Enhanced π conjugation around a porphyrin[6] nanoring

Article abstract of DOI:10.1002/anie.200801188

(Figure Presented) Strong cycle: The cyclic hexamer-template complex 3 obtained through template-directed trimerization of a porphyrin dimer 2, using a hexapyridyl template 1, is extremely stable (K_f=7×10 ³⁸m^-1), but the free macrocycle 4 can be liberated using amine ligands. Spectroscopic data and DFT calculations show that the cyclic hexamer is more conjugated than its linear analogue.

Full text of DOI:10.1002/anie.200801188

Angewandte
Chemie
DOI: 10.1002/anie.200801188
Strained Molecular Wires
Enhanced p Conjugation around a Porphyrin[6] Nanoring**
Markus Hoffmann, Joakim Kärnbratt, Ming-Hua Chang, Laura M. Herz, Bo Albinsson, and
Harry L. Anderson*
Dedicated to Professor Jeremy Sanders on the occasion of his 60th birthday
Belt-shaped chromophores provide fascinating insights
into electronic p delocalization over curved surfaces
with radially oriented p orbitals.[1] Examples include
the
cyclic
para-phenylacetylenes[2]
and
the
[46]paracyclophanedodecayne of Tsuji and coworkers,[3]
as well as fullerenes and carbon nanotubes. A variety of
belt-shaped porphyrin arrays have been synthesized;[4]
however, the vast majority of them lacks a complete p-
conjugation pathway around the whole macrocycle.
Recently we reported the synthesis of a belt-shaped
D8h symmetric porphyrin[8] nanoring on an octaden-
tate template.[5] Herein we present an efficient synthesis
of an even more strained p conjugated D6h porphyrin[6]
nanoring 1, by template-directed trimerization of a
porphyrin dimer 2 on a hexapyridyl template 3
(Scheme 1). This route is more direct than the synthesis
of the cyclic octamer since both starting materials, 2 and
3, are readily accessible. The cyclic hexamer complex
1·3 is phenomenally stable (Kf = 7 ” 1038 mꢀ1; EM =
340m) but the free macrocycle can be liberated from
the 1·3 complex with amines such as quinuclidine. The
UV/Vis/NIR absorption and emission show that there
is efficient p conjugation around the porphyrin[6]
nanoring 1, and that its S0–S1 gap is even smaller than
that of the corresponding linear porphyrin hexamer 4;
this conclusion is supported by time-dependent density
functional (TD-DFT) calculations.
Scheme 1. Synthesis of porphyrin[6] nanorings 1a and 1b. Reagents:
a) [PdCl2(PPh3)2], CuI, I2, iPr2NH, air, 608C; (b) DABCO.
The key to the synthesis of the hexamer nanoring is
the design of a complementary template (Scheme 2).
The hexadentate template 3 has a calculated nitrogen–
[*] M. Hoffmann, Prof. H. L. Anderson
Department of Chemistry, Oxford University
Chemistry Research Laboratory
12 Mansfield Road, Oxford OX13TA (UK)
Fax: (+44)1865-28-5002
E-mail: harry.anderson@chem.ox.ac.uk
Homepage: http://users.ox.ac.uk/~hlagroup
M.-H. Chang, Dr. L. M. Herz
Department of Physics, Oxford University
J. Kärnbratt, Prof. B. Albinsson
nitrogen distance of 20.1 Š, which is a good fit for the cavity
of cyclic hexamer 1 with a zinc–zinc ring diameter of 24.2 Š
Department of Chemical and Biological Engineering
Physical and Organic Chemistry, Chalmers University of Technology
Kemivägen 10, 41296 Göteborg (Sweden)
ꢀ
(assuming a Zn N bond length of 2.2 Š). The template was
synthesized by a six-fold Suzuki coupling of 4-pyridineboronic
acid with hexakis(4-bromophenyl)benzene[6] in 50% yield
(Supporting Information). Two versions of cyclic hexamer 1
were synthesized—1a·3 with tert-butyl side chains in 44% and
1b·3 with octyloxy side chains in 33% yield—by oxidative
[**] This work was supported by EPSRC. We thank Johannes K. Sprafke
for preliminary experiments on the synthesis of the template, and
the EPSRCMass Spectrometry Service (Swansea) for mass spectra.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200801188.
Angew. Chem. Int. Ed. 2008, 47, 4993 –4996
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4993

