A self-assembled aluminium(III) porphyrin cyclic trimer

Article abstract of DOI:10.1039/b717017n

A self-assembled trimeric ring-shaped aggregate of an aluminium(iii) porphyrin bearing a benzoic acid in one meso-position has been characterized by NMR-spectroscopy and MALDI-TOF spectrometry. The Royal Society of Chemistry 2008.

Full text of DOI:10.1039/b717017n

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
A self-assembled aluminium(III) porphyrin cyclic trimer†‡
Gerald A. Metselaar,a Jeremy K. M. Sandersb and Javier de Mendoza*a
Received 2nd November 2007, Accepted 5th November 2007
First published as an Advance Article on the web 23rd November 2007
DOI: 10.1039/b717017n
A self-assembled trimeric ring-shaped aggregate of an alu-
minium(III) porphyrin bearing a benzoic acid in one meso-
position has been characterized by NMR-spectroscopy and
MALDI-TOF spectrometry.
Al(III) porphyrins with simultaneous axial binding to a carboxylate
and a pyridine reveal that the C–O–Al angle is typically ∼150◦.6
This means that compound 1, an Al(III)porphyrin with a para-
benzoate in one meso-position, should self-assemble after removal
of the axial acetate ligand into an almost perfect triangle 13 in the
presence of a pyridine derivative (Fig. 1).
Cyclic multiporphyrin systems are interesting assemblies because
of their resemblance to the natural light-harvesting system.1 Both
covalent2 as well as supramolecular approaches3 have been used
to construct such cyclic porphyrin arrays. Many examples are
reported in the literature in which metal porphyrins endowed
with nitrogenous bases self-assemble to form either polymeric
structures or well-defined oligomeric aggregates.4 Inducing self-
assembly of metal porphyrins bearing O-containing substituents
like alcohols or carboxylic acids, however, has rarely been reported.
Wojaczyn´ski and coworkers found that Fe(III)-, Mn(III)-, and
Ga(III) tetraphenylporphyrins bearing an OH-group in the b-
pyrrolic 2-position self-assemble to form homometallic cyclic
trimeric structures.5a,c In the case of the Fe(III)-porphyrin, the
structure was confirmed by X-ray crystallography.5d In addition,
using mixtures of the three metal porphyrins, also heterometallic
cyclic trimeric aggregates could be obtained.5d
Porphyrin 1 was readily obtained from the free base methyl
ester 2,7 which by addition of excess trimethylaluminium and
subsequent quenching with methanol afforded 3 in 90% yield.
Treatment of 3 with LiOH in a methanol–water mixture promoted
formation of the lithium salt 4 by simultaneous ester hydrolysis and
replacement of the OMe axial ligand by OH. Then, addition of an
excess acetic acid and evaporation in vacuo overnight resulted in
formation of the desired aggregate.
It is likely that formation of the Al-carboxylate aggregate
proceeds by a similar mechanism as that proposed for the
coordination of carboxylates to Sn(IV) porphyrins.8 The first step
is the formation of a hydrogen bond between the Sn–OH and
the carboxylic acid in its protonated form. The Sn-carboxylate
complex is subsequently formed by release of water (Fig. 2). Hence,
it is necessary to have the benzoic acid of the aluminium porphyrin
in its protonated form for self-assembly to take place.
As no suitable crystals for X-ray diffraction were available, the
9
aggregation of 1n was studied in solution by NMR spectroscopy,
as well as by MALDI-TOF mass spectrometry. A solution of 1n
in deuterated chloroform gave rise to a broad and complicated 1H
NMR-spectrum, while a purple solid developed and precipitated
from solution. Presumably this material was an oligomeric or
polymeric undefined assembly. However, if an excess of tert-
butylpyridine or 1-methylimidazole was added to the CDCl3
solution of 1n, or when the spectrum was recorded in pyridine-
d5, a well-defined 1H-NMR spectrum was obtained (Fig. 3). This
difference in aggregation behaviour with or without a nitrogenous
base could be rationalised by looking at the crystal structures of
aluminium(III) porphyrins bearing benzoate as axial ligand.6 In
the absence of tert-butylpyridine, the plane of the benzene ring of
the benzoate is not perpendicular to the plane of the porphyrin
but it is twisted by ∼135◦ instead. On the contrary, when pyridine
is coordinated as a second axial ligand, the benzoate is nicely
oriented perpendicular to the porphyrin plane (Fig. 1a).
