Design of protonated polyazamacrocycles based on phenanthroline motifs for selective uptake of aromatic carboxylate anions and herbicides

Article abstract of DOI:10.1002/chem.200800993

Three novel large polyazamacrocycles containing two 1,10-phenanthroline (phen) units connected by two polyamine spacers of different length, [32]phen₂N₄, [30]phen₂N₆ and Me ₂[34]phen₂N₆, have been synthesised and their protonated forms used as receptors for binding studies with several aromatic carboxylate anions (benzoate (bzc^-), 1-naphthalate (naphc ^-), 9-anthracenate (anthc^-), pyrene-1-carboxyl-ate (pyrc^-), phthalate, (ph^2-), isophthalate (iph ^2-), terephthalate (tph^2-), 2,5-dihydroxy-1,4- benzenediacetate (dihyac^2-) and, 1,3,5-benzenetricarboxylate (btc^3-)) and three herbicides (4-amino-3,5,6-trichloropyridine-2- carbox-ylate (ATCP^-), dichlorophenoxyacetate (2,4-D^-) and glyphosate (PMG^2-)) in water solution. The [30]phen₂N ₆ receptor was found to be the most suitable for binding the anions considered in a 1:1 stoichiometry. The three receptors exhibit a remarkable binding selectivity towards the extended aromatic anion pyre^- at low pH values. Their binding affinities for the monocarboxylate anions decrease with the extension of the aromatic system in the order pyre^- > anthc^- > naphc^- > bzc^-, which indicates the presence of π-π stacking interactions in the molecular recognition of these anions. Molecular dynamics simulations carried out for the binding of (H₄[30]phen₂N₆}⁴⁺ and {H ₆Me₂[34]phen₂N₆}⁶⁺ with pyre^-, anthc^-, naphc^-, iph^2- and btc^3- in water showed that these receptors adopt a folded conformation with the anion inserted between the two phen heads and that the molecular recognition is governed by π-π stacking interactions and multiple N-H...O=C hydrogen bonds. The binding free energies estimated theoretically are very similar to those found by Potentiometric methods, which supports the proposed binding arrangement.

Full text of DOI:10.1002/chem.200800993

FULL PAPER
DOI: 10.1002/chem.200800993
Design of Protonated Polyazamacrocycles Based on Phenanthroline Motifs
for Selective Uptake of Aromatic Carboxylate Anions and Herbicides
Carla Cruz,[a] Vꢀnia Calisto,[a] Rita Delgado,[b, c] and Vꢁtor Fꢂlix*[a]
Abstract: Three novel large polyaza-
macrocycles containing two 1,10-phen-
anthroline (phen) units connected by
two polyamine spacers of different
length, [32]phen2N4, [30]phen2N6 and
Me2[34]phen2N6, have been synthesised
and their protonated forms used as re-
ceptors for binding studies with several
aromatic carboxylate anions (benzoate
(bzcÀ), 1-naphthalate (naphcÀ), 9-an-
thracenate (anthcÀ), pyrene-1-carboxyl-
ate (pyrcÀ), phthalate, (ph2À), isophtha-
late (iph2À), terephthalate (tph2À), 2,5-
dihydroxy-1,4-benzenediacetate
tate (2,4-DÀ) and glyphosate (PMG2À))
in water solution. The [30]phen2N6 re-
ceptor was found to be the most suita-
ble for binding the anions considered
in a 1:1 stoichiometry. The three recep-
tors exhibit a remarkable binding selec-
tivity towards the extended aromatic
anion pyrcÀ at low pH values. Their
binding affinities for the monocarboxy-
late anions decrease with the extension
of the aromatic system in the order
pyrcÀ >anthcÀ >naphcÀ >bzcÀ, which
indicates the presence of p–p stacking
interactions in the molecular recogni-
tion of these anions. Molecular dynam-
ics simulations carried out for the bind-
ing
{H6Me2[34]
of
{H4[30]phen2N6}4+
phen2N6}6+
with
and
N
pyrcÀ,
anthcÀ, naphcÀ, iph2À and btc3À in
water showed that these receptors
adopt a folded conformation with the
anion inserted between the two phen
heads and that the molecular recogni-
tion is governed by p–p stacking inter-
À
actions and multiple N H···O=C hy-
drogen bonds. The binding free ener-
gies estimated theoretically are very
similar to those found by potentiomet-
ric methods, which supports the pro-
posed binding arrangement.
(dihyac2À) and, 1,3,5-benzenetricarbox-
ylate (btc3À)) and three herbicides (4-
amino-3,5,6-trichloropyridine-2-carbox-
ylate (ATCPÀ), dichlorophenoxyace-
Keywords:
anions
·
molecular
dynamics · molecular recognition ·
pi interactions · receptors
Introduction
life cycles.[3–6] For example, the herbicide 4-amino-3,5,6-tri-
chloropyridine-2-carboxylic acid, H(ATCP), commonly used
AHCTUNGTRENNUNG
Anion recognition by cationic receptors is, nowadays, an
active area of research.[1–3] Aromatic carboxylate anions are
widespread environmental contaminants of agricultural land
causing ground water contamination that disrupts aquatic
to control deeply rooted herbaceous weeds and woody
plants, pastures, and small grain crops, is very toxic and is
rapidly absorbed from ground water.[4]
Thus, the development of receptors capable of sensing
and extracting pollutants containing carboxylate groups with
environmental impact, such as herbicides and pesticides, is
essential.[7–9] However, these types of anions have a variety
of special features, such as charge, size, pH dependence, sol-
vation and geometry, that should be taken into account in
the molecular design of receptors capable of their effective
uptake.[10–12] With this main goal in mind, we have been in-
volved in the development of artificial receptors containing
phenanthroline (phen) units linked through polyaza alkyl
spacers (Scheme 1). The protonated forms of these receptors
[a] C. Cruz, V. Calisto, Prof. V. Fꢀlix
Departamento Quꢁmica
CICECO and Secżo Autꢂnoma de CiÞncias da Safflde
Universidade de Aveiro, 3810-193 Aveiro (Portugal)
Fax : (+351)234-370084
E-mail: vitor.felix@.ua.pt
[b] Prof. R. Delgado
Instituto de Tecnologia Quꢁmica e Biolꢂgica, UNL
Apartado 127, 2781-901 Oeiras (Portugal)
E-mail: delgado@itqb.unl.pt
[c] Prof. R. Delgado
À
offer an apparently rigid structure and several potential N
Instituto Superior Tꢀcnico
Av. Rovisco Pais, 1049-001 Lisboa (Portugal)
Fax : (+351)21-4411277
H binding sites suitable for the uptake of carboxylate anions
with different net charges and geometric binding require-
ments (see Scheme 2). Therefore the binding selectivity is
assisted by a combination of multiple and cooperative inter-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800993.
Chem. Eur. J. 2009, 15, 3277 – 3289
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3277

Scheme 2. Carboxylate and herbicide anions studied in this work.
Results and Discussion
Synthesis of receptors: The macrocycles [32]phen2N4,
[30]phen2N6 and Me2[34]phen2N6 were prepared by [2+2]
cycloaddition reactions of 1,10-phenanthroline-2,9-dicarbal-
dehyde[17] with 1,6-hexanediamine, diethylenetriamine and
N,N-bis(3-aminopropyl)methylamine, respectively, followed
by reduction with sodium borohydride, as described for re-
lated compounds.[18] The macrocycles were isolated as chlo-
ride salts in yields of 80, 60 and 50%, respectively.
A few other related bis-phen macrocycles have been pre-
pared from the same phen-dialdehyde derivative and differ-
ent diamines[18] or by the reaction of 2,9-bis(chloromethyl)-
1,10-phenanthroline and sulfonamide.[19] In the first case
only Schiff bases were prepared, whereas in the second case,
the small macrocycle obtained contains only two tertiary
amines.
Scheme 1. Bis-phenanthroline macrocyclic receptors.
molecular interactions, such as hydrogen bonds and electro-
static and p–p stacking interactions.[13–15]
In a previous study we found that in the solid state the re-
lated dioxatetraaza {H5[30]phen2N4O2}5+ receptor (see
Scheme 1) accommodates chloride and bromide anions in a
“horseshoe” shape conformation with the two phen rings
adopting an almost parallel arrangement.[16] Furthermore,
the results of solution binding studies with some aliphatic
and aromatic anionic substrates suggested that this confor-
mation is adopted in the binding of aromatic anions with ex-
tended p systems (pyrcÀ) or high negative charge (btc3À)
leading to the formation of supermolecules with high associ-
ation constants. By contrast, the affinity of this receptor for
aliphatic anions is much lower, which suggests that other
types of interactions and/or other conformational binding ar-
rangements are adopted.[16] To obtain further insights into
the dynamic binding behaviour, we present herein a com-
prehensive binding study of new receptors of this series, pro-
tonated [32]phen2N4, [30]phen2N6 and Me2[34]phen2N6, with
several aromatic anions (see Scheme 2), including the
common herbicides ATCPÀ and 2,4-DÀ. For comparison pur-
poses, an aliphatic herbicide PMG2À was also considered.
The anion-binding abilities of these receptors were evaluat-
Acid–base properties of the receptors: The acid–base behav-
iour of the three macrocycles was studied by potentiometric
methods in water for [30]phen2N6 and Me2[34]phen2N6 and
in H2O/MeOH (50:50, v/v) for [32]phen2N4 at 298.2 K and
an ionic strength of 0.10m in KCl. The protonation constants
were determined by using the HYPERQUAD program.[20]
The overall protonation constants (logbH A ) are given in
h
a
Table S1 in the Supporting Information and the stepwise
constants are listed in Table 1. Seven protonation constants
were determined for [30]phen2N6 and Me2[34]phen2N6
whereas only five were found for [32]phen2N4 in the 2–11
pH range.
1
ed by potentiometry and H NMR spectroscopy and the as-
sociation constants determined. The binding arrangements
between the {H4[30]phen2N6}4+ and {H6Me2[34]phen2N6}6+
receptors and selected anionic substrates (pyrcÀ, anthcÀ,
naphcÀ, iph2À and btc3À) in water solution were established
by molecular dynamics simulations (MD). The entropic and
enthalpic contributions to the binding free energies were es-
timated and compared with experimental values.
The first four protonation constants for the three com-
pounds are relatively high, with values ranging from 9.81 to
6.03. These values correspond to the successive protonation
of the four secondary nitrogen atoms of the aliphatic amine
linkages contiguous with the phen heads. The following two
constants for the [30]phen2N6 and Me2[34]phen2N6 macrocy-
cles were assigned to the protonation of the remaining two
3278
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3277 – 3289

