Metalloporphyrin anion sensors: The effect of the metal centre on the anion binding properties of amide-functionalised and tetraphenyl metalloporphyrins

Article abstract of DOI:10.1039/b807153e

This article describes the synthesis and anion binding properties of a series of 'picket fence' metalloporphyrin complexes, within which the metal centre is systematically varied. The porphyrin structure contains four amide bonds and is the same for each metal. The anion binding properties of these receptors are further contrasted with those of their tetraphenylporphyrin congeners to elucidate both the effect of the metal centre and the influence of the amide groups on the anion recognition process. Anion binding was demonstrated using UV/visible and ¹H NMR spectroscopies, electrochemistry and luminescence. The metal centre was found to be highly influential in the strength and selectivity of binding; for example, the cadmium and mercury complexes exhibited far greater affinities for anions than the zinc complexes in competitive solvents such as DMSO. The amide functionalities were found to enhance the anion binding process.

Full text of DOI:10.1039/b807153e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Metalloporphyrin anion sensors: the effect of the metal centre on the anion
binding properties of amide-functionalised and tetraphenyl metalloporphyrins†
David P. Cormode,‡a Michael G. B. Drew,b Raymond Jagessar§a and Paul D. Beer*a
Received 6th May 2008, Accepted 11th August 2008
First published as an Advance Article on the web 15th October 2008
DOI: 10.1039/b807153e
This article describes the synthesis and anion binding properties of a series of ‘picket fence’
metalloporphyrin complexes, within which the metal centre is systematically varied. The porphyrin
structure contains four amide bonds and is the same for each metal. The anion binding properties of
these receptors are further contrasted with those of their tetraphenylporphyrin congeners to elucidate
both the effect of the metal centre and the influence of the amide groups on the anion recognition
process. Anion binding was demonstrated using UV/visible and 1H NMR spectroscopies,
electrochemistry and luminescence. The metal centre was found to be highly influential in the strength
and selectivity of binding; for example, the cadmium and mercury complexes exhibited far greater
affinities for anions than the zinc complexes in competitive solvents such as DMSO. The amide
functionalities were found to enhance the anion binding process.
properties of these receptors were contrasted with those of their
Introduction
tetraphenylporphyrin analogues and it was found that the amide
The topic of anion binding is of great current interest due to the
moieties enhance anion binding, as expected. In addition, the
important roles of anions in a variety of chemical, biological and
nature of the metal centre was found to significantly influence
environmental processes.1 Lewis acids are an important class of
anion receptors as they bind anions through their ability to accept
an electron pair. This type of anion receptor was first reported
the strength and selectivity of anion binding.
Syntheses and characterization
by Katz, who synthesized diboron compounds that can chelate
anions such as hydride and chloride.2 Similar mercury and silicon
based receptors have been reported,3 while the concept has been
extended in, for example, Newcomb’s anticrown and anticryptand
stannacycles4 and Shur’s mercury macrocycles that form sandwich
complexes with anions.5
More recently, a large amount of attention has focused on
sensing anions through a change in a measurable physical property
such as redox potential, UV/visible absorption or luminescence
band.6 Metalloporphyrins have proven particularly amenable to
such anion sensing methods.7 We8 and others9 have developed an-
ion receptors based on the tetra-a,a,a,a-2-aminophenylporphyrin
(TAPP) framework, which provides an attractive binding motif
due to the four convergent amide bonds with close proximity to
the metal centre.
3,4-Dimethoxybenzoyl chloride, 2, was synthesized via re-
fluxing 3,4-dimethoxybenzoic acid in thionyl chloride10 which
gave the product as a red powder in near quantitative yield
(Scheme 1). H2TAPP was synthesized via the method of Collman,11
and the a,a,a,a-isomer was isolated via Lindsey’s atropiso-
merization procedure.12 a,a,a,a-Tetra-(2-[3,4-dimethoxyphenyl]-
acetamidophenyl) porphyrin, 1, was prepared via dropwise addi-
tion of a dichloromethane solution of an excess of 2 to a solution
of freshly prepared a,a,a,a-H2TAPP and excess triethylamine in
dichloromethane, cooled in an ice-bath, which gave 1 as a light
purple powder in a 71% yield. The purpose of the methoxy groups
of this compound was to give improved solubility.
The iron(III), cobalt(III), copper(II), zinc(II), cadmium(II)
and mercury(II) complexes of 1 were prepared using litera-
ture methods.13 These complexes are hereafter referred to as
M–1, where M is the chemical symbol of the metal, e.g. Zn–1.
It was not possible to isolate a pure sample of the analogous
nickel(II) compound, possibly because the elevated temperatures
required for this metallation caused decomposition of either the
reactant or the product. All the new diamagnetic metalloporphyrin
receptors were characterized by 1H, 1H–1H COSY and 13C NMR
spectroscopies, mass spectrometry and elemental analysis. In the
case of the paramagnetic iron(III) and copper(II) complexes,
NMR spectroscopies could not be used. Despite the propensity
of cobalt(III) porphyrin complexes to disproportionate into
paramagnetic species,14 the 1H NMR spectrum of Co(PF6)–1 was
of good quality, and thus this receptor was judged to be sufficiently
pure to allow anion sensing investigations.
In order to create anion receptors of increased anion affinity
and novel selectivity we have synthesized a series of amide
functionalized porphyrin anion receptors based on the TAPP
framework that include various metal centres. The anion binding
aDepartment of Chemistry, Inorganic Chemistry Laboratory, University of
Oxford, South Parks Road, Oxford, UK. E-mail: Paul.Beer@chem.ox.ac.uk;
Fax: +44 (0)1865 272690; Tel: +44 (0)1865 272600
bDepartment of Chemistry, University of Reading, Whiteknights, Reading,
UK RG6 6AD
† CCDC reference number 686999. For crystallographic data in CIF or
other electronic format see DOI: 10.1039/b807153e
‡ Current address: Translational and Molecular Imaging Institute, Mount
Sinai School of Medicine, New York, USA
§ Current address: Department of Chemistry, University of Guyana,
Guyana
6732 | Dalton Trans., 2008, 6732–6741
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Scheme 1 Synthesis of metallated porphyrin anion receptors.
Table 1 Wavelength l/nm (molar extinction coefficient e/103 M-1 cm-1)
of the absorbances of receptors Fe(Cl)-1, Co(PF6)-1, Cu–1, Zn–1, Cd–1,
Hg–1. Solvent: DMSO
direction with one hydrogen bond between N(57)–H ◊ ◊ ◊ O(39)(x -
1, y, z) (H ◊ ◊ ◊ O 2.04, N–H ◊ ◊ ◊ O 148, N ◊ ◊ ◊ O 2.81 A). In addition
N(96)–H(96) forms hydrogen bonds to two disordered water
˚
molecules O(200) and O(201).
Receptor
Soret
Q (b)
Q (a)
Fe(Cl)-1a
Co(PF6)-1
Cu–1
428 (146.0)
443 (204.4)
424 (239.1)
434 (242.4)
443 (244.0)
455 (157.0)
521 (9.5)
558 (12.1)
545 (18.2)
564 (19.3)
577 (16.0)
—
592 (4.3)
—
—
602 (4.2)
618 (7.6)
—
UV/Visible anion binding studies
Zn–1
The anion binding properties of the metal complexes of 1 were
assessed by UV-visible spectroscopic methods in acetonitrile
solution, where attention was focused on the Q absorbance bands
of the receptors. For the cadmium and mercury metalloporphyrins,
freshly made up solutions were used for each investigation due
to the well-known photosensitivity of such compounds.18 Anions
were added as their salts of the non-complexing cation tetra-
butylammonium (TBA). The porphyrin compounds responded
differently upon addition of anions depending on the metal centre.