Communications
Very large equilibrium constants for the formation of
complexes—as in the case of 1·3—can be measured by
analyzing the thermodynamics of template displacement with
a competing monodentate ligand.[5] Thus UV/Vis titrations
were used to quantify the displacement of the template 3
from 1b·3, and from the complex of the linear hexamer, 4·3.
All these titrations show simple isosbestic behavior, and the
binding isotherms fit well to the calculated curves for two-
state equilibria with Kb values of 4.5 ” 10ꢀ3 mꢀ5 and 2.4”
104 mꢀ5 for 1b·3 and 4·3, respectively, which correspond to
Kf values of (6.6 ꢁ 4.2) ” 1038 mꢀ1 and (1.4 ꢁ 0.9) ” 1021mꢀ1
,
respectively. Comparison of these two binding constants
implies that the Gibbs energy required to bend the linear
hexamer 4 into a cyclic conformation is 101 kJmolꢀ1. The
complementarity of template 3 to the porphyrin hexamers can
be quantified by the effective molarity (EM) according to
Equation (1), where K0 is the binding constant of one arm of
Scheme 2. Optimized geometries of 23·3 and 1·3, calculated using the
MM+ force field (meso-aryl substituents omitted for clarity).
coupling of porphyrin dimer 2a/b under palladium/copper
catalysis, using iodine and air as oxidants.[7] Size-exclusion
chromatography in the presence of 1,4-diazabicyclo-
[2.2.2]octane (DABCO) gave the template-free nanorings
1a and 1b.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
5
ꢀ6
ð1Þ
EM ¼ Kf K0
1H NMR spectroscopic analysis proves the high symmetry
of 1·3 and 1 (Figure 1). Each compound gives just two sharp
doublets for the b protons on the porphyrin units. The aryl
and tBu resonances in 1a·3 are split as the faces of the
the template for one site of the hexamer. K0 can be
approximated to the binding constant for 4-phenylpyridine
and a 5,10-diethynylporphyrin zinc monomer (K0 = 2.3 ”
104 mꢀ1), which gives EM values of (340 ꢁ 60) and (0.10 ꢁ
0.02)m for 1b·3 and 4·3, respectively. Each of these effective
molarities is an average of five values, for coordination of the
second, third, fourth, fifth, and sixth sites of the template to
the porphyrin hexamer. The effective molarity for forming
1b·3 is an extremely high value for a noncovalent self-
assembly process,[8] and it is consistent with the fact that
pyridine is not able to displace the template.
To test how curvature changes the p conjugation in these
wires, we compared the absorption and fluorescence spectra
of the hexamer nanoring 1, and its template complex 1·3, with
those of the linear hexamer 4 and its template complex 4·3
(Figure 2). The template-bound nanoring 1a·3 exhibits a
sharp structured absorption band with maxima at 772, 809,
and 850 nm, and a shoulder at 905 nm. The template-free
cyclic hexamer 1a has a similar multiplet pattern, although its
spectrum is less well resolved, indicating greater conforma-
tional flexibility. The absorption spectrum of the linear
hexamer 4 is much broader with a maximum at 804nm, but
when coordinated to the template to form 4·3 it splits into four
peaks at 732, 781, 821, and 887 nm. The fluorescence spectra
of 1a, 1a·3, and 4·3 (lmax = 896, 914, and 894 nm, respectively)
are significantly red-shifted compared to that of the linear
hexamer 4 (lmax = 827 nm), showing that the nanoring has a
smaller S0–S1 gap and confirming that p conjugation is
effective around these curved porphyrin wires.
Changing the geometry of the hexamer from linear to
cyclic imposes symmetry-related constraints on the electronic
transitions. This can be understood qualitatively in terms of a
simple exciton model, assigning one transition dipole moment
to each porphyrin (Scheme 3). The lowest-energy transition,
corresponding to a head-to-tail arrangement of transition
moments, is strong for the linear hexamer 4, but forbidden for
the nanoring 1 because the transition moments cancel. This
model is of course very crude, but TD-DFT calculations give a
Figure 1. 1H NMR spectrum of 1a·3 (CDCl3, 298 K, 500 MHz).
porphyrin rings are nonequivalent with one side of each aryl
substituent pointing towards the center of the ring. After
removal of the template the faces of the porphyrin rings
become equivalent, probably because the porphyrin units
rotate rapidly on the NMR time-scale.
Previously, we reported that the template is displaced
from our porphyrin[8] nanoring with pyridine.[5] However,
pyridine does not displace the template from 1a·3 and 1b·3,
even when the complexes are dissolved in neat pyridine
([pyridine] = 12.4m). More strongly coordinating amines,
such as quinuclidine do displace the template from these
complexes, when used in large excess ([quinuclidine] > 0.4m;
250000 equiv).
4994 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 4993 –4996

