Synthesis, binding and fluorescence studies of a new neutral h-bonding receptor for anions based on 3,5-bis(trifluoromethyl)phenylurea

Article abstract of DOI:10.1080/10610271003731146

The synthesis and anion binding studies of the new neutral receptor 1,1'-(2,2'-(4,10-dimethyl-1,4,7,10-tetraazacyclodo-decane-1,7-diyl)bis(2-oxoethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (L1) are reported. L1 is a macro-cyclic ligand con

Full text of DOI:10.1080/10610271003731146

Supramolecular Chemistry
Vol. 22, No. 6, June 2010, 365–379
Synthesis, binding and fluorescence studies of a new neutral H-bonding receptor for anions based
on 3,5-bis(trifluoromethyl)phenylurea
Mauro Formicaa, Vieri Fusia*, Luca Giorgia, Eleonora Macedia, Giovanni Piersantib,
Maria Antonietta Varreseb and Giovanni Zappiab*
b
aInstitute of Chemical Sciences, University of Urbino, Piazza Rinascimento 6, I-61029 Urbino, Italy; Institute of Pharmaceutical
Chemistry, University of Urbino, Piazza Rinascimento 6, I-61029 Urbino, Italy
(Received 11 January 2010; final version received 18 February 2010)
The synthesis and anion binding studies of the new neutral receptor 1,10-(2,20-(4,10-dimethyl-1,4,7,10-tetraazacyclodo-
decane-1,7-diyl)bis(2-oxoethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (L1) are reported. L1 is a macro-
cyclic ligand containing the 3,5-trifluoromethylphenylureido-binding fragment attached as a side arm on the
tetraazacyclododecane. L1 is soluble in numerous organic solvents; the binding properties of L1 towards several simple
anions (G) were investigated by NMR, UV–vis and fluorescence techniques in DMSO and CH3CN solutions. L1 is able to
bind F2, Cl2 and AcO2 in both solvents; in addition, it binds Br2 in CH3CN. Fluoride shows the highest constant values in
the halide series (F2 . Cl2 . Br2) and AcO2 is the most strongly bound among all the anions investigated. L1 is able to
signal the presence of the anions in solution by fluorescence change; in the case of acetate, this occurs in the visible range.
Keywords: urea; H-bonding; fluorescent sensor; macrocycles; anion recognition
Introduction
the anion. A strategy to increase the versatility of this
molecular topology could be the modification of the
aromatic unit linked to the ureido fragment. In the light of
this, 1,10-(2,20-(4,10-dimethyl-1,4,7,10-tetraazacyclodo-
decane-1,7-diyl)bis(2-oxoethane-2,1-diyl))bis(3-(3,5-bis
(trifluoromethyl)phenyl)urea) (L1, Chart 1) has been
synthesised substituting the p-nitrophenyl (PNP) group in
L2 with the 3,5-bis(trifluoromethyl)phenyl (TFP) unit
widely employed in non-covalent organocatalysis (7). The
choice of two trifluoromethyl moieties in place of the nitro
group was based on the following: (i) it is an inert and
powerful electron-withdrawing group able to increase NH
polarisation of the urea; (ii) it was also proposed to rigidify
the structure by polarising the adjacent H atom, which in
turn facilitates a H-bonding interaction with the O atom of
the ureido fragment, thus blocking unfilled H-bonding
valencies; and (iii) it does not contribute to molecule self-
association (7). In this way, the N-[3,5-bis(trifluoro-
methyl)phenyl] substituent could play the dual role to
preventing aggregation/self-association by blocking
H-bond donor group and providing organisation through
intramolecular H-bonding.
The design and synthesis of fluorogenic artificial receptors
for recognition and sensing of anionic species are of
current interest in the field of supramolecular chemistry
(1). There are many factors for the rapid growth of this
field including biological concerns, sensor development,
environmental remediation, selective separation and
extraction of chemical species and catalysis (1b). Several
approaches to the development of anion receptors
involving different interactions such as metal complexes
(2), positively charged hosts (3) and neutral non-metallic
H-bonding-based systems (4) have been reported. Urea
and thio-urea have been successfully reported as neutral
binding sites for anions and are widely recognised as
highly useful templates upon which to build powerful
recognition systems (5). However, N,N0-disubstituted urea
compounds exhibit limited solubility in non-competitive
organic solvents and often self-association through intra-
and inter-molecular interactions of the NH group with the
carbonyl group. Therefore, considerable effort has been
devoted to the development of new urea-based receptors in
order to improve their selectivity and scope (4).