Protonated Polyazamacrocycles
FULL PAPER
Table 1. Protonation constants (logKiH
)
for [32]phen2N4, [30]phen2N6
[a]
occur in the same pD region. The fact that the phen protons
start to shift at pDꢀ3, when the protonations mainly occur
at the central amines, is also an indication that the macrocy-
cle adopts a folded conformation.
Similar protonation schemes were also found for the relat-
ed macrocycles R2[30]phen2N6 (with N-pendant arms on the
middle nitrogen atoms, R=2-(aminoethyl)-2-naphthalen-1-
ylmehyl)[24] and [30]phen2N4O2.[16] The protonation sequence
and the calculated constants are entirely consistent with the
potentiometric data.
and Me2[34]phen2N6 at T=298.2 K and I=0.10m in KCl.
[b]
[c]
[c]
Equilibrium
quotient
[32]phen2N4
[30]phen2N6
Me2[34]phen2N6
[HL]/[L][H]
ACHTUNGTRENNUNG
9.76(1)
8.91(1)
9.70(1) 9.36(3)[d] 9.81(1)
8.86(1) 8.88(4)[d] 9.26(1)
8.16(1) 8.06(6)[d] 8.21(1)
6.03(1) 5.75(7)[d] 6.99(2)
3.27(1) 3.30(5)[d] 6.29(1)
2.56(2) 2.59(4)[d] 5.70(1)
1.95(2) 2.16(4)[d] 1.84(2)
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
G
–
–
E
[a] The values in parentheses are standard deviations on the last signifi-
cant figure. [b] Determined in H2O/MeOH (50:50 v/v). [c] Determined in
aqueous solution. [d] Determined by H NMR titration in D2O (see refer-
Binding studies of the receptors with anionic substrates
1
ences [21,22]).
Potentiometric measurements: The binding constants of the
protonated forms of the macrocycles {Hi[32]phen2N4}i+ (i=
2–5), {Hj[30]phen2N6}j+ and {HjMe2[34]phen2N6}j+ (j=2–7)
as receptors and the anionic substrates represented in
Scheme 2 were determined at 298.2 K and an ionic strength
of 0.10m in KCl by using the HYPERQUAD program.[20]
For the first receptor H2O/MeOH (50:50 v/v) was used as
the solvent for reasons of solubility, whereas the other two
were studied in water. The binding constants determined
and the corresponding equilibria are collected in Table S3 in
the Supporting Information. The values of the protonation
constants reported in the literature for the studied anions
exhibit large discrepancies,[25] so that they were determined
under the experimental conditions reported in this work.
The values for most of the anions in water were presented
earlier[16] and the new ones are compiled in Table S2 in the
Supporting Information.
The receptors are able to interact with the anionic sub-
strates to form several species with different protonation
states, all of them with a 1:1 receptor/substrate stoichiome-
try, which was confirmed by a Job plot[26] (see Figure S3 in
the Supporting Information). This indicates that the recep-
tors and the anions take part in several overlapping protona-
tion equilibria. The stepwise equilibria that effectively
occur, for all those that can be established for each case, are
unambiguously indicated by the plot of the effective binding
constants (Keff) as a function of pH, as described previous-
ly.[16,27–29] Keff is defined as the quotient between the total
amount of supramolecular species formed and the total
amounts of the free receptor and free substrate at a given
aliphatic nitrogen centres located in the middle of the tria-
mine linkages. These constants for [30]phen2N6 are much
lower than those found for Me2[34]phen2N6 due to the short-
er length of the chains between the amine centres in
[30]phen2N6 and the consequent stronger electrostatic repul-
sion between the ammonium ions formed. The last con-
stants, the fifth for [32]phen2N4 and seventh for [30]phen2N6
and Me2[34]phen2N6, have been assigned to the protonation
of a nitrogen donor from a phen unit as observed for the re-
lated receptor [30]phen2N4O2.[16]
The acid–base behaviour of the three macrocycles is illus-
trated by their speciation diagrams[23] shown in Figure S1 in
the Supporting Information. It is possible to see from these
diagrams that the tetraprotonated form, H4L4+, is the main
species in solution at pHꢀ4 for [32]phen2N4 and
[30]phen2N6 whereas, at the same pH, compound
Me2[34]phen2N6 is completely in its hexaprotonated form,
H6L6+. They also highlight the different behaviour of the
three compounds at a pH below 6, which derive mainly
from the low logK5 and logK6 values of [30]phen2N6 and
result in
a narrow pH range (4.0 to 5.2) in which
{H4[30]phen2N6}4+ predominates.
The protonation sequence of [30]phen2N6 was confirmed
1
by the H NMR titration shown in Figure S2. The protona-
tion constants were also determined with the HypNMR pro-
gram[21] and they are compared in Table 1 with those deter-
mined by potentiometric methods. All the proton resonan-
ces (see Scheme 1 for the numbering) were assigned on the
basis of NOESY experiments. Cross-peaks were observed
between the doublet Hc resonance and the singlet Hd and
triplet He resonances, which allowed the doublet at a lower
field to be assigned to the Hb protons. It is apparent from
the titration experiment that the first four successive proto-
nations occur at the secondary amine centres adjacent to the
phen rings, as the Hd, Ha and Hb resonances shift downfield
as well as, but to a lesser extent, the Hc and Hf resonances.
The next three protonations take place almost simultaneous-
ly at the central aliphatic nitrogen centres and at one of the
nitrogen atoms of a phen unit, as revealed by the Hf and He
resonance shifts as well as by the downfield shift of the
three aromatic resonances. In fact, the protonation constants
for the last three protonations are of the same order and
pH: Keff =ꢀACHTNUGTRNE[NNUG HnLA]/ꢀACHTUNTREGNN[GUN HaA]ꢀCAHTNUGTRNE[GUNN HiL] (in our case, n=3–8, a=
0–3 and i=0–7). The plots for the studied systems are
shown in Figure 1 and the effective equilibria for each
system and the corresponding stepwise thermodynamic con-
stants are compiled in Table 2. In fact several equilibria can
be conceived taking into account the different protonated
species available for each receptor and each anion leading
to the same associated species as obtained with the model
of the overall constants presented in Table S3. Therefore, it
is essential to ascertain a criterion to choose the stepwise
equilibria that have chemical significance, which is given by
the Keff, as described in ref. [16].
These results show that {Hi[32]phen2N4}i+ is the weakest
receptor for the anionic substrates studied and
Chem. Eur. J. 2009, 15, 3277 – 3289
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3279

