Highly linear self-assembled porphyrin wires

Article abstract of DOI:10.1021/ic2001255

An efficient noncovalent assembly process involving high geometrical control was applied to a linear bis(imidazolyl zinc porphyrin) 7Zn, bearing C₁₈ substitutents, to generate linear multiporphyrin wires. The association process is based on imi

Full text of DOI:10.1021/ic2001255

ARTICLE
pubs.acs.org/IC
Highly Linear Self-Assembled Porphyrin Wires
Matthieu Koepf,†,‡ Jonas Conradt,§ Je-drzej Szmytkowski,§,^ Jennifer A. Wytko,*,† Lionel Allouche,†
Heinz Kalt,§ Teodor Silviu Balaban,|| and Jean Weiss*,†
†Institut de Chimie, UMR 7177 CNRS-UDS, 1 Rue Blaise Pascal, 67008 Strasbourg, France
§Institut f€ur Angewandte Physik, Karlsruhe Institute of Technology (KIT), 76128 Karlsruhe, Germany
Universitꢀe Paul Cꢀezanne Aix-Marseille III, Institut de Sciences Molꢀeculaires de Marseille, Chirosciences, UMR 6263,
13397 Marseille, France
S Supporting Information
b
ABSTRACT: An efficient noncovalent assembly process invol-
ving high geometrical control was applied to a linear bis-
(imidazolyl zinc porphyrin) 7Zn, bearing C18 substitutents, to
generate linear multiporphyrin wires. The association process is
based on imidazole recognition within the cavity of the phenan-
throline-strapped zinc porphyrin. In chlorinated solvents, dis-
crete soluble oligomers were obtained after (7Zn)n was end-
capped with a terminal single imidazolyl zinc porphyrin deriva-
tive 4Zn. These soluble species, as well as their destabilization
in the presence of protic solvents, were studied by UVꢀvisible
and time-resolved luminescence. In the solid state, assemblies as long as 480 nm, which corresponds to 190 iterative units or a
total of 380 porphyrins, were observed by atomic force microscopy measurements on mica. The length and linearity of the
porphyrin wires obtained illustrate the potential of phenanthroline-strapped porphyrins for the directional control of self-
assembly processes.
’ INTRODUCTION
the largest multiporphyrinic assemblies reported were obtained
by both covalent and self-assembled approaches, using respec-
tively Osuka’s9 iterative oxidative coupling of meso-unsubsti-
tuted porphyrins, or Kobuke’s10 imidazole coordination in self-
complementary zinc porphyrin dimers. These elegant and
powerful approaches led to large multiporphyrin assemblies for
which near-field microscopy imaging was reported.11,12 In both
covalent and self-assembled species, imaging showed that the
linearity of the objects was inversely related to the size of the
linear assembly.
The specific binding of N-unsubstituted imidazoles within the
pocket of phenanthroline-strapped zinc porphyrins has been
established and used as an assembly tool for photonic dyads and
triads over the past few years.13 Coordination of imidazole within
the phenanthroline pocket is enhanced by the combination of
three noncovalent interactions: (1) a metalꢀligand coordination
bond, (2) πꢀπ stacking between the imidazole and the phenyl
substituents on the phenanthroline, and (3) hydrogen bonding
of imidazole’s NH proton to phenanthroline’s nitrogen atoms.
Imidazolyl zinc porphyrins 3Zn and 4Zn assemble into linear
wires under certain conditions on highly ordered pyrolytic
graphite (HOPG).14 To ensure the formation of continuous
linear wires, two of these imidazolyl-porphyrins have now been
In the development of nanomaterials for molecular electro-
nics, porphyrin oligomers have received a great deal of attention
in the past decade. Many examples of photoinduced energy- or
electron-transfer reactions have been reported in geometrically
well-defined porphyrin dyads, triads, and higher oligomers,
providing crucial information on the energy- and electron-
transfer mechanisms involved in these processes.1 To produce
large multiporphyrinic assemblies for future applications, the
strong emergence of self-assembled multiporphyrinic species is
extremely attractive.2 Two main types of arrangements can be
expected for multiporphyrin assemblies, namely, H-aggregates,
with π-stacked porphyrins facing each other, and J-aggregates,
with a partially overlapped side by side arrangement that is
reminiscent of natural light harvesting antennae systems.1,2 Many
reported self-assembled scaffolds, inspired by H-aggregates,
demonstrate the bright future of self-assembly. The mechanical
resistance of assemblies that rely exclusively on aromatic stacking
needs to be reinforced in various manners such as interactions of
lipophilic side chains,3 ion binding,4 or hydrogen bonding.5 For
multiporphyrins of the J-aggregate type, a major issue concerns
the linearity of the scaffolds.
A necessary step prior to the development of materials is the
achievement of a high degree of linearity resembling that
observed in polymer-type assemblies where linearity may origi-
nate from cohesion forces.6,7 Aside from crystalline nanorods,8
Received: January 19, 2011
Published: June 07, 2011
r
2011 American Chemical Society
6073
dx.doi.org/10.1021/ic2001255 Inorg. Chem. 2011, 50, 6073–6082
|

Inorganic Chemistry
ARTICLE
Scheme 1. Synthesis of the Self-Complementary
Bis(imidazolyl porphyrin) 7Zna
Figure 1. Three possible association modes for the homodimers
(3Zn)2 and (4Zn)2. Alkyl chains are omitted for clarity. The terms
“in” and “out” denote respectively whether the imidazole is bound within
the phenanthroline pocket or outside of the phenanthroline pocket, at
the distal site.
Assembly in Solution. Three possible modes of assembly can
be envisioned for (3Zn)2 and (4Zn)2 (Figure 1). These modes
differ by binding of imidazole either within the phenanthroline
strap (denoted as “in”) or on the distal, unhindered side of the
porphyrin (denoted “out”). Semiempirical and molecular me-
chanics calculations in the gas phase indicated that although the
inꢀout configuration is slightly preferred, there is little difference
(<3 kcal) between the relative heats of formation of the three
configurations (see the Supporting Information). Thus, in
theory, three forms of the dimer can be engaged in a configura-
tional equilibrium in solution.