Firstly, perturbations were observed for anion addition to Fe(Cl)-1
for dihydrogen phosphate only. The Q band at ca. 520 nm decreases
in intensity, but it is slightly blue shifted (Fig. 2). There is a general
increase in the intensity in the spectrum in the region from 540 to
650 nm.
No perturbations were seen upon anion addition to Co(PF6)-1 in
this solvent system. However, significant changes were noted upon
anion addition to a solution of Co(PF6)-1 in the less competitive
solvent acetone (data not shown). No perturbations were seen
upon anion addition to Cu–1 in this solvent system also, or in the
less competitive solvent dichloromethane. It seems that the Jahn–
Teller effect makes the coordination of anions unfavourable to this
metalloporphyrin. In contrast, the changes typically seen upon
anion addition to Zn–1 are displayed in Fig. 3, for the example of
TBACl. A bathochromic shift can be seen in the maxima of the
peaks (10–15 nm for this example) and the intensity of the longer
wavelength a band increases relative to that of the b band. Several
isosbestic points were observed as the titration progressed. The
effect of anion addition was also visible by the naked eye. The
acetonitrile solution of Zn–1 is a pink colour, but upon addition
Cd–1
Hg–1
a The notation Q (b) and Q (a) do not apply to iron (III) species and
Fe(Cl)-1 has an additional absorbance at 649 nm (e = 2.8 ¥ 103 M-1cm-1).
For comparative purposes the iron(III), copper(II), zinc(II),
cadmium(II) and mercury(II) metallated tetraphenylporphyrins
(TPP) were also synthesized. TPP was prepared and purified by
standard methods15 and the metallations of TPP were achieved by
a variety of literature methods.16 These compounds are referred
to as M-TPP, where M is the chemical symbol of the metal, from
this point onwards. Unfortunately, Co(PF6)-TPP could not be
isolated, due to the high level of disproportionation into CoII-TPP
and CoIII(PF6)-TPP+ + PF6 that occurs for this compound.14
The metal complexes of 1 all exhibit characteristic porphyrinoid
electronic spectra in which the coordinated metal ion has a great
influence on the wavelengths, extinction coefficients and structure
of the absorbances. The wavelengths and molar extinction coeffi-
cients of the Soret and Q band absorptions of the metal complexes
of 1 in DMSO are summarized in Table 1. The characteristics
observed for these metalloporphyrins agree with those observed
for their TPP congeners and the reported spectral characteristics
for metallated porphyrins.17
∑
-
Crystals of the porphyrin 1 suitable for single crystal X-ray
structure determination were grown by the slow evaporation of
methanol solutions of 1. As expected, the 24 atoms that make
up the central ring in the porphyrin are approximately planar
-
˚
with a root mean square deviation of 0.09 A (Fig. 1), while there
of a strongly binding anion, such as H2PO4 , the solution changes
are four equivalent phenyl groups symmetrically arranged about
the porphyrin core. The molecules are held together along the a
colour to green in the same way as we have previously reported for
other zinc porphyrin receptors.8c
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 6732–6741 | 6733
©

View Article Online
Fig. 1 Two projections of the structure of 1 with ellipsoids at 20% probability. The structure determination shows 4 hydrogens in the centre of the
porphyrin all refined with 50% occupancy. Only the hydrogens on N11 and N23 are shown here.
however, was observed in this case. Lastly, the broad spectrum of
Hg–1 resolves into two well defined Q bands upon anion addition
(Fig. 4).
Fig. 2 UV/Visible titration of Fe(Cl)-1 with TBA H2PO4 in acetonitrile
at 293 K. Addition of 15 aliquots up to 1 equivalent of anion is shown.
Fig. 4 UV/Visible titration of Hg–1 with TBA H2PO4 in acetonitrile at
293 K. Addition of 15 aliquots up to 1 equivalent of anion is shown.
Job plot analyses showed anion binding to be in a 1:1
stoichiometric ratio for each receptor with each anion, where a
spectral response was obtained. The association constants for
19
ꢀ
C
1:1 complexes were calculated using the Specfit program and
are summarized in Table 2. No evidence of perchlorate anion
binding was noted in electronic spectral titrations with any of the
metalloporphyrins.
Receptors Zn–1, Cd–1 and Hg–1 all bind anions strongly in
acetonitrile. The anion binding of Cd–1 and Hg–1 is clearly
enhanced by the replacement of zinc with cadmium or mercury,
due to the greater Lewis acidity of these metals. It is noteworthy
that the halides are all bound more strongly than dihydrogen
phosphate by the cadmium and mercury complexes. This result
contrasts with the selectivity seen for Zn–1, which is related to
the trend of anion basicities. That only the addition of dihydro-
gen phosphate induces changes in the UV/visible spectrum of
Fig. 3 UV/Visible titration of Zn–1 with TBA Cl in acetonitrile at
293 K. Addition of 15 aliquots up to 1 equivalent of anion is shown.
The Q bands are labeled a and b.
The electronic spectral shifts seen for anion addition to Cd–1
were similar to those observed for Zn–1, i.e. a bathochromic shift
in the peak maxima and the intensity of the longer wavelength a
band increases relative to that of the b band. No colour change,
6734 | Dalton Trans., 2008, 6732–6741
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 2 Association constant (log K) data for Fe(Cl)-1, Zn–1, Cd–1 and
Hg–1 in acetonitrile determined from UV/visible titrations at 293 K,
errors < 0.1
observed for the corresponding complexes of 1. As an example, a
titration of Cd-TPP with iodide is displayed below (Fig. 5).
Anion
Fe(Cl)-1
Zn–1
Cd–1
Hg–1
a
a
a
Cl-
5.3
4.5
2.8
> 6
> 6
> 6
4.8
> 6
> 6
> 6
5.0
Br-
I-
-
H2PO4
NO3
4.2
5.1
-
a
a
3.2
2.9
a Perturbations seen were not sufficient to allow association constants to
be calculated.
Table 3 Association constant (log K) data for Cd–1 and Hg–1 in DMSO
and a water:DMSO 5:95 solvent mixture determined from UV/visible
titrations at 293 K, errors < 0.1
Cd–1
Hg–1
Hg–1
Fig. 5 UV/Visible titration of Cd-TPP with TBA I in DCM at 293 K. Ad-
dition of 15 aliquots up to 1 equivalent of anion is shown.
Cl-
4.3
3.6
2.4
3.6
> 6
> 6
> 6
4.4
> 6
> 6
> 6
4.2
Br-
I-
Job plot analyses showed the anion binding to be in a 1:1
ratio for each receptor with each anion. The association constants
for 1:1 complexes were calculated using the Specfit program
-
H2PO4
Solvent
DMSO
DMSO
water:DMSO 5:95
ꢀ
C
(Table 4). In the case of Cd-TPP and Hg-TPP, anion association
constants were too large to be calculated in this solvent, so
titrations were repeated in DMSO and a water:DMSO 5:95 solvent
mixture respectively and association constant values were derived
from the resulting data. Again, no evidence for perchlorate anion
binding was observed with any of the TPP metalloporphyrins.
