The tripodal ligand C60–Tripod,
which is connected to a fullerene moiety through amide
bonding, was synthesized according to Scheme 2. 4-(Triphe-
nylmethyl)benzoic acid was synthesized according to a pub-
lished procedure.[19] The iodination of 4-(triphenylmethyl)-
benzoic acid was carried out by using bis(trifluoroacet-
A
to give a mixture of 7 and tetraiodo-substituted 7 in a molar
ratio of 1:1. When smaller equivalents of bis(trifluoroacetox-
y)iodobenzene (2.5 equiv) and I2 (2.5 equiv) were used, a
mixture of 7 and the meta-iodo compound (see Scheme S1
in the Supporting Information) was formed in a molar ratio
Figure 2. Structures of tripodal ligands 2, 3, and C60–Tripod.
Chem. Eur. J. 2008, 14, 2827 – 2841ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2829

Y. Kobuke et al.
Scheme 1. Synthetic routes for trisporphyrin 1–Zn, reference dimer N-(5–Zn)2, and monomer 6–Zn.
ture of 7 and tetraiodo-substitut-
ed 7 in a molar ratio of 1:1,
which indicated that the iodina-
tion of meta-iodo-substituted
phenyl occurred selectively. In
this step, the use of an excess
amount of bis(trifluoroacetoxy)-
iodobenzene and I2 is necessary
because the meta-iodo-substi-
tuted compound cannot be re-
moved even after the next cou-
pling reaction. The coupling re-
action of the mixture of 7 and
tetraiodo-substituted 7 with 4-
ethynylpyridine was carried out
in the presence of [Pd(PPh3)2Cl2]
A
and CuI.[20] Tetraiodo-substituted
7 was also coupled with 4-ethy-
nylpyridine to afford the corre-
sponding
tetrapyridyl
com-
pound, which could be easily re-
moved by means of silica gel
column chromatography. The
Figure 3. a) 1H NMR spectrum of 1 in CDCl3 at room temperature. Magnifications of the signals for Ph-2 (b),
Im-4 and Im-5 (c), and Im N-Me (d) are also shown.
desired tripyridyl derivative
3
was isolated in a yield of 57%
based on 7 as the starting mix-
ture. The C60–Tripod ligand was
of 1:1. Further iodination of this mixture with bis(trifluoroa-
cetoxy)iodobenzene (3 equiv) and I2 (3 equiv) gave a mix-
obtained in a yield of 21% by condensation of 3 with pyrro-
lidine–C60[21] by using BOP as the coupling agent.
2830
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2827 – 2841

Supramolecular Porphyrin Macrocycle
FULL PAPER
and larger ring compounds to
converge into a single product
that shows a sharp signal at
12.8 min in the gel permeation
chromatogram
When the reorganization was
performed under higher
(Figure 4b).[15]
a
methanol composition (CHCl3/
MeOH 7:3)[11c] or at lower tem-
perature (ꢁ178C), another as-
sembly appeared at 12.3 min as
a small signal in the gel perme-
ation chromatogram (Figure 4b,
inset). To obtain the single
product appearing at 12.8 min,
careful control of the solvent
composition (methanol and
chloroform) and temperature is
required.
To determine the number of
1–Zn units in the assembly, the
MALDI-TOF mass spectrum of
the assembly was measured.
Even though the assembly was
very stable in noncoordinating
solvents, the assembled struc-
ture did not appear under the
mass spectrometer conditions
and predominantly showed a
signal for monomeric 1–Zn. To
prevent the dissociation of the
assembly under the mass spec-
trometer conditions, multisite
ring-closing metathesis reac-
tions[11c,18] of the meso-olefinic
Scheme 2. Synthetic route of C60–Tripod. BOP: 1H-benzotriazol-1-yloxytris-(dimethylamino)phosphonium hex- groups were carried out by
afluorophosphate.
using first-generation Grubbs
catalyst.[23] The MALDI-TOF
mass spectrum of the covalently
Reorganization of 1–Zn into a macrocycle: Treatment of 1
with Zn(OAc)2 in CHCl3 quantitatively afforded the corre-
linked assembly showed a signal corresponding to C-(1–Zn)3
(calcd for [M+H]: m/z: 5816.5; found: 5816.4), which is the
covalently linked trimer of 1–Zn (Figure S1in the Support-
ing Information). The coincidence between the theoretical
and the observed molecular weights further supports the hy-
pothesis that this assembly has no terminal olefinic groups,
that is, it is the ring structure. The gel permeation chroma-
tography (GPC) trace of C-(1–Zn)3 also gave a sharp signal
at a slightly longer retention time (12.9 min) than that of N-
(1–Zn)3, which reflects the loss of three ethylenic parts by
the metathesis reaction (Figure S2 in the Supporting Infor-
mation).
ACHTREUNG
sponding zinc complex 1–Zn (Scheme 3). During zinc inser-
tion, insoluble material that resulted from a polymeric spe-
cies of 1–Zn was partially generated. This insoluble material
was dissolved in CHCl3, in which it was solubilized by the
addition of a mixture of pyridine and toluene (1:2), and
evaporated. This procedure converted the insoluble polymer
into oligomeric assemblies of smaller molecular weights
(Figure 4a). To convert the linear oligomers and any large
ring compounds into macrocyclic species, the as-prepared
mixture of 1–Zn was subjected to a reorganization proce-
dure in which a solution of the as-prepared mixture was di-
luted to 0.02 mm in CHCl3/MeOH (9:1, v/v) and allowed to
stand at 278C for one day in the dark. The solvent was care-
fully evaporated at 308C under reduced pressure.[22] The
equilibrium of the complementary coordination of imidazol-
yl to zinc(II) under dilute conditions drives linear oligomers
GPC analysis of the macrocycles and noncyclic oligomers:
The GPC retention time gives information about the hydro-
dynamic volume, which reflects the molecular shape. To
obtain further evidence for the ring structure of N-(1–Zn)3,
a series of noncyclic porphyrin arrays was prepared and
Chem. Eur. J. 2008, 14, 2827 – 2841ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2831

Y. Kobuke et al.
Scheme 3. Formation of the self-assembled macrocycle.
signal in Figure 5a corresponds to noncyclic arrays 9 (n=0,
1, 2, 3, and 4). Similarly, arrays 10 (n=0, 1, and 2) were ob-
served separately in Figure 5b. Figure 6 shows logarithmic
*
plots of the molecular weights of the cyclic ( ) and noncy-
~
~
clic ( and ) porphyrin arrays as a function of their reten-
tion times. The authentic cyclic gable pentamer and hexa-
mer[11a,b] were also plotted ( ). A dotted straight line was
*
drawn connecting the open and filled triangles. Two series
of the noncyclic arrays are on the straight line, which indi-
cates that the hydrodynamic volume behaves similarly for
different porphyrin ends. Cyclic arrays (N-(1–Zn)3, its cova-
lently linked derivative C-(1–Zn)3, and the authentic gable
pentamer and hexamer) are located above the line in an
almost linear fashion. This result indicates that the cyclic
arrays have smaller hydrodynamic volumes than the noncy-
clic arrays. The longer retention times observed for N-(1–
Zn)3 and C-(1–Zn)3 compared with the noncyclic arrays (9
and 10) are fully consistent with the general tendency of the
cyclic structures.[6g,24,25]
Figure 4. Gel permeation chromatograms of a) the as-prepared mixture
and b) the reorganized sample of trisporphyrinatozinc(II) 1–Zn treated
with CHCl3/MeOH=9:1and [ 1–Zn]=0.02 mm at 278C. The inset shows
the reorganized sample of 1–Zn treated with CHCl3/MeOH=9:1and [ 1–
Zn]=0.02 mm at 178C.
compared with the cyclic porphyrins by using GPC. Two
types of noncyclic porphyrin arrays were prepared, array 9,
which has dimeric porphyrin units at both ends, and array
10, which has monomeric units at the ends. They were pre-
pared by mixing N-(1–Zn)3 with N-(5–Zn)2 or monoimida-
zolyl gable porphyrin 8[14] in pyridine, followed by solvent
evaporation. Dissolution of the residue in chloroform af-
forded a series of noncyclic porphyrin arrays (Scheme 4).
