Giant meso-meso-linked porphyrin arrays of micrometer molecular length and their fabrication

Article abstract of DOI:10.1002/chem.200401306

On the basis of the Ag^I-promoted coupling reaction of zinc(II)-5,15-bis(3,5-dioctyloxyphenyl)porphyrin Z1, chain elongation has been attempted by using a stepwise doubling approach, which provides Z2, Z4, Z8, Z16, Z32, Z64, Z128, Z256, Z384, and Z512. The porphyrin arrays up to Z128 are sufficiently soluble in CHCl₃ and THF despite their very long molecular lengths and rodlike structures, while the arrays over Z128 show a significant drop in solubility and stability. The discrete porphyrin arrays thus isolated were characterized by means of ¹H NMR spectroscopy, matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry, UV/Vis spectroscopy, gel-permeation chromatography (GPC), cyclic voltammetry (CV), single-crystal X-ray crystallography, scanning tunneling microscopy (STM), and atomic force microscopy (AFM). Contrary to expected linear conformations of the arrays Z n (where n is the number of porphyrins), the single molecular images of Z128, Z256, and Z512 revealed largely bent structures; this finding indicates the substantial conformational flexibility of Zn. We also exploited an effective synthetic route by means of which Zn can be fabricated with a thiol-protected aryl group to provide ZnS₂ through ZnBr₂, by bromination with N-bromosuccinimide and subsequent Pd-catalyzed Suzuki-Miyaura arylation. Finally, the reaction of Z256 provided Z512, Z768, and Z1024. Collectively, this work provides an important milestone in the preparation of sub-microscale discrete organic molecules and the fabrication of molecular-based materials, hence significantly contributing to device applications.

Full text of DOI:10.1002/chem.200401306

FULL PAPER
Giant meso–meso-Linked Porphyrin Arrays of Micrometer Molecular Length
and Their Fabrication
[
a]
[b]
[b]
[b]
Naoki Aratani, Akihiko Takagi, Yoshiki Yanagawa, Takuya Matsumoto,*
[
b]
[c]
[c]
[a]
Tomoji Kawai,* Zin Seok Yoon, Dongho Kim,* and Atsuhiro Osuka*
I
Abstract: On the basis of the Ag -pro-
moted coupling reaction of zinc(ii)–
5
flight (MALDI-TOF) mass spectrome-
try, UV/Vis spectroscopy, gel-permea-
tion chromatography (GPC), cyclic vol-
tammetry (CV), single-crystal X-ray
crystallography, scanning tunneling mi-
croscopy (STM), and atomic force mi-
croscopy (AFM). Contrary to expected
linear conformations of the arrays Zn
(where n is the number of porphyrins),
the single molecular images of Z128,
Z256, and Z512 revealed largely bent
structures; this finding indicates the
substantial conformational flexibility of
Zn. We also exploited an effective syn-
thetic route by means of which Zn can
be fabricated with a thiol-protected
aryl group to provide ZnS2 through
,15-bis(3,5-dioctyloxyphenyl)porphy-
rin Z1, chain elongation has been at-
tempted by using a stepwise doubling
approach, which provides Z2, Z4, Z8,
Z16, Z32, Z64, Z128, Z256, Z384, and
Z512. The porphyrin arrays up to Z128
are sufficiently soluble in CHCl3 and
THF despite their very long molecular
lengths and rodlike structures, while
the arrays over Z128 show a significant
drop in solubility and stability. The dis-
crete porphyrin arrays thus isolated
were characterized by means of
ZnBr , by bromination with N-bromo-
2
succinimide and subsequent Pd-cata-
lyzed Suzuki–Miyaura arylation. Final-
ly, the reaction of Z256 provided Z512,
Z768, and Z1024. Collectively, this
work provides an important milestone
in the preparation of sub-microscale
discrete organic molecules and the fab-
rication of molecular-based materials,
hence significantly contributing to
device applications.
Keywords: C–C coupling · cross-
coupling · nanostructures · oligome-
rization · porphyrinoids
1
H NMR spectroscopy, matrix-assisted
laser desorption ionization time-of-
Introduction
derstanding of their properties, and practical applications in
[1–7]
electronic and optical devices.
Even though both the the-
In the last two decades, remarkable progress has been made
in the synthesis of electrochemically and/or photochemically
active oligomeric and polymeric materials, which has led to
the exploration of new molecular architectures, better un-
ories and calculations have been explored based on ideal
[8]
polymers,
experiments on monodisperse oligomers, in
which no ensemble averaging is needed, can provide more
direct and insightful information and are thus of great im-
Yonsei University, Seoul 120–749 (Korea)
Fax : (+82)22-123-2434
E-mail: dongho@yonsei.ac.kr
[
a] Dr. N. Aratani, Prof. Dr. A. Osuka
Department of Chemistry, Graduate School of Science
Kyoto University, Sakyo-ku, Kyoto 606–8502 (Japan)
Fax : (+81)75-753-3970
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author. Synthetic proce-
dures for the new compounds, the recycling GPC chromatograms for
E-mail: osuka@kuchem.kyoto-u.ac.jp
[
b] Dr. A. Takagi, Y. Yanagawa, Prof. Dr. T. Matsumoto,
Prof. Dr. T. Kawai
1
the separation of the Z8 reaction mixture, ROESY H NMR spectrum
1
of Z4, H NMR spectra of Hn, ESR spectrum of the radical cation of
The Institute of Scientific and Industrial Research (ISIR)
Osaka University, 8–1, Mihogaoka, Ibaraki 567–0047 Osaka (Japan)
Fax : (+81)66-875-2440
E-mail: matsumoto@sanken.osaka-u.ac.jp
kawai@sanken.osuka-u.ac.jp
zinc(ii)–5,15-bis(3,5-di-tert-butylphenyl)porphyrin complex formed by
electrochemical oxidation, and UV/Vis absorption and fluorescence
data of Zn.
[
c] Z. S. Yoon, Prof. Dr. D. Kim
Center for Ultrafast Optical Characteristics Control and
Department of Chemistry
Chem. Eur. J. 2005, 11, 3389 – 3404
DOI: 10.1002/chem.200401306
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3389

portance. In this context, the synthesis of monodisperse
macromolecular rods of precise length and constitution has
attracted keen attention in light of their potential applica-
tion in molecular-scale electronics, optical devices, solar
the meso–meso coupling reaction is always quite high, 2) the
molecular length grows rapidly at the rate of 2 , where n is
n
the number of iterations, 3) the orthogonal conformations of
the porphyrin array products causes them to be highly solu-
ble, which allows the manipulation of very long porphyrin
arrays, 4) the separation of the coupling products is easily
carried out by means of recycling preparative gel-permea-
tion chromatography HPLC (GPC-HPLC), due to the large
difference in the molecular weights of the porphyrin prod-
ucts, and finally 5) the long coupling products still bear two
free meso-positions that are available for the next reaction
as long as the porphyrin array is soluble in the reaction sol-
vent. Taking advantage of these features, we synthesized a
series of meso–meso-linked porphyrin arrays up to 128-mer,
[1–3]
energy conversion, and so forth,
because such discrete
oligomers can also offer a good opportunity for systematic
studies on structure–property relationships that will be
useful for predicting specific electronic, photonic, and mor-
phological property data.
Through recent extensive synthetic efforts, the range of
molecular lengths of linear monodisperse p-conjugated
oligomers has now reached over 10 nm. With respect to
oligo-p-phenylene, up to 16 p-phenylene rings were cova-
lently connected by use of the repetitive approach by
[9]
Hensel and Schlꢁter, who also synthesized a cyclotetraico-
which was, at that time, the longest discrete molecule
[10]
[1j,20b]
saphenylene based on the same strategy. Meier et al. syn-
thesized monodisperse oligo(1,4-phenylene ethenylene)s up
made.
In this respect, further extension of this synthetic
strategy is challenging in its own right.
[11]
to 12-mer as a more planar p-conjugated oligomer. Tour
et al. developed an approach to growing molecular systems
by means of an iterative divergent/convergent method,
which resulted in the synthesis of phenylethynyl 18-mer
bearing protected thiol groups for connection onto gold
Here, we report the synthesis of a series of much longer
meso–meso-linked porphyrin arrays. These arrays are partic-
ularly attractive due to their linear rodlike structure, ample
electronic interaction between neighboring porphyrins, and
unprecedented giant molecular size. These properties are in-
teresting from a photonic-molecular-wire viewpoint, because
1) the large Coulombic interactions between the neighboring
porphyrins are favorable for rapid excitation energy hop-
[12]
electrodes. This method, when combined with solid-phase
synthesis, furnished a block-alternating 23-mer consisting of
oligo(1,4-phenylene ethynylene)s and oligo(2,5-thiophene
ethynylene)s, the molecular length of which reached about
[22]
ping,
and 2) each porphyrin unit retains its individual
[13]
160 ꢂ.
character, presumably because of the orthogonal geometry,
[23]
Conversely, oligo- and polythiophenes have been actively
thus preventing the formation of an energy sink. More-
over, it has been demonstrated that these porphyrin arrays
can be effectively transformed into completely p-conjugat-
ed, fused porphyrin arrays, which exhibit an extremely low
optical HOMO–LUMO gap, down to approximately
0.16 eV at the stage of 12-mer, and hence, encouraging the
potential conversion of a photonic wire to an electronic
exploited for molecular devices, such as organic conductive
materials, electroluminescence (EL) devices, organic light-
emitting diodes (OLEDs), field-effect transistors (FETs),
[14–17]
and so forth.
Roncali et al. reported the synthesis and
characterization of monodisperse soluble oligothienylenevi-
[15]
nylenes with chain lengths up to 10 nm. A series of long
discrete oligothiophene molecules was synthesized by
[24]
wire.
[
17]
Otsubo and co-workers,
whose work culminated in the
To use these porphyrin arrays in a variety of applications,
it is highly desirable to fabricate them with a thiol or car-
boxylic acid group, which will allow a chemical bond to be
formed between the array and a metal surface. In this re-
spect, the free end meso-position of the meso–meso-linked
porphyrin arrays is quite attractive. Thus, we examined the
reactivity of the end meso-position and developed a useful
and complete conversion into meso-mercaptophenylporphyr-
in arrays.
[17a]
synthesis of a discrete 96-mer.
Yet it still remains as a
great synthetic challenge to explore discrete functional mol-
ecules with well-defined structures far beyond the above
achievements.
Among the molecular modules suitable for use as a con-
struction element of supramolecular rods, porphyrins are
one of the most attractive building blocks, because they
offer a variety of desirable features, such as rigid planar ge-
ometry, high stability, intense electronic absorption and
emission, a small HOMO–LUMO energy gap, and flexible
tunability of their optical and redox properties by appropri-
In the present study, we isolated a meso–meso linked por-
phyrin 1024-mer, which is, to the best of our knowledge, the
longest (ca. 850 nm=0.85 mm) monodisperse rodlike man-
[18]
[1j]
ate metalation. In fact, extensive efforts have been made
to prepare porphyrin arrays with a view to their use in real-
made molecule,
which has a molecular formula of
C65536H83970N4096O4096Zn1024
.
[19]
izing various functional units and molecular devices. How-
ever, these studies have often been hampered by poor solu-
bility, difficult separation, and demanding characterization,
all of which are intrinsic problems of covalently linked,
large porphyrin arrays.
Results
Synthesis and separation: meso–meso-Coupled porphyrin
oligomers were prepared by the Ag -promoted oxidative
coupling reaction. One of the key advantages of this syn-
I
I
Recently, we reported the Ag -promoted meso–meso cou-
II
[20,21]
[20]
pling reaction of a Zn –5,15-diarylporphyrin.
This reac-
tion has the following advantages: 1) the regioselectivity of
thetic strategy is the high solubility of long porphyrin prod-
3390
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3389 – 3404