V. Fꢀlix et al.
Note that, in spite of the larger overall basicity of
{HjMe2[34]phen2N6}j+, which can be protonated at all of the
six amine centres at pHꢀ4.5, this receptor does not exhibit
the largest association constant with any of the studied
anions. In contrast, {Hj[30]phen2N6}j+, which at the same pH
has only four ammonium centres, has the best affinity with
almost all of the anions. This is a clear indication that elec-
trostatic interactions, although important for the overall
binding interaction of the anions, are not the most important
contribution to the recognition of these partners. In addi-
tion, all three receptors exhibit larger affinities for three of
the studied anionic substrates, pyrcÀ, btc3À and anthcÀ. The
{Hj[30]phen2N6}j+ receptor also exhibits large affinities for
tph2À and ATCPÀ at pH<3.5 and for iph2À and dihyac2À at
pHꢀ5.
The differences in binding interaction of the three recep-
tors for the same anion are illustrated in Figure 2 for three
anionic substrates, pyrcÀ, btc3À and ATCPÀ, versus pH. The
binding affinities for pyrcÀ with each of the three receptors
are higher in the low pH range (up to 4), decreasing at
higher pH values. In contrast, btc3À becomes the preferred
substrate
for
the
{Hj[30]phen2N6}j+
and
{HjMe2[34]phen2N6}j+ receptors at pH>4.5. On the other
hand, the {HjMe2[34]phen2N6}j+ receptor exhibits similar ef-
fective binding constants for btc3À, pyrcÀ and anthcÀ at pH
values between 4.5 and 6.5, but above pH 6.5 the affinity of
this receptor for the two extended aromatic anions is defi-
nitely more favourable (see Figure 1). These results suggest
p–p stacking interactions between the phen fragments of
the receptors and the aromatic moiety of the anions are im-
portant interactions in the whole recognition process.
On the other hand, the {HjMe2[34]phen2N6}j+ receptor
presents not only lower binding constants with the studied
anions when compared with {Hj[30]phen2N6}j+, but also it
poorly discriminates several pairs of anions (see Figure 1c),
such as pyrcÀ/btc3À, btc3À/ph2À, pyrcÀ/anthcÀ or anthcÀ/
naphcÀ, which can be readily separated by {Hj[30]phen2N6}j+
(at pHꢀ3, 5, 2 and 2, respectively). These conclusions con-
cerning the selectivity of the receptors can be better visual-
ised by competitive binding diagrams,[29] which are presented
as selectivity plots of the overall percentages of the associat-
ed species as a function of pH for systems containing equi-
molecular amounts of one of the receptors and several
anions.
Figure 1. Plots of logKeff values versus pH for the systems
a) {Hi[32]phen2N4A}i+
,
b) {Hi[30]phen2N6A}i+
(i=3–7)
and
c) {HjMe2[34]phen2N6A}j+ (j=3–8), with A the anionic substrates pre-
sented in Scheme 1 (CL =CA =2”10À3 m). The figure is presented in
colour in the Supporting Information.
The cases of {Hj[30]phen2N6}j+ in solutions containing
four anions, namely, a) btc3À, pyrcÀ, ATCPÀ and PMG2À and
b) pyrcÀ, anthcÀ, naphcÀ and bzcÀ, are shown in Figure 3.
The behaviour of the other two receptors in solutions con-
taining the same mixtures of anions is shown in Figure S4 in
the Supporting Information.
{Hj[30]phen2N6}j+ is the best. This clearly shows the impor-
tance of the presence of the triammonium binding sites due
to the higher electrostatic charge complemented by the suit-
À
able positions of the N H groups for the establishment of
For these two mixtures of anionic substrates,
{Hj[30]phen2N6}j+ is the most appropriate for the selective
uptake of btc3À and pyrcÀ from the mixture (a) and for the
selective uptake of pyrcÀ from the mixture (b). Indeed, the
plot for this receptor for the mixture (a) displays an isose-
lectivity point at pH 4.5 at which the amounts of btc3À and
pyrcÀ bound to {Hj[30]phen2N6}j+ are equal. The uptake of
j+
À
N H···O hydrogen bonds. The {Hj[30]phen2N6} receptor
presents remarkably high binding affinities (logKeff >7 at
pH 3 with pyrcÀ and 5.5 with btc3À at pHꢀ4) and good se-
lectivity between pyrcÀ and all the other anions studied,
even btc3À or anthcÀ. This receptor also discriminates rather
well the related substrates anthcÀ and naphcÀ (see below).
3280
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3277 – 3289

Protonated Polyazamacrocycles
FULL PAPER
Table 2. Stepwise binding constants (logKH L A ) for the indicated equilibria.[a]
h
1
a
Reaction[b]
bzcÀ
naphcÀ
anthcÀ
pyrcÀ
ph2À
iph2À
tph2À
btc3À
ATCPÀ
2,4-DÀ
PMG2À
dihyac2À
[c]
[32]phen2N4
H4L+H3AQH7LA
H4L+H2AQH6LA
H4L+HAQH5LA
H4L+AQH4LA
H3L+HAQH4LA
H3L+AQH3LA
–
–
–
–
–
–
–
–
–
–
–
3.25(2
3.36(1)
3.75(1)
3.65(1)
–
–
–
–
–
–
–
–
2.81(2)
3.04(1)
3.20(1)
–
2.39(2)
2.46(2)
2.74(1)
–
2.67(2)
2.52(1)
2.80(1)
–
2.36(3)
–
–
2.59(3)
–
3.09(2)
3.05(2)
–
1.87(4)
1.63(5)
–
2.15(6)
2.21(4)
–
3.26(4)
3.13(3)
–
3.82(3)
3.42(2)
–
2.67(4)
2.46(4)
–
2.38(1)
–
2.11(3)
–
1.56(6)
2.06(5)
2.83(4)
2.99(3)
2.73(2)
2.14(3)
2.50(2)
2.91(2)
2.27(6)
3.08(1)
[d]
[30]phen2N6
H6L+HAQH7LA
H5L+H2AQH7LA
H5L+HAQH6LA
H4L+H2AQH6LA
H4L+HAQH5LA
H5L+AQH5LA
H4L+AQH4LA
H3L+H2AQH5LA
H3L+HAQH4LA
H3L+AQH3LA
–
–
–
–
4.62(4)
–
4.42(9)
–
–
–
4.13(3)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
3.19(6)
–
3.90(3)
4.00(3)
3.16(3)
–
2.79(4)
–
2.56(4)
–
–
3.88(2)
–
–
–
3.15(3)
–
–
7.27(5)
–
6.23(3)
6.96(3)
3.79(2)
–
3.09(2)
–
2.81(3)
–
–
5.02(5)
–
3.81(4)
–
3.63(2)
–
–
5.45(1)
5.82(1)
–
5.75(1)
–
2.18(5)
2.81(4)
–
3.88(1)
–
4.06(1)
–
–
–
–
4.58(1)
3.68(1)
–
3.13(2)
3.04(1)
–
–
–
–
4.52(4)
–
3.74(2)
–
–
–
3.40(4)
2.95(3)
–
–
–
–
–
–
2.29(4)
1.93(7)
3.24(3)
2.17(4)
Me2[34]phen2N6
2.71(2)
2.45(5)
4.21(1)
2.47(2)
[d]
H6L+H2AQH8LA
H6L+HAQH7LA
H6L+AQH6LA
H5L+HAQH6LA
H5L+AQH5LA
H4L+HAQH5LA
H4L+AQH4LA
H3L+AQH3LA
–
–
–
–
–
–
–
–
–
–
–
–
–
–
2.14(1)
2.68(1)
–
2.07(1)
–
1.98(1)
–
–
–
–
2.93(2)
2.32(2)
–
2.39(1)
–
3.96(1)
4.73(1)
–
4.68(1)
–
5.55(1)
4.53(1)
–
4.47(1)
–
2.55(2)
3.67(1)
–
3.38(1)
–
2.58(5)
3.31(1)
–
3.16(2)
–
4.27(1)
4.81(1)
–
4.72(1)
–
2.14(3)
–
2.05(4)
–
–
–
3.51(1)
–
3.11(1)
–
2.44(2)
–
2.83(1)
–
2.84(1)
–
2.38(2)
–
2.51(2)
–
2.03(5)
–
2.28(3)
–
3.26(1)
–
3.06(1)
–
2.82(1)
–
2.20(2)
–
4.35(1)
3.64(1)
4.18(1)
3.58(2)
2.82(2)
2.28(2)
2.74(2)
2.43(3)
3.83(2)
2.66(4)
[a] L=[32]phen2N4, [30]phen2N6 and Me2[34]phen2N6; A=anion. I=0.10m in KCl at 298.2 K. [b] Charges omitted for clarity. [c] Determined in H2O/
MeOH (50:50 v/v). [d] Determined in H2O.
btc3À is around 80% at pH values >4.5, whereas at pH 3.0
the uptake of pyrcÀ by this receptor is around 90% from the
anionic mixture. On the other hand the uptake of pyrcÀ by
{Hj[30]phen2N6}j+ in the mixture (b) is almost 100% at
pH 2. Through similar diagrams it is possible to see that
{Hj[30]phen2N6}j+ is the receptor that better discriminates
between tph2À and ph2À (almost 80% at pH 3.5) or between
anthcÀ and naphcÀ (about 80% at pH 2).
The pD was kept at 4 for {H4[32]phen2N4}4+ and ꢀ5 for the
other two during the titrations, within Æ0.25 pD units, at
which the three receptors are in the protonated forms
{H4[32]phen2N4}4+
,
{H4[30]phen2N6}4+
and
{H6Me2[34]phen2N6}6+, respectively, and most of the anions
are deprotonated.
The titrations of the three receptors with the btc3À and
ATCPÀ substrates for the Ha and Hd resonances are shown
in Figure 5. All the curve profiles are consistent with a 1:1
receptor/anion stoichiometry and this was confirmed by the
NMR measurements: The binding affinity process between
{H4[32]phen2N4}4+
,
{H4[30]phen2N6}4+
and
Job
plots.[26]
The
corresponding
plots
for
{H6Me2[34]phen2N6}6+ with the bzcÀ, naphcÀ, ph2À, iph2À,
tph2À, btc3À, ATCPÀ, 2,4-DÀ and PMG2À anionic substrates
was also evaluated by 1H NMR titrations. Separated
1H NMR signals were found for the receptor and the sub-
strate protons upon formation of the receptor–substrate
entity, which indicates a fast exchange between the free and
associated species on the NMR timescale.
{H6Me2[34]phen2N6}6+/btc3À and {H4[30]phen2N6}4+/ph2À are
included in Figure S3 in the Supporting Information.
The binding constants determined from the H NMR ti-
1
trations by using the HypNMR program[21] are compiled in
Table 3. The values found are in good agreement with the
corresponding constants obtained by potentiometric meas-
urements, taking into account the fact that these titrations
were performed in D2O instead of water, the ionic strength
was not controlled and no buffer was used to maintain the
pH.
An illustrative example is shown in Figure 4 for the inter-
action of {H4[30]phen2N6}4+ with btc3À at pD=4.95. The for-
mation of the supramolecular entity is accompanied by a
significant broadening of the signals of the receptor and the
anion. Moreover, the addition of the anion causes significant
upfield shifts of the Ha, Hb and Hd resonances in relation to
the free receptor, only minor changes in the Hc doublet and
the two triplets from He and Hf move downfield. The
1H NMR chemical shifts (Dd=dRÀdobs) that result from the
addition of solutions of the studied anions to the three re-
ceptors are listed in Table S4 in the Supporting Information.
2D NMR experiments: The associations of bzcÀ, naphcÀ,
ph2À, iph2À, tph2À, btc3À, ATCPÀ, 2,4-DÀ and PMG2À with the
{H4[30]phen2N6}4+ receptor were also studied by NOESY
experiments. The supramolecular entities formed exhibit
cross-peaks between the Ha, Hb, Hc and Hd resonances and
all the resonances of tph2À (singlet), bzcÀ (two triplets and
doublet) and iph2À (singlet, doublet and triplet). With the re-
Chem. Eur. J. 2009, 15, 3277 – 3289
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3281