As shown in Figure 2, the dimer (3Zn)2 exhibited a rather
simple set of 1H NMR resonances that can only emerge from a
unique and highly symmetrical arrangement. An inꢀout assem-
bly was therefore eliminated as a possible conformation in
solution. Compared to the free base porphyrin 3, the doublet
of the ortho (Ho) protons of the phenyl ring of the phenanthro-
line strap was shifted upfield from 6.68 ppm in the free base to
6.38 ppm in the zinc porphyrin. This shift, which results from a
πꢀπ interaction between the imidazole and the phenyl ring
bearing Ho, is characteristic of imidazole binding within the
phenanthroline strap.19 Only the inꢀin conformer accounts for
the upfield shift of the Ho peak.
a (i) (n-Bu)4NF, THF, 50 ꢀC; 3, 56%; 4, 63%; 7, 48%. (ii) Zn(OAc)2,
THF, 60 ꢀC, quantitative. (iii) (a) NBS, CHCl3, 0 ꢀC; (b) acetone;
100%; (c) 1,4-phenyldiboronic acid, Pd(PPh3)4 K2CO3, toluene,
MeOH, H2O, 60 ꢀC; 6, 65%.
linked by a phenyl spacer in the bis-porphyrin 7. In the presence
of zinc, after assembly at one end, the other end is still available to
continue the oligomerization/polymization process. Solubility of
the assemblies will be the size-limiting factor. Dissociation
studies, photophysical measurements, and atomic force micro-
scopy (AFM) images demonstrate the existence of 7Zn as self-
assembled filaments.
The apparent discrepancy between the rather imprecise
theoretical calculations and the preferred configuration observed
by NMR needs higher level calculations to be clarified, in view of
the small differences involved. Particularly intriguing is that an
X-ray structure of a phenanthroline-strapped zinc porphyr-
inꢀimidazole complex showed that the imidazole’s NH hydro-
gen formed a bifurcated hydrogen bond with both nitrogen
atoms of the phenanthroline.19 In the X-ray analysis, this
particular NH was positioned on the basis of the presence of
electronic density; therefore, in our calculations, the prediction
that the imidazole’s NH hydrogen prefers a single hydrogen
bond to only one of the phenanthroline’s nitrogen atoms may be
an artifact of the force field calculation. However, the preferred
inꢀin configuration observed by NMR is easily explained by the
formation of two hydrogen bonds that are highly favored by the
use of chloroform as solvent.
’ RESULTS AND DISCUSSION
Synthesis. Compounds 1 and 2 (Scheme 1) were prepared14
by combining two synthetic approaches that lead to soluble
phenanthroline-strapped porphyrins bearing long alkyl chains
and to self-complementary imidazolyl-substituted porphyrins.
Bromination15 of the unsubstituted meso position afforded 5
quantitatively. A final coupling of 2 equiv of 5 with 1,4-phenyldi-
boronic acid afforded the protected bis-porphyrin 7. Removal of
the CH2OCH2CH2SiMe3 (SEM) protecting groups on 6 gave
compound 7 in 48% yield.
1
Altogether, H NMR clearly indicates that the self-comple-
mentary dimer (3Zn)2 suffers some conformational mobility
restraints due to stacking of the phenyl groups of both phenylꢀ
ethynyl spacers. In the absence of zinc, the phenyl spacer
located at the porphyrin’s meso position in the free base
monomer showed broad signals at 8.4 and 7.6ꢀ7.8 ppm, which
suggest a precluded or slow free rotation around the imidazoleꢀ
porphyrin axis. After zinc insertion and formation of the inꢀin
dimer, the nearly face-to-face position of the two phenylꢀ
ethynyl groups (see Figure 2 or the Supporting Information) is
now fixed. The phenyl protons appear as two, well-defined sets of
Metalation of 7 with zinc acetate resulted in the precipitation
of the deep green zinc derivative 7Zn. Despite the presence of
two long alkyl chains, this compound was insoluble in chlori-
nated solvents, toluene, alcohols, dimethylformamide (DMF),
and dimethyl sulfoxide (DMSO). Compound 7Zn was partially
soluble in pyridine. For similar self-assembling systems, this type
of behavior was attributed to the formation of very large species
that precipitate at a critical size.16ꢀ18 Mass spectroscopy of dilute
samples of 7Zn showed no trace of nonmetalated porphyrin;
therefore, metalation was assumed to be quantitative.
6074
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082

Inorganic Chemistry
ARTICLE
Figure 2. 1H NMR spectra of 3 (bottom, 300 MHz) and 3Zn (top, 500 MHz) in CDCl3 (∼5 ꢁ 10ꢀ4 M). The phenyl protons of the phenyl-ethynyl
bridge are boxed for the free base 3.
in Supporting Information Figure S3). Consequently, on the basis
0
0 0
of the COSY NMR correlations, Hm and Hm were assigned to
the doublets at 6.98 and 6.78 ppm, respectively. The larger
00
00
upfield shifts of Ho and Hm are explained by molecular
modeling that clearly places these protons at the van der Waals
contact of the slightly offset neighboring phenyl linker (see
Figure 2 and the Supporting Information). Particularly close
00
00
proximities were calculated for Ho and Hm of one molecule
0 0
00
respectively with Hm and Ho of the second molecule (respective
distances of ca. 3.43 and 3.46 Å). All distances between other
protons on opposite molecules are greater than 4.5 Å.
The formation of only a dimeric species was confirmed by
DOSY 1H NMR for 5 ꢁ 10ꢀ4 M solutions of 3Zn and 4Zn in
CDCl3 (D = 300 μm2/s for 3Zn and 250 μm2/s for 4Zn, 298
K).14 These dimers are extremely stable with a self-association
constant of 109 Mꢀ1 in CHCl3 at 298 K. Dimers of 3Zn and 4Zn
assemble into J-aggregates in n-heptane solution at 10ꢀ5 M.14
The absorption spectrum of 4Zn dissolved in a minimum of
dichloromethane and then diluted with n-heptane is shown in
Figure 3. The spectra shown in Figure 3 were measured at
relatively high concentration (1 ꢁ 10ꢀ5 M) to favor aggregation;
nevertheless, the absorptions remain in the linear response domain
of the detector. The broad Soret band has a λmax at 443 nm with a
shoulder around 456 nm. The former absorption was attributed
to the dimer and the latter to J-aggregates. To dissociate the
dimers and aggregates, methanol was added (1 pipet drop or
∼12 000 equiv), causing the Soret band to split into three distinct
components with λmax at 428, 446, and 456 nm. The absorption
at 428 nm became more intense upon the addition of more MeOH
(∼36 000 equiv total), suggesting that this band corresponds to
monomers coordinated by MeOH. The lower energy absorptions at
446 and 456 nm were thus respectively attributed to dimeric
species and to aggregates. Assignment of these bands was con-
firmed by time-resolved photoluminescence studies (vide infra).
The inꢀin mode of coordination in (4Zn)2 was also assumed
for solution assembly of the bis(imidazolyl-porphyrin) 7Zn. Due
to its insolubility in noncoordinating solvents, 1H NMR (300 MHz)
of the bis-porphyrin 7Zn could not be obtained in chlorinated
solvents. However, 7Zn was soluble in pyridine, which is a
Figure 3. Absorption spectra of 4Zn (dissolved in CH2Cl2 and then
diluted with n-heptane to a concentration of 1 ꢁ 10ꢀ5 M) and changes
upon addition of MeOH: red line, ∼12 000 equiv of MeOH; blue line,
∼36 000 equiv of MeOH plus enough CH2Cl2 (DCM) to obtain a
homogeneous solution.