Herein, we document the efficient synthesis of the new
receptor L1, as well as the binding studies involving a
variety of simple anions in DMSO and CH3CN,
demonstrating the ease of handling and the versatility of
N-3,5-bis(trifluoromethyl)phenyl variations. The adapta-
bility of L1 in revealing anions in different solvents is
Recently, we reported a new host (L2, Chart 1) able to
strongly interact with acetate and simple anions such as
chloride in DMSO solution (6). However, L2 versatility is
hampered by the scarce solubility in most of the common
solvents and low optical response to the interaction with
*Corresponding authors. Email: vieri.fusi@uniurb.it; giovanni.zappia@uniurb.it
ISSN 1061-0278 print/ISSN 1029-0478 online
q 2010 Taylor & Francis
DOI: 10.1080/10610271003731146
http://www.informaworld.com

366
M. Formica et al.
F3C
CF3
O2N
NO2
1
2
CF3
F3C
13
3
4
HN
HN
HN
NH
NH
NH
NH
5
7
O
O
O
O
6
HN
O
O
O
O
8
9
CH3
N
CH3
N
N
N
N
N
N
N
10
11
CH3
CH3
12
L1
L2
Chart 1. Ligands with labels for the NMR experiments.
highlighted by UV–vis and fluorescence in addition to
classical NMR experiments.
The same synthetic scheme was applied for the
preparation of L1 in large scale to have enough material
for the present and further studies; unforeseen, the overall
yield did not exceed 20%, probably due to the reduction of
the yield in the coupling process. This result prompted us
to explore new and simple procedures to increase the
yields for the formation of the amide bond (8).
Results and discussion
Synthesis
During these studies, we became interested in the use
of acyl fluorides as fast-acting acylating agents,
introduced by Carpino et al. (9), which react more like
activated esters than acid halides (Cl, Br and I). Other
advantages of the acyl fluorides include their great
stability in the presence of water, including moisture in
the air, and their relative lack of conversion to the
corresponding oxazolones.
Scheme 1 outlines the synthesis of the new multidentate
ligand L1. In a previous paper, we discussed the
preparation of the ligand L2, which was based on the
coupling of a protected form of glycine with 1,7-dimethyl-
1,4,7,10-tetraazacyclododecane, followed by the removal
of the primary amine protecting group and then by
reaction with 4-nitrophenylisocyanate, with an overall
yield of 32% (6).
F
N
F
NHCbz
O
NHCbz
N
N
N
N
HN
NH
O
(2 mmol)
N
N
F
3
F
N
N
CbzHN
CbzHN
COOH
Py, CH2Cl2
O
1
4
2
CF3
F3C
F3C
CF3
i) H2, 10% Pd/C,
EtOH abs
NH
NH
HN
HN
O
O
ii) F3C
CF3
O
O
N
N
N
N
5
NCO
L1
Scheme 1. Synthesis of L1.

Supramolecular Chemistry
Stimulated by the report by Georg et al. (10) on one- Solution studies
367
1
pot transformations of carboxylic acids to amides
mediated by Deoxo-Fluor reagent, we explored the
direct formation of the bis-amide by the reaction of
N-Cbz-glycine (1) and 1,7-dimethyl-1,4,7,10-tetraazacyclo-
dodecane (3) in the presence of Deoxo-Fluor or diethyl-
aminosulphurtrifluoride reagent. In both cases, negligible
yields (10–15%) were obtained, and, in particular, with the
Deoxo-Fluor reagent, the bis(methoxyethyl)amide was
obtained (75% yield) as the main product by the competitive
reaction of N-Cbz-glycine fluoride with bis(methoxyethyl)
amine. We then turned our attention to the original report by
Carpino for the preparation of acid fluorides using cyanuric
fluoride to give Z-gly-F (2), which was coupled with 1,7-
dimethyl-1,4,7,10-tetraazacyclododecane (3) in an aqueous
biphasic system (CH2Cl2/H2O–NaHCO3) yielding 89%. A
slightly better yield (91%) was obtained when the coupling
reaction was performed in DMF, but the former conditions
were selected for their simplicity.