For Zn-TPP, dihydrogen phosphate is bound most strongly, then
chloride, and it is clear that for Zn-TPP the trend of anion binding
strength being associated with the anion basicity, as observed
for the receptor Zn–1, is continued. Anion association constants
determined for Zn-TPP by other workers show similar trends to
those quoted above, although differences in solvent do not allow
for direct comparison of the values.21 Zn-TPP was not observed
to bind anions in acetonitrile. For Cd-TPP, the halides are bound
in line with the trend in anion basicities and in accordance with
previously published values.22 However, for Hg-TPP there was a
very interesting result, where the halides were bound in a selectivity
opposite to anion basicity, i.e., iodide > bromide > chloride >
dihydrogen phosphate, which is presumably due to the ‘soft’ nature
of the mercury Lewis acid metal centre. Attempts to carry out
titrations with Fe(Cl)-TPP in acetonitrile to allow comparison
with the results for Fe(Cl)-1 were foiled as Fe(Cl)-TPP was not
sufficiently soluble.
Fe(Cl)-1 indicates that this anion is the only one capable of
displacing chloride as the axial ligand in this complex.
The extremely large association constants found for Cd–1 and
Hg–1 in acetonitrile necessitated further titrations to be carried
out in the more competitive solvent medium DMSO and also in a
water:DMSO 5:95 solvent mixture for Hg–1. The results of these
titrations are summarized in Table 3.
Zn–1 does not bind anions at all in DMSO, but Cd–1 and
Hg–1 still bind anions very strongly. In this solvent, Cd–1 binds
the halide anions in the order of their basicities, i.e. chloride >
bromide > iodide, whereas dihydrogen phosphate is bound
relatively weakly, which is a different selectivity to Zn–1. The
metalloporphyrin Hg–1 binds the halides very strongly in DMSO
and also in the more competitive water:DMSO 5:95 solvent
mixture. All the halides are bound distinctly more strongly than
dihydrogen phosphate by this mercury complex, which contrasts
with the differing selectivity of Zn–1. Attempts to differentiate
the halide affinities of Hg–1 in more competitive solvents failed
due to solubility problems. It should be noted, however, that
dihydrogen phosphate anions can deprotonate in water:DMSO
mixtures making the exact determination of binding affinities for
this anion in this kind of solvent system challenging.
The absorption spectrum characteristics of the metal complexes
of TPP are well established.20 Anion titrations were carried
out with the TPP complexes of Fe(Cl), Ni, Cu, Zn, Cd and
Hg in dichloromethane. Only addition of dihydrogen phosphate
produced changes in the UV/visible spectrum of Fe(Cl) complex
of TPP (Fe(Cl)-TPP). In the case of Ni-TPP, no changes were
observed upon anion addition. It seems that the square planar
geometry favoured by Ni in strong field ligands such as the
porphyrin macrocycle precludes the formation of Ni-TPP–anion
complexes. As with Cu–1, no changes were observed upon anion
addition to Cu-TPP. Lastly, the changes that were seen upon anion
addition to Zn-TPP, Cd-TPP and Hg-TPP were similar to those
Table 4 Association constant (log K) data for Fe(Cl)-TPP, Zn-TPP, Cd-
TPP and Hg-TPP determined from UV/visible titrations at 293 K, errors
<
0.1
Fe(Cl)-TPP
Zn-TPP
Cd-TPP
Hg-TPP
a
a
a
Cl-
Br-
I-
H2PO4
NO3
3.3
3.9
3.4
3.3
5.0
5.5
> 6
a
a
-
4.0
4.3
3.7
4.0
-
a
a
a
a
Solvent
DCM
DCM
DMSO
Water:DMSO 5:95
a Spectral perturbations insufficient for an association constant to be
calculated.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 6732–6741 | 6735
©

View Article Online
Table 6 Summary of cathodic shifts (mV) in the first oxidation potential
of the porphyrin ring produced by addition of 5 equiv. of the TBA salt of
the anion. Solvent: 0.1 M TBA BF4 in a MeCN:DCM 1:1 solvent mixture,
T= 293 K
Comparing the anion association constants shown in Tables 2
and 3 to those of Table 4, it seems that the incorporation of amides
into the porphyrin structure greatly enhances anion recognition
for the zinc and iron complexes. Zn–1 binds anions in acetonitrile,
a much more competitive solvent than dichloromethane, while
Zn-TPP does not. Fe(Cl)-1 recognizes dihydrogen phosphate with
a larger association constant value in acetonitrile than Fe(Cl)-
TPP does in dichloromethane. Furthermore, the incorporation
of amides into the porphyrin structure largely enhances anion
recognition for the cadmium and mercury complexes. For Cd–1
the association constant is higher for chloride than for Cd-TPP,
whereas that of iodide is lower. Perceivably, while the ‘picket fence’
of amide groups aid anion complexation in the case of chloride,
they disfavour it for iodide, perhaps through unfavourable steric
interactions, for this compound. The association constants of
Hg–1 are uniformly higher than or the same as those for Hg-TPP.
-
-
Receptor
Cl-
H2PO4
NO3
a
b
Fe(Cl)-1
Co(PF6)-1
Cu–1
40
60
60
60
85
70
120
80
130
120
170
110
20
105
90
Zn–1
Cd–1
Hg–1
170
a Peak obscured by Cl- oxidation, b No peak observed.
Electrochemical anion binding studies
Electrochemistry is an anion sensing technique23 complementary
to UV/visible studies and porphyrins have well defined redox
characteristics. Consequently, the electrochemical properties of
the metal complexes of 1 were investigated using cyclic and
square wave voltammetry and they are summarized in Table 5.
For solubility reasons a dichloromethane:acetonitrile 1:1 solvent
mixture was used, with 0.1 M TBA tetrafluoroborate as the
supporting electrolyte. The first oxidation wave was assigned as
P/P+, which is definitely the case for the Cu, Zn, Cd and Hg
complexes,24,18 but may be a metal based oxidation for Fe(Cl) and
Co(PF6) complexes.25 The P+/P2+ waves were almost completely
irreversible for these receptors and hence only the Epa values are
noted.
Thesquare wavevoltammograms were recordedbefore andafter
addition of 5 equivalents of the TBA salt of the anion in question.
Significant cathodic shifts were produced upon anion addition
(Fig. 6) and they are displayed in Table 6.
Chloride, dihydrogen phosphate and nitrate anions all produced
cathodic shifts (some very large) in the P/P+ couple while there was
no shift for perchlorate. For receptors Zn–1, Cd–1 and Hg–1 the
data broadly supports that found via UV/visible investigations,
i.e., chloride and dihydrogen phosphate induced large shifts
in the redox couples while large association constants were
found for these anions via UV/visible titration studies. Nitrate
Fig. 6 Square wave voltammograms of Hg–1 before and after addition
of 5 equivalents of TBA H2PO4 (DHP) at 293 K. Solvent: 0.1 M TBA BF4
in a MeCN:DCM 1:1 solvent mixture.
produced smaller perturbations, which correlates with relatively
small association constants.
Fe(Cl)-1 showed a 40 mV shift upon addition of nitrate, but
unfortunately redox potential peaks could not be discerned upon
addition of chloride or dihydrogen phosphate. Co(PF6)-1 shows
similar results to Zn–1, Cd–1 and Hg–1. It is of interest that Cu–1
gives large shifts upon anion addition yet does not respond to
anions via UV/visible spectroscopic methods.
