Anion-induced urea deprotonation

Article abstract of DOI:10.1002/chem.200401049

The urea-based receptor 1 (l-(7-nitrobenzo[l,2,5]oxadiazol-4-yl)3-(4- nitrophenyl)urea, L-H), interacts with X ions in MeCN, according to two consecutive steps: 1) formation of a hydrogen-bond complex [L-H...X] ^-; 2) deprotonation of L-H to giv

Full text of DOI:10.1002/chem.200401049

FULL PAPER
Anion-Induced Urea Deprotonation
Massimo Boiocchi,[b] Laura Del Boca,[a] David Esteban-Gꢀmez,[a] Luigi Fabbrizzi,*[a]
Maurizio Licchelli,[a] and Enrico Monzani[a]
Abstract: The urea-based receptor 1
(1-(7-nitrobenzo[1,2,5]oxadiazol-4-yl)-
ments. Step 2) takes place with more
basic anions (fluoride, carboxylates, di-
hydrogenphosphate), while less basic
induce proton transfer. On crystallisa-
ꢀ
tion from a solution containing L H
ꢀ
3-(4-nitrophenyl)urea, L H), interacts
and excess Bu4NF, the tetrabutylammo-
nium salt of the deprotonated urea de-
rivative (Bu4N[L]) was isolated and its
crystal and molecular structure deter-
mined.
ꢀ
ꢀ
with Xꢀ ions in MeCN, according to
anions (Clꢀ, NO2
, NO3 ) do not
two consecutive steps: 1) formation of
ꢀ
ꢀ
a hydrogen-bond complex [L H···X] ;
Keywords: acid–base reactions
anions hydrogen bonds
supramolecular chemistry · urea
·
·
ꢀ
ꢀ
2) deprotonation of L H to give L
·
and [HX2]ꢀ, as shown by spectrophoto-
metric and 1H NMR titration experi-
Introduction
tive atoms, for example, F (fluoride) and O (carboxylates,
inorganic oxoanions), and that display the most pronounced
basic tendencies in the chosen solvent. On the other hand,
in most cases electron-withdrawing substituents have been
There exists a large interest on the development of artificial
neutral receptors for anions.[1–3] Such molecules should be
able to behave as hydrogen-bond donors towards the envis-
aged anion, giving rise to a complex that is stable in so-
lution. One of the most frequently employed fragments is
urea. In fact, urea can donate two hydrogen bonds in a par-
allel fashion, and so it is complementary to Y-shaped oxoan-
ions such as carboxylates. One of the first examples of anion
complexation by urea-based receptors was reported by
Wilcox, who characterised the receptor–analyte interaction
ꢀ
appended to the urea framework in order to polarise the N
H bonds and to increase its hydrogen-bond donor tenden-
cies. Therefore, the strongest interactions are expected to
occur between highly basic anions and highly acidic urea
fragments; this opens the way to the potential occurrence of
an acid–base process, with neat proton release from urea to
the anion. Very interestingly, the hydrogen-bonding interac-
tion has been recently defined as an “incipient” and
“frozen” proton-transfer process.[13] On this basis, we were
interested to verify whether urea, in presence of suitable
anions, may undergo deprotonation in solution.
1
through H NMR experiments and determined the associa-
tion constant of the complex by spectrophotometric titra-
tions.[4] A variety of receptors containing one or more urea
subunits have been designed and tested for anion recogni-
tion and sensing over the past years.[5–12] A major element of
selectivity is given by the energy of the receptor–anion inter-
action: in fact, the strongest hydrogen-bond interactions are
established with anions that contain the most electronega-
In this context, we studied the interaction of the urea de-
rivative
1 (1-(7-nitrobenzo[1,2,5]oxadiazol-4-yl)-3-(4-nitro-
phenyl)urea), with some commonly investigated anions (hal-
ides, carboxylates, phosphate). In addition, two electron-
withdrawing substituents, 4-nitrophenyl and nitrobenzofura-
zan, were appended to the 1- and 3-positions of the urea
[a] L. Del Boca, Dr. D. Esteban-Gꢀmez, Prof. L. Fabbrizzi,
Prof. M. Licchelli, Dr. E. Monzani
Dipartimento di Chimica Generale
Universitꢁ di Pavia, viale Taramelli 12
27100 Pavia (Italy)
Fax : (+39)0382-528-544
E-mail: luigi.fabbrizzi@unipv.it
[b] Dr. M. Boiocchi
Centro Grandi Strumenti, Universitꢁ di Pavia
via Bassi 21, 27100 Pavia (Italy)
Chem. Eur. J. 2005, 11, 3097 – 3104
DOI: 10.1002/chem.200401049
ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3097

subunit, in order to enhance both hydrogen-bond donor ten-
dencies and acidity of the receptor. Moreover, nitrobenzo-
furazan is a classical chromogenic fragment,[14,15] and its
colour and absorption spectrum is expected to change, fol-
lowing the interaction with the analyte. Noticeably, append-
ing a chromophore (typically exhibiting electron-withdraw-
ing properties) to the urea subunit is a normal practice for
generating colorimetric anion sensors, as the occurrence of
the hydrogen-bond interaction may alter the intensity of the
chromophore dipole and modify its absorption spec-
trum.[16,17] The interaction of 1 with anions was investigated
1
through UV-visible spectrophotometric and H NMR spec-
troscopic titrations in MeCN. Such an aprotic medium was
chosen in order to preclude the competition of the solvent
as a hydrogen-bond donor.