Long Porphyrin Arrays
FULL PAPER
ucts, even in their 128-mer stage, which allows for their iso-
lation and characterization. The synthesis of zinc(ii)–5,15-
bis(3,5-dioctyloxyphenyl)porphyrin (Z1) is shown in
Scheme 1. 3,5-Dihydroxybenzoic acid (1) was esterified in
merized, trimerized, and tetramerized compounds were ob-
tained constantly in 20–30%, ~13%, and ~6% yields, re-
spectively, along with the recovery of starting materials (50–
55%). Use of purified CHCl as the reaction solvent and
3
keeping the reaction temperature strictly at 308C were cru-
cial to attaining constant results, because the coupling reac-
tion of Z1 is sensitive to additives and temperatures. Addi-
tion of small amounts (0.5–3%) of N,N-dimethylformamide
(
DMF), N,N-dimethylacetamide (DMA), or hexamethyl-
phosphoramide (HMPA) accelerated the reaction to give
II
[27]
poly(Zn –porphyrinylene),
while addition of small
amounts of amine resulted in complete retardation of the
[28]
coupling reaction. Slight heating (40–508C) also accelerat-
ed the coupling reaction to provide a wider range of large
porphyrin arrays in an uncontrollable manner.
The use of recycling preparative GPC-HPLC was crucial
for the isolation of porphyrin products. Figure 1 shows
GPC-HPLC chromatographs (column conditions: JAI-GEL
Scheme 1. Synthesis of Z1. Ar=3,5-dioctyloxyphenyl.
5
H-AF, 4H-AF, and 3H-AF with THF) of the reactions of
Z1, Z2, Z4, and Z8 (Figure 1left), Z16, Z32, and Z64 (Fig-
ure 1right). The coupling products were nicely separated by
means of GPC-HPLC in all cases, after several recycling
separations owing to large differences in molecular weight,
thus allowing isolation of pure, long porphyrin arrays. In
each step, strict separation of the doubling product is cru-
cially important for further chain elongation, as even a small
amount of contamination will lead to serious problems in
the longer array. During many repeated preparations, we
also isolated Z3, Z5, Z6, Z7, Z10, Z12, Z20, Z24, Z40, Z48,
Z80, Z96, and Z192, and accumulated more than 100 mg of
Z128.
Additionally, Zn can be transformed into the correspond-
ing free-base porphyrin arrays Hn (where n represents the
number of porphyrins) in good yields, which can in turn be
metalated with a variety of metal ions to give metalated
meso–meso-linked arrays.
methanol with a catalytic amount of concentrated H SO to
2
4
form methyl 3,5-dihydroxylbenzoate (2). Alkylation of the
ester 2 was performed by a Williamson-type reaction to give
methyl 3,5-dioctyloxybenzoate (3), which was reduced with
LiAlH to 3,5-dioctyloxybenzyl alcohol (4). 3,5-Dioctyloxy-
4
benzaldehyde (5) was obtained by oxidation of 4 with pyri-
dinium chlorochromate in CH Cl . The porphyrin Z1 was
2
2
[25]
prepared from 5 and bis(2-pyrryl)methane (6)
yield under the standard conditions.
in 44%
[26]
The chain elongation strategy is quite simple, that is, by
repeating dimerization reactions from Z1 to Z2, Z2 to Z4,
Z4 to Z8, Z8 to Z16, Z16 to Z32, Z32 to Z64, and Z64 to
Z128 (Scheme 2). The reaction of Z1 with AgPF (1.2 equiv
6
in CHCl ) at 308C for 10 h, followed by preparative SEC
3
(
6
size-exclusion chromatography) gave Z2 (22–27%), Z3 (3–
%), Z4 (2%), and a small amount of higher oligomers
The Z128 array has a molecular length of about 106 nm in
its linear form, but is sufficiently soluble in THF and CHCl3.
Thus, the coupling reaction of Z128 was attempted, although
the coupling reaction of such an extremely long molecule
should be entropically very unfavorable because of only two
reactive free meso-positions. Z128 was reacted with AgPF6
along with the recovery of Z1 (50–55%). In each step, di-
(
12 equiv in CHCl ) for 11 h at room temperature, and the
3
reaction mixture was analyzed by GPC-HPLC, which indi-
cated the formation of Z256, Z384, and Z512 (Figure 2left).
These large porphyrin arrays were actually isolated by the
preparative recycling GPC-HPLC in yields of 26, 11, and
4
1
%, respectively. Isolated arrays eluted at 14.8 min (Z256),
4.1 min (Z384), and 13.6 min (Z512) on the present column
setup. While Z128 can be stored in a refrigerator for a long
time without serious deterioration and can be manipulated
essentially in a similar manner to normal, small organic
compounds, the manipulation of the larger arrays (Z256,
Z384, and Z512) is more difficult, because, although freshly
separated samples are reasonably soluble, they change to an
insoluble material even upon storage in a refrigerator in the
dark. Once these porphyrin arrays become insoluble, it is
Scheme 2. Synthesis of Zn. Ar=3,5-dioctyloxyphenyl.
Chem. Eur. J. 2005, 11, 3389 – 3404
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3391

T. Matsumoto, T. Kawai, D. Kim, A. Osuka et al.
Figure 2. Left: GPC-HPLC chromatograms of the reaction from Z128:
a) Z128, b) reaction mixture of Z128, c) purified Z256, d) purified Z384,
and e) purified Z512. The GPC setup is the combination of JAI-GEL
5
H-AF, 4H-AF, and 3H-AF. Right: GPC-HPLC chromatograms of the
reaction from Z256: a) Z256, b) reaction mixture of Z256, c) purified
Z512, d) purified Z768, and e) purified Z1024. The GPC column setup is
the combination of JAI-GEL 5H-AF and 4H-AF.
ferent analytical column setup to get better separation for
larger arrays: JAI-GEL 5H-AF and 4H-AF with THF. The
preparative GPC column showed clearer separation
(
Figure 3). These reaction products were separated with
I
Figure 1. GPC-HPLC chromatograms of the Ag -promoted coupling reac-
tion detected by absorbance at 413 nm. Each chromatogram is normal-
ized to the maximum intensity. Left: The reaction a–d) from Z1, e–
g) from Z2, h–j) from Z4, and k–m) from Z8; a) Z1, b) after 9 h, c) puri-
fied Z2, d) purified Z3, e) after 8 h, f) purified Z4, g) purified Z6, h) after
4
h, i) purified Z8, j) purified Z12, k) after 8 h, l) purified Z16, and m) pu-
rified Z24. Right: The reaction n–q) from Z16, r–t) from Z32, and u–
w) from Z64; n) Z16, o) after 9 h, p) purified Z32, q) purified Z48,
r) after 8 h, s) purified Z64, t) purified Z96, u) after 11 h, v) purified
Z128, and w) purified Z192. The GPC column setup is the combination
of JAI-GEL 5H-AF, 4H-AF, and 3H-AF.
very difficult to redissolve them in common organic solvents.
This tendency becomes more serious upon increasing the
array size.
Finally, we examined the coupling reaction of Z256 under
similar conditions. If this trial were to go well, the longest
molecule would be formed. Actually this attempt afforded
Z512, Z768, and Z1024, as judged from the GPC-HPLC
chart of the reaction mixture (Figure 2right). We used a dif-
Figure 3. Preparative GPC-HPLC chromatogram of the reaction from
Z256.
3392
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3389 – 3404