The electrochemistry of TPP complexes has already been
thoroughly investigated.26 The effect of anion addition to the TPP
complexes was investigated in an acetonitrile:dichloromethane 1:1
solvent mixture so as to be comparable to the results for the metal
complexes of 1. Unfortunately, the solubilities of Ni-TPP and Hg-
TPP were too poor in this solvent mixture for electrochemical
experiments to be carried out. Substantial cathodic shifts in the
P/P+ potential were seen upon anion addition and are displayed
in Table 7.
Table 5 Summary of data from the cyclic voltammograms of receptors
Fe(Cl)-1, Co(PF6)-1, Cu–1, Zn–1, Cd–1 and Hg–1 using a Ag/Ag+
reference electrode. Solvent: 0.1 M TBA BF4 MeCN:DCM 1:1 solvent
mixture. Scan rate= 100 mV/s
Table 7 Summary of cathodic shifts (mV) in the first oxidation potential
of the porphyrin ring of the TPP complexes produced by addition of
5 equiv. of the TBA salt of the anion. Solvent: 0.1 M TBA BF4 in a
MeCN:DCM 1:1 solvent mixture, T= 293 K
P/P+
P+/P2+
Epa (V)
Epc (V)
DEp (V)
Ipa/Ipc
Epa (V)
-
-
Receptor
Cl-
H2PO4
NO3
Fe(Cl)-1
Co(PF6)-1
Cu–1
0.915
0.755
0.730
0.595
0.530
0.625
0.800
0.690
0.660
0.530
0.460
0.555
0.115
0.065
0.070
0.065
0.090
0.070
2.3
1.4
1.5
1.2
3.4
4.9
1.335
1.325
a
a
a
1.340
Fe(Cl)-TPP
Cu-TPP
Zn-TPP
120
45
170
145
a
Zn–1
60
195
150
Cd–1
Hg–1
1.040
1.145
195
170
Cd-TPP
a Wave broad and indistinct.
a Peak unclear when anions added.
6736 | Dalton Trans., 2008, 6732–6741
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 9 Wavelength l/nm (intensity/AU) of the emissions of receptors
Zn–1 and Cd–1 in DMSO. lex was the peak maximum of the Soret band
Large shifts were seen upon addition of chloride, dihydrogen
phosphate and nitrate while perchlorate induced no shift at all.
The large shifts seen for chloride and dihydrogen phosphate
corroborate the results from UV/visible spectroscopy. The large
shift seen for nitrate, however, is unexpected, as little binding
was observed with this anion in UV/visible studies. One possible
explanation is that the oxidation of the porphyrin system ‘switches
on’ nitrate binding in some way. That Cu-TPP gives shifts upon
anion addition is also surprising as no anion binding was seen
for this receptor via UV/visible experiments, even in very non-
competitive solvents. However, Zn-TPP and Cd-TPP give much
larger shifts than Cu-TPP, supporting the UV/visible result that
Zn-TPP and Cd-TPP bind anions much more strongly than
Cu-TPP.
The magnitudes of the cathodic shifts seen for complexes of TPP
are somewhat higher than those seen for complexes of 1, which may
at first seem counterintuitive given the higher binding constants
of complexes of 1, as compared with those of TPP, derived from
UV/visible titration studies. However the observed anion induced
cathodic shift is proportional to the natural logarithm of the
ratio of the association constants of the oxidized and reduced
host system and hence cannot be regarded as a reflection of the
magnitude binding strength of the neutral receptor.
of each receptor (Table 1). [receptor]= 10-5
M
Receptor
Q¢ (b)
Q¢ (a)
Zn–1
Cd–1
613 (585)
627 (155)
664 (620)
680 (100)
constants to be calculated while the compounds were insufficiently
soluble in acetonitrile for titrations to be performed. Titrations
could not be carried out with receptors Fe(Cl)-1 and Cu–1 as they
are paramagnetic. It has been reported that the addition of anions
to Zn-TPP did not result in any changes in the 1H NMR spectrum
of Zn-TPP,8a therefore, no investigations were made with the TPP
complexes using 1H NMR spectroscopy.
Luminescence anion binding studies
The luminescence properties of receptors Zn–1 and Cd–1 were
investigated in DMSO (only the zinc and cadmium porphyrins
luminesce with sufficient intensity for their luminescence prop-
erties to be investigated28). Each receptor was excited at the
wavelength of its Soret band to allow the data to be comparable.
The wavelengths and intensities of the emission bands of the
receptors are displayed in Table 9.
1H NMR anion binding studies
The heavy atom effect operates for porphyrin fluorescence
spectra (i.e. ISC becomes more efficient due to spin–orbit coupling
being stronger for heavier atoms).29 Therefore the fluorescence
intensity for heavy atoms is less than that for light atoms as the
S1 excited state becomes swiftly depopulated, which is why the
intensity of emission for Cd–1 is less than that for Zn–1.
Anion binding studies were carried out with receptors Zn–1
and Cd–1 in acetonitrile and DMSO solutions. Large changes
were seen in the emission spectra of these receptors upon anion
addition; for example, the effect of chloride addition is shown in
Fig. 7. It can be seen that the overall intensity of the spectrum
drops, the intensity of the b peak decreases relative to the intensity
of the a peak, there is a bathochromic shift in the peak maxima
and three isosbestic points appear.
A 1:1 CD3CN:CDCl3 solution of receptor Zn–1 was titrated
with aliquots of the TBA salts of various anions, whereupon
the amide resonance was seen to shift significantly downfield.
Similar titrations were attempted with receptor Co(PF6)-1, but
unfortunately, this receptor was not sufficiently soluble in this
solvent mixture to allow titrations to be carried out.
Job plot analysis of the data showed that the anion:host binding
was in a 1:1 stoichiometry. EQNMR,27 a least squares fitting
computer program, was used to calculate association constants for
the anion binding, using a 1:1 model to fit the data, and the values
thus derived are displayed in Table 8. Negligible perturbations
of the amide resonance were observed upon addition of the
perchlorate anion, which indicates no anion binding in this case.
The anion association constant for chloride is larger than can
be calculated reliably by EQNMR and unfortunately, peak broad-
ening prevented an accurate value for dihydrogen phosphate being
calculated. However, it can be seen that the anions are bound, in
order of decreasing anion strength: chloride > bromide > iodide >
nitrate which is in very good agreement with the trend seen from
the UV/visible titration studies, a pleasing result.
NMR spectroscopic methods were unsuitable for investigating
the anion binding properties of the other receptors. Addition
of anions to Cd–1 and Hg–1 in d6-DMSO produced only small
shifts in the amide peaks, which did not allow anion association
Table 8 Association constant data (log K) for Zn–1 determined from 1H
NMR titrations at 293 K in a CD3CN:CDCl3 1:1 solvent mixture, errors
<
10%
Fig. 7 Luminescent titration of Zn–1 with TBA Cl in acetonitrile at
Cl-
Br-
4.0
I-
NO3
2.2
H2PO4
-
-
293 K. Addition of 29 aliquots up to 10 equivalents of anion are shown.
Anion
a
Association constant
> 4.0
2.3
Job plot analyses showed the binding to be in a 1:1 ratio for
both receptors with each anion. The association constants for 1:1
a Peak broadening prevented quantitative analysis.