AX peaks at 8.34, 7.79, 6.98, and 6.78 ppm. Whereas the positions
of the first two doublets are nearly unchanged, the latter two
doublets are shifted upfield compared to the broad peaks in the
free base precursor. The COSY NMR spectrum showed one
correlation for the doublets at 8.34 and 6.98 ppm and another for
the doublets at 7.79 and 6.78 ppm (see the Supporting In-
formation). Thus, the first two doublets correspond to protons
on the same side of the phenyl ring and the latter two correspond
to the protons on the opposite side of the phenyl ring. The peaks
0
00
at 8.34 and 7.79 ppm were attributed to Ho and Ho on the basis
of a correlation in the NOESY spectrum between both of these
doublets and the β1 protons of the porphyrin ring (see the Sup-
porting Information). The peak that is more upfield, at 7.79 ppm,
0 0
was assigned to Ho because this proton is shielded by the electronic
density of the phenyl ring of the second molecule in the dimer
(see schematic representation in Figure 3 and the molecular model
6075
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082

Inorganic Chemistry
ARTICLE
298 K), which corresponds to a monomer of 7Zn end-capped
with 4Zn, i.e., 4Zn(7Zn)4Zn.
In hopes of favoring oligomers shorter than 4Zn(7Zn)74Zn,
which might be observed by DOSY NMR, more 4Zn was added,
for a total of 22.5% of 4Zn with respect to 7Zn. The 1H NMR
(CDCl3) spectrum of this solution, which corresponds to 4Zn-
(7Zn)44Zn on the basis of stoichiometry, was similar to that of
4Zn(7Zn)74Zn. Unfortunately, precipitation began within 20
min of preparing the CDCl3 solution of 4Zn(7Zn)44Zn. Thus, a
DOSY experiment was not carried out. More dilute solutions
remained homogeneous for longer periods of time and were used
for photophysical studies.
Photophysics. To verify the assignment of the absorption
bands of 3Zn or 4Zn, time-resolved photoluminescence studies
were carried out on solutions of 3Zn or 4Zn that were dissolved
in a minimum amount of CH2Cl2 and then diluted with n-
heptane to favor the formation of J-aggregates. The behavior of
both compounds was similar; therefore, results representative of
both compounds are described only for 4Zn. Time-resolved
spectra and decay-associated emission spectra (DAES) of 3Zn
are available as Supporting Information.
Upon excitation at 440 nm, two emission bands were observed
at 620 and 672 nm (Figure 5a). The emission dynamics were best
fit using the decay-associated emission spectra with two spectral
components with different associated decay times (Figure 5b).
The slow component decayed with τ1 = 950 ps and the fast one
with τ2 = 250 ps. On the basis of other time-resolved
investigations,22,23 the former should correspond to the fluores-
cence of smaller species and the latter to fluorescence of larger
aggregates. These decay times have to be compared to the
reported 2.2 ns of an imidazole inclusion complex with a simple
phenanthroline-strapped zinc porphyrin in pure CH2Cl2.24 In
4Zn, the observed slow component is more than two times faster
than that of the simple phenanthroline-strapped zinc porphyrin;
therefore, the 950 ps decay time is more likely due to a dimeric
form of 4Zn. The two spectral components of the DAES
(Figure 5b) for these two decay times are of nearly equal
amplitude, meaning that both dimers and oligomers/aggregates
contribute equally to the fluorescence. These findings seem
reasonable given that both dimers and oligomers/aggregates
absorb at the excitation wavelength of 440 nm.
Upon excitation at 445 nm, the emission maxima remained the
same at 620 and 670 nm (Figure 5c). A biexponential decay best
fit the time-resolved spectrum. The decay times (see Table 1) were
similar to those for excitation at 440 nm and were assigned in the
same manner. In the DAES (Figure 5d), the spectrum correspond-
ing to the fast component (aggregates) is more intense than that of
the slow component (dimer). The reason for this difference is that
at 445 nm the aggregates absorb more than the dimers and thus
the aggregates’ emission is more intense upon excitation at this
wavelength. This finding of an amplified intensity supports the
assignment of the red-shifted absorption peak to aggregates.
The existence of aggregates of 4Zn was further verified by
adding methanol and exciting at 445 nm, where the aggregates
absorb more than the dimers. Three pipet drops (∼36 000
equiv) of MeOH were added to the solution, and time-resolved
fluorescence spectrum was recorded (Figure 5e). Again, two-
component DAES gave the best fit. For excitation at 445 nm, a
smaller difference in amplitude between the two components of
the DAES (Figure 5f) was observed than in the situation
displayed in Figure 5d. Compared to the spectrum of the solution
containing no methanol (Figure 5d), the amplitude increased for
Figure 4. Dissociation of polymers of 7Zn by addition of 4Zn.
competitor for axial binding to zinc. The 1H NMR (500 MHz)
spectrum of 7Zn in pyridine-d5 showed sharp peaks (see Sup-
porting Information) and only monomeric 7Zn was observed in
DOSY experiments of this sample.
To establish the existence of 7Zn as self-assembled polymers,
an experiment was carried out to dissociate the presumably long
species into smaller segments by adding 4Zn as an end-capping
group, as schematically depicted in Figure 4. Bis(porphyrin) 7Zn
was dissolved in pyridine, and a small quantity (0.05 or 0.025
equiv) of 4Zn was added. The solution was then equilibrated at
80 ꢀC for 30 min. After removal of the pyridine under reduced
pressure and vacuum drying for 12ꢀ15 h, 0.4 mL of CDCl3 was
added. This process was repeated until the solid completely
dissolved in the 0.4 mL of CDCl3. A soluble oligomer was
obtained after the addition of 0.125 equiv of 4Zn. The stoichi-
ometry reached at this point for the dissolved species was
theoretically seven molecules of 7Zn to two molecules of 4Zn.
The 1H MNR (300 MHz, CDCl3) of this mixture of 7Zn/4Zn
showed numerous broad peaks as well as some sharper peaks that
could be assigned (see the Supporting Information). By compar-
ing the integration of the meso porphyrin peak (10.23 ppm) of
4Zn to that of the imidazole’s H5 (4.31 ppm) present in both
4Zn and 7Zn, the ratio of 7Zn/4Zn was estimated to be seven
molecules of 7Zn for two molecules of 4Zn. Thus, the NMR
estimation was in agreement with the initial stoichiometry, and
the average distribution of end-capped oligomers observed in
solution was centered around scaffolds with formula of 4Zn-
(7Zn)74Zn comprising 16 porphyrins. As shown by the spec-
trum of the soluble material, residual pyridine seemed to be
absent from the solution. Although traces of pyridine may be
considered, in dichloromethane, the large difference between the
log Kim = 6.1 and log Kpy = 3.3 on a phenanthroline-strapped zinc
porphyrin20,21 renders the binding of pyridine ineffective at
NMR concentrations.