Long Porphyrin Arrays
FULL PAPER
considerable difficultly, because their retention times are
quite close to the exclusion limit. But after repeated recy-
cling separations over the preparative GPC-HPLC, we iso-
lated Z512 (13%), Z768 (5%), and Z1024 (2%). These iso-
lated long arrays exhibited the retention times of 9.9, 9.6,
and 9.5 min, respectively. The longest molecule, Z1024, has
a molecular length of 0.85 mm in its linear form and a molec-
ular formula of C65536H83970N4096O4096Zn1024
.
The molecular weights of these porphyrin arrays, up to
Z128, were carefully determined by matrix-assisted laser de-
sorption ionization time-of-flight (MALDI-TOF) mass spec-
trometry, and the results are summarized in Table 1. The
mass spectra of Z64, Z96, and Z128 revealed the parent
peaks at 66323, 99050, and 130295, respectively, along small
dicationic peaks, clearly indicating their clean monodispersi-
ty (Figure 4). Unfortunately, however, the parent ion peaks
of porphyrin arrays higher than Z128 could not be detected.
Further information on the monodispersity of the separat-
ed porphyrin arrays was given from their GPC-HPLC elut-
ing features, as all of the isolated porphyrin arrays exhibited
clear elution bands on the GPC-HPLC charts, and their re-
tention times showed a good continuous relationship, similar
to that of polystyrene standards, versus the actual molecular
weights. Zn arrays and polystyrene standards eluted in a
similar manner against their molecular weights in the
Figure 4. MALDI-TOF mass spectra (9-nitroanthracene was used as a
matrix). a) Z64, b) Z96, and c) Z128.
logM <4 region, while Zn arrays eluted much faster than
w
polystyrene standards of the same molecular weight in the
molecular weights versus analytical GPC molecular weights
estimated with respect to polystyrene standards.
logM >4 region, and this discrepancy became more signifi-
w
cant with increasing molecular weight (Figure 5a). Figure 5b
demonstrates the good continuous relationship of the actual
1
1
H NMR spectroscopy: H NMR spectra of Z1–Z8, taken in
CDCl at room temperature, are shown in Figure 6. Assign-
3
ments were performed by
Table 1. Zn compounds list. Observed and calculated molecular weights and GPC retention times [min].
means
of
comprehensive
[
a]
[b]
[c]
ROESY experiments (the des-
ignation of the protons is given
in Scheme 3. See also Support-
ing Information). In the spec-
trum of Z1, a singlet for the
Compound
Molecular formula
M
c
M
o
Retention time
Z1
Z2
Z3
Z4
Z5
Z6
Z7
Z8
Z10
Z12
Z16
Z20
Z24
Z32
Z40
Z48
Z64
Z96
Z128
Z192
Z256
Z384
Z512
Z768
Z1024
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
64
H
84
N
4
O
4
Zn
1 037
2075
3114
4146
5186
6222
7259
8296
1 036
2076
3112
4145
5183
6214
7253
8286
10351
12416
16566
20769
24465
33292
41177
49497
66323
99050
130295
n.d.
26.1
25.4
24.9
24.5
24.2
23.9
23.6
23.4
22.9
22.5
21.8
21.2
20.7
19.9
19.2
18.7
17.8
16.8
16.1
15.3
14.8
14.1
13.6
n.d.
n.d.
128
192
256
320
384
448
512
640
768
H
H
H
H
H
H
H
H
H
166
248
330
412
494
576
658
822
986
N
N
N
N
N
N
N
N
N
8
O
8
Zn
2
12
16
20
24
28
32
40
O
12Zn
16Zn
20Zn
24Zn
28Zn
32Zn
40Zn10
48Zn12
64Zn16
3
4
5
6
7
8
O
O
O
O
O
O
O
m
meso protons (H ) appears at
1
0.31 ppm and two doublets for
the b protons appear at 9.43
1
2
10369
12443
16590
20736
24884
33178
41470
49764
66350
99527
132709
199052
265416
398122
530830
796242
1061658
(H ) and 9.26 ppm (H ). The
48
spectrum of Z2 exhibits a sin-
1024
1280
1536
2048
2560
3072
4096
6144
8192
H
H
H
H
H
H
H
H
H
1314
1642
1970
2626
3282
3938
5250
7874
N
64
80
96
O
O
O
m
glet for H at 10.38 ppm, two
N
N
N
N
N
N
N
80Zn20
96Zn24
doublets for the outer b protons
1
128
160
192
256
384
O
128Zn32
160Zn40
192Zn48
256Zn64
384Zn96
512Zn128
O768Zn192
O
O
O
O
O
at 9.49 and 9.28 ppm (H and
2
O
O
O
O
H , respectively), and two dou-
blets for the inner b protons at
3
4
8
.82 (H ) and 8.10 ppm (H ),
10498
N
512
O
which are characteristic of a
meso–meso-linked diporphyrin.
The inner b protons are lying in
the shielding region of the
neighboring porphyrin. Similar-
ly, the spectrum of Z3 exhibits
[
d]
12288
16384
24576
32768
49152
65536
H
15746
20994
31490
41986
62978
83970
N
N
N
N
N
N
768
H
H
H
H
H
1024
1536
2048
3072
4096
1024Zn256
1536Zn384
2048Zn512
3072Zn768
4096Zn1024
n.d.
n.d.
n.d.
n.d.
n.d.
m
a singlet for H , two doublets
[
[
a] Calculated molecular weight. [b] Observed molecular weight detected by MALDI-TOF mass spectroscopy.
c] GPC-HPLC analysis (JAIGEL-5H-AF, 4H-AF, and 3H-AF columns in series). [d] n.d.=not determined. for H and H , and four dou-
1
2
Chem. Eur. J. 2005, 11, 3389 – 3404
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3393

T. Matsumoto, T. Kawai, D. Kim, A. Osuka et al.
Figure 5. a) Relationship of the retention time versus log M
and standard polystyrene (&). b) The estimated molecular weights
number-average, M ), determined by GPC, versus actual molecular
W
for Zn (*)
(
n
weights of Zn (data points) relative to polystyrene standards (dashed
line).
1
3
Figure 6. a)–h) H NMR spectra, in CDCl , of Z1–Z8, respectively.
3
4
5
6
blets for the inner b protons H , H , H , and H . The
1
1
H NMR spectra of Z1–Z8 clearly indicate totally symmetric
H NMR spectra of Z16, Z32, Z64, and Z128, taken in
structures with straight, rodlike conformations. The
CDCl at room temperature, were shown in the previous
3
Scheme 3. Structure of Z12 and proton designation. Ar=3,5-dioctyloxyphenyl.
3394
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3389 – 3404

Long Porphyrin Arrays
FULL PAPER
[
20b]
paper.
Similar to these, the array Z256 also displays a rel-
1
m
atively well-resolved H NMR spectrum, in which H ap-
pears as a singlet at 10.42 ppm, and H and H appear as
two doublets at 9.53 and 9.31 ppm, respectively (Figure 7). It
1
2
1
Figure 7. H NMR spectrum of Z256 in CDCl
3
.
m
1
2
is noteworthy that H , H , H , and many inner b protons
3
8
(
H –H , etc.) appear at nearly the same chemical shifts for
all the arrays, independent of concentration, indicating no
significant aggregation of the arrays under these rather con-
1
centrated H NMR conditions. The free-base porphyrin
arrays Hn show similar H NMR spectra. The inner NH
Figure 8. Crystal structure of H1. a) The ORTEP drawing with thermal
ellipsoids at 50% probability. Hydrogen atoms are omitted for clarity.
Crystal packing structure: b) side view and c) top view.
1
protons of the porphyrin appear at ꢀ3.15 ppm for H1,
ꢀ
ꢀ
ꢀ
2.44 ppm for H2, ꢀ1.65 (2H) and ꢀ2.40 ppm (4H) for H3,
1.62 (4H) and ꢀ2.39 ppm (4H) for H4, and ꢀ1.57 (4H),
1.60 (4H), and ꢀ2.38 ppm (4H) for H6. Thus, the signals
of those inner protons are shifted to lower field, reflecting
the ring current of neighboring and other close porphyrins
(
Supporting Information).
Single-crystal X-ray diffraction analysis: The molecular
structure of 5,15-bis(3,5-dioctyloxyphenyl)porphyrin H1 was
determined by X-ray crystal analysis (Figure 8). The por-
phyrin rings show rather planar structures with small devia-
tions (0.023 ꢂ) from the mean planes, which consist of 24
core atoms. Single crystals of Z2 were obtained from a
THF/2-propanol solution, although unfortunately the dif-
[29]
fraction data was not suitable for analysis (not shown).
Through many attempts to make single crystals of meso–
meso-coupled porphyrin arrays, single crystals suitable for
X-ray structure analysis were obtained for triporphyrin ZP3
(
all aryl substituents of the porphyrin array were phenyl
groups), which was the first crystal structure of a meso–
meso-linked triporphyrin (Figure 9). In the solid state, tri-
P
Figure 9. Crystal structure of Z 3. a) The ORTEP drawing with thermal
ellipsoids at 50% probability. Solvent molecules are omitted for clarity.
b) The space-filling drawing. c) The crystal packing structure.
[34]
porphyrin Z 3 displays a straight and orthogonal structure,
P
with dihedral angles between the directly linked porphyrins
of 88.38 and meso–meso bond lengths of 1.50(2) ꢂ. This
bond length is within the range of a typical CꢀC single
These structural features suggest a weak CH–p interaction
between the porphyrin p plane and the parallel octyloxy
side chain; this interaction leads to the formation of an in-
finite, one-dimensional network (Figure 8c). On the other
bond.
The packing structures provide information on molecular
interactions in the solid state. In the crystal structure of H1,
both faces of each porphyrin are covered by the octyloxy
side chains of the neighboring porphyrins in a parallel
manner with a distance between carbon atoms and porphyr-
in plane of about 3.5 ꢂ (Figure 8b). Interestingly, the non-
interacting octyloxy side chains on the opposite side take an
oblique orientation with respect to the porphyrin plane.
hand, the porphyrin trimer Z 3 possesses a one-dimensional
P
network with different structural motifs; orthogonally linked
triporphyrin subunits are nicely packed, with weak CH–p in-
teractions between the porphyrin p plane and the aromatic
[31]
CH group. The distance between the mean p plane and
the phenyl p-carbon atom is estimated to be about 3.23 ꢂ
Chem. Eur. J. 2005, 11, 3389 – 3404
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3395