The anion binding abilities of L1 were evaluated by H,
19F NMR, UV–vis and fluorescence spectroscopic
titrations with anions (G) such as the halide series and
acetate in both DMSO and CH3CN.
L1, unlike L2, is soluble in other solvents in addition to
DMSO, thus widening the field of application. Table 1
reports the association constants of the adducts formed,
obtained by processing the suitable spectroscopic titrations;
in DMSO, only the NMR allows the evaluation of the
association constants, while in CH3CN, they can be obtained
by processing the UV–vis and fluorescence titrations. The
NMR is not suitable since the association constants with
these anions are too high in this medium (11).
NMR experiments
The interaction between L1 and the guest anions (G ¼ F2,
Cl2, Br2, I2 and AcO2) was investigated in DMSO-d6–
0.5% water solution at 298 K. The addition of F2, Cl2 or
AcO2 to L1 gave rise to a significant 1H NMR shift of the
side-arm resonances, mainly of the ureido protons H5 and
H7, while the spectra remain substantially unchanged after
the addition of Br2 and I2 (Figure 1). In addition, the
spectrum obtained by adding a large excess of NBu4OH as
a base clearly indicates the deprotonation of the urea
groups affording the complete disappearance of the ureido
protons as well as the upfield shift of the phenyl protons
H1 and H3. Notably, receptor L2 in the presence of a small
amount of F2 (0.5 equivalent with respect to L2) gave rise
to the disappearance of both ureido proton signals due to
the deprotonation process (12); this does not occur with L1
in which both signals, although broad, are visible in the 1H
NMR spectrum even when more than 2 equivalents of the
anion were added (Figure 1). In addition, the peak
resulting from the formation of the [FHF]2 species,
usually confirming the deprotonation process (12), appears
only in the presence of a very large excess of F2 as a broad
The bis-carbamate (4) was then converted directly into
the final ligand L1 using a one-flash procedure.
Although, in principle, the formation of the bis-urea
system in L1 could be realised by hydrogenation of the
Cbz-protecting group, isolation of the corresponding free
bis-amine and final reaction with 3,5-bis(trifluoromethyl)-
phenyl isocyanate (5), we explored a one-pot procedure to
reach the same goal.
Thus, hydrogenation (H2, 10% Pd/C) at atmospheric
pressure for 4 h, followed by direct addition of isocyanate
(5), produced compound L1 at a satisfying yield of 61%,
as an off-white solid.
Notably, L1 is soluble in most organic solvents such as
CH3CN, CHCl3 and CH2Cl2, but also slightly in methanol,
ethanol, acetone and others, and no aggregation problems
are observed.
Table 1. Logarithms of the association constants of the guests (G ¼ F2, Cl2, Br2 and AcO2) to L1, determined by UV–vis,
fluorescence and NMR titration in DMSO-d6–0.5% water solution or CH3CN at 298 K.
Log K
CH3CN
DMSO
AcO2
Reaction
AcO2
F2
Cl2
Br2
L þ G ¼ LG
4.5(1)a
6.5(1)b
6.7(1)c
5.4(1)b
5.6(1)c
4.6(1)b
4.7(1)c
3.4(1)b
3.4(1)c
LG þ G ¼ LG2
2.1(1)a
3.9(1)b
4.0(1)c
5.1(1)b
5.3(1)c
3.4(1)b
3.4(1)c
2.9(1)b
2.8(1)c
Note: Values in parentheses are the standard deviation to the last significant figure.
a
Values obtained by NMR.
b
Values obtained by UV–vis.
c
Values obtained by fluorescence experiments.