The GPC charts of 9 and 10 are shown in Figure 5. Each
Structural analysis of the macrocycle by NMR spectroscopy:
The unambiguous proof of the cyclic structure of N-(1–Zn)3
was obtained by NMR spectroscopy analysis. Although the
proton NMR spectrum of noncovalently linked cyclic trimer
N-(1–Zn)3 was complicated, most of the signals were assign-
able with support of 2D NMR measurements (COSY,
TOCSY, and HMQC). COSY NMR spectroscopy showed
that one set of four signals was correlated with another set
of four signals (i.e., between d=4.55–4.82 and 1.69–
2832
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2827 – 2841

Supramolecular Porphyrin Macrocycle
FULL PAPER
Scheme 4. Formation of the noncyclic porphyrin array.
Figure 5. Gel permeation chromatograms of a series of noncyclic porphyrin arrays 9 (a) and 10 (b).
1.89 ppm, Figure 7) and HMQC (Figure S3 in the Support-
ing Information) showed that these signals were correlated
with the signals of imidazolyl carbons (d= ꢁ120 ppm). For
all of the Im-4 and Im-5 protons, one set of four signals con-
rocycle show three sets of trisporphyrins because the three
trisporphyrins are all in different environments. Thus, the
*
*
one big ( ) and three small ( ) signals in Figure 7 can be
assigned to the C3h-symmetric and asymmetric macrocycles,
*
*
*
*
sisted of one big ( ) and three small ( ) signals with an in-
tegration ratio of 2:1, irrespective of “in” and “out” configu-
rations (Figure 7). This signal integral ratio can be explained
by assuming that there is a mixture of two topological iso-
mers (C3h symmetric[26] and asymmetric) in a ratio of 2:3.
For each in and out proton, the signals of the C3h-symmetric
macrocycle exhibit only a single set of trisporphyrin signals
because the three trisporphyrins are in the same environ-
ment. On the other hand, the signals of the asymmetric mac-
respectively. The ratio ( /3 ) corresponds to the formation
ratio of the asymmetric and symmetric macrocycles.
The key signals that prove the cyclic structure are the imi-
dazolyl protons inside and outside the ring. In the reference
dimer, N-(5–Zn)2, the Im-4 protons were observed at
around 2.1ppm because of the shielding effect of the coor-
dinating porphyrin. For N-(1–Zn)3, the Im-4 protons ap-
peared in two different positions for the symmetric and
asymmetric ring isomers. One of the groups of Im-4 protons
Chem. Eur. J. 2008, 14, 2827 – 2841ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2833

Y. Kobuke et al.
0.1ppm ( d=2.2–2.3 ppm; Figure 7). The larger shift range
for the inner protons reflects the fact that the inner protons
are influenced more strongly by the macrocycle topology. In
addition, all of the outside N-Me protons of C3h-symmetric
and asymmetric macrocycles appeared at the same position
(d=1.69 ppm), which reflects the weak influences from the
macrocycle topology. The difference in magnitude of the
topological influence between inside and outside is also ob-
served in the b-pyrrole proton adjacent to the imidazolyl
group and the Im-5 protons.[15] The clear differentiation of
these protons into a single signal or three different signals
allows the assignment of the ring structures as C3h-symmet-
ric and asymmetric topological isomers, respectively.
Figure 6. Plots of the retention time versus log Mw for cyclic and noncy-
clic porphyrin arrays. The dashed line is configured by a least-mean-
Absorption and fluorescence properties of the macrocycle
and complexation with a tetrapodal ligand: The absorption
spectrum of macrocycle N-(1–Zn)3 in toluene is shown in
Figure 8 (solid line), and the peak positions of the Soret and
Q bands are listed in Table 1.
~
)
squares approximation of the plots of noncyclic porphyrin arrays 9 (
~
and 10 ( ). The hexameric and pentameric macrocycles of gable porphyr-
ins[11a,b] are plotted as authentic samples of the cyclic structure.
The Soret band exhibited
a
large split (at 411 and 443 nm;
DE=1760 cmꢀ1) by dipole–
dipole exciton coupling[27] from
slipped-cofacial and m-phenyl-
ene-bridged interactions. The
shape of the Q band was close
to the sum of N-(5–Zn)2 and 6–
Zn in a ratio of 1:1, and a negli-
gibly small excitonic coupling
was observed (Figure S4 in the
Supporting Information) be-
cause the Q bands of porphyrin
have much smaller oscillator
strengths than the Soret
band.[28] The peaks at 565 and
619.5 nm correspond to the co-
ordination dimer parts, whereas
the peak at 556 nm is contribut-
ed from the uncoordinated
Figure 7. Correlation of the proton signals for Im-4 and Im-5 in N-(1–Zn)3 on the COSY spectrum recorded at
600 MHz in CDCl3.
appeared at d=2.26 ppm in the C3h-symmetric ring and at
d=2.30, 2.23, and 2.20 ppm in the asymmetric ring, and the
other Im-4 protons appeared at d=1.72 ppm in the C3h-sym-
metric and d=1.89, 1.86, and 1.69 ppm in the asymmetric
rings. Whereas the former set of Im-4 protons was shifted to
a similar position to that of N-(5–Zn)2, the latter protons
were further shifted to higher fields. Therefore, the former
and latter signals can be assigned to the Im-4 protons out-
side and inside the ring, respectively. The inner protons
should receive further upfield shifts of d=0.4–0.5 ppm by
anisotropic effects from other porphyrins of the ring compo-
nents. A similar behavior was observed for the Im-5, N-Me,
and b-pyrrole protons adjacent to the imidazolyl group.
Four signals for the inner Im-4 protons ranged over ap-
Figure 8. UV/Vis absorption spectra of N-(1–Zn)3 (c) and N-(1Zn)3
with of 2 (1.5 equiv; gray line) in toluene (5.6”10ꢀ7 m). The inset shows a
magnification of the Q band region.
proximately 0.2 ppm (d=1.7–1.9 ppm) in contrast to the
outside protons, in which the range is approximately
2834
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2827 – 2841

Supramolecular Porphyrin Macrocycle
FULL PAPER
Table 1. Absorption and fluorescence spectral data in toluene.
Absorption [nm]
quantitatively under dilute con-
1
ditions.[34] A H NMR spectros-
Fluorescence[a] [nm]
(0,0) (0,1)
Storks shift
[cmꢀ1
ff[b]
[%]
Soret
Q
G
Q
U
Q
U
Q
G
A
]
copy titration of 2 with N-(1–
6–Zn
423
592
601651253
3.4
Zn)3 was also performed in
6–Zn/Im[c]
N-(5–Zn)2
N-(1–Zn)3
N-(1–Zn)3/2
4315669210
413.5, 437
411, 443
412.5, 446
238
3.7
CDCl3
([N-(1–Zn)3]=2.6”
566.5
556, 565
567
618.5
619.5
620
622
623
624
681
680
679
91
91
103
4.9
5.1
5.0
10ꢀ4 m) to find the structure of
the composite of 2 and N-(1–
Zn)3. However, the NMR spec-
tra became very complex and
broad upon the addition of 2
and it became difficult to assign
the signals.
[a] Excited at 550 nm. [b] ff values were determined by using ZnTPP ff =3.3% as the standard value.[29,30]
[c] In the presence of an excess amount of N-methylimidazole.
monomeric porphyrin units. The shift of the peak at 556 nm
to a longer wavelength can be useful in monitoring coordi-
Complexation of the macrocy-
nation on the free porphyrinatozinc.
cle with C60–Tripod in benzonitrile: Photoinduced-electron-
transfer experiments are usually conducted in polar solvents,
such as benzonitrile, to stabilize the charge-separated state.
Macrocycle N-(1–Zn)3 is extremely, stable even in benzoni-
trile, as determined by the fact that the UV/Vis spectral
shapes were unchanged in the concentration range of 10ꢀ6
to 10ꢀ8 m at 258C even after standing for several days. There-
fore, it was not necessary to covalently connect the trispor-
phyrin units by a metathesis reaction for the following pho-
tochemical measurements.
