Host-guest interactions in acid-porphyrin complexes

Article abstract of DOI:10.1039/c2cc33668e

In this report we use the weak interactions of acid-porphyrin complexes to selectively bind competing acids to the faces of a rigid cyclic porphyrin dimer, and characterise the resulting interactions by NMR spectroscopy and nano-electrospray ionisation spectrometry.

Full text of DOI:10.1039/c2cc33668e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 9358–9360
www.rsc.org/chemcomm
COMMUNICATION
Host–guest interactions in acid–porphyrin complexesw
´
Matthew J. Webb,z Stephanie Deroo,y Carol V. Robinsonz and Nick Bampos*
Received 22nd May 2012, Accepted 30th July 2012
DOI: 10.1039/c2cc33668e
In this report we use the weak interactions of acid–porphyrin
complexes to selectively bind competing acids to the faces of a
rigid cyclic porphyrin dimer, and characterise the resulting
interactions by NMR spectroscopy and nano-electrospray ioni-
sation spectrometry.
This report addresses the protonation of a freebase cyclic
porphyrin butadiyne linked dimer (1) used previously as a
potential ‘catalytic host’ (Scheme 1).1 Metallation with zinc
allowed Sanders and co-workers to accelerate Diels–Alder and
acyl transfer reactions within the cavities of the conformation-
ally restricted cyclic dimer and related trimer ‘hosts’.2 In more
recent work, the oxophilic nature of Sn(IV) porphyrins was
used to direct carboxylic acids, with varying structural and
Fig. 1 Dimer
1 with a schematic representation of the lower
electronic properties, to the two faces (‘interior’ and ‘exterior’)
of the cyclic butadiyne porphyrin dimers and trimers.3 Mizuno
and Aida have also been able to show that their p-xylylene
linked cyclic porphyrin dimer enhances circular dichroism
activity.4 We have found that, in the absence of metals, 1 also
presents an ‘interior’ and an ‘exterior’ face for selective
recognition of fluorinated acids.
porphyrin viewed side-on with hexyl and methyl groups omitted for
clarity. The distance between the two porphyrin planes is 11.5 A
(Fig. S1, ESIw). Also shown are the six fluorinated acids used in this
investigation.
The experiments outlined below focus on the site specific
protonation of the freebase form of the cyclic porphyrin host.
Since realising of the potential of acid–porphyrin complexes
for molecular assembly, our interest has been directed towards
exploiting the acid–porphyrin binding motif for cavity specific
recognition.5
A selection of fluorinated mono- and diacids has been added
to 1, and the resulting assemblies have been characterised
1
using H and 19F NMR spectroscopy, and nano-electrospray
ionisation spectrometry. Six fluorinated acids (four mono and
two di-acids) were chosen for this study (Fig. 1). Trifluoro-
acetic acid was chosen as the ‘standard’ acid because it readily
protonates the porphyrin and can be easily accommodated
inside the cavity of the host. Pentafluoropropionic (PFPA)
and heptafluorobutyric (HFBA) acids offer longer alkyl chains
with accompanying steric demands and a corresponding degree
of flexibility, while pentafluorobenzoic acid (PFBA) provided
contrasting rigidity in the form of the aromatic ring. Two
dicarboxylic acids, hexafluoroglutaric acid (HFGA) and tetra-
fluoroterephthalic acid (TFTA), were chosen for their ability to
bind cooperatively inside the cavity of 1.
Scheme 1 Representation of the conversion of the freebase dimer 1
to the bis acid–porphyrin dimer complex with TFA (1ꢀTFA4).