V. Fꢀlix et al.
Figure 3. Overall
percentages
of
the
associated
species
of
{Hj[30]phen2N6}j+ as a function of pH in mixtures containing the follow-
ing anions: a) pyrcÀ, anthcÀ, naphcÀ and bzcÀ and b) btc3À, pyrcÀ, ATCPÀ
and PMG2À. The concentration of the receptor and each anion is 2”
10À3 m. The figure is presented in colour in the Supporting Information.
In contrast, the interaction between {H4[30]phen2N6}4+
and the aliphatic herbicide PMG2À is characterised by strong
intermolecular NOE signals between the two singlets of the
anion and the Hd, He and Hf resonances of the aliphatic part
of the receptor, which suggests that PMG2À interacts with
Figure 2. Plots of logKeff versus pH for {Hi[32]phen2N4}i+ (3),
{Hj[30]phen2N6}j+
(
)
and {HjMe2[34]phen2N6}j+
(
with a) pyrcÀ,
*
*
)
b) btc3À and c) ATCPÀ. The figure is presented in colour in the Support-
ing Information.
À
the macrocycle through N H···O hydrogen bonds. Some of
the NOESY spectra are shown for this receptor in Figure S5
in the Supporting Information. Figures S6 and S7 also show
the NOESY spectra of the associated entities formed by
{H4[32]phen2N4}4+ with PMG2À and naphcÀ, and
{H6Me2[34]phen2N6}6+ with PMG2À and 2,4-DÀ.
maining aromatic anions (naphcÀ, btc3À, ph2À and 2,4-DÀ)
only strong cross-peak signals occur with the aromatic pro-
tons (Ha, Hb and Hc). These results suggest that these anions
and the receptor are involved in p–p stacking interactions,
which is consistent with a binding arrangement in which the
anion is inserted between the two phen moieties, as shown
below.
X-ray single-crystal structure of {H6[30]phen2N6}6+ with the
phthalate anion: The single-crystal X-ray diffraction struc-
ture of the supramolecular aggregate formed between
{H6[30]phen2N6}6+ and the phthalate anion shows that its
3282
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3277 – 3289

Protonated Polyazamacrocycles
FULL PAPER
Table 3. Binding
constants
(logK)[a]
for
{H4[32]phen2N4}4+
,
{H4[30]phen2N6}4+ and {H6Me2[34]phen2N6}6+ with the studied anions de-
termined in D2O at 300 K.
Anions {H4[32]phen2N4}4+ {H4[30]phen2N6}4+
{H6Me2[34]phen2N6}6+
ACHTUNGTRENNUNG
bzcÀ
1.92(2)
3.20(3)
3.54 (4)
3.33(4)
4.12(5)
3.83(3)
1.97(2)
2.60(1)
3.01(2)
3.40(3)
3.26(2)
4.55(5)
2.85(1)
2.70(1)
2.16(3)
naphcÀ 2.14(2)
ph2À
3.14(1)
2.51(2)
2.72(1)
3.84(5)
iph2À
tph2À
btc3À
>5.0
ATCPÀ 2.47(1)
2,4-DÀ 2.50(1)
3.99(5)
3.08(5)
2.65(3)
Figure 4. 1H NMR spectra of A) the free receptor {H4[30]phen2N6}4+ in
D2O and B) upon the addition of 0.2 equiv and C) 1.0 equiv of btc3À solu-
tion at pD=4.95 (CR =2.50”10À3 m, CS =2.65”10À2 m).
PMG2À
–
[a] Values in parentheses are standard deviations in the last significant
figures given directly by the program.[21]
crystal structure is built from an asymmetric unit composed
of one half of the macrocyclic cation, two aromatic anions
and four water molecules. The receptor exhibits a crystallo-
found in the solid state of other dimers composed of benzoic
acid derivatives.[30] Hence, the molecular formula of the
supramolecular compound described is definitively
À
graphic centre of symmetry with all six aliphatic N H bind-
[{H6[30]phen2N6}][phACTHNGUTERNUN(G m-H)ph]2·8H2O, with a receptor/anion
ing groups protonated. Furthermore, the charge balance re-
quires an extra proton, crystallographically independent, lo-
cated between the carboxylate groups of the two phthalate
ratio of 1:4, which is different to the equimolar stoichiome-
try found in the experimental binding studies and suggests
that packing effects play an important role in the crystal
structure. Nevertheless, suitable crystals for X-ray diffrac-
tion studies could only be grown from a solution containing
an excess of phthalic acid. The overall structure of
anions, which leads to the formation of the dimer [phACTHNUTRGNEUNG(m-
H)ph]3À with O···H distances of 1.20 and 1.26 Š and a O···O
short distance of 2.45 Š. The O···H···O angle is 1738. The di-
mensions of these structural distances are similar to those
[{H6[30]phen2N6}][phACTHNUTRGNE(UNG m-H)ph]2 is shown in Figure 6.
Figure 5. 1H NMR titrations of {H4[32]phen2N4}4+, {H4[30]phen2N6}4+ and {H6Me2[34]phen2N6}6+ with a) btc3À and b) ATCPÀ anions in D2O; Dd for Ha
6+
(left) and Hd (right) as a function of the number of equivalents of substrate added. The symbols are: {H4[30]phen2N6}4+
(
~
), {H6Me2[34]phen2N6}
^
( )
and {H4[32]phen2N4}4+
(
). The figure is presented in colour in the Supporting Information.
*
Chem. Eur. J. 2009, 15, 3277 – 3289
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3283

V. Fꢀlix et al.
À
O···H···O=C bridge, the longer C O distances involve two
hydrogen-bonded oxygen atoms.
Two different views of the crystal packing diagram of
[{H6[30]phen2N6}][phACHTNUTRGNE(UNG m-H)ph]2·8H2O are presented in
Figure 7. The first one (top) illustrates the p–p stacking in-
teractions between the phen regions of the receptor and one
aromatic ring of the [phACHTNUTGRNEUNG
(m-H)ph]3À dimer units established
along the [100] crystallographic direction. The correspond-
ing interplanar distances between the centroid of the ph ring
and the two phen mean planes are 3.48 and 3.73 Š. The
crystal structure can be described as intercalated layers of
ACTHNUTRGNEUNG
{H6[30]phen2N6}6+ and [ph(m-H)ph]3À connected by water
À
À
molecules through extensive N H···O and O H···O hydro-
gen-bonding interactions (see Figure 7, bottom view) leading
to the formation of open channels that accommodate several
solvent water molecules.
Figure 6. PLUTON view of [{H6[30]phen2N6}][phACHTNUGTRNEUNG(m-H)ph]2 showing its
overall structure. The figure is presented in colour in the Supporting In-
formation.
The {H6[30]phen2N6}6+ recep-
tor and the two [phACTHNUTRGNEUNG
(m-H)ph]3À
dimers are held together in a
centrosymmetric binding ar-
rangement by multiple and co-
À
operative N H···O hydrogen
bonds with H···O distances
ranging from 1.85 to 2.55 Š.
Furthermore, the two [phACTHNUTRGENUG(N m-
H)ph]3À dimer units are clearly
located outside of the macrocy-
clic cavity with the aromatic
ring of one of the ph2À units of
these two supramolecular enti-
ties and the phen rings of the
receptor adopting roughly par-
allel dispositions at interplanar
distances of 3.48 Š (see below),
which seems to suggest that p–
p stacking interactions may also
contribute to the stabilisation
of the supramolecular associa-
tion described. The bond
lengths and angles found for
the receptor are within the
range of expected values and
the endocyclic torsion angles
are consistent with a conforma-
tion of ladder-type shape. For
both ph2À anions, each carbox-
ylate group displays two non-
À
equivalent C O distances that
vary between 1.230(2) and
1.288(2) Š for the first anion
and 1.230(2) and 1.265(2) Š for
the second one. In agreement
Figure 7. Crystal-packing diagram of [{H6[30]phen2N6}][ph
tercalated between {H6[30]phen2N6}6+ molecules (top) and the formation of the open channels along the [100]
H)ph]3À dimer through a C= crystallographic direction (bottom). The figure is presented in colour in the Supporting Information.
(m-H)ph]3À units in-
ACHUTGTNRNE(NGU m-H)ph]2·8H2O showing the [phCAHTUNGTRENNUGN
with the formation of a [phACTHNUTRGENUG(N m-
3284
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3277 – 3289