Results and Discussion
The interaction with halides: On addition of Fꢀ, the pale
ꢀ
yellow solution of 1 (=L H) in MeCN took an intense red
colour.
Figure 1 shows the spectra recorded in the course of the
titration. In particular, Figure 1 (top) displays the spectra
taken during the addition of the first equivalent of fluoride:
the band at 428 nm decreases and disappears, while a new
band develops at 540 nm, to stop after addition of one
equivalent. A definite isosbestic point at 452 nm is observed.
On addition of the second equivalent of Fꢀ, a new band de-
velops at 382 nm, as shown in Figure 1 (bottom). We suggest
Figure 1. Family of spectra taken in the course of the titration of a so-
lution of 1 in MeCN (4.11ꢃ10ꢀ5 m) with a standard solution of [Bu4N]F,
at 258C. Top: Addition of Fꢀ up to 1.0 equiv; inset: titration profile of
the band at 540 nm, which corresponds to the hydrogen-bond complex
[1·F]ꢀ. Bottom: Addition of Fꢀ from 1.0 to 5.0 equiv.
that a first equilibrium is established [see Eq. (1)], leading
ꢀ
ꢀ
to the formation of the hydrogen-bond complex [L H···F] ,
to which the band at 540 nm corresponds. Then, a second
equilibrium takes place, which is tentatively represented
through Equation (2), in which a further Fꢀ ion abstracts a
proton from [L H···F] , with formation of L and HF2ꢀ. In
particular, the band at 382 nm should correspond to the de-
protonated receptor Lꢀ.
ꢀ
ꢀ
ꢀ
LꢀH þ Xꢀ Ð ½LꢀH ꢁꢁꢁ Xꢂꢀ
½LꢀH ꢁꢁꢁ Xꢂꢀ þ Xꢀ Ð Lꢀ þ HX2
ð1Þ
ð2Þ
ꢀ
Nonlinear least-squares treatment of spectral data gave
the following values for stepwise equilibrium logK1 >6,
logK2 =4.2ꢃ0.2. Most interestingly, on slow evaporation of
a solution of 1 and excess of [Bu4N]F in MeCN, red crystals
of a salt of formula [Bu4N]L, suitable for X-ray diffraction
studies, were obtained. The corresponding ORTEP diagram
is shown in Figure 2. It is observed that a proton has been
ꢀ
abstracted from the NH group close to the more electron-
withdrawing nitrobenzofurazan moiety.
Figure 2. An ORTEP view of the [Bu4N]L salt: thermal ellipsoids are
drawn at the 30% probability level, only the hydrogen atom bonded to
the N(5) atom is shown. Selected bond lengths (ꢄ) and angles (8) involv-
Deprotonation induces significant structural modifications
in the urea derivative 1. In fact, previously reported struc-
tural evidence showed that 1,3-diphenyl-substituted urea
molecules are completely flat, the two phenyl rings and the
ꢀ
ꢀ
ꢀ
ing the ureate group: C(1) O(1) 1.210(3), C(1) N(1) 1.387(3), C(1)
ꢀ
ꢀ
N(5) 1.373(3), N(1) C(2) 1.317(3), N(5) C(8) 1.393(3); N(1)-C(1)-O(1)
126.7(2), N(5)-C(1)-O(1) 122.9(2), N(1)-C(1)-N(5) 110.2(2), C(1)-N(1)-
C(2) 119.9(2), C(1)-N(5)-C(8) 128.7(2).
ꢀ
ꢀ
HN(CO)NH fragment lying on the same plane, in ab-
3098
ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3097 – 3104

Acid–Base Chemistry
FULL PAPER
sence of steric effects between adjacent molecules.[18] On the
other hand, in the Lꢀ ion, the benzofurazan group forms a
dihedral angle of 42.4(1)8 with the plane containing the ni-
by 1H NMR titration experiments. Figure 4 displays the
spectra recorded when a 3.00ꢃ10ꢀ3 m CD3CN solution of 1
was titrated with [Bu4N]F.