The steady-state fluorescence spectrum of N-(1–Zn)3 was
similar to that of dimer N-(5–Zn)2 (Table 1). The fluores-
cence quantum yields (Ff) of monomer 6–Zn, dimer N-(5–
Zn)2, and macrocycle N-(1–Zn)3 were 3.4, 4.9, and 5.1%, re-
spectively. The quantum yield of N-(5–Zn)2 is higher than
that of 6–Zn, and its Stokes shift is smaller (91cm ꢀ1 for N-
(5–Zn)2 and 253 cmꢀ1 for 6–Zn), which indicates that the co-
ordination dimer provides a more rigid structure and higher
fluorescence intensity compared with 6–Zn. The fluores-
cence quantum yield of N-(1–
Zn)3 was still maintained even
Table 2. Absorption and fluorescence spectral data in benzonitrile.
Absorption [nm] Fluorescence [nm]
(1,0) (0,0) (0,1)
though three monomeric por-
phyrins and three dimeric por-
phyrins were accumulated in
close proximity. The high fluo-
rescence quantum yield is fa-
vorable for the long-range exci-
tation-energy transfer among
the macrocycles by a singlet–
singlet energy-transfer process.
We previously synthesized
tetrapodal ligand 2 and studied
its complexation with N-(1–
ff[a]
[%]
Binding constant[b]
[mꢀ1
Soret
Q
G
Q
U
Q
N
Q
A
N
]
N-(1–Zn)3
N-(1–Zn)3/2
N-(1–Zn)3/3
N-(1–Zn)3/C60–Tripod
412.5, 445
413, 445
413, 445.5
414, 446
565
567
567
567
619.5
619.5
619.5
620
624[c]
–
675[c]
–
3.9
–
–
1.2”107
2.1”10 7
3.1”10 8
624[d]
625[c]
677[d]
676[c]
3.9[e]
0.11[e]
[a] Fluorescence quantum yields were determined by using ZnTPP ff =3.3% as the standard value.[29,30] [b] Es-
timated from UV/Vis titration. [c] Excited at 427 nm. [d] Excited at 566 nm. [e] These values were calculated
by assuming complexation of 100%.
Zn)3. Judging from the CPK model[31] of the composite N-
(1–Zn)3/2, three pyridyl arms of 2 can adjust to fit the cavity
of N-(1–Zn)3 in both C3h-symmetric and asymmetric isomers
because of the flexibility of the carbon core that connects
the four pyridyl arms. During a UV/Vis titration of 2 with
N-(1–Zn)3 in toluene, the Soret and Q bands were shifted
towards longer wavelengths and the peak at 556 nm derived
from the uncoordinated porphyrin disappeared.[15] This be-
havior indicates that 2 was accommodated in the ring of N-
(1–Zn)3 by three coordination bonds between pyridyl and
zinc (Figure 8, gray line). The clear bending behavior ob-
The UV/Vis spectrum of N-(1–Zn)3 in benzonitrile
showed slight redshifts of the Soret and Q bands (Table 2)
compared with those in toluene and disappearance of the
peak at 556 nm, which indicated that benzonitrile is coordi-
nated to the uncoordinated porphyrinatozinc(II) sites.[35] Ti-
tration of 2 with N-(1–Zn)3 in benzonitrile showed that the
Soret and Q bands were slightly redshifted with a decrease
in the intensities. The binding constant was estimated to be
1.2”107 mꢀ1 by curve fitting of the plot of change in absorb-
ance at 444 nm as a function of the ratio of [2]/[N-(1–Zn)3]
A
(Figure S5 in the Supporting Information). Tripodal ligand
3, which is the precursor for C60–Tripod, was also titrated
with N-(1–Zn)3 in benzonitrile. The binding constant was es-
timated from the UV/Vis spectra to be 2.1”10 7 mꢀ1. This
value is of the same order of magnitude as that of 2. The ex-
tremely high affinity of both 2 and 3, even in the polar ben-
zonitrile environment, must be achieved by three-point
binding between pyridyl and zinc(II). Furthermore, the com-
plexation of 3 did not affect the fluorescence of N-(1–Zn)3
served at the point of [2]/[N-(1–Zn)3]=1in the titration
A
study and the Job plot[32] of the UV/Vis analysis also indicat-
ed the formation of a stable 1:1 complex between 2 and N-
(1–Zn)3 (Figure 1and Figure S6 in ref. [15]). No heterogene-
ity of topological isomers was observed in these titrations.
The binding constant was estimated to be 8”108 mꢀ1 in tolu-
ene by curve-fitting analysis of the UV/Vis titration.[33] This
large binding constant enables us to obtain the complex
Chem. Eur. J. 2008, 14, 2827 – 2841ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2835

Y. Kobuke et al.
(Table 2), which indicated that the tripyridyl ligand moiety
did not work as an acceptor for the excited energy of N-(1–
Zn)3.
109 sꢀ1 and 0.92% from the shorter tf component of the C60–
Tripod /N-(1–Zn)3 composite according to Equations (1) and
(2):
Incorporation of C60–Tripod into the ring of N-(1–Zn)3
was examined by UV/Vis absorption and fluorescence titra-
tions. The UV/Vis absorption spectral changes by complexa-
tion of C60–Tripod with N-(1–Zn)3 are shown in Figure 9.
The Soret and Q bands were slightly redshifted with a de-
crease in intensities by the addition of C60–Tripod. The fluo-
rescence of N-(1–Zn)3 was efficiently quenched by 85%
with the addition of an equimolar amount of C60–Tripod,
and by 95% with four equivalents of C60–Tripod
(Figure 9).[36] The binding constant of C60–Tripod to N-(1–
Zn)3 was estimated to be 3.1”10 8 mꢀ1 from UV/Vis absorp-
tion and 3.4”108 mꢀ1 from fluorescence titrations, respective-
ly.[37] The value of 3”108 mꢀ1 is approximately ten times
larger than the values for tetrapodal ligand 2 and tripodal
ligand 3 without the fullerene moiety, which suggests that
the fullerene moiety significantly contributes to enhancing
the binding of C60–Tripod in N-(1–Zn)3.
kq ¼ ð1=tfÞcompositeꢀð1=tfÞmacrocycle
ð1Þ
ð2Þ
ð1=tfÞcompositeꢀð1=tfÞmacrocycle
Fq ¼
ð1=tfÞcomposite
The quantum yield of the steady-state fluorescence
quenching was estimated to be 0.97, which is compatible
with the value of 0.92 calculated from the fluorescence
decay when considering the errors associated with the re-
spective measurements. The slow fluorescence decay was
observed under the conditions in which complexation must
be almost complete, which suggests that empty N-(1–Zn)3
was regenerated under the high-laser-power conditions used
for the lifetime measurements.
The nanosecond transient absorption spectra of N-(1–
Zn)3 in the absence and presence of C60–Tripod in benzoni-
trile are shown in Figures S7 and S8 in the Supporting Infor-
mation, respectively. The spectrum of N-(1–Zn)3 after 50 ms
(Figure S7) exhibited an absorption peak at 700 to 900 nm,
which corresponded to the triplet excited state of zinc por-
phyrin. The nanosecond transient absorption spectra (Fig-
ure S8) of the complex between N-(1–Zn)3 and C60–Tripod
did not show any clear peaks in the region from 700 to
1200 nm, which indicated that species from charge-separated
and excited states were not detected on this timescale. To
detect the short charge-separated state, picosecond transient
absorption measurements of the complex were also exam-
ined in benzonitrile (Figure S9). However, the transient ab-
sorption spectra did not show apparent charge-separated
species of either a porphyrin cation radical or a fullerene
anion radical, but a broad and featureless absorption band
was observed in the 600 to 800 nm region that extended
over 1000 nm. The decay–time profile of the absorbance at
600 to 750 nm showed a life-
Time-resolved photophysical properties of the macrocycle
and C60–Tripod: To investigate the photophysical dynamics
of the macrocycle/C60–Tripod composite, time-resolved fluo-
rescence decay and transient absorption were measured.