Department of Chemistry, University of Cambridge, Lensfield Road,
Cambridge, CB2 1EW, UK. E-mail: nb10013@cam.ac.uk
w Electronic supplementary information (ESI) available: General
experimental information, synthetic procedures and additional figures
and spectra. See DOI: 10.1039/c2cc33668e
z Current address: Department of Chemistry, Uppsala University,
Protonation of 1 (in d-chloroform) with TFA gave a statistical
mixture of assemblies (as evident from the 1H NMR data outlined
below) before the stoichiometry reached 4 : 1 (acid : dimer)
(Scheme 1). The protonation of one porphyrin of the dimer
occurred independently of the second, and reflected all the
characteristic features of the comparable protonation of a
BMC, Husargatan 3, Box 576, Uppsala 75123, Sweden.
y Current address: Laboratory for Structure and Function of Bio-
logical Membranes, Center for Structural Biology and Bioinformatics,
Universite Libre de Bruxelles, CP206/2, Blvd. du Triomphe, 1050
´
Brussels, Belgium.
z Current address: Department of Physical Chemistry, University of
Oxford, Oxford, OX1 3QZ, UK.
c
9358 Chem. Commun., 2012, 48, 9358–9360
This journal is The Royal Society of Chemistry 2012

View Article Online
Fig. 3 1H NMR spectrum (400 MHz, d-chloroform, 223 K) of the
porphyrin dimer 1 with 4.0 equivalents of PFBA.
acid size (TFA > PFPA > HFBA). This behaviour may be
the result of the increased lability (exchange) of the larger
anions as a result of their greater steric ‘repulsion’ within the
restricted dimer cavity, or the re-adjustment of the geometry to
accommodate the larger anion.
Fig. 2 1H NMR spectra (400 MHz, d-chloroform, 253 K) of (i) 1,
and (ii) 1 with 4.0 equivalents of TFA (1ꢀTFA4). The labels of the
resonances match the proton environments in Scheme 1.
1
The H NMR spectrum (Fig. 3) of the dimer system after
the addition of four equivalents of PFBA (see also Fig. S9,
ESIw) exhibited more peaks than observed previously with the
less ‘bulky’ acids at the same stoichiometry. Multiple meso CH
(a0) and core NH (b0 and c0) signals suggested the presence of
multiple species in solution. However, the relative intensities
and chemical shifts indicated that the appearance of the
spectrum was the result of a single unsymmetrical species,
with the two core NH signals (b0 and c0, in a ratio of 1 : 2)
assigned according to Scheme 2, with the freebase NH reso-
nance (b0) shifted downfield as a result of the close proximity
to the fluoroaromatic anion. Protonation of 1 with the bulky
anion of pentafluorobenzoic acid (PFBA) allowed only one of
the two porphyrins of 1 to generate an acid–porphyrin
complex despite the presence of four equivalents of PFBA.
Protonation and subsequent complexation at the inner face of
the cavity by PFBA blocked any further access to the dimer
core and, as a result, inhibited protonation and complexation
of the second porphyrin altogether.
porphyrin monomer – change in colour from purple to green;
distortion of the porphyrin plane to accommodate the four
protons within the core of the heterocycle; and stabilisation of
the resulting charge by the ‘complexation’ of the anion of the
protonating acid.
Prior to addition of TFA to 1, the 1H NMR spectrum
(Fig. 2(i), acquired at 253 K to minimise any exchange effects)
exhibited diagnostic resonances for the symmetrical cyclic
dimer with a characteristic core NH signal (f, ꢁ2.85 ppm).
Only one species was evident after TFA was added to achieve
a 4 : 1 stoichiometry of TFA : 1 (see also Fig. S2, ESIw), as
characterised by the single meso proton peak (a0, 10.01 ppm)
and the single core NH peak (f0, ꢁ1.43 ppm) (Fig. 2(ii)).
Integration of the peaks with respect to each other indicated
the formation of the bis acid–porphyrin dimer (1ꢀTFA4).
1
It was not possible by H NMR spectroscopy to discriminate
between the core NH protons directed inside and those
directed outside the distorted dimer cavity, all of which were
observed as a broad signal.