The length of these species could not be confirmed by DOSY
experiments, however, because initially homogeneous solutions
of 4Zn(7Zn)74Zn in CDCl3 precipitated within 22 h. Consider-
ing that the assembly is an equilibrium process, the solution of
7Zn/4Zn does not exist as only end-capped hexamers or
heptamers. Rather, oligomers or polymers of various lengths
coexist in solution and species longer than approximately 16
porphyrins precipitate, which displaces the equilibrium toward
the formation of insoluble oligomers 4Zn(7Zn)mþn4Zn and
shorter, soluble species (see Supporting Information for a schematic
representation). The DOSY experiment did show the existence
of a species with a diffusion coefficient of 120 μm2/s (CDCl3,
6076
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082

Inorganic Chemistry
ARTICLE
Figure 5. Time-resolved spectra (a, c, e) and their corresponding DAES (b, d, f) of 4Zn dissolved in a minimum of CH2Cl2 and diluted with n-heptane:
(a and b) λex = 440 nm; (c and d) λex = 445 nm; (e and f) in the presence of ∼36 000 equiv of MeOH at λex = 445 nm.
Table 1. Emission Maxima and Fluorescence Decay times
emission maxima of 4Zn(7Zn)44Zn were considerably red-
shifted compared to those of 4Zn. Kobuke observed similar
red-shifted emission bands in self-assembled imidazolyl zinc
porphyrins and attributed the red-shift to exciton interactions.10
Two samples of different concentrations in dichloromethaneꢀ
cyclohexane (Figure 6) were studied. The exact molarity of each
sample was unknown; however, on the basis of the absorbance of
the Soret band at 434 nm, one sample was five times more con-
centrated than the other. Time-resolved spectra (Figure 6a,c) were
best fit by the DAES method using two different decay constants. In
the more concentrated sample, decay times of 5.16 ns and 400 ps
were observed (Figure 6d). These decay times are much longer than
those of covalently linked multiporphyrin arrays. In general, such a
significant increase can be ascribed to the delocalization of the
excitation energy over several chromophores.25 Small to moderate
increases have been reported for small noncovalent cyclic species or
geometrically less defined covalent species. The significant changes
observed in this work suggest a more efficient delocalization along
either longer or more linear species for example. Upon dilution, the
shorter and longer decay times remained approximately constant
(Figure 6b).
λex (nm)
τ1 (ps)
τ2 (ps)
3Zn a
440
445
445
440
445
445
700
700
700
950
1100
900
150
200
200
250
250
300
3Zn a þ MeOH b
4Zn a
4Zn a þ MeOH b
4Zn(7Zn)44Zn c
concentrated
dilute
434
434
5160
5340
400
400
a Dissolved in CH2Cl2 and diluted with n-heptane. ∼36 000 equiv
b
of MeOH. c Dissolved in CH2Cl2 and diluted with cyclohexane.
the longer decay time’s DAES (dimer component). Such a pre-
sence of more dimers is expected because the addition of methanol
dissociates aggregates into dimers.
Time-resolved fluorescence measurements were also per-
formed on coordination oligomers of 7Zn end-capped with
22.5% of 4Zn, which corresponds to oligomers with an average
formula of 4Zn(7Zn)44Zn. Upon excitation at 434 nm, the
As for porphyrins 3Zn and 4Zn, the shorter decay time (400 ps)
of 4Zn(7Zn)44Zn was attributed to longer oligomeric species.
6077
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082

Inorganic Chemistry
ARTICLE
Figure 6. Time-resolved spectra (a, c) and their corresponding DAES (b, d) of 4Zn(7Zn)44Zn in a dilute sample (a þ b) and a sample five-times more
concentrated (c þ d). The compound was dissolved in dichloromethane and diluted with cyclohexane.
Figure 7. AFM images of 7Zn on mica: (a) 1 μM solution of 7Zn in THF drop-cast onto freshly cleaved mica (vertical scale = 6 nm); (b and c) 1 μM
solution of 7Zn in THF drop-cast onto mica that was first covered with a thin film of heptane [(b) vertical scale = 6 nm]; (c) close view of laterally packed
chains observed on the same sample in an area of higher surface coverage density; vertical scale = 5 nm; (d) 1 μM solution of 7Zn in pyridine drop-cast
onto mica (vertical scale = 5 nm).
The amplitude of both components was nearly equal in the
concentrated sample. In the more dilute sample, the associated
spectrum of the longer lived component was more intense than
that of the shorter lived component. The longer lived component
was assigned to emission from the end caps. Thus, as intuitively
expected, 7Zn tends to form longer oligomers at higher con-
centration. Upon dilution, these longer species are dissociated
into shorter segments.
AFM Studies of Assembly in the Solid State. Atomic force
microscopy (AFM) studies were carried out to investigate the
self-assembly of 7Zn on surfaces. A rapid screening of substrates,
solvents, and deposition conditions was initially conducted.
Typically, 1 μM solutions of 7Zn in CHCl3, toluene, heptane,
tetrahydrofuran (THF), or pyridine were prepared by diluting a
100 μM stock solution of 7Zn in pyridine with the desired
solvent. The dilute solutions were deposited on mica, HOPG,
native silicon (SiO2), or molybdenum disulfide (MoS2) sub-
strates by dip-coating or drop-casting. On HOPG, MoS2, and
SiO2 substrates, ill-structured aggregates or filmlike deposits
were mainly observed under the conditions tested (not shown).
THF and pyridine solutions of 7Zn led to the formation of
filament-like structures on mica, as shown in Figure 7.
Drop-casting of this THF solution onto freshly cleaved
mica led principally to the dispersion of 7Zn into nonorga-
nized aggregates on the surface, as shown in Figure 7a.
However, beside the presence of some ill-defined aggregates,
elongated structures were observed when the THF solution
of 7Zn was drop-cast onto mica that had first been covered
with a thin film of heptane (Figure 7b,c and the Supporting
Information).