Angewandte
Chemie
S2 transition. Several experimental observations corroborate
this model. First, the total lack of mirror symmetry between
the absorption and emission spectra of 1a·3 points to a
complex absorption band composed of several electronic
transitions rather than a vibronic progression. Second, the
fluorescence quantum yield Ff drops from 0.08 to 0.002 when
comparing 4 and 1a and the radiative rate constant kf
decreases by nearly two orders of magnitude (Table 1),
indicating a dramatic reduction in the S0–S1 oscillator strength
in the nanoring. Conformational flexibility evidently relaxes
this symmetry-related selection rule: the nanoring–template
complex 1a·3 with the most rigid symmetrical structure has
the lowest quantum yield and shows the slowest radiative
decay; both parameters increase in the more flexible free
nanoring 1, and in the less symmetrical complex 4·3.
Further insights into the electronic structure are provided
by comparing the observed radiative rate constants with those
estimated from the absorption spectra through the Strickler–
Berg relation,[9] where n is the transition frequency and e is the
Z
kf ¼ 2:88 ꢂ 10ꢀ9n2 eðnÞdn
ð2Þ
Figure 2. Vis/NIR absorption spectra (a) and emission spectra (b) of
1a·3 (red, plain) excited at 850 nm, 1a (blue, dashed) excited at
870 nm, 4 (black, dotted-dashed) excited at 499 nm and 4·3 (green,
bold) excited at 499 nm in toluene. 1% pyridine was added to 1a and
4 to prevent aggregation. The weak emission spectra of 1a·3 and 1a
are slightly distorted from scattering of the excitation light. The small
peak at 760 nm in the emission spectrum for 4·3 is due to a trace of a
more fluorescent impurity. e: extinction coefficient; Inorm: normalized
photoluminescence.
molar absorptivity. The radiative rate constants, kf, were
estimated from the absorption spectra first assuming that the
whole Q-band at 650–950 nm corresponds to the S0–S1
transition (Table 1). The results from the Strickler–Berg
relation are in agreement with the observed radiative rate
constant for the linear hexamer, but are greatly overestimated
for the other compounds. A much better agreement for the
cyclic structures is achieved if only the weak most red-shifted
peak in the absorption spectrum is considered.[10] This
confirms that the S0–S1 transitions for the cyclic structures
are weakly allowed and lie at around 900 nm, as suggested by
the exciton model and the TD-DFT calculations.
In summary, we have synthesized a strained p conjugated
porphyrin nanoring by template-directed trimerization of a
porphyrin dimer. The effective molarity (ca. 300m) for
forming the complex 1b·3 is exceptionally high. Absorption
and emission spectra show that p conjugation is more
effective in the nanoring than in its linear analogue. Thus
the S1!S0 fluorescence band shifts from 827 nm in the linear
hexamer to ca. 900 nm in the cyclic hexamers. DFT calcu-
lations corroborate the enhanced p conjugation in the nano-
ring, and show that the S0–S1 transition is forbidden,
accounting for the low radiative rate constant of the nanoring.
The enhanced conjugation in the nanoring probably results
from its rigid geometry and from the lack of end-effects,
perhaps with a contribution from out-of-plane distortion of
Scheme 3. Schematic representation of the transition dipole moment
arrangement in the lowest S0!S1 transition of the cyclic hexamer 1
and linear hexamer 4. The cancellation of the transition moments in 1
makes the lowest electronic transition symmetry-forbidden.
similar picture (Supporting Information), predicting a for-
bidden S0!S1 transition followed by a strongly allowed S0!
Table 1: Quantum yields Ff, fluorescence lifetimes tf, and radiative rate constants kf.
Ff
tf [ps]
kf [sꢀ1 [a]
]
kf [sꢀ1 [b]
]
kf [sꢀ1 [c]
]
4
0.08
650
500
340
460
1.3”108
1.2”107
2.9”106
4.3”106
2.7”108
1.9”108
2.2”108
1.9”108
4·3
1a·3
1a
0.006
0.001
0.002
2.1”107
7.7”106
8.1”106
[a] Calculated from Ff/tf. [b] Calculated from Equation (2) with an integrated absorptivity between 650 and 950 nm. [c] Calculated from Equation (2)
where only the most red-shifted peak in the absorption is considered.[10]
Angew. Chem. Int. Ed. 2008, 47, 4993 –4996
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 4995