368
M. Formica et al.
3
1
8
OH–
3
1
8
5
7
AcO–
3
8
1
F–
5
7
3
3
8
8
5
1
7
7
Cl–
Br–
1
1
5
5
3
12
8
7
L1
14.0 13.0 12.0 11.0 10.0 9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0 1.0
(ppm)
Figure 1. 1H NMR spectra of L1 in DMSO-d6–0.5% water solution at 298 K by adding 3 equivalents of the guest with respect to L1,
with the exception of OH2 added in large excess; the addition of I2 does not modify the L1 spectrum.
triplet at 16.1 ppm. In other words, the urea fragment is
less acidic in L1 than in L2. Unfortunately, the broad lines
of the ureido resonances in the L1/F2 system did not allow
the obtainment of reliable association constants for this
system.
The 1H NMR experiments carried out in CD3CN
solution showed a similar behaviour, although the spectra
resulted just a bit sharper and shifted; in addition, Br2 also
affects the 1H NMR spectrum in a similar way to the other
anions (Figure 2).
Another aspect to be highlighted by NMR spectra of
L1 is the opposite shift exhibited by H3 with respect to H1
protons in the presence of the interacting anions (Figure 1).
For example, at the end of the titration with AcO2, H3 is
0.28 ppm downfield shifted while H1 is 0.06 ppm upfield
shifted with respect to the free L1; on the contrary, by
adding an excess of OH2, i.e. deprotonating the urea units,
both H1 and H3 exhibit upfield shifts of 0.82 and
0.19 ppm, respectively. The different behaviour of H3 in
the presence of interacting or deprotonating anions is due
to a different process occurring at the TFP–ureido
group. In the adduct with anion, the shift is in agreement
with the statement that the host–guest interaction occurs at
the ureido group; this mainly affects the hydrogen atoms in
the ortho position to the TFP group via a through-space
interaction (13), a similar trend is exhibited by H3 in the
presence of F2 and Cl2, with downfield shifts of 0.14 and
0.06 ppm, respectively. The deprotonation of the ureido
fragment mainly affects the hydrogen atoms H1 in the
para position to the TFP group due to a mesomeric effect
favoured by the deprotonation of the nitrogen atom closer
to the TFP group.
It is interesting to observe that the ligand L1 exhibits a
complex number of signals in both solvents. In particular, in
CD3CN, it displays three resonances for the two methyl
groups (H12) of the macrocyclic base (Figure 2), which,
together with the number of all aliphatic resonances in the
13C NMRspectrum(seeExperimental Sectionand FigureS2
of the Supporting Information, available online), suggests
the existence of two slowly interchanging conformers in
solution, on the NMR time scale. The 1:1 mixture of
conformers is produced by the rotational restriction of
the amide groups of the tetraaza base (14). Looking at the
aromatic H1 and H3 as well as at the ureido H5 and H7 and
methylene H8 protons, they all exhibit twice the number of
signals expected by their equality on the NMR time scale.
Assuming that only two distinct resonances are
observed for each side arm vs. a total of a possible four,
two possible explanations for this pattern of signals can be
hypothesised. In the first one, the two side arms are
equivalent inside the same conformer but not equivalent
among the two conformers (a); the second one is contrary,
i.e. the two side arms are equivalent among the two
conformers but not equivalent inside the same

Supramolecular Chemistry
369
3
1
OH–
3
1
7
AcO–
5
3
7
5
1
F–
3
7
5
1
Cl–
3
7
1
5
Br–
12
3.0
8
3
1
7
5
L1
13.0 12.0 11.0 10.0
9.0
8.0
7.0
6.0
5.0
4.0
2.0
1.0
(ppm)
Figure 2. 1H NMR spectra of L1 in CD3CN solution at 298 K by adding 3 equivalents of the guest with respect to L1, with the exception
of OH2 added in large excess; the addition of I2 does not modify the L1 spectrum.
conformer (b); both hypotheses are schematically reported
in Figure 3.
adducts with acetate as well as with fluoride gives rise to a
small downfield shift of the resonances which collapse in
only one singlet at 261.69 and 261.53 ppm for AcO2 and
1H NMR experiments at variable temperatures in both
solvents were carried out. In DMSO-d6, the resonances of
the side arms collapse in only one pattern of signals at
T ¼ 333 K (Figure 4); at this temperature, the resonances
of the macrocyclic base are still broad. The resonances of
the side arms preserve the same linewidth at higher
temperature while the macrocyclic resonances collapse in
a sharp pattern of signals, resulting from a C2v symmetry
mediated on the NMR time scale, at 353 K (Figure 4).