6078
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082

Inorganic Chemistry
ARTICLE
Figure 8. Space filling representation of optimized models of dimers of 7Zn (in blue and red) including two 4Zn end-capping units (in gray): (a and b)
view along the axis of the porphyrin assembly; (c and d) bird’s eye view of the assemblies. Specific dimensions are indicated for two possible orientations
of the assembly on a mica surface. Model optimized on MOPAC2009 26 using the PM6 level.27
The drop-casting deposition method resulted in an inhomo-
geneous dispersion of the material on the sample substrate’s
surface. In areas presenting a low surface coverage, isolated
filament-like structures of homogeneous height and width were
clearly observed (Supporting Information Figure S12a,b) and
were attributed to isolated self-assembled porphyrin chains.
These structures were sometimes decorated with small aggre-
gates, which gave them an irregular appearance. In some cases the
observation of loops and twisted structures suggested the folding
of cohesive structures, i.e., chains of coordination polymers
(Figure 7b and Supporting Information Figure S12a,b). In areas
presenting a higher surface coverage, more complex patterns
were observed; however, the filamentous nature of the under-
lying structures was still distinguishable even in large assemblies
(Figure 7b,c and Supporting Information Figure S12c). The
reproducibility of these observations was confirmed by exploring
three different samples at multiple positions. Precoating the mica
with heptane was found to be essential to promote the formation
of such filamentous structures. It is proposed that heptane acts as
a temporary passivation layer preventing the direct contact of the
monomeric/oligomeric species initially present in the solution
with the mica surface, thus permitting growth of the porphyrin
chains before they are trapped on the surface. Further investiga-
tions are necessary to fully understand the role of the heptane
layer in the self-assembly process.
is measured from the top of one phenanthroline strap to the
bottom of the phenanthroline strap in a second porphyrin. In this
case, the porphyrins would lie parallel to the mica surface. The
second calculated distance that corresponds to the measured
height is the 1.6 nm width of a porphyrin between the outermost
carbon atoms of the meso-phenyl groups (Figure 8b). In this
scenario, the porphyrins would lie on their sides, perpendicular to
the surface, as previously hypothesized for assemblies of dimers
of 3Zn and 4Zn on mica.14 The precision of the measurements
does not permit discrimination between the diverse orientations
possible for the porphyrin chains on the surface, and the
variations observed can indicate the presence of different orien-
tations of these chains on the mica surface.
In the area exhibiting a high density of 7Zn, lateral association
of some filament-like structures was observed, as shown in
Figure 7c (circled area). In the more closely, laterally packed
assemblies, a center to center distance of 6.3 ( 1.2 nm was
measured between the subsequent filaments. This value is close
to the theoretical widths calculated for fully extended C18 alkyl
chains, i.e., 5.2ꢀ5.6 nm, as shown in Figure 8. Thus, no or very
little interdigitation of the C18 alkyl chains is observed and the
linear substructures most likely correspond to single strands of
(7Zn)n chains assembled side-by-side.
Samples prepared by drop-casting of a micromolar pyridine
solution of 7Zn onto mica were more heterogeneous, presenting
mostly dense films in which fibrous substructures could be
distinguished (see Supporting Information). In one case isolated
linear, filament-like structures were seen (Figure 7d). These
structures appeared to be lying on top of a poorly organized layer
of material that covered the mica surface and made it difficult to
measure the height of the filament-like structures. The micro-
molar pyridine solution probably contained mostly monomeric
or small oligomeric species due to the coordinating nature of this
solvent. Therefore, the observed structures are assumed to have
The isolated filaments were 9ꢀ16 nm wide and 1.62 ( 0.52 nm
high. Taking into account the radius curvature (e10 nm) of the
AFM probe, the observed widths were consistent with single
strands of porphyrin assemblies. The height of the filaments does
not correspond to the previously calculated and observed
1.25 nm height of the phenanthroline-strapped porphyrin from
the porphyrin ring to the top of the phenanthroline strap.14 The
measured height is compatible with two possible calculated
widths, as shown in Figure 8a,b. In Figure 8a, the 1.8 nm distance
6079
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082

Inorganic Chemistry
ARTICLE
formed as the pyridine evaporated. The average, uncorrected
width of the filaments was 10ꢀ15 nm, which was consistent with
single stands of self-assembled 7Zn, as described above for the
THF solutions. These filaments were as long as 480 nm, which
translates to 190 repeating bis(porphyrin) units for a total of 380
porphyrins. Further experiments are necessary to refine the
deposition method to favor the formation of these assemblies.
during the measurement, the solutions were continuously pumped
through a quartz cuvette in a closed loop. The following three-dimen-
sional function,
n
i
Iðλ, tÞ ¼
AiðλÞ eꢀt=τ
∑
i¼1
was used to fit the recorded data. I(λ,t) is the wavelength λ and time t
dependent experimental fluorescence intensity; Ai and τi are the weights
and the global decay times, respectively. This applied fitting technique
allows the observed fluorescence intensity to be decomposed into a sum
of wavelength-dependent components Ai that each decay with wave-
length-independent lifetimes τi. The determined set of curves Ai(λ),
called the decay-associated emission spectra, can be graphically pre-
sented to show the extracted decay times τi for each amplitude Ai. To
compare the quality of the different DAES fits, a χ2-value was calculated
using the following formula:
’ CONCLUSIONS
The self-complementary imidazolyl zinc porphyrins 3Zn and
4Zn exist as dimers and/or J-aggregates, as demonstrated by
absorption spectroscopy and time-resolved photoluminescence
measurements. Assembly into linear structures was achieved with
the bis(imidazolyl zinc porphyrin) 7Zn. In organic halogenated
solvents, 7Zn self-assembles into insoluble oligomers or poly-
mers that are at least 18 porphyrins long. Linear filaments or
fibers up to 480 nm long were observed by near-field microscopy,
starting from dilute pyridine solutions, in which only isolated
species were present. This work demonstrates the potential of
noncovalent interactions to assemble appropriately designed
building blocks into long, linear structures.
1
1
2
χ2ðfitÞ ¼
2 ð
ðIsignalðn, mÞ ꢀ Ifitðn, mÞÞ Þ
∑
n
maxmmax
ðIfit,max
Þ
all pixelsðn,mÞ
where (n,m) is the index for the pixel map. χ2 is normalized to the size of
the pixel map (nmaxmmax) and to the maximum value of the fit squared
(Ifit,max)2. All presented fits had a χ2-value below 10ꢀ3
.
Steady-state absorption spectra were recorded with a Varian-Cary
500 spectrophotometer.
’ EXPERIMENTAL SECTION
Calculations. Molecular mechanics calculations with the MMþ
force field, semiempirical calculations at the PM3 level were performed
within the HyperChem program package,28 and semiempirical calcula-
tions at the PM6 level27 (Figure 8) were peformed with MOPAC2009.26
Geometry optimizations were done by alternating the molecular me-
chanics and semiempirical calculations, the latter providing the atomic
charges used by the former without cutoffs. Convergence was usually
attained for different dimers after 24 h on an IBM computer equipped
with a Pentium processor and an 850 GHz clock.