ꢀ
C
complexes were calculated using the Specfit program and are
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 6732–6741 | 6737
©

View Article Online
1H NMR spectra were recorded on a Varian Mercury spec-
trometer operating at 300 MHz, 13C NMR at 75.48 MHz, and
31P NMR at 282.46 MHz. Absorption spectra were obtained on
a Perkin Elmer Lambda 6 UV/visible spectrophotometer. Cyclic
voltammetry was performed on an Eco Chemie mAutolab Type
II combined with General Purpose Electrochemical Software
version 4.9. A glassy carbon working electrode, a Ag/Ag+
reference electrode and a platinum counter electrode were used
in the instrumental setup. TBA BF4 was used as the supporting
electrolyte. Mass spectrometry was performed in the Inorganic
Chemistry Department at the University of Oxford. Stephen
Boyer of the SACS at London Metropolitan University performed
elemental analyses.
Table 10 Association constant (log K) data for Zn–1 and Cd–1 deter-
mined from luminescence titrations at 293 K, errors < 0.1
Zn–1
Cd–1
Cd–1
Cl-
5.1
5.3
5.0
5.3
> 6
> 6
> 6
5.7
4.1
4.4
3.7
3.7
Br-
I-
-
H2PO4
Solvent
MeCN
MeCN
DMSO
summarized in Table 10. No evidence of perchlorate or nitrate
anion binding was noted in luminescence titrations with either of
these compounds.
These studies indicate that Cd–1 binds anions much more
strongly than Zn–1 and that anions are bound strongly in DMSO,
which is in agreement with the UV/visible studies. Surprisingly, in
contrast to the results of the UV/visible studies, the luminescence
studies indicate that Zn–1 demonstrates no particular selectivity
for the binding of one anion over another. In addition, the order
of halide binding preference of Cd–1 in DMSO, is at odds with
the UV/visible studies, with bromide being bound most strongly.
The source of the discrepancies between the UV/visible and
luminescence studies is unclear, but it could be due to the halides
affecting the luminescence of the porphyrin ring via the heavy
atom effect.30
Syntheses
3,4-Dimethoxybenzoyl chloride (2). 5 g (27.4 mmol) of 3,4-
dimethoxybenzoic acid was dissolved in 20 ml of thionyl chloride
and this solution was refluxed overnight. The solvents were
removed under reduced pressure, via secondary trap leaving a red
solid (5.42 g, 27.0 mmol, 98.5%). Electrospray M.S. m/z: 165.05
(M - Cl)+ 1H NMR (300 MHz, CDCl3) d: 7.84 (1H, dd, J= 8 and
.
2 Hz, aryl H), 7.53 (1H, d, J= 2 Hz, aryl H), 6.94 (1H, d, 8 Hz,
aryl H), 3.98 and 3.94 (3H, s, CH3).
cis-5,10,15,20-Tetrakis(2-[(3,4-dimethoxybenzoamido)phenyl])-
porphyrin (1). 1.28 g (1.90 mmol) of a,a,a,a-H2TAPP was
dissolved in 50 ml of dry CH2Cl2. 3 ml (excess) of NEt3 was
added and the stirred solution was placed in an ice bath. 5.42 g
(27.0 mmol) of 2, dissolved in 20 ml of dry CH2Cl2, was added
dropwise. After 18 hours, the solvents were removed in vacuo,
50 ml of H2O was added to the resulting purple residue, the
suspension thus produced was stirred for 20 minutes and filtered.
The solid was washed with 3 ¥ 30 ml portions of MeOH and 2 ¥
30 ml portions of Et2O, to leave a fine purple powder (1.80 g,
The lack of luminescence response on the addition of TBA
salts to solutions of Zn TPP and Cd TPP in dichloromethane and
DMSO respectively confirmed the necessity of amide groups for
mediating the anion binding event.
Conclusions
A variety of metals have been successfully inserted into an amide
functionalized porphyrin structure. The anion binding of these
receptors has been compared with that of their tetraphenylpor-
phyrin analogues, where in each case the most effective anion
receptors were based on the Group 12 elements, zinc, cadmium and
mercury. The cadmium and mercury receptors have been shown to
bind anions strongly in highly competitive solvent mixtures and
to have differing selectivity to that found for the zinc receptors.
The receptors based on the first row transition metals iron, cobalt,
nickel and copper have been shown to be poor anion receptors.
The amide groups of the metal complexes of 1 have been shown
to enhance anion binding. This work lays the foundations for
future porphyrin based anion receptors where the metal centre
will be cadmium or mercury in order to recognize anions in more
competitive solvent systems.
1.35 mmol, 71.1%). Electrospray M.S. m/z: 1331.50 (M + H)+
.
1H NMR (300 MHz, CDCl3) d: 8.98 (8H, s, pyrrole H), 8.74, (4H,
d, J = 8 Hz, aryl H), 7.86 (12H, m, aryl H and C(O)NH), 7.53 (4H,
t, J = 8 Hz, aryl H), 6.34 (4H, s, aryl H), 5.95 (4H, d, J = 9 Hz,
aryl H), 5.54 (4H, J = 9 Hz, aryl H), 3.11 and 2.71 (12H, s, OCH3),
-2.58 (2H, s, NH pyrrole). UV/Vis Maxima found at: 426, 519,
554, 592 and 651 nm (DMSO).
Iron 5,10,15,20-tetrakis(2-[(3,4-dimethoxy-benzoamido)phenyl])-
porphyrin chloride (Fe(Cl)-1). 0.13 g (0.1 mmol) of 1 was taken up
in 30 ml of a CH2Cl2: acetic acid 1:1 solvent mixture. To this stirred
solution were added 10 drops of 2,6-lutidine and 0.5 g (excess) of
FeCl2.4H2O. The reaction was monitored by UV spectroscopy
and after 7 hours the reaction appeared to be complete. The
solvents were removed in vacuo and the residue was taken up
in 50 ml of CH2Cl2. This brown solution was washed with 50 ml
Experimental
Materials and general procedures
R
of H2O, 2 ¥ 50 ml of brine and filtered through Celiteꢀ . The
Solvents (HPLC grade) were purchased from Rathburn Chemi-
cals, Alfa Aesar and Fisher Scientific. Solvent drying was carried
out by degassing via sparging with dinitrogen and then passing
through a column of activated alumina using Grubbs apparatus.31
Solvent thus dried was used immediately. The starting materials
were used as received (Acros, Aldrich) without any further
purification except for pyrrole which was distilled under nitrogen
just before use.
solvents were removed in vacuo and the product was precipitated
from a saturated CH2Cl2 solution using 50 ml of Et2O to give
a brown solid (71 mg, 0.05 mmol, 50%). Silica gel thin layer
chromatography of this product showed only one spot using a
MeOH:CH2Cl2 5:95 solvent mixture as eluent. MALDI-TOF m/z:
1384.61 (M - Cl)+. Elemental (%) calculated for C80H64N8O12FeCl:
C(67.6), H(4.5), N(7.8), found: C(67.6), H(4.5), N(7.8). UV/Vis
maxima found at: 428, 521, 592 and 649 nm (DMSO).
