Synthesis and crystal structure of a push-pull quinoidal porphyrin: A nanoporous framework assembled from cyclic trimer aggregates

Article abstract of DOI:10.1039/b502776d

A quinoidal porphyrin has been synthesised with such a curved π-system that π-π stacking leads to the formation of cyclic trimer aggregates in the crystal, which pack to generate cylindrical channels with an internal diameter of 1.0 nm. The Royal Society

Full text of DOI:10.1039/b502776d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Synthesis and crystal structure of a push–pull quinoidal porphyrin: a
nanoporous framework assembled from cyclic trimer aggregates{
Martin J. Smith,a William Clegg,b Kiet A. Nguyen,c Joy E. Rogers,c Ruth Pachter,c Paul A. Fleitzc and
Harry L. Anderson*a
Received (in Cambridge, UK) 24th February 2005, Accepted 22nd March 2005
First published as an Advance Article on the web 13th April 2005
DOI: 10.1039/b502776d
A quinoidal porphyrin has been synthesised with such a curved
p-system that p–p stacking leads to the formation of cyclic
trimer aggregates in the crystal, which pack to generate
cylindrical channels with an internal diameter of 1.0 nm.
Push–pull quinoidal chromophores generally exist in resonance
with aromatic-zwitterionic canonical forms, often resulting in
strong second-order nonlinear optical behaviour,1 and leading to
Fig. 1 Otsubo’s push–pull diphenoquinoid 1.
charge-transfer absorption bands at visible to near infrared
wavelengths.2,3 For example, Otsubo’s diphenoquinoid 1 (Fig. 1)
shows a strong absorption band at 1045 nm, a remarkably long
wavelength for a p-system of this size.2 These results motivated us
to explore the behaviour of push–pull quinoidal systems with
porphyrinoid cores, as a strategy for distorting the electronic
structure of the porphyrin, shifting the absorption further into the
infrared and enhancing nonlinear optical effects.4 Here we report
the first synthesis of a push–pull quinoidal porphyrin (M2 where
M 5 2H, Ni and Zn, Scheme 1). In contrast to Otsubo’s
diphenoquinoid 1, the aromatic-zwitterionic canonical form makes
little contribution in these quinoidal porphyrins, and the HOMO–
LUMO gaps of the push–pull systems M2 are only slightly smaller
than those of the pull–pull analogues M3.5 The crystal structure of
Zn2 has several extraordinary features: the chromophore is so
highly curved that p–p stacking generates cyclic trimer aggregates,
˚
which pack to create a honeycomb of 10 A diameter channels.
Zinc porphyrin macrocycles line the walls of each channel, with
chains of H-bonded methanol molecules, some with partial
occupancy, streaming into the channel from each Lewis-acidic
zinc site.
Scheme 1 Synthesis of the quinoidal push–pull porphyrins (Ar 5 3,5-
But2C6H3). Reagents and conditions: (a) 1,2-benzenedithiol, TiCl4,
Na2CO3, 92%; (b) NaCH(CN)2, [Pd2(dba)3], CuI, PPh3, 42%; (c) DDQ,
94%; (d) H2SO4, TFA, 85%; (e) Zn(OAc)2, 90% yield.
free-base H22, providing access to other metal complexes such as
Zn2. The quinoidal characters of Ni2, H22 and Zn2 are
immediately apparent from the lack of porphyrin aromatic ring
current effects in their 1H NMR spectra; the b-protons resonate at
6.8–7.6 ppm (cf. 8.2–9.5 ppm in a simple porphyrin) and the NH
protons of H22 appear at 11.9 ppm (cf. 23.0 ppm in a typical free-
base porphyrin).
The push–pull quinoidal porphyrin Ni2 was synthesised from
the 5-bromo-15-formyl porphyrin Ni46 as shown in Scheme 1, by
thioacetal formation, Takahashi coupling5,7 and DDQ oxidation.