ꢀ
trophenyl urea subunit. Moreover, the N(1) C(2) distance
ꢀ
(1.317(3) ꢄ) is significantly shorter than the N(5) C(8) dis-
tance (1.393(3) ꢄ), and also than that observed in other di-
arylureas (varying over the range 1.340–1.377 ꢄ).[18] This
suggests the occurrence of a partial delocalisation of the
negative charge of the nitrogen atom towards the nitroben-
zofurazan group, which is described by the resonance repre-
sentation reported in Scheme 1 (structures a and b).
Scheme 1. A resonance representation of the deprotonated form of 1.
In particular, on deprotonation, the phenyl ring of the fur-
azan subunit partially loses its aromaticity, being so exclud-
ed from the extended p-system responsible for diphenylurea
coplanarity. The nitrobenzofurazan fragment is then free to
rotate, probably in order to minimise steric repulsions with
the carbonyl oxygen atom.
Quite interestingly, each ureate ion establishes weak hy-
drogen-bond interactions with another close ureate ion, to
give the dimer shown in Figure 3, in which each nitrobenzo-
Figure 4. 1H NMR spectra obtained in the course of the titration of a so-
lution of 1 in CD3CN (3.00ꢃ10ꢀ3 m) with Fꢀ.
Figure 3. An ORTEP view of the ureate dimer (dashed lines show the
ꢀ
weak N H···O interactions and atomic labels are reported only for the
ꢀ
ꢀ
atoms involved in these interactions). Features of the N H···O interac-
Two effects are expected to originate on C H protons,
ꢀ
tion: N(5) H(5N) 0.89(2) ꢄ, H(5N)···O(3)’ 2.252(25) ꢄ, N(5)···O(3)’
following the hydrogen-bond formation between the urea
subunit and the anion: 1) the increase of electron density in
the phenyl rings, with a through-bond propagation—this
causes a shielding effect and should promote an upfield
3.111(3) ꢄ, N(5)-H(5N)···O(3)’ 162.7(23)8, symmetry code (’)=ꢀx+1,
ꢀy, ꢀz+2.
ꢀ
furazan moiety faces the same subunit belonging to the
other anion. In particular, hydrogen-bond interactions form
shift; 2) the polarisation of the C H bonds, induced by a
through-space effect, electrostatic in origin—the partial posi-
tive charge created onto the proton causes a de-shielding
effect and promotes a downfield shift. In fact, on addition of
the first equivalent of Fꢀ, Ha and Hb undergo upfield shift,
which reflects the increase of electron density due to a
through-bond effect. This supports the hypothesis that the
ꢀ ꢀ
between each N H fragment and the oxygen atom of the
nitro group of the other ureate ion; such an atom has as-
sumed an especially high partial negative charge due to the
p-delocalisation that follows deprotonation (as emphasized
by the limiting formula b in the resonance representation in
Scheme 1).
Fꢀ ion preferably interacts with the NH group linked to
ꢀ
The X-ray structural characterisation unambiguously dem-
onstrates the fluoride-induced deprotonation of the urea
fragment, and the occurrence of the stepwise equilibria
given in Equations (1) and (2) has been fully corroborated
the nitrobenzofurazan substituent. The Hc proton is, there-
fore, subject to a dominant electrostatic effect that induces
deshielding and causes downfield shift. On the other hand,
urea deprotonation, obtained on further addition of Fꢀ, in-
Chem. Eur. J. 2005, 11, 3097 – 3104
www.chemeurj.org
ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3099

L. Fabbrizzi et al.
duces a moderate upfield of the Ha and Hb protons (in-
creased through-bond charge delocalisation) and has no
effect on the Hc protons. In fact, the polarisation effect is no
longer present, since the anion does not remain in close
proximity to the receptor. It should be noted that the signals
arrangement (see the limiting formula b in Scheme 1). This
alters the energy of p orbitals and ultimately leads to a red-
shift in the optical transition. It should be noted that the
spectrophotometric pattern is complicated by the presence
of the nitrobenzene substituent at the other side of the urea
subunit, whose optical properties may be affected, even if to
a lesser extent, by complexation and deprotonation.
ꢀ
pertinent to the N H groups of 1 are observed neither in
the absence nor in the presence of fluoride. This may be as-
cribed to the occurrence of a fast proton exchange with
water molecules that are present in solution as impurities. A
similar behaviour has been observed in DMSO. Moreover,
attempts were carried out to follow the formation of the
[HF2]ꢀ complex through 19F NMR spectroscopy. In particu-
lar, receptor 1 was added to a Bu4NF solution in CD3CN.