Most remarkably, the 1H-NMR spectrum of 1n in the presence
of nitrogenous bases was characterised by the presence of four well
separated signals for the protons of the benzoate ring. For other
aromatic acids axially bound to metallated porphyrins only two
sets of signals are typically observed, for the two ortho- and two
meta-protons, respectively. The observed differences in chemical
shifts between the four protons are due to differences in shielding
as a result of their distance to the porphyrin plane. Although these
four different environments and distances are clearly seen in the
crystal structure of a benzoate axially coordinated to an Al(III)
Recently, the use of diamagnetic aluminium(III) porphyrins as
supramolecular building blocks able to strongly bind axially to
a carboxylate ligand has been described.6 A pyridine can then
be coordinated as a second axial ligand. Crystal structures of
aInstitute of Chemical Research of Catalonia (ICIQ), Avgd. Pa¨ısos Catalans
16, 43007, Tarragona, Spain. E-mail: jdemendoza@iciq.es; Fax: +34 977
920226; Tel: +34 977 920220
bUniversity of Cambridge, Department of Chemistry, Lensfield Road,
Cambridge, UK CB2 1EW, England. E-mail: jkms@cam.ac.uk; Fax: +44
1223 336411; Tel: +44 1223 336017
† The HTML version of this article has been enhanced with colour images.
‡ Electronic supplementary information (ESI) available: Experimental
procedures, DOSY-NMR, MALDI-TOF-MS, calculation hydrodynamic
radii. See DOI: 10.1039/b717017n
588 | Dalton Trans., 2008, 588–590
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Fig. 1 a) The crystal structure of a meso-tetraphenyl aluminium(III) porphyrin bearing benzoic acid and tert-butylpyridine as axial ligands.6 b) Optimized
structure of trimer 13 bearing pyridine as additional axial ligand.11
for the carboxylate ortho-, meta- and para-protons in the 1H-NMR
spectrum.10 Therefore, the frozen conformation in the case of 1n
constitutes strong evidence for a well-organised, cyclic structure.4e
The optimised trimer structure shown in Fig. 1b,11 based on the
coordinates of the X-ray structure of Fig. 1a, fully agrees with the
angles and distances found in the latter, without any substantial
distortion of the three porphyrin rings.
Further evidence for the formation of a well-defined aggregate
1
came from H DOSY NMR experiments (Table 1). For 0.02 M
solutions of 1n and of its methyl ester 3 in pyridine-d5, D =
(2.06
0.10) ×10−10 m2 s−1 and D = (3.84
0.10) × 10−10
m2 s−1 values were found, respectively. In both cases, the diffusion
coefficient found for pyridine itself was D ∼ 1.38 × 10−9 m2 s−1,
suggesting that the viscosities of the solutions were similar, so
comparison of the obtained hydrodynamic radii was justified.12
The hydrodynamic radii (rexp) could be estimated by the Stokes–
Einstein equation, assuming the species in solution to be spherical.
Comparison of the calculated average radius (rcalc = average of
x, y and z values of the corresponding gas phase-minimised
structures)13 with the experimental value suggests that monomer
Fig. 2 Protonation of Li-carboxylate and subsequent self-assembly of
metal porphyrins bearing a carboxylic acid by removal of acetic acid.
˚
3 is significantly smaller than expected: rexp (5.9 0.2 A) vs. rcalc
˚
(6.9 A). This difference could be ascribed to the non-spherical,
disk-shaped structure of the monomeric porphyrin 3. The trimer
13, on the other hand, has a more globular shape and the calculated
˚
radius (rcalc = 11.3 A) corresponds remarkably well with that found
14
˚
experimentally (rexp = 11.0 0.2 A).
Table 1 Diffusion coefficients determined by DOSY and effective exper-
imental (rexp) and calculated (rcalc) hydrodynamic radii for 3 and 13
a
D/1× 10−10 m2 s−1
r
exp/A
r
calc/A
˚
˚
Fig. 3 1H NMR spectrum of 13 in CDCl3 with 3 equivalents of
1-methylimidazole with the Dd-values relative to 3 (with OMe as axial
ligand) indicated above the corresponding signals.
Monomer 3
Trimer 13
3.84 0.10
2.06 0.10
5.9 0.2
11.0 0.2
6.9
11.3
a Calculated using the Stokes–Einstein equation: R = KbT/6pgD where Kb
is the Boltzmann constant and T is the absolute temperature. The viscosity
of pyridine g(20 ◦C) = 0.95 cp was used.
porphyrin (Fig. 1a), the rotational freedom of the benzoate in
solution results in the observation of only the three typical signals
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 588–590 | 589
©