This chemistry proceeds most cleanly with the nickel(II) complexes,
and the product Ni2 is readily demetallated with acid to afford the
The absorption and emission spectra reveal several interesting
differences in electronic structure between the push–pull M2 and
pull–pull M3 systems. For example, the spectra of the zinc
complexes are compared in Fig. 2. All of the observed absorption
bands correlate well with those predicted by DFT calculations8
(see calculated transitions in Fig. 2 and MOs in Fig. 3). The longest
{ Electronic supplementary information (ESI) available: details of spectro-
scopic data and DFT calculations. See http://www.rsc.org/suppdata/cc/b5/
b502776d/
*harry.anderson@chem.ox.ac.uk
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 2433–2435 | 2433

View Article Online
…
Fig. 4 Structure of Zn2, showing the O O distances in a chain of
H-bonded MeOH (omitting aryl substituents; 40% probability ellipsoids).
The other two molecules of Zn2 in the asymmetric unit have similar
geometries, although in these cases the methanol is more disordered.
Fig. 2 Experimental and calculated absorption spectra of Zn2 (plain)
and Zn3 (dashed), and the fluorescence spectrum of Zn2, in benzene
containing 1% pyridine; calculated DFT excitation wavelengths and
becoming progressively longer on moving away from the
Lewis-acidic zinc site, as illustrated in Fig. 4. The three crystal-
lographically distinct molecules of Zn2 adopt the same conforma-
tion, and the bond lengths in the porphyrinoid core are very
similar to those in Zn3,5 showing that Zn2 is quinoidal rather than
aromatic. A contribution from the aromatic-zwitterionic canonical
of Zn2 would result in an elongation of the CLC(CN)2 and CMN
bonds,3,9 but comparison of the crystal structures of Zn2 and Zn3
gives no indication of an effect of this type. Evidently the aromatic
stabilisation in the zwitterionic canonical form of Zn2 is
insufficient to compensate for the energy of charge separation.
The conjugated p-system of Zn2 is severely nonplanar due to
steric clashes between the b-hydrogens and sulfur atoms (mean
oscillator strengths (fcalc
) are denoted by black and white circles
respectively. The emission spectrum of Zn2 (excitation at 655 nm) is
plotted on an arbitrary vertical scale. The insert shows an expansion of the
absorption spectrum of Zn3.
wavelength (S0–S1) absorption of Zn3 is virtually forbidden
(fcalc 0.0008; lcalc 813 nm), and indeed this band is so weak that
at first we overlooked it,5 but it appears reproducibly at 770–
820 nm (e 1000 M21 cm21, see insert in Fig. 2). The low oscillator
strength of this transition explains why Zn3 is non-fluorescent. In
contrast, Zn2 gives a clear fluorescence band (lmax 795 nm)
because its S0–S1 transition is allowed (fcalc 0.16), due to its lower
symmetry. The HOMO and LUMO of both molecules are
delocalised over the donor and acceptor groups, so these orbitals
show the greatest change in energy, comparing Zn2 and Zn3. The
calculated increase in HOMO and LUMO energy, and reduction
…
˚
S distance: 2.67 A cf. 2.91 A for sum of van der Waals radii) at
˚
H
one side, and between the b-hydrogens and nitrile carbons (mean
…
˚ ˚
C distance: 2.57 A cf. 2.87 A for sum of van der Waals radii)
H
in HOMO–LUMO gap, in Zn2 are consistent with the first
Red
at the other. Similar distortions characterise the solid-state
conformations of other quinoidal porphyrins such as Zn3,5,10
and of closely related push–pull anthraquinoids.3 The angle
between the mean plane of the thioacetal benzene ring and that
of the porphyrin core is about 120u (129.1u, 123.6u and 122.8u for
the three molecules), so p–p stacking between these planes leads to
the formation of a cyclic trimeric aggregate, as shown in Fig. 5.