Protonated Polyazamacrocycles
FULL PAPER
Molecular modelling studies: Further insights into the bind-
ing molecular recognition between the bis-phen macrocyclic
receptors {H6Me2[34]phen2N6}6+ and {H4[30]phen2N6}4+ and
the anions pyrcÀ, anthcÀ, naphcÀ, btc3À and iph2À were ob-
tained by molecular dynamics (MD) simulations with the
GAFF force field[31] within the AMBER 9 software[32] using
a water solvent explicit model. As reported above, in solu-
tion these two receptors and anions form supramolecular
entities of different protonated states. However, the molecu-
lar modelling studies were carried out with only the hexa-
and
tetraprotonated
forms
of
the
receptors
{H6Me2[34]phen2N6}6+ and {H4[30]phen2N6}4+, respectively,
and the fully deprotonated forms of the anions, which are
the dominant species in the pH range 4.5–5.0. The anions
were selected by taking into account the extension of their
aromatic systems and the number of carboxylate groups to
evaluate the role of the net charge of the anions, as well as
the hydrogen bonding and p–p stacking interactions in the
dynamics of the molecular association process.
The docking studies between the receptor and the anion
in an equimolar ratio were carried out in the gas phase by
MD quenching runs (see the Experimental Section). Two
distinct binding scenarios were found. The lowest-energy
structures
for
the
associated
entity
with
{H6Me2[34]phen2N6}6+ display an open conformation with
the two phen moieties twisted with respect to each other
(arrangement A). The other significant binding arrange-
ment, B, with a higher energy, adopts a folded conformation
with the anion inserted into the macrocyclic cavity between
the two phen units, establishing p–p stacking interactions
with both phen heads. These two binding arrangements are
Figure 8. Two alternative binding models found for the docking of
{(H6Me2[34]phen2N6)ACHTNUGRTNEUNG
(pyrc)}5+: arrangement A (top) and arrangement B
(bottom). The figure is presented in colour in the Supporting Informa-
tion.
illustrated in Figure 8 for {(H6Me2[34]phen2N6)ACHTNUTRGNEUNG
(pyrc)}5+. For
the polyaromatic anions, the energy difference between the
B and A arrangements (EBÀEA) decreases with increasing
extension of the anion aromatic system, being 31.08, 26.20
and 12.43 kcalmolÀ1 for naphcÀ, anthcÀ and pyrcÀ, respec-
tively. For the btc3À and iph2À anions this energy difference
is 16.68 and 8.80 kcalmolÀ1, respectively. As found in our
preceding studies for {(H5[30]phen2N4O2)(X)}4+ (X=pyrcÀ
and naphcÀ) complexes,[16] the folded binding arrangement
becomes progressively more stable with the enlargement of
the aromatic ring, which gives a clear indication that the p–
p stacking interactions play an important role in the molecu-
lar recognition of the aromatic carboxylate anions (repre-
sented in Scheme 2) by the synthetic receptors based on two
phen moieties (outlined in Scheme 1). In line with this ob-
servation, note that for the {H4[30]phen2N6}4+ supramolec-
ular entities, the arrangement B is systematically favoured
with the naphcÀ, anthcÀ and pyrcÀ anions by 8.45, 10.08 and
28.31 kcalmolÀ1, respectively. For anions with a single aro-
matic ring this arrangement is stabilised relative to the ar-
rangement A by 2.89 kcalmolÀ1 for btc3À and by 28.43 kcal
molÀ1 for iph2À.
ing results, for the larger receptor two independent simula-
tions were performed by using as starting models the bind-
ing arrangements A and B, whereas for the 30-membered
macrocycle only the arrangement B was considered. The
binding geometry B, with the anion encapsulated between
the
two
phen
moieties,
is
favoured
for
all
{H6Me2[34]phen2N6}6+ associated entities, reversing the
trend observed in the gas phase. The conversion of A into B
was evaluated by measuring the distance between the pyrcÀ
anion and the two phen rings, as shown in Figure 9.
It is evident that the open binding conformation (grey
line) is retained during the first 0.5 ns of the simulation.
After that period, the binding arrangement A is converted
into B, which subsequently remains stable until the end of
the simulation. On the other hand, the folded binding ar-
rangement B (black line) is stable over the entire course of
the simulation. A similar dynamic binding behaviour was
observed {(H6Me2[34]phen2N6)
for
(anthc)}5+
and
{(H6Me2[34]phen2N CHTUNGTRENNNUG
6)A
Subsequently, the dynamic behaviour of the associations
of {H6Me2[34]phen2N6}6+ and {H4[30]phen2N6}4+ with the
pyrcÀ, anthcÀ, naphcÀ, btc3À and iph2À anions was investigat-
ed in water solution for 8 ns. As a consequence of the dock-
occurs after 2.5 and 4 ns of simulation, respectively. This
suggests that the stability of the open binding arrangement
increases with a reduction in the extension of the aromatic
system of the substrate, which underlines the role of the aro-
Chem. Eur. J. 2009, 15, 3277 – 3289
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3285

V. Fꢀlix et al.
chanics energy and the unfavourable polar solvation free
energy.
As would be expected, the values of the energetic term
TDS indicate that the binding of the anions is entropically
disfavoured leading to the binding free energies listed in
Table 4. The binding free energies calculated for the associa-
tions of pyrcÀ, anthcÀ, btc3À and iph2À anions with
{H4[30]phen2N6}4+ are overestimated by À3.0, À2.1, À2.3
and À3.4 kcalmolÀ1, respectively, when compared with those
obtained from potentiometric data. The unique exception is
the binding association with naphcÀ for which the theoreti-
cal binding free energy is only 0.1 kcalmolÀ1 higher than the
experimental one. The theoretical binding affinity of
{H6Me2[34]phen2N6}6+ with btc3À, naphcÀ and iph2À is under-
estimated by 2.2, 1.6 and 3.3 kcalmolÀ1, respectively, where-
as the energy difference between the theoretical and experi-
mental binding free energies for pyrcÀ and anthcÀ is À2.4
and À0.6 kcalmolÀ1, respectively. Hence, the binding affinity
of both receptors for mono-charged aromatics follows the
order pyrcÀ >anthcÀ >naphcÀ, which indicates that the
strength of the binding interaction decreases with a decrease
in the extension of the substrate aromatic system. These re-
sults show, undoubtedly, the role of p–p stacking interac-
tions in the molecular recognition process of this type of
anion. Furthermore, it is also evident that supramolecular
association with {H4[30]phen2N6}4+ is more suitable for ac-
commodating all anions in a folded arrangement.
Figure 9. Distances between the centroids defined by the carbon atoms of
the phen units for the system {(H6Me2[34]phen2N6)
(pyrc)}5+ in arrange-
ments A (grey line) and B (black line). The figure is presented in colour
in the Supporting Information.
AHCTUNGTRENNUNG
matic system of the anion in the molecular recognition pro-
cess. Analysis of the hydrogen bonds between
{H6Me2[34]phen2N6}6+ and {H4[30]phen2N6}4+ and the five
anions (arrangement B) for 8 ns of simulation using a cut-
À
off of 2.5 Š for the H···O distances and 1208 for the N
H···O angles showed that the anion-binding is assisted by
À
multiple and cooperative N H···O hydrogen-bonding inter-
À
actions. Furthermore, the N H binding groups of the recep-
Finally, when these calculations were repeated for the
open binding arrangement A using the portion of the simu-
lation in which this scenario is retained (see above), positive
unfavourable values of DGbind were obtained. This result
leads us to the conclusion that the experimental data corre-
sponds to the folded binding arrangement B, with the encap-
sulation of the aromatic anions in the macrocyclic cavity.
tor as well as the oxygen donors of the carboxylate groups
involved in the molecular recognition process vary during
the molecular dynamics simulation, yielding low occupancies
À
for all potential N H···O hydrogen-bonding interactions, as
exemplified for the association of {(H6Me2[34]phen2N6)-
ACHTUNGTRENNUNG .
(pyrc)}5+ [33]. However, in all cases, at least one hydrogen
bond is observed throughout the entire simulation period.
Subsequently the binding free energies of the supramolec-
ular associations were estimated in water solution by the
MM-PBSA methodology (molecular mechanics/Poisson–
Boltzmann surface area).[34] Simulations were also undertak-
en for the isolated receptors and anions. Thus, three inde-
pendent trajectories were used in the MM-PBSA calcula-
tions to allow the entropic and internal energetic contribu-
tions associated with the receptor conformational changes
that occur upon binding to be taken into account. The
values of the enthalpic and entropic terms of the binding
free energies for the geometric arrangement B were calcu-
lated with snapshots taken throughout the entire simulation,
as described in the Experimen-
Conclusion
This work clearly shows that of the three studied receptors,
{Hj[30]phen2N6}j+, with a smaller cavity size, is the best for
the selective binding of aromatic carboxylate anionic sub-
strates, including the herbicide ATCPÀ. This receptor is par-
ticularly suitable for the uptake of extended aromatic
anions, such as pyrcÀ and anthcÀ, or of highly charged
anions (btc3À). Indeed, it is capable of sensing and recognis-
ing these substrates from a complex mixture of anions that
includes other aromatic anions.
Table 4. Energetic contributions to the binding free energies (in kcalmolÀ1) for the association between the re-
ceptors {H6Me2[34]phen2N6}6+ and {H4[30]phen2N6}4+ and five selected anions.
tal Section, and they are listed
in Table 4. The individual con-
tributions to the enthalpic term
for each binding association are
given in Tables S5 and S6 in the
Supporting Information. The
enthalpy is basically dictated by
the balance between the favour-
H
{H4[30]phen2N6}4+
TDS s*
Anion
DH
s*
DH
s*
DGbind
s*
pyrcÀ
À27.17 0.19 À17.84 0.01 À9.33
À24.50 0.19 À17.44 0.01 À7.06
0.19 À26.07 0.18 À15.11 0.01 À10.96 0.18
anthcÀ
0.19 À24.91 0.18 À15.25 0.01
0.19 À18.42 0.17 À13.92 0.01
À9.66 0.18
À4.50 0.17
naphcÀ À18.73 0.19 À17.01 0.02 À1.72
btc3À
iph2À
À23.47 0.18 À19.29 0.02 À4.18
0.18 À25.83 0.17 À15.62 0.01 À10.21 0.17
À19.15 0.18 À17.69 0.02 À1.46
0.18 À24.19 0.13 À15.32 0.01
À8.87 0.13
able internal molecular me- [a] s* represents the standard deviation from the mean.
3286
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3277 – 3289