On titrant addition, the rather broad signal at ꢀ150 ppm un-
derwent further broadening and disappearance. This may be
due to the occurrence of a rather fast exchange equilibrium.
Hydrogen bonding has been defined as an incipient
proton-transfer process from the hydrogen-bond donor to
the acceptor.[13] The strong interaction, documented by the
high value of logK1, indicates that the proton transfer has
been “frozen” in a rather advanced state. Most interestingly,
In a recent paper,[20] the stepwise interaction of a cyclic di-
amide with fluoride (anion complexation and deprotona-
tion) was investigated through NMR studies. In particular,
both complexation and deprotonation equilibria took place
in the timescale of several minutes, as demonstrated by the
simultaneous appearance of distinct signals for each species
present at the equilibrium. Sluggishness has to be ascribed
to the occurrence of conformational rearrangements of the
constrained cyclic framework of the receptor. In the present
study, the anion–receptor interaction was too fast, relative
to the timescale of the NMR experiment, to allow the ap-
pearance of distinct signals for species at the equilibrium.
This indicates the kinetically uncomplicated nature of the
process.
the addition of a second equivalent of acceptor induces a
When receptor 1 was titrated with chloride, different spec-
troscopic patterns were obtained. Figure 5 displays the
family of UV-visible spectra recorded in the course of the ti-
tration of a 4.11ꢃ10ꢀ5 m solution of 1 in MeCN. Decrease
of the band at 428 nm and development of the band at
540 nm were observed, as found for the previously investi-
gated anions. However, no band development at 382 nm was
detected, even after the addition of a large excess of chlo-
ride. This suggests that 1) only the equilibrium in Equa-
ꢀ
neat and definitive proton transfer, with formation of HF2
.
ꢀ
This is due to the very high stability of the HF2 ion, which,
among 1:1 hydrogen-bond complexes, shows the highest for-
mation energy in the gas phase (39 kcalmolꢀ1).[19]
At this stage, we will try to explain the spectrophotomet-
ric behaviour observed in the course of the titration of 1
with fluoride. The nitrobenzofurazan chromophore typically
presents two main absorption bands: one band centred at
300–350 nm, assigned to a p-p* transition of the aromatic
group, and another band at higher wavelength, to be ascri-
bed to the charge-transfer transition, which takes place from
the linked donor group to the nitro acceptor group in the
phenyl ring.[14] When the donor group is poorly donating
(e.g., an amide group), the band is located between 400 and
450 nm (in MeCN). These two bands are observed in the
spectrum of 1 in MeCN.
tion (1) takes place and 2) the hydrogen-bond complex that
ꢀ
ꢀ
forms, [L H···X] , is stable with respect to the proton ab-
straction from the urea subunit. The logK value for the asso-
ciation equilibrium, as determined through nonlinear least-
squares fitting of the smooth titration profile, is 4.05ꢃ
0.01.
On addition of the first equivalent of fluoride, the [1···F]ꢀ
complex forms, in which the Fꢀ ion interacts with the N H
ꢀ
fragment close to the benzofurazan ring. Due to the strong
hydrogen-bond interaction, the negative charge on the
donor group is substantially increased, which explains the
remarkable bathochromic shift to 540 nm. On addition of
the second equivalent of Fꢀ, the proton is released from the
ꢀ
N H group. Apparently, this does not induce any further in-
crease of negative charge on the donor group and does not
alter the intensity of the dipole: the CT band remains un-
modified. On the other hand, still on addition of the second
equivalent of Fꢀ, the band at 330 nm (assigned to a p–p*
transition, in the phenyl ring of the benzofurazan moiety)
undergoes a bathochromic shift to 380 nm. The molecular
ꢀ
ꢀ
structure of the ureate anion L has shown that N H depro-
tonation induces drastic structural and electronic changes to
the receptor. In particular, p delocalisation onto the benzo-
furazan moiety provokes a loss of the aromatic character of
the phenyl ring, which moves towards a quinoid electronic
Figure 5. Family of spectra taken in the course of the titration of a so-
lution of
1 a standard solution of
in MeCN (4.11ꢃ10ꢀ5 m) with
[Bz(Et)3N]Cl, at 258C; inset: titration profile of the band at 540 nm,
which corresponds to the hydrogen-bond complex [1·Cl]ꢀ.
3100
ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3097 – 3104

Acid–Base Chemistry
FULL PAPER
The formation of a hydrogen-bond complex of 1:1 stoichi-
ometry has been confirmed by 1H NMR titration experi-
ments; recorded spectra are shown in Figure 6. Even on
excited state, following the formation of a 1:1 hydrogen-
bond complex); development of the band at 382 nm, on
addition of the second equivalent of Xꢀ, pertinent to the
deprotonated receptor.