The aggregate has virtual C3h symmetry, with each benzene ring
stacking directly over the Zn–N–C–C–C–N ring of an adjacent
porphyrin core, as illustrated in Fig. 5B. The shortest contact in
oxidation and reduction potentials (Zn3: E1
5 20.71 V,
Ox
E1 5 0.87 V; Zn2: E1
Red
Ox
5 21.15 V, E1 5 0.28 V, all data
recorded in CH2Cl2 with 0.1 M Bu4NBF4 relative to ferrocene;
scan rate 0.1 V/s).
Crystals of the push–pull porphyrinoid Zn2 were grown by
vapour diffusion of methanol into a chloroform solution.{ The
asymmetric unit contains three molecules of Zn2, and 21 molecules
of methanol. Each zinc atom is ligated by a chain of hydrogen-
…
bonded methanol molecules, with the O–H O hydrogen bonds
Fig. 3 Frontier orbitals of Zn3 (left) and Zn2 (right) from time-dependent DFT calculations (using Becke’s three-parameter hybrid functional, B3LYP/6-
31G(d)).8 Solid and dotted arrows denote the major and minor contributions of one-electron excitations respectively.
2434 | Chem. Commun., 2005, 2433–2435
This journal is ß The Royal Society of Chemistry 2005

View Article Online
In summary, we report the first synthesis of a quinoidal push–
pull porphyrin. The NMR spectra, electronic absorption spectra,
DFT calculations and crystal structure of this macrocycle show no
significant contribution from the aromatic-zwitterionic canonical.
The curved structure of the chromophore results in the formation
of unusual cyclic trimer aggregates in the solid state, which pack
into a remarkable honeycomb structure. The channels in this
framework appear to be just large enough to accommodate a
˚
string of C60 molecules (van der Waals radius 10.5 A).
We thank EPSRC, CCLRC and EOARD for support, the
EPSRC Mass Spectrometry Service (Swansea) for mass spectra,
and Oxford University for a studentship for MJS.
Martin J. Smith,a William Clegg,b Kiet A. Nguyen,c Joy E. Rogers,c
Ruth Pachter,c Paul A. Fleitzc and Harry L. Anderson*a
aDepartment of Chemistry, University of Oxford, Chemistry Research
Laboratory, Mansfield Road, Oxford, UK OX1 3TA.
E-mail: harry.anderson@chem.ox.ac.uk; Fax: +44(0)1865 285002;
Tel: +44(0)1865 275704
bDepartment of Chemistry, University of Newcastle, Newcastle upon
Tyne, UK NE1 7RU, and CCLRC Daresbury Laboratory, Warrington,
UK WA4 4AD
cAir Force Research Laboratory, AFRL/MLPJ, 3005 Hobson Way,
Wright-Patterson Air Force Base, Dayton, Ohio, 45433-7702, USA
Notes and references
Fig. 5 Two views of the (Zn2)3 aggregate, omitting coordinated
methanol and aryl substituents; 50% probability ellipsoids.
{ Crystal data for Zn2: crystals were grown from CHCl3/CH3OH and the
structure was solved using synchrotron X-rays at Daresbury Station 9.8.
¯
C58H54N6S2Zn?6.5CH4O, M 5 1172.8, triclinic, space group P1,
˚
a 5 20.792(3), b 5 24.707(3), c 5 25.188(3) A, a 5 63.913(2),
3
˚
˚
b 5 84.532(2), c 5 81.473(2)u; V 5 11486(3) A , Z 5 6, l 5 0.6932 A,
m 5 0.42 mm21, T 5 150 K, R 5 0.0807 for 18810 observed reflections [I .
2s(I)] and Rw 5 0.2381 for all 32471 unique reflections. CCDC 264994. See
http://www.rsc.org/suppdata/cc/b5/b502776d/ for crystallographic data in
CIF or other electronic format.