Atomic Force Microscopy. AFM investigations were carried out
with a commercial Nanoscope IIIa setup that controlled a Dimension
3100 scan-head (Digital Instruments, Santa Barbara, CA). All experi-
ments were carried out in Tapping Mode under ambiant conditions.
N-doped silicium cantilevers were used for surface imaging
(NanoAndMore Gmbh): NanoSensors PointProbePlus NCH (typical
tip radius, <10 nm; force constant, 42 N/m; resonance frequency, 330
kHz; back side reflex coating). For HOPG, SPI SP2 or SPI SP1 grade
plates (1 ꢁ 1 cm2) were used as supports and freshly cleaved with
adhesive tape just before to use. For mica, synthetic fluorophlogopite
plates (1 ꢁ 0.5 cm2) were used. They were cut from a 1 ꢁ 1 cm2 plate
(with scissors) and cleaved with adhesive tape just before use. Images
were generated using NanoScope software (version 6.14, Digital In-
struments). A third-order leveling was applied to correct the pictures for
the background. No other treatments were applied to the data before
analysis.
Photophysical Studies. The tunable Ti:sapphire laser (Coherent,
Mira) delivering sub-150 fs pulses with a repetition rate of 76 MHz was
used for the excitation of the samples. To obtain an excitation within the
Soret band of the investigated porphyrins (425ꢀ445 nm), the light was
frequency-doubled by using beta barium borate (BBO) crystal. A
combination of a spectrometer (Jobin-Yvon HR460, 100 lines/mm
grating) and a streak camera (Hamamatsu C5680) was used to record the
fluorescence from the samples. The streak camera was operated in the
analog integration mode. The temporal and spectral resolutions of
fluorescence spectra were 8 ps and 2 nm, respectively. The polarizer
oriented at an angle of 54.7ꢀ relative to the polarization of the excitation
was utilized to compensate for the effect of molecular rotation on the
fluorescence decay. Scattered light from the excitation beam was blocked
by a color glass filter (SCHOTT OG570). To minimize photodegradation
’ SYNTHETIC PROCEDURES
General Method. Compounds 1ꢀ4, 3Zn, and 4Zn were prepared
according to ref 14. Reagents and solvents of reagent grade were
purchased and used without further purification. Anhydrous Na2SO4
was used as a drying agent after aqueous workup. Evaporation and
concentration in vacuo were carried out at H2O-aspirator pressure.
Column chromatography was performed with alumina from Merck
(aluminum oxide 90 standardized, activity IIꢀIII according to
Brockmann). Mass spectra were performed by le Service de Spectro-
mꢀetrie de Masse de l’Institut de Chimie, Universitꢀe Louis Pasteur.
Elemental analyses were performed by le Service d’Analyse Elementaire
de l’Institut Universitaire de Technologie, Strasbourg, Sud, and le
Service d’Analyse Elementaire de l’Institut de Chimie, Universitꢀe Louis
Pasteur. 1H NMR spectra were recorded on Bruker Advance 300 (300
MHz) or 500 (500 MHz) spectrometers. DOSY spectra were recorded
on a Bruker Avance 500 (500 MHz) spectrometer. Chemical shifts were
determined by taking the solvent as a reference: CHCl3 (7.26 ppm);
pyridine-d5 (7.19, 7.55, 8.71 ppm).
Synthesis of 5. To a solution of 2 (94 mg, 0.06 mmol) in THF
(10 mL), zinc(II) acetate (130 mg, 0.6 mmol) was added. The mixture
was stirred at 60 ꢀC for 2 h. After evaporation of the solvent, the residue
was dissolved in CHCl3/H2O medium. The organic layer was carefully
washed with water. After evaporation of CHCl3, the metalated porphy-
rin was dried by azeotrope distillation with toluene. Crude product
(quantitative yield) was used without further purification in the next
step. The stoichiometry of the brominating agent was calculated on the
basis of the initial quantity of free base porphyrin 2.
To a solution of crude metalated 2 in chloroform (20 mL) was added
N-bromosuccinimide (11 mg, 0.06 mmol). After 20 min at 0 ꢀC, the
reaction was quenched with acetone (5 mL). After evaporation of the
solvent, the mixture was dissolved in CH2Cl2 (200 mL) and trifluor-
oacetic acid (10 mL) was added. The solution was stirred for 10 min at
room temperature and then washed twice with 2 M Na2CO3(aq). The
organic layer was dried over Na2SO4, and the solvent was removed. The
mixture of porphyrins was purified by column chromatography over Al2O3
in CH2Cl2/cyclohexane (1/1) to afford 5 (97 mg, 0.06 mmol, 100%) as
a glassy purple solid. mp > 250 ꢀC. 1H NMR (CDCl3, 300 MHz): δ 9.61
6080
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082

Inorganic Chemistry
ARTICLE
(d, J = 4.8 Hz, 2H), 8.97 (d, J = 4.8 Hz, 2H), 8.67 (d, J = 4.8 Hz, 2H), 8.59
(d, J = 4.8 Hz, 2H), 8.47 (br s, 1H), 8.20 (d, J = 2.6 Hz, 2H), 7.98 (d, J =
8.4 Hz, 2H), 7.81 (br s, 3H), 7.77 (d, J = 8.5 Hz, 2H), 7.54 (d, J = 8.4 Hz,
2H), 7.51 (s, 2H), 7.47 (dd, J1 = 2.6 Hz, J2 = 8.5 Hz, 2H), 7.20 (d, J =
1.3 Hz, 1H), 7.16 (d, J = 1.3 Hz, 1H), 6.78 (d, J = 8.4 Hz, 4H), 6.45 (d, J =
8.4 Hz, 4H), 5.54 (s, 2H), 4.28 (m, 4H), 3.67 (t, J = 7.7 Hz, 2H), 1.99 (m,
4H), 1.2ꢀ1.5 (m, 60H), 1.01 (t, J = 7.7 Hz, 2H), 0.88 (m, 6H), 0.01
(s, 9H), ꢀ2.42 (s, 2H).
a schematic representation of the equilibrium process of 4Zn-
(7Zn)mþn4Zn, additional AFM images of 7Zn on mica. This
material is available free of charge via the Internet at http://
pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: jweiss@unistra.fr (J.W.), jwytko@unistra.fr (J.A.W.).
Synthesis of 6. To a suspension of 5 (95 mg, 57 μmol), 1,4-di-
phenylboronic acid (4.35 mg, 26 μmol), and K2CO3 (82 mg, 0.6 mmol)
in toluene (15 mL) were added water (0.3 mL) and methanol (2 mL).