Protonated Polyazamacrocycles
FULL PAPER
Me2[34]phen2N6:
1,10-Phenanthroline-2,9-dicarbaldehyde
(0.65 g,
This comprehensive study also shows that the dipropyl-
enetriammonium spacers, in spite of their higher positive
net charge (+3), are less appropriate than diethylenetriam-
monium ones (+2) for the design of receptors for the selec-
tive binding of aromatic carboxylate anions. Furthermore,
experimental binding studies coupled with MD simulations
have allowed us to establish that the folded conformation,
with the aromatic anion inserted between the phen rings, is
the most probable dynamic structure adopted in solution.
This arrangement is stabilised by multiple and cooperative
2.75 mmol) was dissolved in hot methanol (60 mL) and then cooled to
RT and then added dropwise into a methanol solution (30 mL) containing
N,N-bis(3-aminopropyl)methylamine (0.40 g, 2.75 mmol) over 10 min.
The resulting mixture was stirred for six days and then concentrated to
30 mL under vacuum and sodium borohydride (0.3 g, 8.0 mmol) was
added in small portions when the mixture had been cooled in ice bath.
After that the solution was stirred for 12 h and then heated at reflux for
5 h. The resulting solution was evaporated, NaOH was added until pH>
8 and then the solution was extracted with chloroform (5”30 mL). The
brown oil was dried, precipitated as the hydrochloride salt and recrystal-
lised from hot methanol to give a yellow solid that was dried and stored.
Yield: 50%; m.p. 238–2398C; 1H NMR (400 MHz, D2O, 25 8C): d=2.36
(m, 8H; NCH2CH2CH2N), 2.96 (s, 6H; NCH3), 3.49 (t, 16H;
NCH2CH2CH2N), 3.53 (t, 16H; NCH2CH2CH2N), 4.71 (s, 8H;
phenCH2N), 7.83 (d, 4H; phen), 8.02 (s, 4H; phen), 8.58 ppm (d, 4H;
phen); 13C NMR (400 MHz, D2O, 25 8C, dioxane): d=22.7
(NCH2CH2CH2N), 41.5 (NCH3), 46.3 (NCH2CH2CH2N), 53.6
(NCH2CH2CH2N), 54.6 (phenCH2N), 124.6, 128.7, 130.5, 140.6, 145.7,
152.7 ppm (phen); MS (ESI): m/z: 699.46 [L+H]+; elemental analysis
calcd (%) for C42H78Cl6N10O9: C 46.71, H 7.28, N 12.97; found: C 46.55,
H 7.66, N 13.27.
À
interactions, N H···O hydrogen bonds complemented by p–
p stacking interactions.
Experimental Section
General: Microanalyses were carried out by the ITQB Microanalytical
1
Service. The H and 13C NMR spectra were recorded with a Bruker CXP
Crystals of [{H6[30]phen2N6}][phACTHNUGRTNEG(UN m-H)ph]2·8H2O: Potassium hydrogen
400 spectrometer.
phthalate (2.5 equiv, 40 mL, 7”10À2 m) in D2O was added to a solution of
the receptor {H4[30]phen2N6}4+ (600 mL, 5.5”10À3 m) in D2O and the pD
adjusted to 3.2. The solution was left in the NMR tube at RT and brown
crystals were obtained in 3 days.
1,10-Phenanthroline-2,9-dicarbaldehyde was prepared by the reaction of
2,9-dimethyl-1,10-phenanthroline with selenium dioxide.[17] 2,2’-Oxydie-
thylamine was purchased from Acros and hexane-1,6-diamine, diethyl-
enetriamine, N,N-bis(3-aminopropyl)methylamine and sodium borohy-
dride from Aldrich. All chemicals were of reagent grade and used as sup-
Preparation of the anionic substrates: A 1.0m aqueous potassium hydrox-
ide solution (1.0–3.0 equiv depending on the number of acidic functions)
was added to a stirred aqueous solution of the acid form of the substrates
(10.0 mmol, 20 mL). The solvent was then evaporated and the salts were
recrystallised from acetone and dried under vacuum.
1
plied. The reference used for the H NMR spectra in D2O was the [D4]3-
(trimethylsilyl)propanoic acid sodium salt. For 13C NMR spectroscopy, di-
oxane was used as the internal reference.
[32]phen2N4: 1,10-Phenanthroline-2,9-dicarbaldehyde (0.45 g, 1.93 mmol)
was dissolved in hot methanol (45 mL) and then cooled to RT. This solu-
tion was added dropwise to a methanol solution (25 mL) containing
hexane-1,6-diamine (0.28 g, 2.41 mmol) over 10 min. The resulting mix-
ture was stirred overnight. The white solid formed was filtered off and
dissolved in ethanol (30 mL). Sodium borohydride (0.3 g, 7.93 mmol) was
then added in small portions to the cooled mixture (in an ice bath). Then
the solution was stirred for 12 h at RT. The solvent was removed under
reduced pressure and the resulting residue was treated with water and re-
peatedly extracted with chloroform (6”30 mL). The organic phases were
combined and completely evaporated under vacuum and then dissolved
in ethanol. The pure compound was precipitated from hydrochloric acid
solutions as a green powder easy to filter after 12 h at about 48C. Yield:
Potentiometric measurements
Reagents and solutions: The potentiometric titrations were carried out in
H2O or H2O/MeOH (50:50, v/v) solutions at 298.2Æ0.1 K using KCl
(0.10m) as the supporting electrolyte. The solutions of the anions were
standardised by titration with a standard HCl solution. Carbonate-free
solutions of KOH were freshly prepared in H2O or H2O/MeOH (50:50,
v/v) by dilution of concentrated solutions of titrisol ampoule (Merck),
maintained in a closed bottle and discarded when the percentage of car-
bonate was about 0.5% of the total amount of base, as verified by the
Gran method.[35] The demineralised water used was obtained from a Mil-
lipore/Milli-Q system.
Equipment and working conditions: The equipment used has been de-
scribed previously.[36] The temperature was kept at 298.2Æ0.1 K. Atmos-
pheric CO2 was excluded from the cell during the titration by passing pu-
rified nitrogen across the top of the solution in the reaction cell.
1
80%; m.p. 209–2108C; H NMR (400 MHz, D2O, 25 8C): d=1.46 (m, 8H;
NCH2CH2CH2), 1.81 (m, 8H; NCH2CH2CH2), 3.24 (m, 8H;
NCH2CH2CH2), 4.67 (s, 8H; phenCH2N), 7.78 (d, 4H; phen), 7.92 (s,
4H; phen), 8.49 ppm (d, 4H; phen); 13C NMR (400 MHz, D2O, 25 8C, di-
oxane): d=26.9 (NCH2CH2CH2), 27.1 (NCH2CH2CH2), 49.1
(NCH2CH2CH2), 51.5 (phenCH2N), 124.6, 128.6, 130.4, 140.5, 144.1,
152.8 ppm (phen); MS (ESI): m/z: 641.4 [L+H]+; elemental analysis
calcd (%) for C40H65Cl5N8O6: C 51.59, H 7.04, N 12.23; found: C 51.32, H
7.17, N 12.28.
Measurements: The [H+] of the solutions was determined by the mea-
surement of the electromotive force of the cell, E=E’A+Qlog[H+]+
ACTHNUTRGNEUNG
EjE’A. Q, Ej and Kw =[H+][OHÀ] were obtained as described previous-
ly.[36] The term pH is defined as Àlog[H+]. The value of Kw was found to
be equal to 10À13.80 m2 in aqueous solution and 10À13.91 m2 in H2O/MeOH
(50:50, v/v).[37] The measurements were carried out by using 20.00 mL of
ꢀ2.0”10À3
m receptor solutions and the substrate concentration was
[30]phen2N6: Diethylenetriamine (0.382 g, 3.70 mmol) was dissolved in
methanol (35 mL) and then slowly added dropwise to a cooled methanol
(80 mL) solution of 10-phenanthroline-2,9-dicarbaldehyde (0.87 g,
3.68 mmol). The white solid of the macrocyclic Schiff base formed was
filtered off and dissolved in ethanol (30 mL). Then the same procedure
as described for [32]phen2N4 was used. The rose-coloured pure com-
pound was precipitated as the hydrochloric salt from methanol solution
at a low temperature (48C). Yield: 60%; m.p. 229–2308C; 1H NMR
(400 MHz, D2O, 25 8C): d=3.21 (t, 8H; NCH2CH2N), 3.48 (t, 8H;
NCH2CH2N), 4.52 (s, 8H; phenCH2N), 7.28 (d, 4H; phen), 7.63 (s, 4H;
phen), 7.95 ppm (d, 4H; phen); 13C NMR (400 MHz, D2O, 25 8C, diox-
ane): d=47.9 (NCH2CH2N), 48.1 (NCH2CH2N), 54.0 (phenCH2N), 125.1,
129.7, 131.4, 141.6, 145.9, 153.5 ppm (phen); MS (ESI): m/z: 615.4
[L+H]+; elemental analysis calcd (%) for C36H57Cl5N10O5: C 48.45, H
6.07, N 15.64; found: C 48.74, H 6.48, N 15.79.
varied from 2”10À3 to 5”10À3 m. At least three titrations were performed
in the 2–10 pH range at receptor (R)/anion (A) concentration ratios of
1:1, 1:2 and 1:3.
Calculation of equilibrium constants: The protonation constants of the
three macrocycles and of all the substrates, bH L =[HhLl]/[H]h[L]L, were
h
l
determined from the experimental data by using the HYPERQUAD pro-
gram.[20] All these constants were taken as fixed values in order to obtain
the equilibrium constants of the associated species from the experimental
data corresponding to titrations of different R/A ratios. The different ti-
tration curves of the same system were first considered as a single set
and finally all the data were merged together and simultaneously evaluat-
ed to give the final model of binding constants. In addition, back titra-
tions (from alkaline to acid pH) were also performed to check the rever-
sibility of the reactions. The initial computations were obtained in the
Chem. Eur. J. 2009, 15, 3277 – 3289
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3287