T. Matsumoto, T. Kawai, D. Kim, A. Osuka et al.
(
Figure 9c). This CH–p network structure cannot be suitable
HOMO in Z2 is close to that in Z1, while EOX2 of Z1 was
slightly higher than its EOX1. The one-electron oxidation po-
tential of Z3 was detected at 0.29 V, but the first oxidation
potential waves of Z4 and Z8 could not be identified clearly.
Despite the poorly resolved voltammograms, it was possible
to assign the first one-electron oxidation at around 0.3 V for
for Zn owing to m-dioctyloxy side chains.
Electrochemical properties: The one-electron oxidation po-
tentials of the meso–meso-linked Zn –porphyrin arrays are
also an important functional parameter for them, as well as
for an understanding of the mechanism of the Ag -promoted
meso–meso coupling. Therefore, the electrochemical prop-
erties, particularly the one-electron oxidation potentials,
were studied by cyclic voltammetry (CV) in dry CHCl con-
II
I
II
Z4 and Z8. Therefore, it may be concluded that these Zn –
[32]
porphyrin arrays (Zn) have almost the same first one-elec-
tron oxidation potentials, which are independent of the
number of porphyrin rings.
3
taining
0.1m
tetrabutylammonium
tetrafluoroborate
(
nBu NBF ) as a supporting electrolyte. At the time, we
UV/Vis absorption and fluorescence spectroscopy: The opti-
cal properties of Zn, up to Z128, were reported previous-
ly and are summarized here. Characteristically, Zn exhibit
4
4
could not precisely measure waves associated with electro-
chemical reduction. The cyclic voltammograms of the Zn –
II
[23]
porphyrin arrays are shown in Figure 10. The monomer Z1
underwent a reversible first oxidation at 0.31 V (versus fer-
rocene/ferrocenium), while two reversible oxidation waves
were detected in Z2 at 0.28 and 0.43 V; these have been as-
signed as split first oxidation waves (one electron per por-
phyrin) judging from the results of other electronically cou-
large splitting of Soret bands due to strong exciton coupling.
With an increase in the number of porphyrins, the low-
energy Soret band is shifted to longer wavelengths, while
the high-energy Soret band remains at nearly the same posi-
tion, resulting in a progressive increase in the splitting
energy. The observed Soret-band splitting can be described
[24b]
[36]
pled diporphyrins.
each Zn –porphyrin should arise from the influence of the
positive charge of first-generated Zn –porphyrin radical cat-
This split one-electron oxidation for
well by the exciton-coupling theory. The energy difference
II
between high- and low-energy Soret bands becomes larger
as the array becomes longer, leading to an asymptotic fea-
II
ꢀ
1
ions. EOX1 of Z2 was similar to that of Z1, indicating the
ture (DEffi4615 cm ). On the other hand, the shifts in the Q
bands are not so significant, but show a gradual increase in
intensity. UV/Vis absorption spectra of Z128, Z256, and
Z512 are shown in Figure 11a. A plot of the molar extinc-
tion coefficients of Zn as a function of the number of por-
phyrin units exhibits a good linear correlation, indicating
the higher absorptivity as the array becomes longer. In
other words, each porphyrin moiety in the array acts as an
independent light-absorbing unit.
On the other hand, the fluorescence quantum yields of
Zn increase up to Z16, but then continuously decrease as
the array becomes longer than Z16. The decreased fluores-
cence quantum yields of longer porphyrin arrays indicate
that each porphyrin unit in the array cannot act as an indi-
vidual fluorophore. As shown in Figure 11b, the total fluo-
rescence intensity decreases from Z128 to Z256, but not to
such a great extent from Z256 to Z512. The fluorescence
decays of Zn also show a similar trend to the fluorescence
quantum yields. As the porphyrin arrays become longer,
their average fluorescence lifetimes decrease consistently up
to Z512. In the arrays shorter than Z16, the fluorescence
decays exhibit single exponential behavior. Conversely, as
the array becomes longer than Z16, the fluorescence tempo-
ral profiles start to show double exponential decay, in which
the contribution of the fast-decay component increases grad-
ually with array length, up to Z512 (Table 2). It should be
noted that the fluorescence-decay profiles depend on the
probe wavelength, especially in Z256 and Z512 (Supporting
Information). Specifically, as the probe wavelength moves
towards red, the overall fluorescence decay becomes slower
in Z256 and Z512. In shorter arrays, however, the fluores-
cence decays do not show this probe-wavelength depend-
ence. This feature indicates that the excitation-energy migra-
tion occurs from the initially excited, shorter segments of
+
Figure 10. Cyclic voltammograms (vs Fc/Fc ) for the oxidation of a)–
d) Z1–Z4, respectively, and e) Z8, in CHCl
3
using a platinum working
electrode and 0.1m nBu NBF as a supporting electrolyte.
4
4
3396
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3389 – 3404

Long Porphyrin Arrays
FULL PAPER
The fluorescence excitation polarization of Z1 was nearly
1
/7 regardless of the excitation wavelength, which is typical
when both absorption and emission oscillators are degener-
ate and polarized in the same plane.For Zn, in contrast, neg-
ative anisotropy values were observed in the fluorescence
excitation anisotropy spectra at around 413 nm, which corre-
sponds to the high-energy Soret band, and positive anisotro-
py values were observed from 450 to 500 nm, which corre-
sponds to the low-energy Soret band. The anisotropy values
above ~450 nm for Zn became larger, reaching a maximum
at Z16, and then decreased for the longer arrays. The maxi-
mum anisotropy value of ~0.3 above ~450 nm in Z16,
which is still slightly smaller than 0.4 for perfect in-plane
orientation between absorption and emission dipoles, indi-
cates a nearly parallel alignment between absorption and
emission dipoles upon photoexcitation at the low-energy ex-
citon split Soret and Q bands. These features indicate that
the overall geometries of Zn remain linear in form when
Zn is shorter than Z16 and thereafter start to deviate from
linearity.
Conformational investigation by means of pulsed-field-gra-
dient NMR (PFG-NMR) spectroscopy: To obtain informa-
II
tion on the molecular motion of meso–meso-linked Zn –
porphyrin arrays in solution, the self-diffusion coefficients
(
D) were measured based on pulsed-field-gradient NMR
[34,35]
spectroscopy.
Usually D values become smaller (mole-
cules diffuse more slowly) with increasing molecular size.
Given that diffusion coefficients reflect effective hydrody-
namic radii, not only the molecular size, but also the molec-
ular shape will determine the value of D. Therefore, we may
acquire information on the molecular shape of the porphyr-
in arrays in solution from their D values.
Figure 11. a) UV/Vis and b) fluorescence spectra of Z128 (c), Z256
(
g), and Z512 (b) in THF.
As noted above, all of these porphyrin arrays exhibit a
[
a]
m
Table 2. Fluorescence lifetimes of Zn.
[ns]
singlet signal of H at around 10.4 ppm. Taking advantage
1
t
1
t
2
[ns]
t
av [ns]
of this H-NMR characteristic signal, the D values (in
2
ꢀ1
m s ) were determined on the basis of the PFG method at
Z1
Z2
Z3
Z4
Z6
Z8
Z12
Z16
Z24
Z32
Z48
Z64
Z96
Z128
Z256
Z512
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
2.62
1.83
1.72
1.65
1.59
1.55
1.52
1.50
1.43
1.40
0.93
0.51
0.50
0.49
0.40
0.26
ꢀ
10
208C in CDCl and are as follows: 9.7ꢃ10
for Z3, 3.4ꢃ
3
ꢀ10
ꢀ10
ꢀ10
1
0
for Z8, 1.3ꢃ10
for Z20, and 1.2ꢃ10
for Z24.
Thus, it is actually shown that the D values become smaller
with increasing molecular length (Figure 12).
Structure investigation by scanning tunneling microscopy
STM) and atomic force microscopy (AFM): In recent
1.49 (95.4)
1.48 (92.8)
1.47 (61.1)
1.47 (31.2)
1.45 (27.7)
1.42 (21.8)
1.05 (28.8)
0.92 (17.7)
0.14 (4.6)
0.33 (7.2)
0.08 (38.9)
0.08 (68.8)
0.14 (72.3)
0.22 (78.2)
0.14 (71.2)
0.12 (82.3)
(
years, the determination of molecular shape with a certain
molecular size has become more and more feasible by
[36–40]
means of scanning tunneling microscopy.
Even atomic
force microscopy can be effective in resolving a molecular
shape and/or packing pattern of a sufficiently big mole-
[40]
cule. Therefore, an STM micrograph of Z6 adsorbed on a
Cu(100) surface was taken; it shows an image of paired
dots, having a constant spacing of about 2.00Æ1.04 nm be-
[
a] Numbers given in parentheses are percentages. t
1
, t
2
are slow and fast
decaying components of the fluorescence decay analyzed in terms of
double exponential function, and tav is the averaged fluorescence life-
time.
[39a]
tween the dots (Figure 13a).
This observation was as-
signed to Z6 having a bent conformation on the Cu surface
II
porphyrin oligomer to the longer ones; this migration be-
havior is commonly observed in the photoexcitation dynam-
ics of polymers.
with the Zn –porphyrin at each end, which were imaged
strongly by STM, attached parallel to the metal surface.
Here we report the STM images of Z9, Z48, and Z128 on a
Chem. Eur. J. 2005, 11, 3389 – 3404
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3397

T. Matsumoto, T. Kawai, D. Kim, A. Osuka et al.
formation with the porphyrins intermediate between the
two end porphyrins being detached from the metal surface.
This assignment suggests conformational flexibility of meso–
meso-linked porphyrins. This feature is highlighted by the
STM images of Z48 and Z128. Among many unclear images
that arose from significant entanglement or aggregation, a
clear image of a single molecule of Z48, taken at V =2.0 V
s
and I=0.03 nA, was obtained, which displays a severely
bent conformation on the plane of the metal surface (Fig-
ure 13c). Significantly, the molecular length of Z48 is esti-
mated to be 43.5 nm from this STM image, which is exactly
its predicted value. We also obtained a clear image of a
single molecule of Z128 (Figure 13d), taken at V =0.5 V
s
and I=0.05 nA, which also shows a snakelike, largely bent
structure with a molecular length of 117 nm that again
agrees well with its predicted molecular length.
The detection of long porphyrin arrays on a single-mole-
cule level has also been attempted by means of AFM on a
highly oriented pyrolytic graphite (HOPG) substrate. In
these experiments, a very dilute solution of porphyrin array
3
Figure 12. Diffusion coefficients of Z3, Z8, Z20, and Z24 in CDCl . (n is
the number of porphyrins.)
in CHCl was spread onto a freshly cleaved HOPG surface
3
followed by slow evaporation of the solvent. Figures 14a and
b displays the tapping-mode AFM images of Z96, in which
the molecules are aligned along the graphite crystal lattice.
It is interesting to note that many Z96 molecules are ob-
served with an extended-wire shape and with nearly the
same molecular length of 87.3Æ14.5 nm. Subsequently,
single molecules of Z256 and Z512 were observed on a
clean sapphire-Al O (0001) surface (Figures 14c and d, re-
2
3
spectively). The single molecules were well dispersed and
showed a clear zig-zag shape, with molecular lengths that
were fully consistent with the structures. The molecular
Figure 13. STM images of a) Z6, b) Z9, c) Z48, and d) Z128.
Cu(111) substrate. Sample deposition was performed using
the pulse injection method under high vacuum
ꢀ
6
[41]
(
10 mbar).
The observed STM images are shown in
Figure 13. As for Z6, the micrograph of the array Z9, which
was taken at V =2.0 V and I=0.03 nA (V and I are sample
s
s
bias voltage and tunneling current, respectively), shows a
pair of images that are nearly constantly spaced at a dis-
tance of 55 ꢂ (Figure 13b). As bright images usually stem
from strong adsorption of p-conjugated molecules on a
metal surface, the bright pair-based images seen here may
be assigned to the adsorption of the two end porphyrins
with their planes parallel to the metal substrate. Thus, it
may be suggested that Z9, similar to Z6, takes a bent con-
Figure 14. AFM images of a) Z96 on HOPG, b) Z96 on sapphire, c) Z256
on sapphire, and d) Z512 on sapphire.
3398
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3389 – 3404