Time-resolved fluorescence decay profiles of N-(1–Zn)3 ([N-
(1–Zn)3]=1.4”10ꢀ6 m) in the absence and presence of C60–
Tripod (4 equiv) are shown in Figure S6 in the Supporting
Information. Under these conditions, it is estimated that
99.92% of the macrocycle exists as the complex with C60–
Tripod. The fluorescence–time profile of N-(1–Zn)3 decayed
monoexponentially with a lifetime of 2100 ps, whereas that
of the macrocycle/C60–Tripod composite decayed biexponen-
tially with time constants of 164 ps (62%) and 2100 ps
(38%). A significant contribution of the longer lived compo-
nent from free N-(1–Zn)3 was observed. The quenching rate
(kq) and quantum yield (Fq) were evaluated to be 5.6”
time of 77 ps (k=1.3”1010 sꢀ1,
Figure S10), which may corre-
spond to the charge-recombina-
tion rate. Although the steady-
state fluorescence and the time-
resolved fluorescence decay in-
dicated an efficient quenching
of the singlet excited state of
N-(1–Zn)3, neither the nanosec-
ond nor picosecond transient
absorption spectra showed any
detectable peaks of the radical
ions, which suggested that the
charge-recombination rate is as
Figure 9. a) UV/Vis spectroscopy titration for the binding of N-(1–Zn)3 with C60–Tripod in benzonitrile. [N-(1–
fast as the charge-separation
rate.
Zn)3]=7.2”10ꢀ8 m, [C60–Tripod]=0–1.8, 4 equiv, 258C. The inset shows the titration plot at 412.5 nm ( ) and
*
the theoretical curve for K=3.1”10 8 mꢀ1 (c). b) Steady-state fluorescence quenching of N-(1–Zn)3 by C60–
Tripod in benzonitrile. Excited at 427 nm. [N-(1–Zn)3]=7.2”10ꢀ8 m, [C60–Tripod]=0–1.8, 4 equiv, 258C. The
*
inset shows a plot of the change in the fluorescence intensity with the signal integration (580–750 nm; ) and
the theoretical curve for K=3.4”108 mꢀ1 (c).
2836
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2827 – 2841

Supramolecular Porphyrin Macrocycle
FULL PAPER
Discussion
affords an ideal complementarity to form a dimer with a
binding constant that reaches 1011 mꢀ1 in toluene. Even
though the dimer is stabilized by complementary coordina-
tion coupled with p–p interactions in nonpolar solvents, the
structure can be easily dissociated into the monomer by
adding coordinating solvents, such as MeOH or pyridine.
The combination of formation and dissociation enables us to
construct macrocyclic structures by a reorganization process.
Because zinc insertion into 1
Atropisomers of 1: The 1H NMR spectrum of 1 exhibited
nonequivalent signals for imidazolyl and phenyl moieties
(Figure 3). The number of the signals for the imidazolyl part
(four) and the Ph-2 part (six) and their integration ratios are
reasonably explained by considering the atropisomers
(Figure 10): Two imidazolyl groups of 1 are separated at the
was performed at a relatively
high
concentration
([1]=
1.44 mm),[15]
self-assembled
polymers of 1–Zn were formed
along with cyclic species. To
eliminate the polymers, we ap-
plied the reorganization process
by using a mixed solvent of
CHCl3/MeOH (9:1, v/v) and a
low concentration ([1–Zn]=
0.02 mm). Formation of poly-
mers is entropically unfavora-
ble, and the polymers turn into
smaller molecular species under
dilute conditions. In addition,
the cyclic structures satisfy all
of the complementary coordi-
nations of imidazolyl to zinc
and are enthalpically favorable.
Combined factors drive small
oligomers into cyclic structures.
When the reorganization pro-
cess was performed at a re-
latively
low
temperature
(ꢁ178C), GPC analysis showed
a small signal at 12.3 min along
with a major signal for N-(1–
Zn)3 (Figure 4b, inset). The
signal at 12.3 min is assigned as
the cyclic tetramer of 1–Zn (N-
Figure 10. Schematic structure of atropisomers in 1. The isomers ZZ, ZE (EZ), and EE were named according
to the conformation of the imidazolyl moiety (Z and E).
terminal ends and the conformational differences at the op-
posite terminal imidazolyl group are negligible. Thus, the
imidazolyl environments are affected by the relative confor-
mation of N-Me to the central porphyrin (Z and E) and of
the two terminal porphyrin planes (chair and boat), which
results in four different environments (Z and E for the boat
and chair conformers) for the imidazolyl group (Im-4, Im-5
and N-Me). In the case of the Ph-2 protons, their environ-
ments are affected by the two N-Me directions at both ends
to afford six combinations of ZZ, ZE (=EZ), EE (1:2:1) of
chair and boat conformers. Although 1 has several atro-
pisomers as a result of their slow exchange rate on the
NMR timescale, isomerizations among atropisomers took
place under reorganization conditions in the ring-forming
procedure.[11b]
(1–Zn)4).[38] When a higher temperature (278C) was used in
the reorganization process, formation of N-(1–Zn)4 was sup-
pressed and cyclic trimer N-(1–Zn)3 was formed exclusively.
Negative enthalpy and entropy changes (DH and DS) are
expected on formation of macrocycles from 1–Zn. Because
the larger cyclic oligomers presumably have a more negative
DS than the smaller macrocycles, the smaller rings are pref-
erentially obtained, which was demonstrated in the case of
ferrocene-bridged trisporphyrin.[25] It is the same reason that
a higher temperature makes the formation of N-(1–Zn)4 un-
favorable by the more negative entropy term (ꢀTDS) than
that of N-(1–Zn)3. In this case, the formation of a cyclic
dimer was not observed because of unfavorable strain ener-
gies.
Cyclic trimer N-(1–Zn)3 possesses two topological iso-
mers, C3h symmetric and asymmetric, in a ratio of 2:3. The
statistical distribution should give a ratio of 1:3 for the C3h-
symmetric and asymmetric isomers, which means that the
Selective formation of the macrocycle: The coordination of
imidazolyl to zinc(II) in imidazolyl-substituted porphyrins
Chem. Eur. J. 2008, 14, 2827 – 2841ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2837

Y. Kobuke et al.
C3h-symmetric isomer is favored by a factor of two. Of
course, the C3h-symmetric isomer is free from angle strain
for the formation of the cyclic trimer. Because these isomers
behave similarly when accommodating tetrapodal and tripo-
dal ligands, the structural heterogeneity of the isomers does
not seem to produce any problems.
The binding constant of C60–Tripod to N-(1–Zn)3 in ben-
zonitrile was determined to be 3”108 mꢀ1, which was approx-
imately ten times larger than the values for tetrapodal
ligand 2 and tripodal ligand 3 without the fullerene moiety
(Table 2). This larger binding constant indicates that the
binding mode of C60–Tripod to N-(1–Zn)3 is different from
that expected for three-point coordination from pyridyls to
uncoordinated porphyrinatozinc sites. These results suggest
that the fullerene moiety in C60–Tripod has fallen down into
the inner surface of N-(1–Zn)3 to make direct contact be-
tween the fullerene and the porphyrin through p–p interac-
tions,[42] which is a structure that is acceptable according to
the CPK model (Figure 11b). In this conformation the fuller-
ene moiety of C60–Tripod is in direct contact with and
strongly interacts with one or two porphyrins of N-(1–Zn)3.
Actually, Hirsch et al. reported porphyrinatozinc(II)–fuller-
ene p-stacked dyads, which formed the charge-transfer state
by strong p–p interactions.[43,44] Their transient absorption
spectra showed broad and featureless absorption bands in
the wide range of 600 to 1000 nm with a decay time of 38 ps
in benzonitrile, which is attributed to the exciplex[43a] and is
comparable to the lifetime (77 ps) of the charge-separated
state or exciplex of N-(1–Zn)3 and C60–Tripod. Furthermore,
the fluorescence quenching rate of N-(1–Zn)3 by C60–Tripod
does not appreciably change in toluene and benzonitrile.