2) 1H NMR spectra: marked upfield shift of Ha and Hb pro-
tons and downfield shift of Hc protons on addition of the
first equivalent of anion; moderate upfield shift of Ha
and Hb protons on further addition.
Figure 7 shows the absorption spectra taken during the ti-
tration with [Bu4N]CH3COO of a 4.11ꢃ10ꢀ5 m solution of 1.
1
Figure 8 displays the family of H NMR spectra obtained on
titrating with acetate a 5.00ꢃ10ꢀ3 m solution of 1 in CD3CN.
Figure 6. 1H NMR spectra obtained in the course of the titration of a so-
lution of 1 in CD3CN (3.00ꢃ10ꢀ3 m) with Clꢀ. Further addition of Clꢀ
after 1.5 equiv did not cause any further spectral change.
excess addition of [BzEt3N]Cl, only a very moderate upfield
of Hb, which reflects a weak hydrogen-bond interaction and
a small through-bond effect, and a moderate downfield shift
of Hc, which indicates a not too pronounced polarisation,
were observed. The above behaviour has to be ascribed to
the much lower basicity and hydrogen-bond acceptor prop-
erties of Clꢀ with respect to Fꢀ. In particular, Clꢀ in unable
to form a stable HX2ꢀ complex.
The interaction with oxoanions: On titration of a solution of
1 in MeCN with [Bu4N]CH3COO, [Bu4N]C6H5COO and
[Bu4N]H2PO4, a behaviour similar to that of fluoride was
observed:
Figure 7. Family of spectra taken in the course of the titration of a so-
lution of
1 a standard solution of
in MeCN (4.11ꢃ10ꢀ5 m) with
[Bu4N]CH3COO, at 258C. Top: Addition of CH3COOꢀ up to 1.0 equiv;
inset: titration profile of the band at 540 nm, which corresponds to the
hydrogen-bond complex [1·CH3COO]ꢀ. Bottom: Addition of CH3COOꢀ
from 1.0 to 5 equiv.
1) UV-visible spectra: bathochromic shift of the band at
428 nm (due to the stabilisation of the charge transfer
Chem. Eur. J. 2005, 11, 3097 – 3104
www.chemeurj.org
ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3101

L. Fabbrizzi et al.
The constants of equilibria given in Equations (1) and (2)
involving the investigated anions are reported in Table 1.
Table 1. Constants for the equilibria given in Equations (1) and (2). The
standard deviation on the last figure is given in parentheses. LogK values
were calculated through nonlinear least-squares treatment of spectropho-
tometric titration data.
Xꢀ
logK1
logK2
F
>6
>6
>6
4.2 (2)
3.8 (1)
4.0 (1)
3.83 (5)
–
CH3COOꢀ
C6H5COOꢀ
ꢀ
H2PO4
5.40 (2)
Clꢀ
4.05 (1)
3.82 (1)
1.99 (3)
ꢀ
NO2ꢀ
–
–
NO3
ꢀ
ꢀ
On the other hand, on titration of 1 with NO3 and NO2
,
patterns similar to that observed for chloride were obtained.
Figure 9 shows the family of UV-visible spectra obtained on
titration of 1 with [Bu4N]NO2: simultaneous disappearance
of the band at 428 nm and development of a band at 540 nm
were observed, with formation of a clear isosbestic point,
even after addition of a large excess of anion.
Figure 8. 1H NMR spectra obtained in the course of the titration of a so-
lution of 1 in CD3CN (5.00ꢃ10ꢀ3 m) with CH3COOꢀ.
The above evidence indicates that the interaction of re-
ceptor 1 with acetate, benzoate and dihydrogenphosphate
proceeds according to the stepwise equilibria [Eqs. (1) and
ꢀ
(2)], that is, formation of two directional N H···O hydrogen
bonds and deprotonation of the urea subunit, which leads to
the formation of the hydrogen-bond complex involving the
anion and its conjugated acid: CH3COOꢀ···CH3COOH,
ꢀ
C6H5COOꢀ···C6H5COOH and H2PO4 ···H3PO4. The hypoth-
esised geometrical arrangement of the acetic acid/acetate
complex is sketched below.
Figure 9. Family of spectra taken in the course of the titration of a so-
lution of
1 a standard solution of
in MeCN (4.11ꢃ10ꢀ5 m) with
[Bu4N]NO2, at 258C; inset: titration profiles of the bands at 540 and
350 nm, which correspond to the hydrogen-bond complex [1·NO2]ꢀ.