The integrity of 1ꢀPFBA2 (Scheme 2) was confirmed by
19F NMR spectroscopy in which a set of resonances was
recorded at approximately ꢁ144 (ortho), ꢁ159 (meta) and
ꢁ163 (para) ppm, as compared to ꢁ136, ꢁ144 and ꢁ159 ppm,
respectively, for the corresponding resonances of the ‘free’ acid
(Fig. S10(i), ESIw). The formation of 1ꢀPFBA2 provides an
interesting contrast to the observations reported when two
equivalents of a series of aromatic carboxylic acids were added
to the di-tin metallated analogue of 1.3 Tin porphyrins form
The 19F NMR spectrum, recorded after the addition of four
equivalents of TFA at 253 K (Fig. S3, ESIw), did however
identify only two signals (ꢁ78.7 and ꢁ79.5 ppm) in a ratio
of 1 : 1, neither of which corresponded to free TFA in solu-
tion (ꢁ75.8 ppm), but both of which coalesced to a peak at
ꢁ78.8 ppm at room temperature. The low temperature reso-
nances were assigned to the trifluoroacetate anions bound to
the ‘exterior’ and ‘interior’ faces of the dimer, respectively, as
the anion inside the cavity would be influenced by the additive
shielding effect of the adjacent porphyrin and appear at the
lower (more shielded) chemical shift.
Titrations involving the incrementally longer alkyl fluori-
nated acids, PFPA and HFBA, exhibited similar features by
1H NMR spectroscopy described for TFA (Fig. S4 and S5,
ESIw), whereby the bis acid–porphyrin dimers were generated
and identified upon addition of four equivalents of acid, and
for fewer than four equivalents of acid a statistical mixture of
complexes was identified. Analysis of the 19F NMR spectra of
the respective acid–porphyrin dimer complexes (Fig. S6–S8,
ESIw) suggested that the degree of shielding (displacement from
‘free’ chemical shift) experienced by each of the acid resonances,
and the resolution of the respective signals, was dependent on the
Scheme 2 Schematic representation of the unsymmetrical acid–
porphyrin dimer 1ꢀPFBA2. Only two acid molecules participate in
complexation with the host.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 9358–9360 9359

View Article Online
The length of the dicarboxylic acids was such that proto-
nation at one of the ‘interior’ faces preorganised the proto-
nation of the second ‘interior’ face (especially in the case of the
more rigid TFTA) such that the intramolecular cooperativity
led to cavity-specific recognition of the dicarboxylic acids.
This left the smaller mono-functional TFA to satisfy the
requirement for protonation and anion complexation at the
‘exterior’ faces.
Further characterisation of the acid–porphyrin host–guest
complexes was provided by nano-ESI mass spectrometry. Two
peaks were recorded for 1 at 902 and 1803 m/z for the 2+ and
1+ charge states (Fig. S17, ESIw). Peaks corresponding to the
freebase dimer 1 were again recorded in the presence of TFA
together with additional signals (Fig. S18, ESIw). In this case,
no signals were observed for the 1ꢀTFA4 complex unlike the
assemblies characterised by NMR spectroscopy (Fig. 2). How-
ever, under the same conditions, the MS results of the dimer
with HFGA and TFTA (with TFA to cap the ‘exterior’ faces)
were able to support the NMR characterisation. The spectrum
of the acid–porphyrin dimer with HFGA (1ꢀHFGA) was
dominated by a peak at 1022 m/z, assigned to the 2+ charge
state of the complex (Fig. S19, ESIw). The equivalent
experiment with TFTA generated a spectrum for 1ꢀTFTA
also dominated by the diprotonated state of the host–
guest complex at 1021 m/z (Fig. S20, ESIw). These two
results with the diacids are consistent with the structures
shown in Scheme 3 in the absence of the exterior bound
TFA molecules.
Scheme 3 Two schematic representations of the 1ꢀHFGA–TFA2 (left)
and 1ꢀTFTA–TFA2 (right) porphyrin dimer complexes.
strong bonds with the carboxylates of the acids, and stacking
interactions operate in the cavity between two neighbouring
aromatic groups bound to the two ‘interior’ faces of the
porphyrins. In the protonation mechanism5 the plane of one
porphyrin must distort in order to accommodate the extra
protons, and the additional distance required to form the
(weak) interaction with the carboxylate anion creates a micro-
environment that inhibits protonation with the second porphyrin
of the dimer.