This would be compatible with the explanation that charge
separation occurs without bridging solvent molecules, that
is, through a direct pathway between the porphyrin and the
fullerene.[43a]
We have reported that the excitation energy hopping
times through m-phenylene linkages were 8.0 and 5.3 ps in
pentameric and hexameric macrocycles of gable porphyrin,
respectively.[14] In the complex of N-(1–Zn)3 and C60–Tripod,
comparable energy hopping times are expected because the
coordinated monomeric porphyrin has a similar excited
energy level to that of the dimeric porphyrin. Actually, on
complexation of C60–Tripod, the fluorescence of N-(1–Zn)3
from both monomeric and dimeric porphyrin units was com-
pletely quenched, which demonstrated that the excitation
energy migrates rapidly among the porphyrins and then con-
verges on the fullerene moiety by electron transfer.
Complexation of the tripodal ligand with the macrocycle:
We previously reported that a slipped-cofacial dimer at-
tached to a fullerene afforded a stable charge-separated
state by photoexcitation of the porphyrins.[39] Because the
first one-electron oxidation potential of pyridine-coordinat-
ed porphyrinatozinc is similar to that of the complementary
coordination dimer[40] and the energy gap between the
ground and the lowest excited states is also almost the
same,[41] both of the excited porphyrin units in the composite
of N-(1–Zn)3 and C60–Tripod are considered to have a simi-
lar ability for electron donation. Thus, the fullerene unit in
C60–Tripod can accept an electron from all of the porphyrin
components in the ring. The center-to-center distance from
each porphyrin in N-(1–Zn)3 to the fullerene moiety of C60–
Tripod in the complex is estimated to be 15 to 20 Š by as-
suming three-point coordination from the tripodal pyridyl li-
gands to the uncoordinated porphyrinatozinc(II) sites (Fig-
ure 11a). Judging from this distance (15–20 Š), we reasona-
bly expected that the complex of N-(1–Zn)3 and C60–Tripod
would have a long lifetime for the charge-separated state by
photoexcitation of the porphyrins in N-(1–Zn)3. Efficient
fluorescence quenching of N-(1–Zn)3 observed in the
steady-state fluorescence titration studies clearly showed
that fast electron transfer from the excited porphyrin to the
fullerene occurred in the complex in which all of the por-
phyrin pairs act as one fluorescence chromophore. However,
the transient absorption spectra of nano- and picosecond
timescales were featureless, which showed that there was no
definitive charge-separated species nor any triplet excited
state of either porphyrin or fullerene. This strongly indicates
that fast charge recombination to the ground state occurs
after the initial charge separation, which suggests that the
fullerene and the porphyrin are closely positioned in the
macrocycle.
Figure 11. The CPK model of C60–Tripod /N-(1–Zn)3, which was created by using Cerius2 software,[31] based on a) three pyridyl coordinations and b) two
pyridyl coordination and fullerene–porphyrin p–p interactions.
2838
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2827 – 2841

Supramolecular Porphyrin Macrocycle
FULL PAPER
on silica gel 60 F254 TLC plates (Merck). The silica gel utilized for
column chromatography was purchased from Kanto Chemical (Silica Gel
60N (Spherical, Neutral) 60–210 mm).
Conclusion
Cyclic trimer N-(1–Zn)3 could be obtained exclusively from
trisporphyrinatozinc(II) 1–Zn appended with imidazolyl
groups at the terminal porphyrins when the appropriate con-
ditions were applied in the reorganization process. Macrocy-
cle N-(1–Zn)3 accommodated tripodal ligands with two
kinds of binding modes. Model ligand 2 and the tripodal
ligand without fullerene (3) were bound by using three-
point coordinations from the pyridyl ligands to uncoordinat-
ed porphyrinatozinc(II) sites of N-(1–Zn)3, the binding con-
stants of which are estimated to be 1–2”107 mꢀ1 in benzoni-
trile. The fullerene–tripodal ligand (C60–Tripod) was bound
by using two-point coordinations from the pyridyl ligands to
uncoordinated porphyrinatozinc(II) sites of N-(1–Zn)3 and
fullerene–porphyrin p–p interactions, the binding constant
of which was estimated to be 3”108 mꢀ1 in benzonitrile.
Direct contact of the fullerene moiety to porphyrin pro-
duced a binding constant ten times larger than the values
for three-point coordinations, and complete quenching of
the fluorescence of N-(1–Zn)3. The large binding constant
enabled us to quantitatively obtain the complex of C60–
Tripod with N-(1–Zn)3 by the addition of equivalent tripodal
ligands under dilute conditions (ꢁ10ꢀ6 m) even in benzoni-
trile, and to utilize C60–Tripod as a fluorescence quencher
for investigations into the energy-transfer process among
the macrocycles. Because displacement of the fourth arm of
the tripodal ligand is easily achieved, the present method is
applicable for introducing various types of functional
groups, and to construct their composites with porphyrin
macrocycles.
4-Tris(4-iodophenyl)methylbenzoic acid 7: 4-(Tripenylmethyl)benzoic
acid[19] (123 mg, 3.18”10ꢀ4 mol), bis(trifluoroacetoxy)iodobenzene
(684 mg, 1.59”10ꢀ3 mol) and iodine (404 mg, 1.59”10ꢀ3 mol) were dis-
solved in CCl4 (3 mL). The mixture was stirred for 3 h at 608C. After the
mixture was cooled to room temperature, the reaction mixture was dilut-
ed with CHCl3 (ꢁ50 mL), and then washed successively with aqueous
sodium bisulfite and water. The organic layer was evaporated to dryness
and the residue was purified by means of a short column (SiO2: CHCl3 to
CHCl3/MeOH 9:1). The invisible second fraction, which eluted with 10%
MeOH/CHCl3, was collected and evaporated to afford a pale yellow
solid (234 mg). The integral values of 1H NMR signals indicated that the
solid was a mixture of 7 (108 mg, 46%) and tetraiodo-substituted 7
(126 mg) in a molar ratio of 1:1. 7: 1H NMR (600 MHz, CDCl3): d=7.98
(d, J=8.5 Hz, 2H; Ph’), 7.61(d, J=8.8 Hz, 6H; Ph), 7.28 (d, J=8.5 Hz,
2H; Ph’), 6.89 (d, J=8.8 Hz, 6H; Ph). Tetraiodo-substituted 7: 1H NMR
(600 MHz, CDCl3): d=8.00 (d, J=8.2 Hz, 2H; Ph’), 7.74 (d, J=8.5 Hz,
1H; Ph-5), 7.67, (d, J=2.2 Hz, 1H; Ph-2), 7.60 (d, J=8.8 Hz, 4H; Ph),
7.26–7.27 (d, 2H; Ph’), 6.90 (d, J=8.8 Hz, 4H; Ph), 6.81ppm (dd, J=8.5,
2.2 Hz, 1H; Ph-6).
4-Tris[4-{2-(4-pyridyl)ethynyl}phenyl]methylbenzoic acid 3: 4-Tris(4-iodo-
phenyl)methylbenzoic acid (32 mg, 4.3”10ꢀ5 mol, the amount of tetraio-
do-substituted
7
(38 mg) was subtracted from 70 mg of the impure
(PPh3)2Cl2
sample), 4-ethynylpyridine[45] (40 mg, 3.9”10ꢀ4 mol), Pd
ACHTREUNG
(13 mg, 1.9”10ꢀ5 mol), CuI (4 mg, 1.9”10ꢀ5 mol), dry Et2NH (0.5 mL),
and dry THF (0.5 mL) were placed in a Schlenk flask under an argon at-
mosphere. The mixture was degassed by freeze–thaw cycles and stirred
for 12 h at room temperature under argon. The reaction mixture was di-
luted with CHCl3 and washed with saturated aqueous NaCl and water.
The solvent was evaporated and the residue was purified by means of
column chromatography (SiO2; CHCl3/MeOH 99:1to 93:7). The third
fraction was collected and evaporated to afford
3 (16.5 mg, 57%).
1H NMR (600 MHz, CDCl3): d=8.62 (d, J=5.9 Hz, 6H; Py), 8.03 (d, J=
8.2 Hz, 2H; Ph’), 7.49 (d, J=8.2 Hz, 6H; Ph), 7.39 (d, J=5.9 Hz, 6H;
Py), 7.33 (d, J=8.2 Hz, 2H; Ph’), 7.25 (d. J=8.2 Hz, 6H; Ph), 1.94 ppm
(brs; COOH); MALDI-TOF: m/z: 668.0 [M+H].