Communications
8927 – 8940; d) Y. Kuramochi, A. Satake, Y. Kobuke, J. Am.
Chem. Soc. 2004, 126, 8668 – 8669; e) T. Hori, N. Aratani, A.
Takagi, T. Matsumoto, T. Kawai, M.-C. Yoon, Z. S. Yoon, S. Cho,
D. Kim, A. Osuka, Chem. Eur. J. 2006, 12, 1319 – 1327.
[5] M. Hoffmann, C. J. Wilson, B. Odell, H. L. Anderson, Angew.
Chem. 2007, 119, 3183 – 3186; Angew. Chem. Int. Ed. 2007, 46,
3122 – 3125.
the p system. The structural similarity between the arrange-
ment of porphyrin chromophores in these nanorings and that
of chlorophyll units in the natural LH2 light-harvesting
system,[11] and the strong electronic coupling between the
porphyrin units, makes it interesting to explore the exciton
delocalization in these nanorings using ultra-fast time-
resolved techniques. These experiments are now in progress.
[6] R. Rathore, C. L. Burns, M. I. Deselnicu, Org. Lett. 2001, 3,
2887 – 2890.
[7] a) J. A. Marsden, J. J. Miller, M. M. Haley, Angew. Chem. 2004,
116, 1726 – 1729; Angew. Chem. Int. Ed. 2004, 43, 1694– 1697;
b) A. S. Batsanov, J. C. Collings, I. J. S. Fairlamb, J. P. Holland,
J. A. K. Howard, Z. Lin, T. B. Marder, A. C. Parsons, R. M.
Ward, J. Zhu, J. Org. Chem. 2005, 70, 703 – 706.
[8] a) P. Ballester, A. I. Oliva, A. Costa, P. M. Deya, A. Frontera,
R. M. Gomila, C. A. Hunter, J. Am. Chem. Soc. 2006, 128, 5560 –
5569; b) V. M. Krishnamurthy, V. Semetey, P. J. Bracher, N.
Shen, G. M. Whitesides, J. Am. Chem. Soc. 2007, 129, 1312 –
1320; c) G. Ercolani, Struct. Bonding (Berlin) 2006, 121, 167 –
215.
[9] S. J. Strickler, R. A. Berg, J. Chem. Phys. 1962, 37, 814– 822.
[10] The integrated absorptivity of the most red-shifted component
of the Q-band was calculated from an estimate of the width of
the band. Since this transition overlaps heavily with the rest of
the Q-band, this estimate may be inaccurate but it is good
enough for the qualitative analysis presented here.
Received: March 12, 2008
Revised: April 26, 2008
Published online: May 28, 2008
Keywords: conjugation · molecular electronics · porphyrinoids ·
.
self-assembly · template synthesis
[1] For reviews see: a) T. Kawase, H. Kurata, Chem. Rev. 2006, 106,
5250 – 5273; b) K. Tahara, Y. Tobe, Chem. Rev. 2006, 106, 5274–
5290; c) L. T. Scott, Angew. Chem. 2003, 115, 4265 – 4267;
Angew. Chem. Int. Ed. 2003, 42, 4133 – 4135.
[2] T. Kawase, H. R. Darabi, M. Oda, Angew. Chem. 1996, 108,
2803 – 2805; Angew. Chem. Int. Ed. Engl. 1996, 35, 2664– 2666.
[3] M. Ohkita, K. R. Ando, T. Tsuji, Chem. Commun. 2001, 2570 –
2571.
[4] a) S. Anderson, H. L. Anderson, J. K. M. Sanders, Acc. Chem.
Res. 1993, 26, 469 – 475; b) S. Rucareanu, A. Schuwey, A.
Gossauer, J. Am. Chem. Soc. 2006, 128, 3396 – 3413; c) J. Li, A.
Ambroise, S. I. Yang, J. R. Diers, J. Seth, C. R. Wack, D. F.
Bocian, D. Holten, J. S. Lindsey, J. Am. Chem. Soc. 1999, 121,
[11] a) G. McDermott, S. M. Prince, A. A. Freer, A. M. Haw-
thornthwaite-Lawless, M. Z. Papiz, R. J. Cogdell, N. W. Isaacs,
Nature 1995, 374, 517 – 521; b) Y. Nakamura, N. Aratani, A.
Osuka, Chem. Soc. Rev. 2007, 36, 831 – 845.
4996 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 4993 –4996