Long Porphyrin Arrays
FULL PAPER
lengths observed by scanning probe microscopy techniques
versus the number of porphyrin units n are plotted in
Figure 15.
Figure 15. Relationship between the number of porphyrin units (n) and
the observed molecular length, measured using AFM (*) or STM (*).
Fabrication of meso–meso-linked porphyrin arrays: Func-
tionalization of one or both ends of the porphyrin array is
the next important step for their useful application. When
the fabrication of porphyrin arrays is possible with a thiol-
containing group, resultant fabricated arrays could be used
for covalent attachment on gold surfaces. For such a pur-
pose, we selected Pd-catalyzed Suzuki–Miyaura arylation of
meso-halogenated porphyrins with an arylboronic acid on
Scheme 4. Ar=3,5-dioctyloxyphenyl.
uct through a short silica gel column without purification
over a GPC column.
Suzuki–Miyaura arylation of Z6Br with 10 (10 mol% Pd-
2
(PPh ) , 3 equiv Cs CO , DMF/toluene, at 908C for 4 h)
3
4
2
3
II
gave thiolated Zn –porphyrin array Z6S in 90% yield.
2
[42]
the basis of several successful precedents. However, usual
thiol protective groups, such as acetyl and benzoyl groups,
were found not to able to tolerate the basic conditions of
Similarly, Z12Br , Z24Br , Z48Br , and Z64Br were effec-
2
2
2
2
tively arylated under the same conditions to furnish Z12S2
(85%), Z24S (79%), Z48S (81%), and Z64S (91%), re-
2
2
2
[
12,43]
the coupling reaction.
To avoid this, a diphenylmethyl
spectively. Here again, the conversion of ZnBr to ZnS2
2
[44]
group was selected as a protective group. The synthesis of
was nearly quantitative and almost-pure ZnS2 products
were isolated by simple precipitation of the product. These
products were also characterized by MALDI-TOF mass
meso-thiophenylated porphyrin arrays is shown in Scheme 4.
4-Bromobenzenethiol (8) was protected with 1,1-diphenyl-
1
methanol (7) with the aid of trifluoroacetic acid (TFA) to
form sulfide 9, which was transformed into boronate 10 by a
Pd-catalyzed reaction with an equivalent of bis(pinacolato)-
spectrometry and H NMR spectroscopy. Figure 16 illus-
1
trates the comparison of the H NMR spectra of Z64,
m
Z64Br , and Z64S . The single H signal in the spectrum of
2
2
[45]
diboron.
Z64 is missing in those of Z64Br and Z64S , and the dou-
2 2
1
Regioselective meso,meso’-dibromination of Z2 was per-
formed with 2.2 equiv N-bromosuccinimide (NBS) to give
blet signal for H is low-field shifted in the spectrum of
Z64Br due to the influence of the meso-substituted bro-
2
zinc(ii)–meso,meso’-dibromoporphyrin Z2Br in quantitative
mine, and high-field shifted in Z64S due to the influence of
2
2
yield. This bromination was similarly effective for Z3, Z4,
Z6, Z8, Z12, Z16, Z24, Z48, and Z64 to afford zinc(ii)–me-
so,meso’-dibromoporphyrin arrays Z3Br2 (95%), Z4Br2
the meso-substituted aryl group. The protecting groups in
ZnS2 can be cleaved by subjecting them to phenol/TFA
(2.5% w/v) solution for a certain time to provide meso,me-
(
87%), Z6Br2 (78%), Z8Br2 (83%), Z12Br2 (92%),
Z16Br (84%), Z24Br (89%), Z48Br (91%), and Z64Br
2
so’-phenylthiol-modified
arrays, the details of which will be reported elsewhere.
meso–meso-linked
porphyrin
2
2
2
(
87%), respectively, with high regioselectivity at the meso-
position despite increasing pyrrolic b-positions in longer
arrays, which are potentially susceptible to NBS bromina-
tion. The ratio of meso- to b-positions was as much as 256 in
Z64. It is noteworthy that the present NBS bromination pro-
ceeded nearly quantitatively for all the Zn examined, which
Discussion
The iterative dimerization strategy has proven to be effec-
tive for the preparation of very long meso–meso-linked
Zn –porphyrin arrays. Although the same starting monomer
II
allowed for the isolation of pure ZnBr by passing the prod-
2
Chem. Eur. J. 2005, 11, 3389 – 3404
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3399

T. Matsumoto, T. Kawai, D. Kim, A. Osuka et al.
phyrin is oxidized to its radical cation, the corresponding a
2u
[46]
orbital becomes a magnetic orbital. Such a radical cation
II
will be attacked by a neutral Zn –porphyrin at its meso-po-
sition, which is, in this case, the most nucleophilic site. In
line with this mechanism, the electrochemically generated
radical cation of the zinc(ii)–5,15-bis(3,5-di-tert-butylphe-
nyl)porphyrin complex showed clear splitting in the ESR
spectrum at g=2.0055, due to the coupling with the four
porphyrinic nitrogen atoms (a =1.5 G), thus indicating that
N
one electron was removed from the a2u orbital (Supporting
Information). A big question may be how two giant mole-
cules can find two reacting sites (the end meso-positions) for
each other in the reaction? Considering the large size of the
porphyrin arrays, the coupling reaction must be entropically
very unfavorable. Despite this, the coupling reactions pro-
ceed to give even longer products. One possible explanation
may be the pre-association of porphyrin arrays in CHCl3,
which might bring two or more porphyrin arrays parallel to
each other and their ends close together.
The Zn series is a rare example of a set of exactly dis-
crete, homologous chain molecules covering a very wide
range of molecular sizes (from 1000 to 1000000 Da). Thus,
this series provides a nice opportunity to examine the rela-
tionship of hydrodynamic volume to molecular shape. As
shown in Figure 5b, the molecular weights of Zn measured
using polystyrene standards are much larger than the actual
1
2 2 3
Figure 16. H NMR spectra of a) Z64, b) Z64Br , and c) Z64S in CDCl .
Z1 can be polymerized to macromolecules with dispersed
molecular weights under the slightly activated reaction con-
[
27]
ditions, the stepwise doubling strategy, together with strict
purification of each doubling product, allows for the discrete
preparation of gigantically long molecules in an unprece-
dented manner. Crucial to this success is the high regioselec-
values. Particularly in the case of Z512, the estimated M is
w
about three times larger than the actual M . This is in line
w
with previous reports that the molecular weights of rodlike
[17b,c,47]
molecules,
when estimated by GPC against random
I
tivity of the Ag -promoted oxidative coupling. In the reac-
coil polystyrene standards, are considerably inflated relative
to the actual values. These results thus indicate that Zn do
not take random coil conformation like the polystyrene
standards, but rather as a whole take rodlike conformation
in solution. This relationship may provide important infor-
tion of Zn, we did not observe the formation of any prod-
ucts that have any other form of connectivity. Equally im-
portant is the high solubility of Zn despite their large mo-
lecular weights; this solubility allows for the manipulation of
these large molecules. We noted, however, marked differen-
ces in solubility and stability between the arrays up to Z128
and those longer than Z128. As noted in the Results section,
Z128 can be stored for a long time without any serious
chemical changes, while Z256, for example, tends to become
insoluble when stored in the solid state for several months.
This Z256 solid is insoluble in organic solvents, precluding
its analysis by GPC or UV/Vis spectroscopy. Another evi-
dent advantage of this doubling strategy is the large differ-
ence between the molecular weights of the reactant and
products, which becomes larger as the reactant array be-
comes larger, thus making the separation of coupling prod-
ucts relatively easy over the preparative GPC-HPLC. De-
spite this advantage, the separation of extremely long mole-
cules, such as Z512 and Z1024, required multiple recycled
separations over the GPC. Lastly, the long arrays, such as
Z128 and Z256, still have a self-coupling reactivity that is
sufficient to produce even longer arrays in a reaction time
comparable to those of the shorter arrays.
mation on the hydrodynamic volume of one-dimensional
5
rodlike molecules of molecular size up to M =5ꢃ10 Da
w
and is actually quite useful in estimating the weight-average
[27]
molecular weights of large poly(porphyrinylene)s.
Information on porphyrin arrays in solution may be
gained from their self-diffusion coefficients (D). The
Stokes–Einstein equation [Eq. (1)] relates D to the molecu-
lar parameters:
D ¼ k
T=f
ð1Þ
B
where k is the Boltzmann constant, T is the absolute tem-
B
perature, and f is the frictional coefficient, a measure of the
forces that retard a moleculeꢄs motion. Equation (2) gives
the expression for the frictional coefficient of a spherical
particle:
f ¼ 6 phr
ð2Þ
H
The high regioselectivity for meso–meso coupling ob-
served for zinc(ii)–5,15-diarylporphyrins may be related to
its favorable a2u HOMO. When a zinc(ii)–5,10,15-triarylpor-
where r is the hydrodynamic radius and h is the solvent vis-
cosity. Equations (1) and (2) predict that the self-diffusion
coefficient is inversely proportional to the hydrodynamic
H
3400
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3389 – 3404