C60–Tripod: 4-Tris[4-{2-(4-pyridyl)ethynyl}phenyl]methylbenzoic acid
(11 mg, 1.7”10ꢀ5 mol) and BOP (7 mg, 1.7”10ꢀ5 mol) were placed in a
20 mL flask and purged with argon gas. CHCl3 (2.5 mL) and pyridine
(0.5 mL) were added to the mixture, and the mixture was stirred for
10 min at room temperature. Pyrrolidine–C60 TFA salt[21] (13 mg, 1.5”
10ꢀ5 mol) was then added and the reaction mixture was stirred at room
temperature. The reaction progress was monitored by MALDI-TOF
mass spectrometry. After stirring for 24 h, further BOP (14 mg, 3.2”10ꢀ5
mol) was added to the reaction mixture to promote the reaction, and stir-
ring was continued for another 24 h. The solvent was evaporated to give
a crude residue. The residue was purified with by column chromatogra-
phy (SiO2; CHCl3/MeOH 95:5) and the brown band was collected. This
fraction was evaporated and further reprecipitated with hexane to give
C60–Tripod as a brown solid (4.8 mg, 21%): Rf =0.6 (SiO2; CHCl3/MeOH
9:1); 1H NMR (600 MHz, CDCl3): d=8.61(brs, 6H; Py), 7.85 (d, J=
8.5 Hz, 2H; Ph’), 7.51(d, J=8.5 Hz, 6H; Ph), 7.46 (d, J=8.5 Hz, 2H;
Ph’), 7.37 (brd, J=4.9 Hz, 6H; Py), 7.30 (d, J=8.5 Hz, 6H; Ph),
5.56 ppm (brs, CH2); 13C NMR (150 MHz, CDCl3): d=169.8, 149.8,
148.6, 147.4, 146.43, 146.40, 146.2, 145.6, 145.53, 145.45, 145.40, 144.5,
143.2, 143.1, 142.8, 142.2, 142.1, 142.0, 140.2, 133.2, 131.6, 131.3, 131.2,
131.0, 128.1, 125.5, 120.6, 93.3, 87.3 ppm; UV/Vis (CHCl3): lmax (e): 432
(2900), 698.5 nm (290 molꢀ1 dm3 cmꢀ1); MALDI-TOF: m/z: 1414.0
[M+H].
Experimental Section
General procedure: The syntheses of porphyrins 1, N-(1–Zn)3, C-(1Zn)3,
2, 4, 5, N-(5–Zn)2, 6, and 6–Zn have been previously reported.[15] All
chemicals and solvents were of commercial reagent quality, and used
without further purification unless otherwise stated. Dry THF was pre-
pared by distillation over benzophenone-Na. Dry DMF, Et2NH, and
Et3N were prepared by distillation over CaH2. First generation Grubbs
catalyst (benzylidene-bis(tricyclohexylphosphine)dichlororuthenium) was
purchased from Aldrich. 1H NMR spectra were recorded by using a
JEOL ECP-600 (600 MHz) spectrometer and chemical shifts were re-
corded in parts per million (ppm) relative to tetramethylsilane. UV/Vis
absorption spectra were recorded by using a Shimadzu UV-3100PC spec-
trometer. Steady-state fluorescence emission spectra were recorded by
using a Hitachi F-4500 spectrometer and corrected for the response of
the detector system. The fluorescence intensities were normalized at the
absorption of their excitation wavelength. UV/Vis lmax values are report-
ed in nm. Fluorescence quantum yields were determined by corrected in-
tegrated ratios of steady-state fluorescence spectra relative to that of tet-
raphenylporphyrinatozinc (ZnTPP; Ff =3.3%).[29] UV/Vis and fluores-
cence titrations were performed by adding a solution of pyridyl ligand to
a solution of N-(1–Zn)3 in a quartz cuvette (1cm path length) by using
microliter syringes. MALDI-TOF mass spectra were obtained by using a
PerSeptive Biosystems Voyager DE-STR instrument with dithranol (Al-
drich) as the matrix. Analytical gel permeation chromatograms were ob-
tained by using a Hewlett Packard HP1100 series instrument equipped
with an analytical JAIGEL 3HA column (Japan Analytical Industry,
8 mm”500 mm, exclusion limit=70,000 Da). Reactions were monitored
Noncyclic porphyrin array 9 (10): Pyridine (0.5 mL) was added to a mix-
ture of N-(1–Zn)3 (1.2 mg, 2.0”10ꢀ7 mol) and N-(5–Zn)2 (1.0 mg, 6.0”
10ꢀ7 mol) in CHCl3 (10 mL), and then the solvents were evaporated to
dryness under reduced pressure. The residue was dissolved in CHCl3 and
subjected to GPC analysis. The preparation of series 10 were performed
with the same procedure as that used for 9 except 8 was used instead of
N-(5–Zn)2.
Chem. Eur. J. 2008, 14, 2827 – 2841ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2839

Y. Kobuke et al.
Time-resolved emission and transient absorption measurements: The
fluorescence lifetimes were measured by using an argon-ion pumped Ti/
sapphire laser (Tsunami) and a streak scope (Hamamatsu Photonics).
The nanosecond transient absorption spectra in the NIR region were
measured by means of laser-flash photolysis; 565 nm light from Nd/YAG
laser was used as the exciting source, and a Ge-avalanche-photodiode
module was used for detecting the monitoring light from a pulsed Xe
lamp. The picosecond transient absorption spectra were measured by the
pump and probe method by using a Ti/sapphire regenerative amplifier
seeded by the SHG of an Er-doped fiber laser (Clark-MXR CPA-2001
plus). The details of the experimental setup are described elsewhere.[46]
Ahn, D. M. Ko, A. Takagi, T. Kawai, T. Matsumoto, D. Kim, A.
Osuka, J. Am. Chem. Soc. 2005, 127, 236; k) M. Hoffmann, C. J.
Wilson, B. Odell, H. L. Anderson, Angew. Chem. 2007, 119, 1 ;
Angew. Chem. Int. Ed. 2007, 46, 3122.
[7] For recent reports of noncovalently linked cyclic array, see: a) C. M.
Drain, K. C. Russell, J.-M. Lehn, Chem. Commun. 1996, 337; b) S.
Knapp, J. Vasudevan, T. J. Emge, B. H. Arison, J. A. Potenza, H. J.
Schugar, Angew. Chem. 1998, 110, 2537; Angew. Chem. Int. Ed.
1998, 37, 2368; c) C. Ikeda, N. Nagahara, E. Motegi, N. Yoshioka, H.
Inoue, Chem. Commun. 1999, 1759; d) R. A. Haycock, A. Yartsev,
U. Michelsen, V. Sundstrçm, C. A. Hunter, Angew. Chem. 2000, 112,
3762; Angew. Chem. Int. Ed. 2000, 39, 3616; e) R. A. Haycock, C. A.
Hunter, D. A. James, U. Michelsen, L. R. Sutton, Org. Lett. 2000, 2,
2435; f) A. Tsuda, T. Nakamura, S. Sakamoto, K. Yamaguchi, A.
Osuka, Angew. Chem. 2002, 114, 2941; Angew. Chem. Int. Ed. 2002,
41, 2817; g) T. Kamada, N. Aratani, T. Ikeda, N. Shibata, Y. Higuchi,
A. Wakamiya, S. Yamaguchi, K.-S. Kim, Z.-S. Yoon, D. Kim, A.
Osuka, J. Am. Chem. Soc. 2006, 128, 7670.
[8] D. M. Guldi, M. Prato, Acc. Chem. Res. 2000, 33, 695.
[9] For reviews of covalently linked porphyrin and fullerene conjugates,
see: a) D. Gust, T. A. Moore in The Porphyrin Handbook, Vol. 8,
(Eds.: K. M. Kadish, K. M. Smith, R. Guilard), Academic Press,
New York, 2000, pp. 153; b) H. Imahori, Org. Biomol. Chem. 2004,
2, 1425.