View Article Online
Finally, additional support for the formation of self-assembled
trimers was obtained by MALDI-TOF mass spectrometry. When
1 was deposited on the MALDI-plate from a pyridine solution, a
relatively large signal for a trimeric assembly 13 was found, with
only a small signal for the dimer 12 and a signal for monomeric
1 as additional peaks. When deposition on the MALDI-plate
from a chloroform solution containing only five equivalents of
tert-butylpyridine, small peaks were also observed for tetrameric
14 and larger oligomeric assemblies in addition to 1 and 12 or
1999, 40, 8347–8350; (c) K.-Y. Tomizaki, L. Yu, L. Wei, D. F. Bocian
and J. S. Lindsey, J. Org. Chem., 2003, 68, 8199–8207; (d) T. Hori,
N. Aratani, A. Takagi, T. Matsumoto, T. Kawai, M. -C. Yoon, Z. S.
Yoon, S. Cho, D. Kim and A. Osuka, Chem.–Eur. J., 2006, 12, 1319–
1327.
3 (a) E. Iengo, E. Zangrando and E. Alessio, Acc. Chem. Res., 2006, 39,
841–851; (b) F. Hajjaj, Z. S. Yoon, M.-C. Yoon, J. Park, A. Satake, D.
Kim and Y. Kobuke, J. Am. Chem. Soc., 2006, 128, 4612–4623; (c) X.
Shi, K. M. Barkigia, J. Fajer and C. M. Drain, J. Org. Chem., 2001, 66,
6513–6522; (d) H. Ohkawa, A. Takayama, S. Nakajima and H. Nishide,
Org. Lett., 2006, 8, 2225–2228.
4 (a) T. Imamura and K. Fukushima, Coord. Chem. Rev., 2000, 198, 133–
156; (b) C. Maeda, T. Kamade, N. Aratani and A. Osuka, Coord. Chem.
Rev., 2007, 251, 2743–2752; (c) J. Wojaczyn´ski and L. Latos-Graz˙in´ski,
Coord. Chem. Rev., 2000, 204, 113–171; (d) A. Satake and Y. Kobuke,
Tetrahedron, 2005, 61, 13–41; (e) J. C. Hawley, N. Bampos and J. K. M.
Sanders, Chem.–Eur. J., 2003, 9, 5211–5222.
5 (a) J. Wojaczyn´ski and L. Latos-Graz˙in´ski, Inorg. Chem., 1995, 34,
1044–1053; (b) J. Wojaczyn´ski and L. Latos-Graz˙in´ski, Inorg. Chem.,
1996, 35, 4812–4818; (c) J. Wojaczyn´ski and L. Latos-Graz˙in´ski, Inorg.
Chem., 1995, 34, 1054–1062; (d) J. Wojaczyn´ski and L. Latos-Graz˙in´ski,
Inorg. Chem., 1997, 36, 4548–4554.
1
13 aggregates. As mentioned already for the H-NMR spectra,
(excess) pyridine is needed to prevent linear aggregation.
In conclusion, introduction of a para-benzoate moiety at one
meso-position of an Al(III) porphyrin results in a self-assembled
cyclic trimer in the presence of nitrogenous bases. It is likely that
playing with angles (such as meta-benzoates) or spacers would
yield a diversity of self-assembled structures based on the Al(III)
porphyrin scaffold. Additionally, use of diverse pyridyl or bipyridyl
axial ligands would expand the supramolecular assemblies into
more complex, hierarchically assembled three-dimensional net-
works. These strategies are currently being developed in our
laboratories.
6 G. J. E. Davidson, L. H. Tong, P. R. Raithby and J. K. M. Sanders,
Chem. Commun., 2006, 3087–3089.
7 C. M. Carcel, J. K. Laha, R. S. Loewe, P. Thamyongkit, K.-H.
Schweikart, V. Misra, D. F. Bocian and J. S. Lindsey, J. Org. Chem.,
2004, 69, 6739–6750.
This work was supported by the Spanish Ministry of
Science and Education (MEC) (projects BQU2002–03536
and CTQ2005–06909-C02–02/BQU). Consolider Ingenio 2010
(Grant CSD2006–0003) and the ICIQ Foundation.
8 (a) J. C. Hawley, N. Bampos, R. J. Abraham and J. K. M. Sanders,
Chem. Commun., 1998, 661; (b) P. R. Brotherhood, R. A. S. Wu, P.
Turner and M. J. Crossley, Chem. Commun., 2007, 225–227.
9 We represent as 1n any aggregate resulting from 1 after removal of acetic
acid.
We thank Dr P. Ballester (ICIQ) for molecular modelling studies
and Dr G. J. E. Davidson (Cambridge) for synthetic advice.
10 Also in the case of coordination of 5-(4-)carboxyphenyl-10,15,20-
tri-p-tolyl porphyrin to 5,10,15,20-(4-)pentylphenyl aluminium(III)
1
porphyrin only two signals in H NMR were observed for the ortho-
and meta-protons.
11 CAChe WorkSystem, Version 6.1.12.33, Fujitsu Limited.
12 E. J. Cabrita and S. Berger, Magn. Reson. Chem., 2001, 39, S142-S148.
13 P. Timmerman, J.-L. Weidmann, K. A. Jollife, L. J. Prins, D. N.
Reinhoudt, S. Shinkai, L. Frish and Y. Cohen, J. Chem. Soc., Perkin
Trans. 2, 2000, 2, 2077–2089.
Notes and references
1 M.-S. Choi, T. Yamazaki and T. Aida, Angew. Chem., Int. Ed., 2004,
43, 150–158.
2 (a) S. Anderson, H. L. Anderson, A. Bashall, M. McPartlin and J. K. M.
Sanders, Angew. Chem., Int. Ed. Engl., 1995, 34, 1096–1099; (b) O.
Mongin, A. Schuwey, M.-A. Vallot and A. Gossauer, Tetrahedron Lett.,
˚
˚
14 For hypothetical 14 and 12 species: rcal = 12.1 A and rcalc = 9.8 A,
respectively. See ESI‡.
590 | Dalton Trans., 2008, 588–590
This journal is
The Royal Society of Chemistry 2008
©