In other words, the side arms become equivalent at lower
temperature than the macrocyclic base, on the NMR time
scale, independent of two conformers, thus the second
hypothesis (b) seems to be the most likely one (see
Figure 3(b)). In addition, the NMR titrations performed
considering all the shifting signals separately yield
comparable stability constants, supporting the previous
statement. The spectra at variable temperatures carried out
in CD3CN showed a similar trend but, due to the lower
boiling point of the solvent, the complete collapse of all
resonances was not achieved.
19F NMR can be carried out to follow the behaviour of
CF3 groups in the presence of anions in both solvents; for
example, in DMSO-d6, they show two singlets at 262.0
and 261.84 ppm due to the non-equivalence of the two
side arms on the NMR time scale. The formation of the
Figure 3. Schematic drawing of the possible conformers.

370
M. Formica et al.
353 K
343 K
333 K
323 K
313 K
303 K
10.0
9.0
8.0
7.0
6.0
4.0
3.0
2.0
(ppm)
Figure 4. 1H NMR spectra of L1 species, recorded in DMSO-d6 solution at different temperatures.
F2, respectively (see Figure S1 of the Supporting
Information). In the case of fluoride, it is possible to
highlight the interaction with L1 following the signal
exhibited by F2 in the 19F NMR spectra; the free F2 shows
a sharp resonance at 297.4 ppm. By adding increasing
equivalents of L1 to the F2 solution, the signal becomes
broader and shifts downfield up to 287.6 ppm in an excess
of the host (all F2 bound; see Figure S1 of the Supporting
Information); however, only one resonance is observable
even at different L1/F2 molar ratios, indicating a fast
exchange between coordinated and free fluoride on the
NMR time scale. A similar behaviour was observed in
CD3CN where the signal of fluoride shifts from 2111.5 to
288.7 ppm in the presence of a large excess of L1.
and the lowest for Br2. A marked downfield shift in the H7
resonance, compared to the halides, was observed in the
presence of AcO2 anion in both media. This accentuated
shift of H7 with AcO2 can be attributed to the different
way that L1 stabilises the two different anions, i.e.
spherical vs. V-shape.
In conclusion, the NMR experiments proved that the
replacement of PNP with TFP in L1 preserves the
behaviour of this class of receptors in binding anions. The
TFP groups decrease the acidity of the ureido fragments
but increase the solubility of the receptor in other solvents
different from DMSO. A remarkable result is the ability of
L1 to form stable adducts with F2 in CD3CN,
unobtainable in DMSO.
1
In H NMR, the interaction between F2 and L1 in
CD3CN broadens the ureido resonances, preventing the
accurate evaluation of the stability constants for the L1/F2
system (see Figure S3 of the Supporting Information).
However, although very broad, the H5 and H7 signals still
appear in the spectrum even when an excess of F2 was
added (Figure 2 and see Figure S3 of the Supporting
Information), suggesting that F2, unlike DMSO, is not
able to deprotonate the TFP–urea group in this medium.
This is also supported by the lack of resonance resulting
from the [FHF]2 species.