1 M. Ravi, D. N. Rao, S. Cohen, I. Agranat and T. P. Radhakrishnan,
Chem. Mater., 1997, 9, 830; Y. Yamashita and M. Tomura, J. Mater.
Chem., 1998, 8, 1933.
2 S. Inoue, Y. Aso and T. Otsubo, Chem. Commun., 1997, 1105.
3 A. S. Batsanov, M. R. Bryce, M. A. Coffin, A. Green, R. E. Hester,
J. A. K. Howard, I. K. Lednev, N. Mart´ın, A. J. Moore, J. N. Moore,
E. Ort´ı, L. Sa´nchez, M. Saviro´n, P. M. Viruela, R. Viruela and T.-Q. Ye,
Chem. Eur. J., 1998, 4, 2580.
4 M. Drobizhev, Y. Stepanenko, Y. Dzenis, A. Karotki, A. Rebane,
P. N. Taylor and H. L. Anderson, J. Am. Chem. Soc., 2004, 126, 15352;
K. J. McEwan, P. A. Fleitz, J. E. Rogers, J. E. Slagle, D. G. McLean,
H. Akdas, M. Katterle, I. M. Blake and H. L. Anderson, Adv. Mater.,
2004, 16, 1933.
Fig. 6 Packing diagram for Zn2 showing the van der Waals surface,
omitting methanol except where coordinated to zinc; an infinite projection
along the a-axis.
5 I. M. Blake, H. L. Anderson, D. Beljonne, J.-L. Bre´das and W. Clegg,
J. Am. Chem. Soc., 1998, 120, 10764.
6 M. Yeung, A. C. H. Ng, M. G. B. Drew, E. Vorpagel, E. M. Breitung,
R. J. McMahon and D. K. P. Ng, J. Org. Chem., 1998, 63, 7143.
7 M. Uno, K. Seto and S. Takahashi, J. Chem. Soc., Chem. Commun.,
1984, 932.
each case is between a pyrrole nitrogen and a carbon atom of the
…
˚
˚
8 R. Bauernschmitt and R. Ahlrichs, Chem. Phys. Lett., 1996, 256, 454;
M. Casida, C. Jamorski, K. C. Casida and D. R. Salahub, J. Chem.
Phys., 1998, 108, 4439; A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
9 S. Inoue, S. Mikami, Y. Aso, T. Otsubo, T. Wada and H. Sasabe,
Synth. Met., 1997, 84, 395.
benzene ring (C N distances: 3.36, 3.43 and 3.44 A, cf. 3.41 A for
the sum of the van der Waals radii). These aggregates pack to
generate infinite methanol-filled channels (Fig. 6). The channels are
˚
large enough to accommodate a continuous cylinder of 10 A
10 I. M. Blake, L. H. Rees, T. D. W. Claridge and H. L. Anderson, Angew.
Chem., Int. Ed., 2000, 39, 1818; I. M. Blake, A. Krivokapic, M. Katterle
and H. L. Anderson, Chem. Commun., 2002, 1662.
11 D. V. Soldatov, I. L. Moudrakovski and J. A. Ripmeester, Angew.
Chem., Int. Ed., 2004, 43, 6308; S. Kitagaw, R. Kitaura and S. Noro,
Angew. Chem., Int. Ed., 2004, 43, 2334; D. Laliberte´, T. Maris and
J. D. Wuest, J. Org. Chem., 2004, 69, 1776; O. M. Yaghi, M. O’Keeffe,
N. W. Ockwig, H. K. Chae, M. Eddaoudi and J. Kim, Nature, 2003,
423, 705.
diameter within the van der Waals surface. The planes of the
porphyrin and the thioacetal benzene ring are parallel to the walls
of the channels (Figs. 5A and 6 have the same orientation), and
contacts between cyclic trimer aggregates are confined to the meso-
3,5-di-tert-butylphenyl substituents. Although many other zeolite-
like nanoporous organic frameworks have been reported, very few
exhibit such large continuous channels.11
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 2433–2435 | 2435