The mixture was Ar flushed (15 min); then Pd(PPh3)4 (5 mol %) was
introduced. After degassing by bubbling the mixture with Ar for another
15 min, the reaction mixture was stirred at 60 ꢀC for 6 h under argon.
The organic phase was washed with saturated NH4Cl(aq) (2 ꢁ 50 mL)
and water (2 ꢁ 50 mL) and then dried over Na2SO4, filtered, and
evaporated to dryness to afford a dark residue. Purification by column
chromatography (Al2O3, C6H12/CH2Cl2, 1/1) was conducted in the
dark to yield the SEM-protected derivative 6 (57 mg, 17 μmol, 65%) as a
violet solid (second eluted product). The compound was used without
further purification for the deprotection step that follows. mp > 250 ꢀC.
1H NMR (CDCl3, 300 MHz): δ 9.18 (d, J = 4.5 Hz, 4H), 8.99 (d, J = 4.5
Hz, 4H), 8.87 (d, J = 4.5 Hz, 4H), 8.75 (d, J = 4.5 Hz, 4H), 8.55 (br s,
2H), 8.29 (d, J = 2.6 Hz, 4H), 8.20 (br s, 2H), 7.97 (d, J = 8.4 Hz, 4H),
7.88 (br s, 8H), 7.82 (d, J = 8.5 Hz, 4H), 7.48ꢀ7.56 (m, 12H), 7.22
(s, 2H), 7.19 (s, 2H), 6.83 (d, J = 8.2 Hz, 8H), 6.55 (d, J = 8.2 Hz, 8H),
5.59 (s, 4H), 4.29 (m, 8H), 3.71 (t, J = 8 Hz, 4H), 1.98 (m, 8H), 1.2ꢀ1.6
(m, 120H), 1.02 (t, J = 8 Hz, 8H), 0.87 (m, 12H), 0.03 (s, 18H), ꢀ2.70
(s, 4H).
Synthesis of 7. A solution of 6 (57 mg, 17 μmol) and (n-Bu4)NF
(0.1 mL of a 1 M solution in THF, 0.1 mmol) in 6 mL of THF was heated
at 50 ꢀC, in the dark, for 4 h and then poured into water. The precipitate
was washed with water and dissolved in CH2Cl2. The organic layer was
washed again with water. Solvent was removed under vacuum, and the
resulting residue was dried by azeotrope distillation. Purification by
column chromatography (Al2O3, C6H12/CH2Cl2, 1/1 to 0/1) was
conducted in the dark to yield 7 (25 mg, 8.2 μmol, 48%) as a glassy
purple solid (third eluted product). mp > 250 ꢀC. 1H NMR (CDCl3, 300
MHz): δ 9.18 (d, J = 4.5 Hz, 4H), 8.99 (d, J = 4.5 Hz, 4H), 8.86 (d, J = 4.5
Hz, 4H), 8.73 (d, J = 4.5 Hz, 4H), 8.51 (br s, 2H), 8.28 (d, J = 2.6 Hz,
4H), 8.18 (br s, 2H), 7.94 (d, J = 8.5 Hz, 4H), 7.84 (br s, 2H), 7.77 (d, J =
8.6 Hz, 4H), 7.66 (br s, 6H), 7.34ꢀ7.56 (m, 16H), 6.74 (d, J = 8.4 Hz,
8H), 6.53 (d, J = 8.4 Hz, 8H), 4.26 (m, 8H), 1.96 (m, 8H), 1.56 (m, 8H),
1.2ꢀ1.5 (m, 112H), 0.83 (m, 12 H), ꢀ2.34 (s, 4H). MS. Calcd for
Mþ: 3061. Found: 3063.650. Anal. Calcd for C212H226N16O4 þ 2H2O þ
2CH2Cl2: C, 78.61; H, 7.39; N, 6.38. Found: C, 78.34; H, 7.24; N, 6.00.
Synthesis of 7Zn. Zinc acetate (7.1 mg, 33 μmol) was added to a
solution of 7 (10 mg, 3.3 μmol) in 5 mL of THF. The mixture was stirred
at 50 ꢀC under argon. A solid precipitated immediately. After 2 h, the
mixture was poured into water. The precipitate was collected by
filtration and washed thoroughly with water and then with methanol
to afford 7Zn (8 mg, 2.5 μmol, 76%) as an insoluble green material. MS
MALDI TOF. Calcd for Mþ: 3183.62. Found: 3183.11.
Present Addresses
‡Department of Chemistry and Biochemistry, Arizona State
University, P. O. Box 871604, Tempe, AZ 85287-1604, USA.
^Department of Chemistry, University of Saskatchewan, Saska-
toon, SK S7N 5C9, Canada. On leave from the Faculty of Applied
Physics and Mathematics, Gdanꢀsk University of Technology,
80-233 Gdanꢀsk, Poland.
’ ACKNOWLEDGMENT
We thank Prof. Jean-Pierre Bucher for providing access to an
AFM instrument. M.K. thanks the Rꢀegion Alsace and the CNRS
for a Ph.D. fellowship. This work was supported by the CNRS
(Grant ACI-NX001), the Research Council of the Universitꢀe
Louis Pasteur (now called the Universitꢀe de Strasbourg), the
DFG Research Center for Functional Nanostructures (CFN)
Karlsruhe (Project C3.5), and a grant from the Ministry of
Science, Research and the Arts of Baden-W€urttemberg (Grant
No. Az. 7713.14-300). J.C. thanks the Karlsruhe School of Optics
and Photonics (KSOP) for financial support.
’ REFERENCES
(1) (a) Wytko, J. A.; Weiss, J. In N-4 Macrocyclic Metal Complexes;
Zagal, J. H., Bedioui, F., Dodelet, J. P., Eds.; Springer: Heidelberg,
Germany, 2006; pp 603ꢀ724. (b) Harvey, P. D. In The Porphyrin
Handbook II; Kadish, K. M., Smith, K. M., Guilard, R., Eds.; Academic
Press: San Diego, CA, 2003; Vol. 18, Chapter 113. (c) Wolffs, M.;
Hoeben, F. J. M.; Beckers, E. H. A.; Schenning, A. P. H. J.; Meijer, E. W. J.
Am. Chem. Soc. 2005, 127, 13484. (d) Morisue, M.; Yamatsu, S.; Haruta,
N.; Kobuke, Y. Chem.—Eur. J. 2005, 11, 5563. (e) Balaban, T. S. Acc.
Chem. Res. 2005, 38, 612. (f) Lo, P.-C.; Leng, X.; Ng, D. K. P. Coord.
Chem. Rev. 2007, 251, 2334.
(2) (a) Maeda, C.; Kamada, T.; Aratani, N.; Osuka, A. Coord. Chem.