UV–vis experiments
The UV–vis absorption spectrum of L1 can be affected by
the host–guest interaction with anions in CH3CN, while it
is not in DMSO. In DMSO solution, only the deprotona-
tion of the TFP–urea groups, obtained by adding a large
amount of hydroxide or fluoride anions, is able to change
the spectral feature of L1. In these cases, the absorption
profile shows two main bands at lmax 261 nm (1 ¼
20,900 cm21 mol21 dm3) and 296 nm (1 ¼ 4800 cm21
mol21 dm3) in the neutral ligand and a band at lmax
305 nm (1 ¼ 30,500 cm21 mol21 dm3) by adding OH2 or
F2 in large amount with respect to L1 due to the
As shown in Figures 1 and 2, the shift of both ureido
protons in the presence of halides is in agreement with the
H-bonding acceptor power of the anion, the highest for F2

Supramolecular Chemistry
371
deprotonated species. On the contrary, also in CH3CN, the
presence of interacting anions affects the UV–vis spectral
profile of free L1; in particular, this occurs with the
addition of acetate, fluoride, chloride and bromide, but not
iodide. Performing titration experiments, it was possible to
spectrophotometrically determine the association con-
stants of G (AcO2, F2, Cl2 and Br2) to L1 (Table 1). An
example of the UV–vis experiments is shown in
Figure 5(a). The free L1 in CH3CN shows a spectrum
with two main bands, with lmax at 246 nm (1 ¼
32,700 cm21 mol21 dm3) and 291 nm (1 ¼ 3000 cm21
mol21 dm3); when increasing amounts of AcO2 were
added to L1, the two lmax shifted towards lower energies
reaching, after the addition of about 4 equivalents of AcO2
(see the inset of Figure 5(a)), the maximum shift at lmax
252 nm (1 ¼ 33,600 cm21 mol21 dm3) and 295 nm (1 ¼
3400 cm21 mol21 dm3). This spectral change is attributed
to the formation of L1–AcO2 adducts able to affect the
Figure 5. (a) UV–vis [L1] ¼ 2.5 £ 1025 M and (b) fluorescence titration [L1] ¼ 2.5 £ 1026 M of L1 with AcO2 in CH3CN solution at
298 K. Inset: variation of the labs at 252 nm and lem at 334 nm as a function of AcO2 added.

372
M. Formica et al.
chromophore in this medium. No deprotonation process
was observed in the presence of a large amount of the guest
by UV–vis experiments confirming the NMR exper-
iments. In this solvent, the deprotonation of the ureido
units can be obtained only by adding a strong base such as
NBu4OH, which yields a decrease in the band near 250 nm
with the simultaneously strong increase in that near
300 nm with an eight-fold increase in absorptivity
(see below).
This different optical response of the TFP–ureido
chromophore to the presence of anions in the two solvents
can be ascribed to the formation of stronger H-bond
interactions in CH3CN than in DMSO due to the different
dielectric constant of the solvents (11) and in agreement
1
with the H NMR results.
The addition of F2, Cl2 and Br2 to the CH3CN solution
containingL1 givesrisetoUV–visabsorptionresultssimilar
to those obtained with AcO2 (Figure 6(a)), with the shift of
Figure 6. (a) UV–vis [L1] ¼ 2.5 £ 1025 M and (b) fluorescence titration [L1] ¼ 2.5 £ 1026 M of L1 with F2 in CH3CN solution at
298 K. Inset: variation of the labs at 252 nm and lem at 355 nm as a function of F2 added.

Supramolecular Chemistry
both bands reaching their maximum shift at lmax 253 nm Fluorescence experiments
373
(1 ¼ 36,200 cm21 mol21 dm3) and 249 nm (1 ¼
32,500 cm21 mol21 dm3), 247 nm (1 ¼ 31,400 cm21
mol21 dm3) and 297 nm (1 ¼ 3800 cm21 mol21 dm3),
293 nm (1 ¼ 3200 cm21 mol21 dm3) and 292 nm (1 ¼
3100cm21 mol21 dm3) for F2, Cl2 and Br2, respectively.
The invariance of the spectra was observed after the addition
of about 3, 5 and 7 equivalents of F2, Cl2 and Br2,
respectively. The shift of lmax, as well as the equivalents of
the anion necessary to saturate the spectra, highlights that the
affinity of L1 towards the halide series depends on their H-
bonding acceptor affinity, the highest for fluoride and the
lowest for bromide. In agreement with the NMR
experiments, no deprotonation process was observed with
the halide series, neither with F2, in CH3CN, even when
adding a large excess of the anion with respect to L1; this
highlights that, in CH3CN, only the association between L1
and the halide series takes place.