6738 | Dalton Trans., 2008, 6732–6741
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Cobalt 5,10,15,20-tetrakis(2-[(3,4-dimethoxybenzoamido)phe-
nyl])porphyrin hexafluorophosphate (Co(PF6)-1). 0.13
(75.43 MHz, CDCl3) d: 165.21, 151.45, 150.73, 148.23, 138.27,
135.80, 132.86, 132.20, 130.00, 126.31, 123.37, 121.42, 118.58,
116.32, 109.85, 109.65, 55.33, 55.19. Elemental (%) calculated for
C80H64N8O12Zn: C(68.9), H(4.6), N(8.0), found: C(68.8), H(4.6),
N(7.9). UV/Vis maxima found at: 434, 564 and 602 nm (DMSO).
g
(0.1 mmol) of 1 was taken up in 30 ml of a CH2Cl2: MeOH 1:1
solvent mixture. To this stirred solution was added 0.5 g (excess)
of CoCl2 and this mixture was heated at 30 ◦C. The reaction
was monitored by UV spectroscopy and after five hours the
reaction appeared to be complete. The solvents were removed
in vacuo and the residue was taken up in 30 ml of CH2Cl2. This
Cadmium 5,10,15,20-tetrakis(2-[(3,4-dimethoxybenzoamido)-
phenyl])porphyrin (Cd–1). 0.13 g (0.1 mmol) of 1 was taken up
in 30 ml of a CH2Cl2: MeOH 1:1 solvent mixture. To this stirred
solution was added 0.5 g (excess) of Cd(acac)2. After 16 hours,
50 ml each of CH2Cl2 and H2O were added to the dark green
solution. The phases were separated and the organics were washed
with 2 ¥ 50 ml of H2O. The solvents were removed in vacuo and
the residue was precipitated from a saturated CH2Cl2 solution
using 30 ml of hexane to give a bright green powder (77 mg,
R
orange-red solution was filtered through Celiteꢀ. The solvents
were removed in vacuo and the product was precipitated from
a saturated CH2Cl2 solution using 30 ml of hexane to give an
orange-brown solid. This was taken up in a MeOH:CH2Cl2 1:4
solvent mixture and 10 ml of sat. NH4PF6(aq) was added. This
biphasic mixture was stirred for 16 hours before the organics were
separated and washed with 2 ¥ 50 ml of H2O. The solvents were
removed in vacuo and the product precipitated from a saturated
CH2Cl2 solution using 50 ml of Et2O, to give a crimson solid
(70 mg, 0.05 mmol, 50%). MALDI-TOF m/z: 1387.42 (M-PF6)+.
1H NMR (300 MHz, d6-DMSO) d: 9.61 (4H, s, NH), 9.17 (8H, s,
pyrrole H), 8.01 (4H, d, J = 8 Hz, aryl H), 7.88 (8H, m, aryl H),
7.66 (4H, t, J = 7 Hz, aryl H), 6.68 (4H, s, aryl H), 6.11 (4H,
d, J = 8 Hz, aryl H), 5.92 (4H, d, J = 8 Hz, aryl H), 3.12 (12H, s,
OCH3), 2.70 (12H, s, OCH3). 13C NMR (75.43 MHz, CDCl3) d:
165.46, 151.76, 148.58, 144.23, 143.64, 138.61, 135.59, 134.92,
133.61, 130.37, 130.21, 125.92, 124.86, 118.59, 110.50, 109.94,
55.94, 54.93. Elemental (%) calculated for C80H64CoF6N8O12P:
C(62.7), H(4.1), N(7.5), found: C(62.5), H(4.5), N(7.8). UV/Vis
maxima found at: 443 and 558 nm (DMSO).
1
0.05 mmol, 50%). MALDI-TOF m/z: 1442.37 (M)+. H NMR
(300 MHz, CDCl3) d: 8.64 (8H, s, pyrrole H), 8.05 (4H, br s,
aryl H), 7.71 (8H, m, aryl H), 7.54 (4H, br s, aryl H), 7.41 (4H,
t, J = 8 Hz, aryl H), 5.57 (4H, br s, aryl H), 5.33 (8H, br m,
aryl H), 3.02 and 2.37 (12H, s, OCH3). 13C NMR (75.43 MHz,
CDCl3) d: 165.15, 151.22, 151.00, 147.63, 137.78, 135.94, 132.73,
132.59, 129.75, 125.50, 123.19, 120.97, 118.59, 116.51, 109.36,
109.12, 55.13, 54.83. Elemental (%) calculated for C80H64N8O12Cd:
C(66.6), H(4.5), N(7.8), found: C(66.8), H(4.4), N(7.7). UV/Vis
maxima found at: 443, 577 and 618 nm (DMSO).
Mercury 5,10,15,20-tetrakis(2-[(3,4-dimethoxybenzoamido)-
phenyl])porphyrin (Hg–1). 0.13 g (0.1 mmol) of 1 was taken
up in 20 ml of a CH2Cl2: MeOH 1:1 solvent mixture. 0.5 g
(excess) of Hg(OAc)2 was added to this stirred solution. After
30 minutes UV spectroscopy showed that the reaction had gone
to completion. 50 ml of CH2Cl2 was added to the reaction
mixture and the resulting green solution was washed with 3 ¥
Copper 5,10,15,20-tetrakis(2-[(3,4-dimethoxybenzoamido)phe-
nyl])porphyrin (Cu–1). 0.27 g (0.20 mmol) of 1 was taken up in
30 ml of CH2Cl2. To this stirred solution was added a slurry of
0.40 g (2.0 mmol) of Cu(OAc)2·H2O and 25 ml of MeOH. The
mixture was heated at reflux for 5 hours, before the solvents
were removed in vacuo. The red residue was taken up in 50 ml
of CH2Cl2, washed with 3 ¥ 50 ml of H2O, dried with Na2SO4,
filtered and the solvents were removed in vacuo. The product
was precipitated from a saturated CHCl3 solution using 50 ml
of Et2O, to leave a red powder (0.20 g, 0.14 mmol, 70.0%).
Silica gel thin layer chromatography of this product showed
only one spot using a MeOH:CH2Cl2 5:95 solvent mixture as
eluent. Electrospray M.S. m/z: 1414.36 (M + Na)+. Elemental
(%) calculated for C80H64N8O12Cu: C(69.0), H(4.6), N(8.0), found:
C(69.0), H(4.5), N(7.9). UV/Vis maxima found at: 424 and 545 nm
(DMSO).
R
50 ml of H2O. The organics were filtered through Celiteꢀ and
the solvents were removed in vacuo from the filtrate. The product
was precipitated from a saturated CH2Cl2 solution using 30 ml
of hexane to give a dull green solid (142 mg, 0.09 mmol, 90%).
1
MALDI-TOF m/z: 1530.76 (M)+. H NMR (300 MHz, CDCl3)
d: 8.89 (8H, s, pyrrole H), 8.67 (4H, d, J = 8 Hz, aryl H), 8.06
(4H, s, NH), 7.77 (8H, m, aryl H), 7.43 (4H, t, J = 8 Hz, aryl H),
6.43 (4H, s, aryl H), 5.94 (4H, d, J = 8 Hz, aryl H), 5.54 (4H,
d, J = 8 Hz, aryl H), 3.07 and 2.84 (12H, s, OCH3). 13C NMR
(75.43 MHz, CDCl3) d: 165.31, 152.38, 151.58, 148.58, 138.66,
136.29, 133.46, 132.16, 130.13, 126.79, 123.35, 121.64, 118.63,
118.54, 117.18, 109.83, 55.40, 55.17. Elemental (%) calculated for
C80H64N8O12Hg:C(62.8), H(4.2), N(7.3), found: C(62.7), H(4.1),
N(7.3). UV/Vis maximum found at: 455 nm (DMSO).