This indicates the occurrence of the simple equilibrium
ꢀ
[Eq. (1)] with formation of the hydrogen-bond complex [L
ꢀ
ꢀ
H···X]ꢀ. Association constants for the NO2 and NO3 hy-
drogen-bond complexes are reported in Table 1.
In fact, it has been stated that the most stable hydrogen-
bond complexes result from the combination of fragments
exhibiting similar pKA values.[13] In particular, the most fa-
vourable case is that represented by complexes of formula
[HX2]ꢀ, interfacing the Xꢀ ion and its conjugate acid HX
(DpKA =0), as those investigated in this work (X=F,
CH3COO, C6H5COO, H2PO4).
The obtained results can be summarised as follows: all
the investigated anions give a hydrogen-bond complex with
receptor 1, whose stability decreases along the series:
ꢀ
CH3COOꢀ ꢄC6H5Oꢀ ꢄFꢀ >H2PO4ꢀ >Clꢀ >NO2ꢀ >NO3
(see logK1 values in Table 1). Such a sequence seems to par-
allel the hydrogen-bond acceptor tendencies of the anion. In
particular, for the three oxoanions, for which the logK1
3102
ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3097 – 3104

Acid–Base Chemistry
FULL PAPER
value could be exactly determined (H2PO4ꢀ, NO2ꢀ, NO3 ),
a linear relationship was found between logK1 and the aver-
age negative charge on each oxygen atom, calculated
through an ab initio method: the higher the negative charge,
the higher the hydrogen-bond acceptor tendencies of the
anion (see Figure 10). The existence of such a relationship
urea subunit. Such a goal can be achieved by appending to
the urea subunit substituents of different electron-withdraw-
ing tendencies.
ꢀ
Conclusion
This work has elucidated the details of the urea–anion inter-
action: 1) on addition of the first anion equivalent, a hydro-
gen-bond 1:1 complex forms, the stability of which is related
to the basicity of the anion (which, for oxoanions, can be ex-
pressed by the partial negative charge lying on oxygen
atoms); 2) on addition of a second anion equivalent, a
ꢀ
proton may be released by an N H group, with formation
of the [HX2]ꢀ hydrogen-bond complex. The occurrence of
the second step depends upon the intrinsic acidity of the
urea subunit. Thus, when designing a urea-based receptor,
one faces a paradoxical situation: the receptorꢅs hydrogen-
bond donor tendencies can be enhanced by linking electron-
withdrawing groups to its framework; however, this also in-
creases its acidity and may lead to a neutralisation process,
which brings the anion out of the control of the receptor.
Therefore, selectivity has to be achieved by incorporating
urea fragments within a concave receptor (e.g., tripod or
cage), whose shape and size match the geometrical features
of the anion. However, the employed urea subunits should
Figure 10. A linear relationship between the logK1 value for the complex-
ation equilibrium: 1+XꢀQ[1·X]ꢀ in MeCN and the average negative
ꢀ
charge on the oxygen atom of the oxoanion Xꢀ (H2PO4ꢀ, NO2ꢀ, NO3
,
gray triangles). Partial charges were calculated through an ab initio
method. LogK1 values for CH3COOꢀ and C6H5COOꢀ (white triangles)
have been obtained from the least-squares straight line. The dashed hori-
zontal line discriminates the effect of excess addition of Xꢀ: anions
above the line induce deprotonation of 1, anions below do not and form
only the hydrogen-bond complex.
ꢀ ꢀ
not be too acidic, in order to avoid N H deprotonation
and escaping of the anion from the receptorꢅs cavity to the
solution (i.e., from the realm of supramolecular chemistry to
the classic domain of Brønsted acid–base equilibria).
indicates the purely electrostatic nature of the receptor–oxo-
anion interaction and leaves out any geometrical effect on
the formation of the hydrogen-bond complex (e.g., matching
of the distance of the N atoms of the urea subunit and the
distance of the O atoms of the oxoanion). On the basis of
the linear relationship, values of logK1 for acetate (7.80,
from a partial charge of 0.799) and for benzoate (7.55, from
a partial charge of 0.790) could be extrapolated from the
least-squares straight line (see open triangles in Figure 10).