[10] For reports of a noncovalently linked porphyrin and fullerene conju-
gate, see: a) N. Armaroli, F. Diederich, L. Echegoyen, T. Habicher,
L. Flamigni, G. Marconi, J.-F. Nierengarten, New J. Chem. 1999, 23,
77; b) F. DꢁSouza, G. R. Deviprasad, M. S. Rahman, J.-P. Choi,
Inorg. Chem. 1999, 38, 2157; c) T. Da Ros, M. Prato, D. M. Guldi, E.
Alessio, M. Ruzzi, L. Pasimeni, Chem. Commun. 1999, 635; d) S. R.
Wilson, S. MacMahon, F. T. Tat, P. D. Jarowski, D. I. Schuster,
Chem. Commun. 2003, 226; e) F. DꢁSouza, S. Gadde, M. E. Zandler,
M. Itou, Y. Araki, O. Ito, Chem. Commun. 2004, 2276; f) F.
DꢁSouza, R. Chitta, S. Gadde, M. E. Zandler, A. S. D. Sandanayaka,
Y. Araki, O. Ito, Chem. Commun. 2005, 1279; g) J. L. Sessler, J.
Jayawickramarajah, A. Gouloumis, T. Torres, D. M. Guldi, S. Maldo-
nado, K. J. Stevenson, Chem. Commun. 2005, 1892; h) A. Trabolsi,
M. Elhabiri, M. Urbani, J. L. D. de la Cruz, F. Ajamaa, N. Solladiꢂ,
A.-M. Albrecht-Gary, J.-F. Nierengarten, Chem. Commun. 2005,
5736; i) L. Sꢃnchez, M. Sierra, N. Martin, A. J. Myles, T. J. Dale, J.
Rebek, Jr., W. Seitz, D. M. Guldi, Angew. Chem. 2006, 118, 4753;
Angew. Chem. Int. Ed. 2006, 45, 4637.
Acknowledgements
This work was supported by Grants-in-Aid for Scientific Research (A)
(Y.K.) from the Japan Society for the Promotion of Science (JSPS).
[1] R. E. Blankenship, Molecular Mechanisms of Photosynthesis, Black-
well Science, Oxford, 2002.
[2] a) S. Karrasch, P. A. Bullough, R. Ghosh, EMBO J. 1995, 14, 631;
b) A. W. Roszak, T. D. Howard, J. Southall, A. T. Gardiner, C. L.
Law, N. W. Isaacs, R. J. Cogdell, Science 2003, 302, 1969.
[3] a) G. McDermott, S. M. Prince, A. A. Freer, A. M. Haw-
thornthwaite-Lawless, M. Z. Papiz, R. J. Cogdell, N. W. Isaacs,
Nature 1995, 374, 517; b) J. Koepke, X. Hu, C. Muenke, K. Schulten,
K. Michel, Structure 1996, 4, 581.
[4] a) S. Bahatyrova, R. N. Frese, C. A. Siebert, J. D. Olsen, K. O.
van der Werf, R. van Grondelle, R. A. Niederman, P. A. Bullough,
C. Otto, C. N. Hunter, Nature 2004, 430, 1058; b) S. Scheuring, J. N.
Sturgis, V. Prima, A. Bernadac, D. Levy, J.-L. Rigaud, Proc. Natl.
Acad. Sci. USA 2004, 101, 11293; c) S. Scheuring, J. N. Sturgis, Sci-
ence 2005, 309, 484.
[5] For recent reviews, see: a) J. K. M. Sanders in Comprehensive Supra-
molecular Chemistry, Vol. 9 (Eds.: J. L. Atwood, J. E. D. Davies,
D. D. MacNicol, F. Vçgtle), Pergamon, Oxford, 1996, pp. 1 31 ; b) J.
Chambron, V. Heitz, J. P. Sauvage in The Porphyrin Handbook,
Vol. 6 (Eds.: K. M. Kadish, K. M. Smith, R. Guilard), Academic
Press, New York, 2000, pp. 1; c) A. K. Burrell, D. L. Officer, P. G.
Plieger, D. C. W. Reid, Chem. Rev. 2001, 101, 2751; d) P. D. Harvey
in The Porphyrin Handbook, Vol. 18 (Eds.: K. M. Kadish, K. M.
Smith, R. Guilard), Academic Press, New York, 2003, pp. 63; e) M.-
S. Choi, T. Yamazaki, I. Yamazaki, T. Aida, Angew. Chem. 2004,
116, 152; Angew. Chem. Int. Ed. 2004, 43, 150; f) Y. Kobuke, Eur. J.
Inorg. Chem. 2006, 2333; g) Y. Nakamura, N. Aratani, A. Osuka,
Chem. Soc. Rev. 2007, 36, 831.
[6] For recent reports of covalent linking cyclic array, see: a) S. Ander-
son, H. L. Anderson, A. Bashall, M. McPartlin, J. K. M. Sanders,
Angew. Chem. 1995, 107, 1196; Angew. Chem. Int. Ed. Engl. 1995,
34, 1096; b) H. A. M. Biemans, A. E. Rowan, A. Verhoeven, P. Va-
noppen, L. Latterini, J. Foekema, A. P. H. J. Schenning, E. W.
Meijer, F. C. de Schryver, R. J. M. Nolte, J. Am. Chem. Soc. 1998,
120, 11054; c) J. Li, A. Ambroise, S. I. Yang, J. R. Diers, J. Seth,
C. R. Wack, D. F. Bocian, D. Holten, J. S. Lindsey, J. Am. Chem.
Soc. 1999, 121, 8927; d) O. Mongin, A. Schuwey, M.-A. Vallot, A.
Gossauer, Tetrahedron Lett. 1999, 40, 8347; e) O. Mongin, N.
Hoyler, A. Gossauer, Eur. J. Org. Chem. 2000, 1193; f) M. Takase,
R. Ismael, R. Murakami, M. Ikeda, D. Kim, H. Shinmori, H. Furuta,
A. Osuka, Tetrahedron Lett. 2002, 43, 5157; g) X. Peng, N. Aratani,
A. Takagi, T. Matsumoto, T. Kawai, L.-W. Hwang, T. K. Ahn, D.
Kim, A. Osuka, J. Am. Chem. Soc. 2004, 126, 4468; h) M. C. Lensen,
S. J. T. van Dingenen, J. A. A. W. Elemans, H. P. Dijkstra, G. P. M.
van Klink, G. van Koten, J. W. Gerritsen, S. Speller, R. J. M. Nolte,
A. E. Rowan, Chem. Commun. 2004, 762; i) A. Kato, K. Sugiura, H.
Miyasaka, H. Tanaka, T. Kawai, M. Sugimoto, M. Yamashita, Chem.
Lett. 2004, 33, 578; j) Y. Nakamura, L.-W. Hwang, N. Aratani, T. K.
[11] a) R. Takahashi, Y. Kobuke, J. Am. Chem. Soc. 2003, 125, 2372;
b) R. Takahashi, Y. Kobuke, J. Org. Chem. 2005, 70, 2745; c) C.
Ikeda, A. Satake, Y. Kobuke, Org. Lett. 2003, 5, 4935.
[12] F. Hajjaj, Z. S. Yoon, M.-C. Yoon, J. Park, A. Satake, D. Kim, Y.
Kobuke, J. Am. Chem. Soc. 2006, 128, 4612.
[13] Y. Kobuke, H. Miyaji, J. Am. Chem. Soc. 1994, 116, 4111.
[14] I.-W. Hwang, M. Park, T. K. Ahn, Z. S. Yoon, D. M. Ko, D. Kim, F.
Ito, Y. Ishibashi, S. R. Khan, Y. Nagasawa, H. Miyasaka, C. Ikeda,
R. Takahashi, K. Ogawa, A. Satake, Y. Kobuke Chem. Eur. J. 2005,
11, 3753.
[15] Y. Kuramochi, A. Satake, Y. Kobuke, J. Am. Chem. Soc. 2004, 126,
8668.