Zinc 5,10,15,20-tetrakis(2-[(3,4-dimethoxybenzoamido)phenyl])-
porphyrin (Zn–1). 0.27 g (0.20 mmol) of 1 was taken up in 30 ml
of CH2Cl2. To this stirred solution was added 0.44 g (2.0 mmol)
of Zn(OAc)2.2H2O in 20 ml of MeOH. After 24 hours, 50 ml
each of CH2Cl2 and H2O were added. The phases were separated
and the organics were washed with 2 ¥ 50 ml of H2O, dried with
Na2SO4, filtered and solvents were removed in vacuo. The product
was precipitated from a saturated CHCl3 solution using 30 ml
of hexane, to leave a purple powder (0.21 g, 0.15 mmol, 75.0%).
Electrospray M.S. m/z: 1414.75 (M + Na)+. 1H NMR (300 MHz,
CDCl3) d: 8.98 (8H, s, pyrrole H), 8.56, (4H, d, J= 8 Hz, aryl
H), 7.82 (12H, m, aryl H and C(O)NH), 7.49 (4H, t, J= 8 Hz,
aryl H), 6.32 (4H, s, aryl H), 5.74 (4H, d, J= 9 Hz, aryl H), 5.50
(4H, J= 9 Hz, aryl H), 3.10 and 2.88 (12H, s, OCH3). 13C NMR
UV/Visible anion titration protocol
In a typical experiment, aliquots of guest (2.5 ¥ 10-5 mol in
5 ml) were added to a 3 ml solution of the host (5 ¥ 10-5 M)
at 293 K. Twenty eight aliquots were added (13 ¥ 2 ml, 1 ¥
4 ml, 6 ¥ 5 ml, 4 ¥ 15 ml, 4 ¥ 60 ml). Spectra were recorded
ꢀ
C
and the data was analysed by the computer program Specfit .
The spectra together with the host and guest concentrations were
read into the program for every titration point and the complex
stoichiometry and whether the components species were coloured
was entered. The parameters were refined by global analysis that
uses singular value decomposition and non-linear modelling by the
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 6732–6741 | 6739
©

View Article Online
Levenberg-Marquardt method. Using the calculated stability con-
stants, the program plots the predicted spectra of the component
species together with the observed and calculated absorption
versus guest concentration at a given wavelength, both of which
reveal the accuracy of the experimental data and the suitability
of the model. The program also gives the best-fit values of the
stability constants together with their errors. The parameters were
varied until the values for the stability constants converged.
2 H. E. Katz, J. Am. Chem. Soc., 1985, 107, 1420–1421; H. E. Katz,
Organometallics, 1987, 6, 1134–1136.
3 J. D. Wuest and B. Zacharie, Organometallics, 1985, 4, 410–411; K.
Tamao, T. Hayashi, Y. Ito and M. Shiro, Organometallics, 1992, 11,
2099–2114.
4 M. Newcomb, A. M. Madonik, M. T. Blanda and J. K. Judice,
Organometallics, 1987, 6, 145–150; M. Newcomb, J. H. Horner, M. T.
Blanda and P. J. Squattrito, J. Am. Chem. Soc., 1989, 111, 6294–6301.
5 A. L. Chistyakov, I. V. Stankevich, N. P. Gambaryan, Y. T. Struchkov,
A. I. Yanovsky, I. A. Tikhonova and V. B. Shur, J. Organomet. Chem.,
1996, 508, 271–273; V. B. Shur, I. A. Tikhonova, A. I. Yanovsky, Y. T.
Struchkov, P. V. Petrovskii, S. Y. Panov, G. G. Furin and M. E. Vol,’pin,
J. Organomet. Chem., 1991, 418, C29–C32; I. A. Tikhonova, F. M.
Dolgushin, K. I. Tugashov, O. G. Ellert, V. M. Novotortsev, G. G.
Furin, M. Y. Antipin and V. B. Shur, J. Organomet. Chem., 2004, 689,
82–87.
6 R. Martinez-Manez and F. Sancenon, Chem. Rev., 2003, 103, 4419–
4476; P. D. Beer and E. J. Hayes, Coord. Chem. Rev., 2003, 240, 167–189;
C. Sukasi and T. Tuntulani, Chem. Soc. Rev., 2003, 32, 192–202.
7 M. J. Gunter, S. M. Farquhar and K. M. Mullen, New J. Chem., 2004,
28, 1443–1449; M. Takeuchi, T. Shioya and T. M. Swager, Angew.
Chem., Int. Ed., 2001, 40, 3372–3376; M. Dudicˇ, P. Lhota´k, I. Stibor,
K. Lang and P. Prosˇkova´, Org. Lett., 2003, 5, 149–152; P. K. Panda and
C.-H. Lee, J. Org. Chem., 2005, 70, 3148–3156; C. Bucher, C. H.
Devillers, J.-C. Moutet, G. Royal and E. Saint-Aman, New J. Chem.,
2004, 28, 1584–1589; S. D. Starnes, S. Arungundram and C. H.
Saunders, Tetrahedron Lett., 2002, 43, 7785–7788.
8 (a) P. D. Beer, M. G. D. Drew and R. Jagessar, J. Chem. Soc., Dalton
Trans., 1997, 881–886; (b) P. D. Beer, D. P. Cormode and J. J. Davis,
Chem. Commun., 2004, 414–415; (c) D. P. Cormode, S. S. Murray,
A. R. Cowley and P. D. Beer, Dalton Trans., 2006, 5135–5140; (d) D. P.
Cormode, J. J. Davis and P. D. Beer, J. Inorg. Organomet. Polym., 2008,
18, 32–40.
9 R. C. Jagessar, M. Shang, W. R. Scheidt and D. H. Burns, J. Am.
Chem. Soc., 1998, 120, 11684–11692; C. Lee, D. H. Lee and J.-I. Hong,
Tetrahedron Lett., 2001, 42, 8665–8668.
10 V. G. Gore and N. S. Narasimhan, J. Chem. Soc., Perkin Trans. 1, 1988,
481–483.
11 J. P. Collman, R. R. Gagne, C. A. Reed, T. R. Halbert, G. Long and
W. T. Robinson, J. Am. Chem. Soc., 1975, 97, 1424–1439.
12 J. S. Lindsey, J. Org. Chem., 1980, 45, 5215–5215.
13 B. Boitrel, V. Baveux-Chambenoˆıt and P. Richard, Helv. Chim. Acta,
2004, 87, 2447–2464; T. Imaoka, H. Horiguchi and K. Yamamoto,
J. Am. Chem. Soc., 2003, 125, 340–341; M. J. Crossley, P. J. Sintic,
R. Walton and J. R. Reimers, Org. Biomol. Chem., 2003, 1, 2777–2787;
M.-C. Wang, L.-C. Sue, B.-C. Liau, B.-T. Ko, S. Elango and J.-H. Chen,
Inorg. Chem., 2001, 40, 6064–6068.
14 J. Huet, A. Gaudemer, C. Boucly-Goester and P. Goucly, Inorg. Chem.,
1982, 21, 3413–3419.