Then, on addition of an excess of Xꢀ, for the more basic
anions (fluoride, carboxylates, dihydrogenphosphate), a fur-
ther process takes place, which involves urea deprotonation
and formation of the [HX2]ꢀ complex. It may be surprising
that proton release takes place in the most stable hydrogen-
bond complexes. However, such an endoergonic term is
more than compensated by the especially exoergonic term
associated to the formation [HX2]ꢀ, whose stability is strictly
related to the basicity of Xꢀ. Whether addition of excess Xꢀ
induces deprotonation or not depends both on the intrinsic
acidity of the urea-containing receptor and on the stability
of the [HX2]ꢀ complex. Discrimination between the two be-
haviours is expressed in Figure 10 by the horizontal dashed
line. Anions above the line promote deprotonation and
[HX2]ꢀ formation, anions below do not. In principle, one
can control the behaviour of the receptor, moving the hori-
zontal line up and down at will, by varying the acidity of the
Experimental Section
4-Amino-7-nitro-2,1,3-benzoxadiazole: 4-Chloro-7-nitro-2,1,3-benzoxadia-
zole (0.023 g, 0.1632 mmol) was disolved in methanol (10 mL). After the
addition of ammonia (0.7 mL), the mixture was stirred for 12 h. The reac-
tion mixture was evaporated to dryness under reduced pressure and the
residue was subjected to chromatography on silica gel with AcOEt/
hexane (3:2) to afford (0.111 g, 62%) an air-stable brown powder.
1H NMR (CDCl3): d=8.46 (d, J=8.0 Hz, 1H), 6.41 ppm (d, J=8.0 Hz,
1H); ESI-MS: m/z: 181 [M+H]+.
1-(7-Nitrobenzo[1,2,5]oxadiazol-4-yl)-3-(4-nitrophenyl)urea (1): 4-Nitro-
phenylisocyanate (0.027 g, 0.1638 mmol) was added into an argon-filled
reactor
containing
4-amino-7-nitro-2,1,3-benzoxadiazole
(0.030 g,
0.164 mmol) in dry dioxane (10 mL). The mixture was heated at 608C
under magnetic stirring for 1 h and then left under magnetic stirring for
12 h. During the reaction, a light orange precipitate appeared, which was
collected by filtration, washed with water and dried under vacuum. The
product (0.020 g, 36%) is air-stable in the solid state and soluble in all
common organic solvents. 1H NMR (CD3CN): d=8.78 (br, 2H; NH),
8.65 (d, J=8.0 Hz, 1H; Hb), 8.29 (d, J=8.0 Hz, 2H; Ha), 8.25 (d, J=
8.1 Hz, 2H; Hd), 7.77 ppm (d, J=8.1 Hz, 2H; Hc); IR (nujol mull): n˜ =
ꢀ
1736 (C=O), 1558 (nas(NO2)), 1310, 1282 (ns(NO2)), 3382, 3325 (nas,ns(N
ꢀ
H)), 1446, 1202 (das,ds(N H)).
General procedures and materials: All reagents were purchased by Al-
drich/Fluka and used without further purification. UV/Vis spectra were
recorded on a Varian CARY 100 spectrophotometer, with a quartz cu-
vette (path length: 1 cm). NMR spectra were recorded on a Bruker
Avance 400 spectrometer, operating at 9.37 T. Spectrophotometric titra-
tions were performed at 258C on 10ꢀ4 and 10ꢀ5 m solutions of 1 in MeCN.
Chem. Eur. J. 2005, 11, 3097 – 3104
www.chemeurj.org
ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3103

L. Fabbrizzi et al.
Typically, aliquots of freshly prepared Bu4NX standard solutions were
added and the UV/Vis spectra of the samples were recorded. All spectro-
photometric titration curves were fitted with the HYPERQUAD pro-
gram.[21] 1H NMR titrations were carried out in CD3CN.
[1] P. A. Gale, Coord. Chem. Rev. 2003, 240, 1–226.
[2] P. D. Beer, P. A. Gale, Angew. Chem. 2001, 99, 502–532; Angew.
Chem. Int. Ed. 2001, 40, 486–516.
[3] F. P. Schmidtchen, M. Berger, Chem. Rev. 1997, 97, 1609–1646.
[4] P. J Smith, M. V. Reddington, C. S. Wilcox, Tetrahedron Lett. 1992,
33, 6085–6088.
[5] E. Fan, S. A. van Arman, S. Kincaid, A. D. Hamilton, J. Am. Chem.
Soc. 1993, 115, 369–370.
[6] J. Scheerder, J. F. J. Engbersen, A. Casnati, R. Ungaro, D. N. Rein-
houdt, J. Org. Chem. 1995, 60, 6448–6454.
[7] M. P. Hughes, M. Shang, B. D. Smith, J. Org. Chem. 1996, 61, 4510–
4511.
[8] M. P. Hughes, B. D. Smith, J. Org. Chem. 1997, 62, 4492–4499.
[9] R. J. Fitzmaurice, G. M. Kyne, D. Douheret and J. D. Kilburn, J.
Chem. Soc. Perkin Trans. 1 2002, 841–864.
[10] A. J. Evans, S. E. Matthews, A. R. Cowley, P. D. Beer, Dalton Trans.