The monometallation of 1 can be achieved by adding less
than one equivalent of a zinc salt but this typically leads to
a mixture of the freebase, mono-zinc, and di-zinc dimers
which are extremely difficult to separate chromatographically.
In our previous work5 we identified the significance of the
role of the anion in the zinc metallation of porphyrins. The
1ꢀPFBA2 system offered the opportunity to investigate
the potential for selectively mono-metallating the dimer by
directing the metallation using the PFBA to ‘activate’ a single
porphyrin core (Fig. S11 and S12, ESIw). Unfortunately,
the lability of the acid–porphyrin complex resulted in scram-
bling of the Zn metallation that ultimately led to the same
unsatisfactory mixture of products according to our NMR
spectroscopy.
In conclusion, we have exploited the restricted cavity of the
freebase porphyrin dimer 1 and its unique capacity for proto-
nation and anion complexation to selectively accommodate
diacid guests into the core whilst using NMR spectroscopy
together with the diamagnetic ring currents of the porphyrins
to better understand the complexation behaviour of (fluori-
nated) acids. The use of nano-ESI mass spectrometry (not
previously reported for comparable complexes) provided an
additional technique for characterisation of more elaborate
complexes using these ‘weak’ interactions. The lability of the
complexation has directed us to explore the cavity of analogous
cyclic hosts as potential ‘reaction centres’.
In order to establish the competitive binding preference of a
mixture of acids, 1 was challenged with a 1 : 1 mixture of TFA
and PFBA. Only a mixture of complexes was identified in
solution, with no preference for the acids for the ‘interior’ or
‘exterior’ faces of the host (Fig. S13, ESIw).
We thank the EPSRC for financial support (studentship
for MJW).
In contrast, mixtures of TFA and dicarboxylic acids such as
HFGA or TFTA did show site-specific preference, and the
complexes (Scheme 3) were characterised by 1H (Fig. S14 and
S15, ESIw) and 19F NMR spectroscopy. For example, a single
peak (indicative of symmetry) was identified for the TFTA
ligand of the 1ꢀTFTA–TFA2 complex at ꢁ148.6 ppm in the
19F NMR spectrum (Fig. S16, ESIw), compared to the chemical
shift of ‘free’ acid at ꢁ141.3 ppm. One single TFA resonance
was observed at ꢁ78.2 ppm. The chemical shift implies that the
anions were bound to the ‘exterior’ faces of the complex whilst
the absence of a second resonance at 253 K (Fig. S3, ESIw)
suggests one chemical environment and assembly symmetry.
Similar structural features were identified for 1ꢀHFGA–TFA2.
Notes and references
1 (a) H. L. Anderson and J. K. M. Sanders, J. Chem. Soc., Perkin
Trans. 1, 1995, 2223–2229; (b) H. L. Anderson, S. Anderson and
J. K. M. Sanders, J. Chem. Soc., Perkin Trans. 1, 1995, 2231–2245;
(c) S. Anderson, H. L. Anderson and J. K. M. Sanders, J. Chem.
Soc., Perkin Trans. 1, 1995, 2247–2254.
2 (a) C. J. Walter, H. L. Anderson and J. K. M. Sanders, J. Chem.
Soc., Chem. Commun., 1993, 5, 458–460; (b) L. G. Mackay,
R. S. Wylie and J. K. M. Sanders, J. Am. Chem. Soc., 1994, 116,
3141–3142.
3 J. C. Hawley, N. Bampos and J. K. M. Sanders, Chem.–Eur. J.,
2003, 9, 5211–5222.
4 Y. Mizuno and T. Aida, Chem. Commun., 2003, 20–21.
5 M. J. Webb and N. Bampos, Chem. Sci., 2012, 3, 2351–2366.
c
9360 Chem. Commun., 2012, 48, 9358–9360
This journal is The Royal Society of Chemistry 2012