[16] a) D. Kuciauskas, P. A. Liddell, S. Lin, T. E. Johnson, S. J. Weghorn,
J. S. Lindsey, A. L. Moore, T. A. Moore, D. Gust, J. Am. Chem. Soc.
1999, 121, 8604; b) A. Nakano, A. Osuka, T. Yamazaki, Y. Nishi-
mura, S. Akimoto, I. Yamazaki, A. Itaya, M. Murakami, H. Miyasa-
ka, Chem. Eur. J. 2001, 7, 3134; c) M.-S. Choi, T. Aida, H. Luo, Y.
Araki, O. Ito, Angew. Chem. 2003, 115, 4194; Angew. Chem. Int. Ed.
2003, 42, 4060.
[17] a) J. S. Lindsey, I. C. Schreiman, H. C. Hsu, P. C. Kearney, A. M.
Marguerettaz, J. Org. Chem. 1987, 52, 827; b) B. J. Littler, Y. Cir-
ingh, J. S. Lindsey, J. Org. Chem. 1999, 64, 2864.
[18] A. Ohashi, A. Satake, Y. Kobuke, Bull. Chem. Soc. Jpn. 2004, 77,
365.
[19] R. K. R. Jetti, F. Xue, T. C. W. Mak, A. Nangia, J. Chem. Soc. Perkin
Trans. 2, 2000, 1223.
2840
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2827 – 2841

Supramolecular Porphyrin Macrocycle
FULL PAPER
[20] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975, 16,
4467.
[21] a) M. Maggini, G. Scorrano, M. Prato, J. Am. Chem. Soc. 1993, 115,
9798; b) T. Da Ros, D. M. Guldi, A. F. Morales, D. A. Leigh, M.
Prato, R. Turco, Org. Lett. 2003, 5, 689.
[22] When the evaporation speed of the solvent became slow during the
concentration process, the solvent tended to increase in methanol
concentration, which resulted in slight dissociation of the cyclic
trimer. Therefore, it is better for the evaporation process to be con-
ducted at high speed so that the solvent composition does not
change significantly.
[38] The fraction of the signal at 12.3 min was separated with a prepara-
tive GPC column (JAIGEL 3H; Japan Analytical Industry) by using
CHCl3 as eluent, and used in the metathesis reaction with Grubbs
catalyst. The MALDI-TOF mass spectrum exhibited a broad signal
at around 7750 Da, which corresponded to the calculated value of
the tetramer (calcd for C432H368N64O32Zn12: 7752.9). Furthermore,
the plot of retention time (12.3 min) against molecular weight was
added to a calibration curve for the series of cyclic compounds
(Figure 6).
[39] H. Nakagawa, K. Ogawa, A. Stake, Y. Kobuke, Chem. Commun.
2006, 1560.
[23] a) S. T. Nguyen, R. H. Grubbs, J. W. Ziller, J. Am. Chem. Soc. 1993,
115, 9858; b) E. L. Dias, S. T. Nguyen, R. H. Grubbs, J. Am. Chem.
Soc. 1997, 119, 3887.
[24] a) J. Roovers, P. M. Toporowski, Macromolecules 1983, 16, 843;
b) H. Jiang, W. Lin, J. Am. Chem. Soc. 2004, 126, 7426.
[25] O. Shoji, S. Okada, A. Satake, Y. Kobuke, J. Am. Chem. Soc. 2005,
127, 2201.
[26] In ref. [15] we reported that the compound was D3h-symmetric and
this has since been corrected to C3h-symmetric.
[27] M. Kasha, H. R. Rawls, M. A. El-Bayoumi, Pure Appl. Chem. 1965,
11, 371.
[28] I. Yamazaki, S. Akimoto, T. Yamazaki, H. Shiratori, A. Osuka, Acta
Phys. Pol. A 1999, 95, 105.
[29] D. J. Quimby, F. R. Longo, J. Am. Chem. Soc. 1975, 97, 5111.
[30] C. A. Parker, W. T. Rees, Analyst 1960, 85, 587.
[31] MATERIAL STUDIO (Version 4.1/Forcite/Force Field UNIVER-
SAL), Accelrys, Inc. (San Diego, CA, USA, 2001).
[40] The first oxidation potentials of the coordinated monomeric and the
slipped-cofacial dimeric porphyrin in dichloromethane were estimat-
ed to be 0.60 and 0.63 V, respectively, see: a) H. Ozeki, A. Nomoto,
K. Ogawa, Y. Kobuke, M. Murakami, K. Hosoda, M. Ohtani, S. Na-
kashima, H. Miyasaka, T. Okada, Chem. Eur. J. 2004, 10, 6393; b) D.
Kalita, M. Morisue, Y. Kobuke, New J. Chem. 2006, 30, 77; further-
more, the electrochemical interaction of the porphyrins through the
m-phenylene linker could be ignored, and therefore, both the dimer-
ic and monomeric porphyrin units in the composite C60–Tripod/N-
(1–Zn)3 would have similar oxidation potentials, see: c) J. L. Sessler,
M. R. Johnson, S. E. Creager, J. C. Fettinger, J. A. Ibers, J. Am.
Chem. Soc. 1990, 112, 9310.
[41] Both absorption and fluorescence spectra of the monomeric por-
phyrins were shifted towards longer wavelengths by complexation of
ligands, and approached those of the slipped-cofacial dimeric por-
phyrin. Thus, the energy gaps between the lowest excited state and
the ground state of the coordinated monomeric and the complemen-
tary coordinated dimeric porphyrins became almost the same
(2.02 eV for 6–Zn/Im and 2.00 eV for N-(5–Zn)2, Table 1).
[42] J.-L. Hou, H.-P. Yi, X.-B. Shao, C. Li, Z.-Q. Wu, X.-K. Jiang, L.-Z.
Wu, C.-H. Tung, Z.-T. Li, Angew. Chem. 2006, 118, 81 0;Angew.
Chem. Int. Ed. 2006, 45, 796.
[32] K. A. Connors, Binding Constants, Wiley, New York, 1987, pp. 24–
28.
[33] The binding constant of the monodentate pyridyl ligand (4-ethynyl-
pyridine) was calculated to be 2”103 mꢀ1 in toluene. This large bind-
ing constant (8”108 mꢀ1) between 2 and N-(1–Zn)3 reflects three-
point binding from pyridyl to the zinc(II) ion.
[43] a) D. M. Guldi, A. Hirsch, M. Scheloske, E. Dietel, A. Troisi, F. Zer-
betto, M. Prato, Chem. Eur. J. 2003, 9, 4968; b) V. Chukharev, N. V.
Tkachenko, A. Efimov, D. M. Guldi, A. Hirsch, M. Scheloske, H.
Lemmetyinen, J. Phys. Chem. B 2004, 108, 16377.
[34] According to the binding constant of 8”108 mꢀ1 in toluene, it is esti-
mated that 97% of 2 is bound to N-(1–Zn)3 in an equimolar mixture
of 2 and N-(1–Zn)3 at 1.0”10ꢀ6 m.
[35] Benzonitrile weakly binds to 6–Zn in toluene and the binding con-
stant was estimated to be ꢁ0.5mꢀ1 in toluene from the UV/Vis titra-
tion.
[44] We also attempted to determine the CT absorption and emission
bands on the steady-state spectra. However, the CT bands were not
observed because their signals were very weak and the noninteract-
ing porphyrins of N-(1–Zn)3 hid them.
[45] L. Yu, J. S. Lindsey, J. Org. Chem. 2001, 66, 7402.
[46] T. Nakamura, M. Fujitsuka, Y. Araki, O. Ito, J. Ikemoto, K. Taka-
miya, Y. Aso, T. Otsubo, J. Phys. Chem. B 2004, 108, 10700.
[36] The titration in toluene and o-dichlorobenzene also showed almost
complete fluorescence quenching (ꢁ98%). However, because a
part of macrocycle N-(1–Zn)3 was precipitated or adsorbed onto the
quartz cell on incorporating C60–Tripod, accurate binding constants
could not be obtained in toluene and o-dichlorobenzene.
[37] A control titration experiment performed by titrating C60 into N-(1–
Zn)3 showed no change in the UV/Vis absorption and fluorescence
measurements under the same titration conditions.
Received: November 1, 2007
Published online: January 28, 2008
Chem. Eur. J. 2008, 14, 2827 – 2841ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2841