2003, 4644–4650.
X-ray crystallographic studies: Single crystals of [Bu4N]L, suitable for X-
ray crystallography, were grown by slow diffusion of diethyl ether into a
solution of 1 in THF in the presence of excess [Bu4N]F. Crystal data for
[Bu4N]L salt: C29H43N7O6, Mr =585.70, T=293 K, crystal dimensions
0.55ꢃ0.10ꢃ0.10 mm3, monoclinic, P21/n (No. 14), a=9.8165(6), b=
16.1456(10), c=20.2528(13) ꢄ, b=102.466(2), V=3134.3(3) ꢄ3, Z=4,
1calcd =1.241 gcmꢀ3, F(000)=1256, m=0.088 mmꢀ1, MoKa X-radiation (l=
0.7107 ꢄ), 2qmax =608, 27066 measured reflections, 9173 independent re-
flections (Rint =0.061), 3765 strong reflections [Io >2s(Io)], 386 refined
parameters, R1=0.0658 (strong data) and 0.1720 (all data), wR2=0.1496
(strong data) and 0.2007 (all data), GOF=0.966, 0.25/ꢀ0.24 max/min re-
sidual electron density. Diffraction data were collected at ambient tem-
perature on
a Bruker-Axs Smart-Apex CCD-based diffractometer.
Omega rotation frames (scan width 0.38, scan time 30 s, sample-to-detec-
tor distance 5 cm) were processed with the SAINT software (Bruker-
Axs) and intensities were corrected for Lorentz and polarisation effects.
Absorption effects were analytically evaluated by the SADABS soft-
ware,[22] and correction was applied to the data (0.96/0.99 min/max trans-
mission factor). The structure was solved by direct methods (SIR-97),[23]
and refined by full-matrix least-square procedures on F2 using all reflec-
tions (SHELXL-97).[24] Calculations were performed with the WinGX
package.[25] All non-hydrogen atoms were refined with anisotropic tem-
perature factors. Hydrogen atoms bonded to carbon atoms were placed
by using the appropriate AFIX instructions, the hydrogen atom bonded
to the N(5) amine was located in the DF map and refined without con-
straint on the atomic coordinate. CCDC 244867 contains the supplemen-
tary crystallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[11] H. Miyaji, S. R. Collinson, I. Prokes, J. H. R. Tucker, Chem.
Commun. 2003, 64–65.
[12] M. Barboiu, G. Vaughan, A. van der Lee, Org. Lett. 2003, 5, 3074–
3076.
[13] T. Steiner, Angew. Chem. 2002, 114, 50–80; Angew. Chem. Int. Ed.
2002, 41, 48–76.
[14] S. Uchiyama, T. Santa, K. Imai, J. Chem. Soc. Perkin Trans. 2 1999,
2525.
[15] B. Ramachandram, A. Samanta, J. Phys. Chem. A 1998, 102, 10579–
10587.
[16] C. Suksai, T. Tuntulani, Chem. Soc. Rev. 2003, 32, 192–202.
[17] R. Martꢆnez-MaÇez, F. Sancenꢀn, Chem. Rev. 2003, 103, 4419–4476.
[18] M. C. Etter, Z. Urbanczyk-Lipkowska, M. Zia-Ebrahimi, T. W. Pan-
unto, J. Am. Chem. Soc. 1990, 112, 8415–8426.
[19] S. Gronert, J. Am. Chem. Soc. 1993, 115, 10258–10266.
[20] A. M. Costero, M. J. BaÇuls, M. J. Aurell, M. D. Ward, S. Argent,
Tetrahedron 2004, 60, 9471–9478.
[21] P. Gans, A. Sabatini, A. Vacca, Talanta 1996, 43, 1739–1753.
[22] G. M. Sheldrick, SADABS, Siemens Area Detector Absorption Cor-
rection Program, University of Gçttingen (Germany), 1996.
[23] A. Altomare, M. C. Burla, M. Cavalli, G. L. Cascarano, C. Giacovaz-
zo, A. Guagliardi, A. G. G. Moliterni, G. Polidori, R. Spagna, J.
Appl. Crystallogr. 1999, 32, 115–119.
Acknowledgements
The financial support of the European Union (RTN Contract HPRN-CT-
2000-00029) and of the Italian Ministry of University and Research
(PRIN—Dispositivi Supramolecolari; FIRB—Project RBNE019H9K) is
gratefully acknowledged.
[24] G. M. Sheldrick, SHELX-97, Programs for Crystal Structure Analy-
sis, University of Gçttingen (Germany), 1997.
[25] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
Received: October 14, 2004
Published online: March 15, 2005
3104
ꢂ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 3097 – 3104