V. Fꢀlix et al.
form of overall stability constants, bH
=[HhLlAa]/[H]h[L]l[A]a. The
were carried out by using the SAINT-NT software package from Bruker
AXS. The structure was solved by direct methods and subsequent differ-
ence Fourier syntheses and refined by full-matrix least-squares on F2 by
using the SHELX-97 suite of programs.[38] Anisotropic thermal parame-
ters were used for all non-hydrogen atoms. Hydrogen atoms bonded to
carbon and nitrogen atoms were placed at calculated positions. The
atomic positions of the hydrogen atoms of the water molecules and the
hLlAa
errors quoted are the standard deviations of the stability constants given
directly by the program for the input data, which include all the experi-
mental points of all the titration curves. The species considered in a par-
ticular model were those that could be justified by the principles of
supramolecular chemistry.
1H NMR spectroscopic titrations
À
À
O ACTHNUTRGENUG(N m-H) O bridging hydrogen atoms were taken from difference Fouri-
Reagents and solutions: The concentrations of [32]phen2N4, [30]phen2N6
and Me2[34]phen2N6 (0.01m) solutions in D2O were determined by titra-
tion with DCl (0.1m in D2O) or CO2-free KOD freshly prepared (0.10m).
À
er maps. The water hydrogen atoms were refined with O H distances re-
strained to 0.83 Š. All hydrogen atoms were refined with Uiso =1.2Ueq of
the parent atom except for the bridging hydrogen, which was refined
with isotropic thermal parameters. The residual electronic density ranging
Solutions
of
{H4[32]phen2N4}4+
,
{H4[30]phen2N6}4+
and
{H6Me2[34]phen2N6}6+ (2.0”10À3 m) were prepared by dissolution of the
macrocycles in 0.5 mL of D2O and then the pD was adjusted to ꢀ5.0.
The anion solutions were prepared at 1.2–2.2”10À2 m by dissolution of
the potassium salts in 0.5 mL of D2O. The potassium salt of each anion
dissolved D2O was added in portions of 0.01 mL to solutions of the pre-
pared receptors in D2O at 298.2 K by using a Hamilton syringe (Microli-
ter 700 series of 25 mL). About 15 additions were necessary for each titra-
tion until no further change in the chemical shift was observed, and each
solution was left 15–20 min to stabilise. No effort was made to maintain a
constant ionic strength, but care was taken to maintain the pD value
within a range of Æ0.25 and to avoid CO2 absorption from the atmos-
phere. The pH*=Àlog[H*] (pH* refers to the value directly measured in
the pH meter of solutions in D2O upon calibration with buffers prepared
in H2O) values were directly measured in the NMR tube with a micro-
electrode Hamilton SpinTrode coupled with an Orion 420A instrument.
The pH meter was calibrated with aqueous buffered solutions at pH 7.20
and 4.00. The final pD was calculated from pD=pH*+(0.40Æ0.02).[22b]
The NOESY experiments were performed by collecting 1024(t2)”512(t1)
data points by using standard Bruker pulse programs. A 5.0 ms pulse
width corresponding to a 908 flip angle and a mixing time of 150 ms was
used for the receptor (pD=4.95) and 500 ms for the supramolecular enti-
ties.
À3
from À0.404 to 0.530 eŠ was within the expected values. Molecular di-
agrams were drawn with PLATON.[39] The crystal data are summarised in
Table S7 in the Supporting Information. CCDC-688493 contains the sup-
plementary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif
Molecular dynamics simulations: Molecular modelling simulations and
free-energy calculations were performed with AMBER 9[32] with the
atomic parameters taken from the GAFF force field.[31] The starting
models for {H6Me2[34]phen2N6}6+ and {H4[30]phen2N6}4+ were generated
from the crystal structure of [30]phen2N4O2.[16] The partial atomic charges
for the receptors and anions (pyrcÀ, naphcÀ, btc3À, iph2À and anthcÀ)
were calculated by using the Gaussian 03 program[40] at the HF/6-31G(d)
level of theory with the RESP methodology.[41] Docking between the re-
ceptor and the anion, with a stoichiometry of 1:1, was investigated by
using molecular dynamics quenching runs as follows: the initial structures
were energy-minimised by molecular mechanics calculations and subse-
quently submitted to a MD run at 2000 K in the gas phase for 1 ns using
a time step of 1 fs. A total of 10000 conformations were generated and
then minimised by MM.
For each supramolecular system two different binding arrangements
were selected and solvated using the TIP3P water model, giving cubic
boxes containing between 2231 and 2832 water molecules. The isolated
anions and receptors were also solvated to give three independent simu-
lations for free-energy calculations. The electrostatic neutrality of the sys-
tems was achieved by the replacement of water molecules by the equiva-
lent number of chloride anions depending on the total charge of the
system. The chloride anions were described with force-field parameters
taken from ref. [42]. The systems were equilibrated with a multistage pro-
tocol composed of two successive MM energy minimisations to eliminate
undesired contacts. An NVT (constant number of particles, volume and
temperature) simulation for 50 ps was run to increase the temperature
from 0 to 300 K and, finally, an NPT (constant number of particles, pres-
sure and temperature) simulation over 150 ps to adjust the systemꢄs den-
sity to the experimental density of liquid water at the considered temper-
ature. After the equilibration process, the dimensions of the boxes were
in the range of 40.9–48.3 Š. Finally, the MD data collection was run for
8 ns using an NPT ensemble and a time step of 2 fs. The bonds involving
hydrogen atoms were constrained by using the SHAKE algorithm, the
long-range electrostatic interactions were described by the Particle Mesh
Edwald method and the non-bonding van der Waals interactions were
subjected to a 12 Š cut-off. Frames were saved every 0.2 ps throughout
the MD simulation.
Determination of the association constants: Changes in the chemical shifts
of all the protons were recorded and the association constants of the vari-
ous species formed in solution were determined from the experimental
data by using HypNMR,[21] which requires as input the concentration of
each component and the observed chemical shift. Initial calculations pro-
vided the overall stability constants bR A (bR A =[RrAa]/[R]r[A]a, with
r
a
r
a
R={H4[32]phen2N4}4+
,
{H4[30]phen2N6}4+ and {H6Me2[34]phen2N6}6+).
Only species with a 1:1 stoichiometry were found in all cases. The errors
quoted are the standard deviations of the overall stability constants given
directly by the program from the input data, which includes the experi-
mental points for most of the resonances of each compound.
Job plots: Stock solutions of the receptors (conc. from 2.00 to 3.00 mm)
and the potassium salts of the substrates (under the same concentrations)
were prepared in D2O. Ten NMR tubes were filled with 500 mL solutions
of the receptor and substrate in the following volume ratios: 50:450,
100:400, 150:350, 200:300, 250:250, 300:200, 350:150, 400:100, 450:50 and
500:0. The changes in the chemical shifts of the resonances for each solu-
tion were measured and then the plot of the product between the incre-
ment in the chemical shift and the receptor concentration as a function
of the molar fraction of the receptor was performed. The maximum of
the curve indicates the stoichiometry of the new entity formed. [C]=
[R]
G
U
dobs is the observed chemical shift, dR is the chemical shift of the free re-
ceptor and dmax is the chemical shift of the associated entity. As
(dmaxÀdR) is a constant, the concentration of the associated entity is pro-
portional to Dd[R]0 [with Dd=(dobsÀdR)]. All these curves exhibit one
maximum at X=0.5, which indicates a 1:1 stoichiometry for the associat-
ed species.[26] The plots of [30]phen2N6 as a function of ph2À, and of
Me2[34]phen2N6 as a function of btc3À are presented in Figure S3 in the
Supporting Information, as typical examples.
The binding free energy between the two receptors and the five aromatic
anions were calculated by post-processing the trajectory files of the indi-
vidual simulations carried out with assembled species and the isolated re-
ceptors and anions by using the MM-PBSA (Molecular Mechanics/Pois-
son–Boltzmann Surface Area) methodology.[34] The enthalpic contribu-
tions were determined by solving the Poisson–Boltzmann equations by
the Delphi approach[43] with a solute interior dielectric constant set to 2.0
and 3.0, which are acceptable values for small organic molecules. Howev-
er, generally, the free binding energies calculated with the smaller value
display a better fitting with the experimental ones. The entropic contribu-
tions were determined through normal mode analysis, as implemented in
AMBER 9.[32] The average binding free energy was calculated by using
Crystallography: The X-ray data for [{H6[30]phen2N6}][phACHTNUGTRNEUNG(m-H)ph]2·8H2O
were collected on a CCD Bruker APEX II instrument at 100(2) K using
graphite monochromatised MoKa radiation (l=0.71073 Š). The crystal
was positioned at 35 mm from the CCD and the spots were measured by
using a counting time of 80 s. Data reduction and empirical absorption
3288
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3277 – 3289