Rev. 2007, 251, 2743. (b) Imamura, T.; Fukushima, K. Coord. Chem. Rev.
2000, 198, 133. (c) Balaban, T. S. In Handbook of Porphyrin Science,
Vol. 1; Kadish, K. M., Smith, K. M., Guilard, R., Eds.; World
Scientific: Singapore, 2010; Vol. 1, Chapter 3.
(3) See for example:(a) Ozawa, H.; Tanaka, H.; Kawao, M.; Uno, S.;
Nakazato, K. Chem. Commun. (Cambridge, U. K.) 2009, 7411. (b) Xiao,
S.; Myers, M.; Miao, Q.; Sanaur, S.; Pang, K.; Steigerwald, M. L.;
Nuckolls, C. Angew. Chem., Int. Ed. 2005, 44, 7390. (c) Zang, L.; Che,
Y.; Moore, J. S. Acc. Chem. Res. 2008, 41, 1596.
(4) Sugimoto, T.; Suzuki, T.; Shinkai, S.; Sada, K. J. Am. Chem. Soc.
2007, 129, 270.
(5) See for example: (a) Johnson, R. S.; Yamazaki, T.; Kovalenko, A.;
Fenniri, H. J. Am. Chem. Soc. 2007, 129, 5735. (b) Jonkheijm, P.;
Hoeben, F. J. M.; Kleppinger, R.; van Herrikhuyzen, J.; Schenning, A. P.
H. J.; Meijer, E. W. J. Am. Chem. Soc. 2003, 125, 15941.
(6) Li, G.; Wang, T.; Schultz, A.; Bhosale, S.; Lauer, M.; Espindola,
P.; Heinzeb, J.; Furhop, J.-H. Chem. Commun. (Cambridge, U. K.)
2004, 552.
(7) Kawao, M.; Ozawa, M.; Tanaka, H.; Ogawa, T. Thin Solid Films
2006, 499, 23.
(8) (a) Yeats, A. L.; Schwab, A. D.; Massare, B.; Johnston, D. E.;
Johnson, A. T.; de Paula, J. C.; Smith, W. F. J. Phys. Chem. C 2008,
’ ASSOCIATED CONTENT
S
Supporting Information. Figures showing molecular
b
modeling of dimer (3Zn)2 or (4Zn)2 without the alkyl chains,
stick models of conformations of the outꢀout dimer of 3Zn or
4Zn, selected atomic distances in the model of the inꢀin dimer
(3Zn)2, COSY and NOESY NMR spectra of (3Zn)2, time-
resolved fluorescence spectra and the corresponding DAES of
1
3Zn, absorption spectrum of 4Zn(7Zn)44Zn, H NMR spec-
trum of 7Zn in pyridine-d5 and of 4Zn(7Zn)74Zn in CDCl3,
6081
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082

Inorganic Chemistry
ARTICLE
112, 2170, and references therein. (b) Ji, H.-X.; Hua, J.-S.; Wan, L.-J.
Chem. Commun. (Cambridge, U. K.) 2008, 2653.
(9) Aratani, N.; Osuka, A.; Kim, Y. H.; Jeong, D. H.; Kim, D. Angew.
Chem., Int. Ed. 2000, 39, 1458.
(10) Ogawa, K.; Kobuke, Y. Angew. Chem., Int. Ed. 2000, 39, 4070.
(11) Aratani, N.; Takagi, A.; Yanagawa, Y.; Matsumoto, T.; Kawai,
T.; Yoon, Z. S.; Kim, D.; Osuka, A. Chem.—Eur. J. 2005, 11, 3389.
(12) Satake, A.; Fujita, M.; Kurimoto, Y.; Kobuke, Y. Chem. Com-
mun. (Cambridge, U. K.) 2009, 1231.
(13) Koepf, M.; Trabolsi, A.; Elhabiri, M.; Wytko, J. A.; Paul, D.;
Albrecht-Gary, A. M.; Weiss, J. Org. Lett. 2005, 7, 1279.
(14) Koepf, M.; Wytko, J. A.; Bucher, J.-P.; Weiss, J. J. Am. Chem. Soc.
2008, 130, 9994.
(15) Nudy, L. R.; Hutchinson, G.; Schieber, C.; Longo, F. R.
Tetrahedron 1984, 40, 2359.
(16) Michelsen, U.; Hunter, C. A. Angew. Chem., Int. Ed. 2000,
39, 764.
(17) Nomoto, A.; Kobuke, Y. Chem. Commun. (Cambridge, U. K.)
2002, 1104.
(18) Nomoto, A. H.; Mitsuoka, H.; Oseki, H.; Kobuke, Y. Chem.
Commun. (Cambridge, U. K.) 2003, 1074.
(19) Froidevaux, J.; Ochsenbein, P.; Bonin, M.; Schenk, K.; Maltese,
P.; Gisselbrecht, J.-P.; Weiss, J. J. Am. Chem. Soc. 1997, 119, 12362.
(20) Paul, D.; Melin, F.; Hirtz, C.; Wytko, J.; Ochsenbein, P.; Bonin,
M.; Schenk, K.; Maltese, P.; Weiss, J. Inorg. Chem. 2003, 42, 3779.
(21) Froidevaux, J. Ph.D. Thesis, Universitꢀe Louis Pasteur, 1998.
(22) Huijser, A.; Marek, P. L.; Savenije, T. J.; Siebbeles, L. D. A.;
Scherer, T.; Hauschild, R.; Szmytkowski, J.; Kalt, H.; Hahn, H.; Balaban,
T. S. J. Phys. Chem. C 2007, 111, 11726.
(23) Yoon, M.-C.; Cho, S.; Kim, P.; Hori, T.; Aratani, N.; Osuka, A.;
Kim, D. J. Phys. Chem. B 2009, 113, 15074.
(24) Leray, I.; Valeur, B.; Paul, D.; Regnier, E.; Koepf, M.; Wytko,
J. A.; Boudon, C.; Weiss, J. Photochem. Photobiol. Sci. 2005, 4, 280.
(25) Yoon, Z. S.; Yang, J.; Yoo, H.; Cho, S.; Kim, D. In Handbook of
Porphyrin Science, Vol. 1; Kadish, K. M., Smith, K. M., Guilard, R.,
Eds.; World Scientific: Singapore, 2010; Vol. 1, Chapter 5.
(26) Stewart, J. J. P. MOPAC2009; Stewart Computational Chem-
istry: Colorado Springs, CO, USA, 2009; http://OpenMOPAC.net.
(27) Stewart, J. J. P. J. Mol. Model. 2007, 13, 1173.
(28) HyperChem, Release 8; Hypercube: Gainesville, FL, 2008.
6082
dx.doi.org/10.1021/ic2001255 |Inorg. Chem. 2011, 50, 6073–6082