Long Porphyrin Arrays
FULL PAPER
radius. Interestingly, it has been predicted that Equation (2)
can be used even for stiff, rodlike macromolecules by using
Equation (3) to calculate rH.
has also been aimed at the engineering of single molecules
and/or atoms and at the exploration of molecular properties
on an individual, nonstatistical basis. In 1996, an IBM re-
II
search group succeeded in obtaining an image of a Cu –tet-
rH ¼ L=2½lnðL=dފ
ð3Þ
rakis-(3,5-di-tert-butylphenyl)porphyrin complex by means
[37a]
of STM.
suitable sampling method, has been applied to the imaging
In recent years, this technique, combined with a
where L and d (L@d) are molecular length and diameter,
[8]
respectively. In the present case, the molecular length of
Zn can be precisely predicted by the number of porphyrins,
provided they have linear rodlike conformations in solution.
Given d=10 ꢂ, their D values lead to an estimate of the ef-
fective L for Z20 and Z24 of 170 and 192 ꢂ, respectively.
These values are consistent with the calculated structures.
[38,39]
of large porphyrin oligomers.
ods, the pulse injection method under high vacuum
Among the imaging meth-
[41]
has
been demonstrated to be particular effective in getting a
good STM image of a large porphyrin array that is decom-
posed at high temperature (>3008C). By employing this
method, we succeeded in acquiring an STM image of Z48
on a Cu(111) surface, which provided strong support for its
discrete structure. However, its largely bent structure was
somewhat unexpected, as a 10,20-meso–meso direct connec-
tion of porphyrins was thought to result in a stiffer, linear
rodlike structure. A similarly bent and more winding struc-
ture was detected for Z128 under comparable conditions. It
is thus conceivable that conformation flexibility of Zn may
be substantially large in solution and may become larger
with increasing molecular size. The observed, roughly
straight, structure of Z96 on HOPG may stem from its fa-
vorable interaction with the graphite crystal lattice, hence
suggesting that the conformational flexibility of porphyrin
arrays depends on their environment.
The rotational correlation time of the molecule (q) is
given by Equation (4):
q ¼ hV=RT
ð4Þ
where R is the gas constant, and V is the volume of the ro-
tating unit. The rotational correlation time is also related to
the diffusion coefficient by q=(6D) . The rotational diffu-
ꢀ
1
sion time, t , is the rotational correlation time per molecule
rot
and is therefore given by Equation (5):
trot ¼ hV=k T
ð5Þ
B
Because the rotational diffusion time can also be mea-
sured by the fluorescence anisotropy decay, we can obtain
similar information about the self-diffusion processes of Zn
in solution from PFG-NMR measurements. For longer
arrays, slower fluorescence anisotropy decay times were ob-
Compared to the chain elongation steps, the purification
issue is more serious for modifications of the porphyrin
arrays. Separation of ZnBr from Zn was practically impos-
2
sible over the preparative GPC because of the small differ-
ence in molecular weight of the products, and serious tailing
of the long arrays on silica gel or alumina columns preclud-
ed the separation of these products. This obstacle has been
circumvented by the quantitative NBS bromination of the
array. This is also a remarkable finding, in that macromole-
cules in general tend to become less reactive upon increas-
ing in size relative to small molecules. The NBS bromination
proceeded in a quantitative manner at least up to Z64, in
which there are only two free meso-positions versus 512 free
pyrrolic b-positions that are all potentially susceptible to
NBS bromination. Nevertheless, NBS bromination occurred
exclusively only at both meso-positions, hence fabricating
synthetic handles just at the edges of long, rodlike mole-
[23]
served, mainly due to an increase in molecular hydrody-
namic volume caused by an increase in the total length of
Zn.
A plot of the molecular extinction coefficient of Zn
versus the number of porphyrins (n) indicated a linear sum-
mation behavior in absorption intensity as the arrays
became longer. On the other hand, in arrays longer than
Z16, the fluorescence quantum yield gradually decreased
and the contribution by the short component became larger,
exhibiting a good correlation with fluorescence quantum
yield. This spectroscopic data has been ascribed to confor-
mational heterogeneity in longer porphyrin arrays, and the
degree of heterogeneity increases as the arrays become
longer. These conformational heterogeneities are believed
to arise from bending of the main chain, as well as dihedral
angle distribution between adjacent porphyrin planes. On
the whole, we may conclude that the meso–meso-linked por-
phyrin arrays take a roughly linear rodlike conformation de-
spite substantial conformation flexibility and heterogeneity.
Based on this, the observed change of freshly separated,
long array samples to insoluble material may indicate possi-
ble entanglement of the main chain. Another explanation
would be the usual aggregation-induced precipitation.
cules. Furthermore, these ZnBr arrays have been quantita-
2
tively arylated with a boronate smoothly by means of a Pd-
catalyzed Suzuki–Miyaura coupling reaction. These both-
ends-manipulated porphyrin arrays will be quite useful in a
variety of embryonic fields, including molecular-wire, sensor,
and molecular electronics.
Conclusion
We have succeeded in synthesizing meso–meso directly
linked zinc(ii)–porphyrin arrays up to 1024-mer (Z1024) in a
Achieving direct access to a single molecule has become a
growing area of contemporary, interdisciplinary research
since the discovery of STM in 1982. Research in this area
I
discrete manner by a repeated Ag –salt-promoted meso–
[
36]
meso coupling reaction. The arrays up to Z128 are fully
Chem. Eur. J. 2005, 11, 3389 – 3404
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3401

T. Matsumoto, T. Kawai, D. Kim, A. Osuka et al.
1
characterized by H NMR spectroscopy, MALDI-TOF mass
Table 3. Summary of crystallographic data for H1 and Z
p
3.
1
spectrometry, and GPC-HPLC. The H NMR spectrum of
H1
Zp3
Z256 was recorded, but its parent-ion peak could not be
found in its MALDI-TOF mass spectrum. The actual molec-
ular lengths of Z48 and Z128 were confirmed by STM, and
those of Z96, Z128, Z256, and Z512 were confirmed by
AFM. We also succeeded in the manipulation of both ends
of the porphyrin arrays through NBS bromination and sub-
sequent Suzuki–Miyaura arylation (Zn to ZnBr to ZnS )
formula
C
975.41
0.7ꢃ0.2ꢃ0.1
triclinic
P1 (#2)
9.150(4)
11.801(8)
13.869(7)
84.42(5)
74.47(4)
76.70(4)
1403(1)
1.154
64
H
86
N
4
O
4
110 66 12 3
C H N O12Zn
1943.94
M
r
crystal size [mm]
crystal system
space group
a [ꢂ]
b [ꢂ]
c [ꢂ]
0.2ꢃ0.15ꢃ0.10
monoclinic
¯
¯
P1 (#2)
12.092(2)
13.123(4)
16.037(4)
78.27(1)
73.07(2)
88.72(2)
2381.9(9)
1.355
1
123
11422
11422
490
2
2
a [8]
b [8]
up to the Z64 series. In solution, the Zn arrays are consid-
ered to overall take linear rodlike structures, but with sub-
stantial conformational flexibility as judged from their
g [8]
V [ꢂ ]
3
1
ꢀ3
H NMR spectra, diffusion coefficients, eluting behavior on
1
calcd [gcm
]
Z
T [K]
no. of reflns
no. of unique reflns
variables
1
123
12808
6404
368
GPC-HPLC, and UV/Vis and fluorescence spectra. The con-
formational flexibility has been shown by STM images of
single molecules on a Cu(111) surface and AFM images of
single molecules on HOPG and Al O surfaces. The chemis-
2
3
try described here reveals a new area for nanoscaled and
sub-microscaled functional materials.
l
MoKa [ꢂ]
q
max
0.71073
55.0
1.091
0.71073
55.0
1.113
2
GOF
max. peak [eꢂ
min. peak [eꢂ
R1 [I>3.0s(I)]
wR2 [I>3.0s(I)]
ꢀ
3
]
0.30
1.29
ꢀ
3
]
ꢀ0.45
ꢀ0.99
0.069
0.113
0.093
0.122
Experimental Section
General procedure: All reagents and solvents were of commercial re-
agent grade and were used without further purification except where
from the Cambridge Crystallographic Data Centre via www.ccdc.cam.
ac.uk/data_request/cif.
noted. Dry CH
2
Cl
H NMR spectra were recorded on JEOL ALPHA-500 and
ECA-600 spectrometers, and chemical shifts were reported in the delta
scale relative to a CHCl internal standard (d=7.26 ppm). Spectroscopic-
2 3
and CHCl were obtained by refluxing and distilling
1
over CaH
2
.
STM and AFM measurement conditions: For STM measurements Cu-
(100) or Cu(111) were used as substrates. Clean, flat surfaces of Cu(100)
+
3
were obtained from a single crystal by Ar sputtering and annealing
grade tetrahydrofuran (THF) and toluene were used as solvents for all
experiments. UV/Vis absorption spectra were recorded on a Shimadzu
UV-2400PC spectrometer. Steady-state fluorescence emission spectra
were recorded on a Shimadzu RF-5300PC spectrometer. Mass spectra
were recorded on a JEOL HX-110 spectrometer by means of the positive
FAB ionization method with an accelerating voltage of 10 kV and a 3-ni-
(5508C) cycles, while substrates of Cu(111) were made on a cleaved mica
surface by thermal evaporation. The arrays were dissolved into the
proper solvent (CHCl or CH Cl ) to an approximate concentration of
3
2
2
ꢀ
5
ꢀ1
10 molL . They were deposited by spraying the solution (ca. 0.5 mL)
ꢀ6
onto the substrate under high vacuum (10 mbar) using the pulse injec-
tion method, which was adequate for the arrays too large to be thermally
evaporated. STM imaging was performed in situ in constant-height mode
trobenzyl alcohol matrix, or on
a Shimadzu/KRATOS KOMPACT
ꢀ
10
MALDI4 spectrometer by using the positive MALDI-TOF method with
a 9-nitroanthracene matrix. Redox potentials were measured by cyclic
voltammetry and differential pulse voltammetry on a BAS electrochemi-
cal analyzer (660 model). GPC-HPLC was performed by using a JASCO
HPLC system with JAIGEL 2.5H-AF, 3H-AF, 4H-AF, and 5H-AF col-
umns in series, and with a multiwavelength detector MD-915. Recycling
preparative GPC-HPLC was carried out with a JAI LC-908 that used
preparative JAIGEL-2.5H, 3H, 4H, and 5H columns. Preparative sepa-
rations were performed by silica-gel flash column chromatography
at room temperature under ultra-high vacuum (<10 mbar), by using a
home-built STM with an electrochemically etched Pt/Ir tip. For AFM
measurements, the arrays were deposited by placing drops of a dilute so-
ꢀ
5
ꢀ1
lution of the array in CHCl (ca. 10 molL ) onto a tilted sapphire sub-
3
strate (Al O (0001)) and leaving to air-dry under ambient conditions.
2
3
AFM imaging was performed by using the intermittent-contact mode
under ambient conditions with a commercially available AFM (Nanosco-
peIII, Digital Instruments Ltd.) with Si cantilevers (resonant frequency
f=300 kHz).
(
Merck Kieselgel 60H Art.7736), silica-gel gravity column chromatogra-
phy (Wako gel C-200), and preparative size-exclusion chromatography
BioRad Bio-Beads S-X1 packed in chloroform in a 4.0ꢃ100 cm HPLC
(
ꢀ1
column; flow rate 3.8 mLmin ).
Acknowledgements
Single-crystal diffraction analysis: Data was collected on a Rigaku R-axis
imaging plate refined by full-matrix least-squares procedures with aniso-
tropic thermal parameters for the non-hydrogen atoms. The hydrogen
atoms were calculated in ideal positions. Solutions of structures were per-
formed by using the “Crystal Structure” crystallographic software pack-
This work was supported by CREST (Core Research for Evolutional Sci-
ence and Technology) of the Japan Science and Technology Agency
(JST). We thank Prof. T. Kato and Dr. K. Furukawa for taking the ESR
measurements. N.A. thanks the JSPS Research Fellowship for Young Sci-
entists and the Ogasawara Foundation for the Promotion of Science &
Engineering.
age (Molecular Structure Corporation). Compound H1 (C64
Crystals were obtained by vapor diffusion of EtOH into a CH
lution of H1; a red-prism crystal of dimensions 0.7ꢃ0.2ꢃ0.1 mm was
used for calculations. Compound Z Zn ): Crystals were
obtained by vapor diffusion of MeOH into a CH Cl solution of Z 3; a
H
86
N
4
O
4
):
2
Cl
2
so-
P
3 (C110
H
66
N
12
O
3
3
2
2
P
red prism crystal of dimensions 0.2ꢃ0.15ꢃ0.1 mm was used for calcula-
tions. The crystallographic details are summarized in Table 3. CCDC-
[1] a) R. E. Martin, F. Diederich, Angew. Chem. 1999, 111, 1440;
Angew. Chem. Int. Ed. 1999, 38, 1350; b) J. M. Tour, Chem. Rev.
1996, 96, 537; c) P. F. H. Schwab, M. D. Levin, J. Michl, Chem. Rev.
1999, 99, 1863; d) J. Cornil, D. Beljonne, J.-P. Calbert, J.-L. Brꢅdas,
2
61869 (H1) and CCDC-261870 (Z
P
3) contain the supplementary crystal-
lographic data for this paper. These data can be obtained free of charge
3402
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3389 – 3404