15 A. D. Adler, F. R. Longo, J. D. Finarelli, J. Goldmacher, J. Assour
and L. Korsakoff, J. Org. Chem., 1967, 32, 476–476; G. H. Barnett,
M. F. Hudson and K. M. Smith, J. Chem. Soc., Perkin Trans. 1, 1975,
1401–1403.
16 A. D. Adler, F. R. Longo, K. Kampas and J. Kim, J. Inorg. Nucl. Chem.,
1970, 32, 2443–2445; E. G. Azenha, A. C. Serra, M. Pineiro, M. M.
Pereira, J. S. de Melo, S. J. Formosinho and A. M. d’A. R. Gonsalves,
Chem. Phys., 2002, 280, 177–190; M. E. Nino, S. A. Giraldo and E. A.
Paez-Mozo, J. Mol. Catal., 2001, 175, 139–151; M. F. Hudson and
K. M. Smith, Tetrahedron, 1976, 32, 597–601; M. F. Hudson and K. M.
Smith, Tetrahedron Lett., 1974, 15, 2223–2226.
17 L. R. Milgrom, The Colours of Life, Oxford University Press, 1997;
M. Gouterman, The Porphyrins, ed. D. Dolphin, Academic Press, New
York, 1978, Vol. 3, Ch 1.
Luminescence anion titration protocol
In a typical experiment, aliquots of guest (1.5 ¥ 10-5 mol in
5 ml) were added to a 3 ml solution of the host (1 ¥ 10-5 M)
at 293 K. Twenty nine aliquots were added (15 ¥ 2 ml and 14 ¥
5 ml). Spectra were recorded and the data was analysed by the
ꢀ
C
computer program Specfit , in a method as described above.
X-Ray crystallography
Diffraction data were measured with MoKa radiation at 298 K
using the MARresearch Image Plate System. The crystal was
positioned at 70 mm from the Image Plate. The crystal diffracted
weakly and only data up to 2q of 45◦ were collected. A total of
95 frames were measured at 2◦ intervals with a counting time of
10 min to give 17241 reflections. Data analysis was carried out
with the XDS program32 to give 9051 independent reflections with
an R(int) of 0.0654. The structure was solved using direct methods
with the SHELXS-97 program.33 The non-hydrogen atoms were
refined with anisotropic thermal parameters. The hydrogen atoms
bonded to carbon were included in geometric positions and given
thermal parameters equivalent to 1.2 times those of the atom to
which they were attached. Four hydrogen atoms were located in a
difference Fourier map bound to the four nitrogen atoms of the
porphyrin core and these were refined with 50% occupancy. There
was one ethanol solvent molecule and three water molecules in
the asymmetric unit all refined with 50% occupancy. There were
appreciable voids in the unit cell but there were no significant
peaks and it can be assumed that they were occupied, if at all, by
disordered solvent at low occupancy. The hydrogen atoms on the
water molecules could not be located. The structure was refined
on F2 using SHELXL-9733 to give R1 0.0944 and wR2 0.2456 for
4906 reflections with I>2s(I).
Crystal Data for 1: C81 H72 N8 O14, M = 1381.47, monoclinic,
spacegroup P21/c, Z = 4, a = 12.727(15), b = 25.271(28), c =
◦
3
23.357(9) A, b = 94.49(1) , U = 7489(5) A , dcalc = 1.225 g cm-3.
˚
˚
The cif file for this structure may be found in the ESI.†
Acknowledgements
We thank the EPSRC for a studentship (DPC) and Dr Jason
Davis for use of his electrochemical equipment. We gratefully
acknowledge Dr Michael Lankshear and Dr Simon Bayly for their
critical reading of the manuscript.
18 K. M. Kadish and L. R. Shuie, Inorg. Chem., 1982, 21, 3623–
3630.
19 R. A. Binstead, A. D. Zuberbuhler and B. Jung, Specfit 3.0.30,
Spectrum Software Associates, 2002.
20 See, for example G. D. Dorough, J. R. Miller and F. M. Huennekens,
J. Am. Chem. Soc., 1951, 73, 4315–4320.
References
21 M. Nappa and J. S. Valentine, J. Am. Chem. Soc., 1978, 100, 5075–
5080; A. S. Hinman and B. J. Pavelich, J. Electroanal. Chem., 1989,
269, 53–61.
1 J. L. Sessler, P. A. Gale and W.-S. Cho, Anion Receptor Chemistry, RSC,
Cambridge, 2006; P. A. Gale and R. Quesada, Coord. Chem. Rev., 2006,
250, 3219–3244; K. Bowman-James, Acc. Chem. Res., 2005, 38, 671–
678; P. A. Gale, Coord. Chem. Rev., 2003, 240, 191–221; P. D. Beer and
P. A. Gale, Angew. Chem., Int. Ed., 2001, 40, 486–516.
22 T. Ozawa and A. Hanaki, Chem. Pharm. Bull., 1983, 31, 2110–
2113.
23 P. D. Beer, Adv. Inorg. Chem., 1992, 39, 79–157.
6740 | Dalton Trans., 2008, 6732–6741
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
24 M. W. Renner, K. M. Barkigia, Y. Zhang, C. J. Medforth, K. M. Smith
and J. Fajer, J. Am. Chem. Soc., 1994, 116, 8582–8592; J. Fajer, D. C.
Borg, A. Forman, D. Dolphin and R. H. Felton, J. Am. Chem. Soc.,
1970, 92, 3451–3459; J. Fajer, D. C. Borg, A. Forman, R. H. Felton, L.
Vegh and D. Dolphin, Ann. N. Y. Acad. Sci., 1973, 206, 349–364.
25 See, for example: M. A. Phillippi, E. T. Shimomura and H. M. Goff,
Inorg. Chem., 1981, 20, 1322–1325; S. Fukuzumi, K. Miyamoto, T.
Suenobu, E. Van Caemelbecke and K. M. Kadish, J. Am. Chem. Soc.,
1998, 120, 2880–2889.
28 G. Szintay and A. Horva´th, Inorg. Chim. Acta, 2001, 324, 278–285;
R. L. Ake and M. Gouterman, Theor. Chim. Acta, 1969, 15, 20–42.
29 A. Harriman, J. Chem. Soc., Faraday Trans. 2, 1981, 77, 1281–1291.
30 D. J. Quimby and F. R. Longo, J. Am. Chem. Soc., 1975, 97, 5111–5117.
31 P. J. Alaimo, D. W. Peters, J. Arnold and R. G. Bergman, J. Chem. Educ.,
2001, 78, 64–64; A. B. Pangborn, M. A. Giardello, R. H. Grubbs, R. K.
Rosen and F. J. Timmers, Organometallics, 1996, 15, 1518–1520.
32 W. Kabsch, J. Appl. Crystallogr., 1988, 21, 916–924.
33 G. M. Sheldrick, SHELXS-97, Program for solution of crystal struc-
tures, University of Go¨ttingen, Germany, 1997; G. M. Sheldrick,
SHELXL-97, Program for refinement of crystal structures, University
of Go¨ttingen, Germany, 1997.
26 See, for example D. Chang, T. Malinski, A. Ulman and K. M. Kadish,
Inorg. Chem., 1984, 23, 817–824.
27 M. J. Hynes, J. Chem. Soc., Dalton Trans., 1993, 311–312.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 6732–6741 | 6741
©