Protonated Polyazamacrocycles
FULL PAPER
4000 binding scenarios extracted with a frequency of 10 from the MD
trajectories. Molecular diagrams were drawn with PyMOL.[44]
[27] M. T. Albelda; M. A. B. E. Garcꢁa-EspaÇa, M. L. Godino-Salido,
V. L. Santiago, M. J. Melo, F. Pina, C. Soriano, J. Chem. Soc. Perkin
Trans. 2 1999, 2545–2549.
[28] J. A. Aguilar, B. Celda, V. Fusi, E. Garcꢁa-EspaÇa, V. L. Santiago,
M. C. Martꢁnez, J. A. Ramꢁrez, C. Soriano, R. Tejero, J. Chem. Soc.
Perkin Trans. 2 2000, 1323–1328.
[29] A. Bianchi, E. Garcia-EspaÇa, J. Chem. Educ. 1999, 76, 1727–1732.
[30] F. H. Allen, Acta Crystallogr. Sect. B 2002, 58, 380–388.
[31] J. Wang, R. M. Wolf, J. W. Caldwell, P. A. Kollman, D. A. Case, J.
Comput. Chem. 2004, 25, 1157–1174.
Acknowledgements
The authors acknowledge the Fundażo para CiÞncia e Tecnologia (FCT)
with co-participation of European Community funds FEDER for finan-
cial support under projects PTDC/QUI/56569/2004 and PTDC/QUI/
68582/2006. C.C. and V.C. thank the FCT for Ph. D. grants (SFRH/BD/
19266/2004 and SFRH/BD/38075/2007, respectively).
[32] D. A. Case, T. A. Darden, T. E. Cheatham III, C. L. Simmerling, J.
Wang, R. E. Duke, R. Luo, K. M. Merz, D. A. Pearlman, M. Crow-
ley, R. C. Walker, W. Zhang, B. Wang, S. Hayik, A. Roitberg, G.
Seabra, K. F. Wong, F. Paesani, X. Wu, S. Brozell, V. Tsui, H.
Gohlke, L. Yang, C. Tan, J. Mongan, V. Hornak, G. Cui, P. Beroza,
D. H. Mathews, C. Schafmeister, W. S. Ross, P. A. Kollman,
AMBER 9, University of California, San Francisco (USA), 2006.
[1] J. W. Steed, D. R. Turner, K. Wallace, Core Concepts in Supramolec-
ular Chemistry and Nanochemistry, Wiley, New York, 2007.
[2] E. Garcꢁa-EspaÇa, P. Dꢁaz, J. M. Llinares, M. P. Clares, A. Bianchi,
Coord. Chem. Rev. 2006, 250, 2952–2986.
[33] For the binding association in {(H6Me2[34]phen2N6)ACTHNUTRGNEUNG
(pyrc)}5+, four
hydrogen bonds between the anion carboxylate group and several
À
[3] A. Bianchi, K. Bowman-James, E. Garcꢁa-EspaÇa, Supramolecular
Chemistry of Anions, Wiley, New York, 1997.
N H groups of the macrocyclic receptor were established during the
simulation (8 ns). Two of the hydrogen bonds had average H···O dis-
À
[4] R. Celis, M. C. Hermosi, L. Cornejo, M. J. Carrizosa, J. Cornejo, Int.
J. Environ. Anal. Chem. 2002, 82, 503–517.
[5] G. Welp, G. W. Brukmmer, Ecotoxicol. Environ. Saf. 1999, 43, 83–
90.
[6] C. M. GonÅalves, J. C. G. E. Silva, M. F. Alpendurada, J. Agric. Food
Chem. 2007, 55, 6227–6235.
[7] H. Schneider, A. K. Yatsimirsky, Chem. Soc. Rev. 2008, 37, 263–277.
[8] P. A. Gale, S. E. Garcia-Garrido, J. Garric, Chem. Soc. Rev. 2008, 37,
151–190.
[9] G. V. Oshovsky, D. N. Reinhoudt, V. Willem, Angew. Chem. 2007,
119, 2418–2445; Angew. Chem. Int. Ed. 2007, 46, 2366–2393.
[10] C. A. Ilioudis, J. W. Steed, J. Supramol. Chem. 2001, 1, 165–187.
[11] C. Anda, M. A. Martꢁnez, A. Llobet, Supramol. Chem. 2005, 17,
257–266.
tances of 1.94 and 1.92 Š, average N H···O angles of 159 and 1628
and occupancies of 48 and 26%, respectively. The remaining two hy-
drogen bonds were considerably weaker with average H···O distan-
À
ces of 2.49 and 2.30 Š, average N H···O angles of 124 and 1588 and
occupancies of 31 and 21%, respectively.
[34] P. A. Kollman, I. Massova, B. Kuhn, S. Huo, L. Chong, M. Lee, T.
Lee, Y. Duan, W. Wang, O. Donni, P. Cieplak, J. Srinivasan, D. A.
Case, T. E. Cheatham III, Acc. Chem. Res. 2000, 33, 889–897.
[35] a) G. Gran, Analyst 1952, 77, 661–671; b) F. J. Rossotti, H. J. Rossot-
ti, J. Chem. Educ. 1965, 42, 375–378.
[36] S. Carvalho, R. Delgado, M. G. B. Drew, V. Calisto, V. Fꢀlix, Tetra-
hedron 2008, 64, 5392–5403.
[37] F. Li, R. Delgado, M. G. B. Drew, V. Fꢀlix, Dalton Trans. 2006,
5396–5403.
[12] K. Wichmann, B. Antonioli, T. Sçhnel, M. Wenzel, K. Gloe, K.
Gloe, J. R. Price, L. F. Lindoy, A. J. Blake, M. Schrçder, Coord.
Chem. Rev. 2006, 250, 2987–3003.
[13] K. Bowman-James, Acc. Chem. Res. 2005, 38, 671–678.
[14] P. A. Gale, R. Quesada, Coord. Chem. Rev. 2006, 250, 3219–3244.
[15] W. J. Belcher, M. Fabre, T. Farhan, J. W. Steed, Org. Biomol. Chem.
2006, 4, 781–786.
[16] C. Cruz, R. Delgado, M. G. B. Drew, V. Fꢀlix, J. Org. Chem. 2007,
72, 4023–4034.
[17] C. Chandler, J. L. W. Deady, J. A. Reiss, J. Heterocycl. Chem. 1981,
18, 599–601.
[18] K. E. Krakowiak, J. S. Bradshaw, W. Jiang, N. K. Dalley, G. Wu,
R. M. Izatt, J. Org. Chem. 1991, 56, 2675–2680.
[19] J. S. J. Keipert, C. B. Knobler, D. J. Cram, Tetrahedron 1987, 43,
4861–4874.
[20] P. Gans, A. Sabatini, A. Vacca, Talanta 1996, 43, 1739–1753.
[21] C. Frassineti, S. Ghelli, P. Gans, A. Sabatini, M. S. Moruzzi, A.
Vacca, Anal. Biochem. 1995, 231, 374–382.
[22] a) R. Delgado, J. J. R. Frafflsto da Silva, M. T. S. Amorim, M. F.
Cabral, S. Chaves, J. Costa, Anal. Chim. Acta 1991, 245, 271–282;
b) H. Galster, pH Measurement, Fundamentals, Methods, Applica-
tions, Instrumentation, VCH, Weinheim, 1991.
[38] G. M. Sheldrick, SHELX-97, University of Gçttingen, Gçttingen,
1997.
[39] L. Spek, PLATON, A Multipurpose Crystallographic Tool, Utrecht
University, Utrecht (Netherlands), 2003.
[40] Gaussian 03, Revision B.04, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomer-
y, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyen-
gar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[23] L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini, A. Vacca,
Coord. Chem. Rev. 1999, 184, 311–318.
[41] C. I. Bayly, P. Cieplak, W. D. Cornell, P. A. Kollman, J. Phys. Chem.
1993, 97, 10269–10280.
[24] J. Pina, J. Seixas de Melo, F. Pina, C. Lodeiro, J. C. Lima, A. J.
Parola, C. Soriano, M. P. Clares, M. T. Albelda; M. A. B. E. Garcꢁa-
EspaÇa, M. L. Godino-Salido, V. L. Santiago, M. J. Melo, F. Pina, C.
Soriano, J. Chem. Soc. Perkin Trans. 2 1999, 2545–2549, R. Aucejo,
E. Garcꢁa-EspaÇa, Inorg. Chem. 2005, 44, 7449–7458.
[25] L. D. Pettit, H. K. J. Powell, IUPAC Stability Constants Database,
Academic Software, Timble, version 5, 2003.
[42] J. R. Blas, M. Mµrquez, J. L. Sessler, F. Javier Luque, M. Orozco, J.
Am. Chem. Soc. 2002, 124, 12796–12805.
[43] W. Rocchia, E. Alexov, B. Honig, J. Phys. Chem. B 2001, 105, 6507–
6514.
[44] W. L. DeLano, The PyMOL Molecular Graphics System DeLano
Scientific, San Carlos, 2002, http://www.pymol.org
[26] a) L. Sebo, F. Diederich, V. Gramlich, Helv. Chim. Acta 2000, 83,
93–113; b) K. Hirose, J. Inclusion Phenom. Macrocyclic Chem.
2001, 39, 193–209.
Received: May 23, 2008
Revised: November 7, 2008
Published online: February 5, 2009
Chem. Eur. J. 2009, 15, 3277 – 3289
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3289