Long Porphyrin Arrays
FULL PAPER
Adv. Mater. 2001, 13, 1053; e) A. Kraft, A. C. Grimsdale, A. B.
Holmes, Angew. Chem. 1998, 110, 416; Angew. Chem. Int. Ed. 1998,
[21] For other examples of meso–meso-linked porphyrin arrays, see a) K.
Susumu, T. Shimidzu, K. Tanaka, H. Segawa, Tetrahedron Lett. 1996,
37, 8399; b) R. G. Khoury, L. Jaquinod, K. M. Smith, Chem.
Commun. 1997, 1057; c) M. O. Senge, X. Feng, Tetrahedron Lett.
1999, 40, 4165; d) J. Wojaczynski, L. Latos-Grazynski, P. J. Chmie-
lewski, P. Van Calcar, A. L. Balch, Inorg. Chem. 1999, 38, 3040;
e) X. Shi, L. S. Liebeskind, J. Org. Chem. 2000, 65, 1655; f) K.
Ogawa, Y. Kobuke, Angew. Chem. 2000, 112, 4236; Angew. Chem.
Int. Ed. 2000, 39, 4070; g) H. Imahori, K. Tamaki, Y. Araki, Y. Seki-
guchi, O. Ito, Y. Sakata, S. Fukuzumi, J. Am. Chem. Soc. 2002, 124,
5165; h) D. Bonifazi, F. Diederich, Chem. Commun. 2002, 2178;
i) M. A. Miller, R. K. Lammi, S. Prathapan, D. Holten, J. S. Lindsey,
J. Org. Chem. 2000, 65, 6634; j) M. Ottonelli, Synth. Met. 2003, 138,
3
7, 402; f) J. S. Moore, Acc. Chem. Res. 1997, 30, 402; g) D. T.
McQuade, A. E. Pullen, T. M. Swager, Chem. Rev. 2000, 100, 2537;
h) M. D. Watson, A. Fechtenkçtter, K. Mꢁllen, Chem. Rev. 2001,
1
01, 1267; i) D. Zhao, J. S. Moore, Chem. Commun. 2003, 807; j) W.-
S. Li, D.-L. Jiang, T. Aida, Angew. Chem. 2004, 116, 3003; Angew.
Chem. Int. Ed. 2004, 43, 2943.
2] a) C. K. Chiang, C. R. Fincher, Jr., Y. W. Park, A. J. Heeger, H. Shir-
akawa, E. J. Louis, S. C. Gau, A. G. MacDiarmid, Phys. Rev. Lett.
[
1
1
977, 39, 1098; b) M. D. McGehee, A. J. Heeger, Adv. Mater. 2000,
2, 1655; c) C. J. Brabec, N. S. Sariciftci, J. C. Hummelen, Adv.
Funct. Mater. 2001, 11, 15.
3] Z. Grubisic, P. Rempp, H. Benoit, J. Polym. Sci. Part B Polym. Lett.
[
1
73.
1
967, 5, 753.
[
[
[
22] N. Aratani, H. S. Cho, T. K. Ahn, S. Cho, D. Kim, H. Sumi, A.
Osuka, J. Am. Chem. Soc. 2003, 125, 9668.
23] Y. H. Kim, D. H. Jeong, D. Kim, S. C. Jeoung, H. S. Cho, S. K. Kim,
N. Aratani, A. Osuka, J. Am. Chem. Soc. 2001, 123, 76.
24] a) A. Tsuda, H. Furuta, A. Osuka, Angew. Chem. 2000, 112, 2649;
Angew. Chem. Int. Ed. 2000, 39, 2549; b) A. Tsuda, H. Furuta, A.
Osuka, J. Am. Chem. Soc. 2001, 123, 10304; c) A. Tsuda, A. Osuka,
Science 2001, 293, 79; d) H. S. Cho, D. H. Jeong, S. Cho, D. Kim, Y.
Matsuzaki, K. Tanaka, A. Tsuda, A. Osuka, J. Am. Chem. Soc. 2002,
[
[
4] P. M. Kovacic, B. Jones, Chem. Rev. 1987, 87, 357.
5] a) G. Wegner, Angew. Chem. 1981, 93, 352; Angew. Chem. Int. Ed.
Engl. 1981, 20, 361; b) J. G. Speight, P. Kovacicand, F. W. Koch, J.
Macromol. Sci. Rev. Macromol. Chem. 1971, C5, 295.
6] a) T. Yamamoto, Bull. Chem. Soc. Jpn. 1999, 72, 621; b) D. Fichou,
in Handbook of Oligo- and Polythiophenes, Wiley-VCH, Weinheim,
[
1
999.
[
[
7] J. M. Tour, Adv. Mater. 1994, 6, 190.
8] H. Yamakawa, in Modern Theory of Polymer Solutions, Harper &
Row, New York, 1971.
9] P. Liess, V. Hensel, A. D. Schlꢁter, Liebigs Ann. 1996, 1037.
10] V. Hensel, A. D. Schlꢁter, Chem. Eur. J. 1999, 5, 421.
1
24, 14642.
[
25] a) P. S. Crezy, G. A. Smith, Aust. J. Chem. 1969, 22, 239; b) Q. M.
Wang, D. W. Bruce, Synlett 1995, 1267; c) B. J. Littler, M. A. Miller,
C.-H. Hung, R. W. Wagner, D. F. OꢄShea, P. D. Boyle, J. S. Lindsey,
J. Org. Chem. 1999, 64, 1391; d) T. E. O. Screen, K. B. Lawton, G. S.
Wilson, N. Dolney, R. Ispasoiu, T. Goodson III, S. J. Martin, D. D. C.
Bradley, H. L. Anderson, J. Mater. Chem. 2001, 11, 312; e) J.-W. Ka,
C.-H. Lee, Tetrahedron Lett. 2000, 41, 4609.
[
[
[
11] U. Stalmach, H. Kolshorn, I. Brehm, H. Meier, Liebigs Ann. 1996,
1
449.
[
12] a) J. M. Tour, Acc. Chem. Res. 2000, 33, 791; b) L. Jones II, J. S.
Schumm, J. M. Tour, J. Org. Chem. 1997, 62, 1388.
13] S. Huang, J. M. Tour, J. Org. Chem. 1999, 64, 8898.
14] A. Cravino, N. S. Sariciftci, J. Mater. Chem. 2002, 12, 1931.
15] a) J. Roncali, Acc. Chem. Res. 2000, 33, 147; b) J. Roncali, Chem.
Rev. 1997, 97, 173; c) I. Jestin, P. Frere, P. Blanchard, J. Roncali,
Angew. Chem. 1998, 110, 990; Angew. Chem. Int. Ed. 1998, 37, 942.
16] a) M. Tachibana, S. Tanaka, Y. Yamashita, K. Yoshizawa, J. Phys.
Chem. B 2002, 106, 3549; b) S. Tanaka, Y. Yamashita, Synth. Met.
[
[
[
[
26] a) J. S. Lindsey, I. C. Schreiman, H. C. Hsu, P. C. Kearney, A. M.
Marguerettaz, J. Org. Chem. 1987, 52, 827; b) S. G. DiMagno, V. S.-
Y. Lin, M. J. Therien, J. Org. Chem. 1993, 58, 5983.
[
[
27] N. Yoshida, N. Aratani, A. Osuka, Chem. Commun. 2000, 197.
28] A. Nakano, H. Shimidzu, A. Osuka, Tetrahedron Lett. 1998, 39,
[
[
9
489.
II
[
29] The X-ray crystal structure of meso–meso-linked Cu –diporphyrin
was reported. N. Yoshida, T. Ishizuka, A. Osuka, D. H. Jeong, H. S.
Cho, D. Kim, Y. Matsuzaki, A. Nagami, K. Tanaka, Chem. Eur. J.
1
995, 69, 599.
17] a) T. Izumi, S. Kobashi, K. Takimiya, Y. Aso, T. Otsubo, J. Am.
Chem. Soc. 2003, 125, 5286; b) T. Otsubo, Y. Aso, K. Takimiya, Bull.
Chem. Soc. Jpn. 2001, 74, 1789; c) K. Inouchi, S. Kobashi, K. Taki-
miya, Y. Aso, T. Otsubo, Org. Lett. 2002, 4, 2533.
18] a) M. R. Wasielewski, Chem. Rev. 1992, 92, 435; b) D. Gust, T. A.
Moore, A. L. Moore, Acc. Chem. Res. 1993, 26, 198; c) D. Gust,
T. A. Moore, A. L. Moore, Acc. Chem. Res. 2001, 34, 40; d) D.
Holten, D. F. Bocian, J. S. Lindsey, Acc. Chem. Res. 2002, 35, 57;
e) A. K. Burrell, D. L. Officer, P. G. Plieger, D. C. W. Reid, Chem.
Rev. 2001, 101, 2751.
2
003, 9, 58.
[
30] For other examples of X-ray crystal structures of singly, directly
linked diporphyrins, see reference [21b] and a) Y. Deng, C. K.
Chang, D. G. Nocera, Angew. Chem. 2000, 112, 1108; Angew. Chem.
Int. Ed. 2000, 39, 1066; b) Y. Nakamura, N. Aratani, A. Tsuda, A.
Osuka, K. Furukawa, T. Kato, J. Porphyrins Phthalocyanines 2003,
[
7
, 264.
31] a) B. Aurivillius, R. E. Carter, J. Chem. Soc. Perkin Trans. 2 1978,
033; b) M. Nishino, H. Hirota, Tetrahedron 1989, 45, 7201; c) H. Ta-
[
1
[
19] a) E. K. L. Yeow, K. P. Ghiggino, J. N. H. Reek, M. J. Crossley, A. W.
Bosman, A. P. H. J. Schenning, E. W. Meijer, J. Phys. Chem. B 2000,
kahashi, S. Tsuboyama, Y. Umezawa, K. Honda, M. Nishio, Tetrahe-
dron 2000, 56, 6185; d) M. Nishio, CrystEngComm 2004, 6, 130.
32] a) T. Ogawa, Y. Nishimoto, N. Yoshida, N. Ono, A. Osuka, Chem.
Commun. 1998, 337; b) T. Ogawa, Y. Nishimoto, N. Yoshida, N.
Ono, A. Osuka, Angew. Chem. 1999, 111, 140; Angew. Chem. Int.
Ed. 1999, 38, 176.
1
04, 2596; b) M.-S. Choi, T. Aida, T. Yamazaki, I. Yamazaki, Chem.
[
Eur. J. 2002, 8, 2668; c) O. Mongin, A. Schuwey, M.-A. Vallot, A.
Gossauer, Tetrahedron Lett. 1999, 40, 8347; d) S. Anderson, H. L.
Anderson, J. K. M. Sanders, Acc. Chem. Res. 1993, 26, 469; e) 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;
f) H. A. M. Biemans, A. E. Rowan, A. Verhoeven, P. Vanoppen, 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;
g) K.-i. Sugiura, Y. Fujimoto, Y. Sakata, Chem. Commun. 2000,
[
[
33] M. Kasha, Rev. Mod. Phys. 1959, 31, 162.
34] C. S. Johnson, Jr., Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 203.
[35] a) B. Olenyuk, M. D. Levin, J. A. Whiteford, J. E. Shield, P. J. Stang,
J. Am. Chem. Soc. 1999, 121, 10434; b) W. H. Otto, M. H. Keefe,
K. E. Splan, J. T. Hupp, C. K. Larive, Inorg. Chem. 2002, 41, 6172;
c) W. S. Jeon, A. Y. Ziganshina, J. W. Lee, Y. H. Ko, J.-K. Kang, C.
Lee, K. Kim, Angew. Chem. 2003, 115, 4231; Angew. Chem. Int. Ed.
2003, 42, 4097.
1
105; h) H. S. Cho, H. Rhee, J. K. Song, C.-K. Min, M. Takase, N.
Aratani, S. Cho, A. Osuka, T. Joo, D. Kim, J. Am. Chem. Soc. 2003,
25, 5849.
1
[
20] For preliminary reports of this work, see a) A. Osuka, H. Shimidzu,
[36] G. Binnig, H. Rohrer, Ch. Gerber, E. Weibel, Phys. Rev. Lett. 1982,
49, 57.
Angew. Chem. 1997, 109, 93; Angew. Chem. Int. Ed. Engl. 1997, 36,
1
35; b) N. Aratani, A. Osuka, Y. H. Kim, D. H. Jeong, D. Kim,
Angew. Chem. 2000, 112, 1517; Angew. Chem. Int. Ed. 2000, 39,
458.
[37] a) T. A. Jung, R. R. Schlittler, J. K. Gimzewski, H. Tang, C. Joachim,
Science 1996, 271, 181; b) J. K. Gimzewski, C. Joachim, Science 1999,
283, 1683.
1
Chem. Eur. J. 2005, 11, 3389 – 3404
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3403

T. Matsumoto, T. Kawai, D. Kim, A. Osuka et al.
[
38] a) H. L. Anderson, Angew. Chem. 2000, 112, 2552; Angew. Chem.
Int. Ed. 2000, 39, 2451; b) N. Bampos, C. N. Woodburn, M. E. Well-
and, J. K. M. Sanders, Angew. Chem. 1999, 111, 2949; Angew. Chem.
Int. Ed. 1999, 38, 2780; c) K.-i. Sugiura, H. Tanaka, T. Matsumoto,
T. Kawai, Y. Sakata, Chem. Lett. 1999, 1193; d) K.-i. Sugiura, T.
Matsumoto, S. Ohkouchi, K. Naitoh, T. Kawai, Y. Takai, K. Ushiro-
da, Y. Sakata, Chem. Commun. 1999, 1957.
Hyslop, M. A. Kellett, P. M. Iovine, M. J. Therien, J. Am. Chem.
Soc. 1998, 120, 12676; f) N. Aratani, A. Osuka, Org. Lett. 2001, 3,
4213.
[43] D. L. Pearson, J. M. Tour, J. Org. Chem. 1997, 62, 1376.
[44] I. Photaki, J. T. -Papadimitriou, C. Sakarellos, P. Mazarakis, L.
Zervas, J. Chem. Soc. C 1970, 2683.
[45] a) T. Ishiyama, M. Murata, N. Miyaura, J. Org. Chem. 1995, 60,
7508; b) M. Murata, S. Watanabe, Y. Masuda, J. Org. Chem. 1997,
62, 6458; c) M. Murata, T. Oyama, S. Watanabe, Y. Masuda, J. Org.
Chem. 2000, 65, 164.
[
[
39] a) A. Takagi, Y. Yanagawa, A. Tsuda, N. Aratani, T. Matsumoto, A.
Osuka, T. Kawai, Chem. Commun. 2003, 2986; b) X. Peng, N. Arata-
ni, A. Takagi, T. Matsumoto, T. Kawai, I.-W. Hwang, T. K. Ahn, D.
Kim, A. Osuka, J. Am. Chem. Soc. 2004, 126, 4468.
40] E. Mena-Osteritz, A. Meyer, B. M. W. Langeveld-Voss, R. A. J. Jans-
sen, E. W. Meijer, P. Bꢆuerle, Angew. Chem. 2000, 112, 2791; Angew.
Chem. Int. Ed. 2000, 39, 2679.
[46] M. Kamo, A. Tsuda, Y. Nakamura, N. Aratani, K. Furukawa, T.
Kato, A. Osuka, Org. Lett. 2003, 5, 2079.
[47] a) D. L. Pearson, J. S. Schumm, J. M. Tour, Macromolecules 1994, 27,
2348; b) J. S. Schumm, D. L. Pearson, J. M. Tour, Angew. Chem.
1994, 106, 1445; Angew. Chem. Int. Ed. Engl. 1994, 33, 1360; c) N.
Sumi, H. Nakanishi, S. Ueno, K. Takimiya, Y. Aso, T. Otsubo, Bull.
Chem. Soc. Jpn. 2001, 74, 979.
[
[
41] H. Tanaka, T. Nakagawa, T. Kawai, Surf. Sci. 1996, 364, L575.
42] a) X. Zhou, K. S. Chan, J. Org. Chem. 1998, 63, 99; b) D. A. Shultz,
H. Lee, R. K. Kumar, K. P. Gwaltney, J. Org. Chem. 1999, 64, 9124;
c) T. Mizutani, K. Wada, S. Kitagawa, J. Am. Chem. Soc. 2001, 123,
Received: December 20, 2004
6
459; d) L. Yu, J. S. Lindsey, Tetrahedron 2001, 57, 9285; e) A. G.
Published online: March 30, 2005
3404
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3389 – 3404