The presence of the TFP groups makes L1 fluorescent in
both DMSO and CH3CN solvents. In DMSO, in agreement
with UV–vis absorption properties, the fluorescence is
only sensitive to the deprotonation process while it is not
to the interaction with the anions. By excitation at 291 nm
in DMSO, the free ligand shows an emission band centred
at 350 nm (quantum yield F ¼ 0.010), while the
deprotonation process gives rise to a new band with
lem ¼ 481 nm having a five-fold enhancement in the
emission when compared to that at 350 nm. The higher
energy band is attributed to the neutral form of L1 while
the lower energy band is attributed to the deprotonated
one.
On the contrary, the fluorescence of L1 in CH3CN is
modified by the presence of the interacting anions. The
free L1 exhibits an emission band with lem at 334 nm
(quantum yield F ¼ 0.013, lex ¼ 296 nm). By adding up
to 10 equivalents of OH2 with respect to L1, a red shift of
the band up to 346 nm, together with the appearance of a
new band at lower energy (lem ¼ 463 nm), can be
observed; however, the emission of this latter remains
lower than that of the former. Adding a large excess of
OH2, the disappearance of the band at higher energy,
together with a further shift of the band at lower energy
(lem ¼ 480 nm), can be observed, with a five-fold
enhancement of the emission with respect to that at
334 nm (Figure 7(b)). The first trend of the titration is
similar to that found for the halides (Figure 6(b)),
confirming, as reported above for the absorption proper-
ties, that OH2 first interacts with L1 via H-bond and then,
only when it is present in large excess, it is able to
deprotonate the TFP–urea units. It must be highlighted
that, in CH3CN, the emission bands of both neutral and
deprotonated ligands are centred at higher energies with
respect to those in DMSO. This means that, being the
absorption transition, an n ! p* charge transfer, the higher
dielectric constant of DMSO better stabilises the excited
state with respect to CH3CN, thus increasing the Stoke
shift.
This is an important aspect which allows, processing
the titration curves, the evaluation of the association
constants of adducts formed with fluoride. The data for all
L1–G adducts formed, obtained by UV–vis titration, are
reported in Table 1 and are discussed below. It is
interesting to underline a further difference of L1 in the
two solvents in the presence of OH2. The addition of OH2
to a DMSO solution of L1 does not significantly affect the
two bands (shift ,1 nm) when adding up to 10 equivalents
of OH2 with respect to L1; moreover, at least 100
equivalents of OH2 are able to deprotonate the urea
group. On the contrary, the titration in CH3CN shows the
red shift of both bands, similar to that exhibited by adding
acetate; they shift from 246 and 291 to 252 and 294 nm,
respectively, in the presence of 10 equivalents of OH2
(Figure 7(a)). In other words, as reported for other anions,
OH2 first interacts with L1 via H-bond and then, only
when it is present in large excess, it deprotonates the TFP–
urea units. Probably, this occurs in both solvents but it is
more evident in CH3CN where, similar to results obtained
with the other anions, the H-bonds formed in the L1–G
adducts are stronger and are thus able to modify the
spectral feature of L1. Unfortunately, either the associ-
ation constant is too high or the deprotonation and
association processes occur simultaneously and they are
not independent since we were not able to safely determine
their constant values by the spectrophotometric titrations.
Performing experiments at a fixed concentration of
OH2, we found that [OH2] $ 1023 M is able to
deprotonate L1 in DMSO while, with [OH2] # 1024 M,
L1 remains in its neutral form. The range found in CH3CN
is 10-fold higher; in fact, only [OH2] $ 1022 M is able to
deprotonate L1 while it is in the neutral form for
[OH2] # 1023 M. This in addition to highlighting the
lower acidity of L1 in CH3CN than in DMSO, as expected,
provides a range of [OH2] in which L1 can be used as a
sensor for a strong base.
The different wavelength of emission exhibited by free
and deprotonated ligands in both solvents is well visible
not only by the spectral profile but also to the naked eyes,
irradiating two different samples with UV light, one
containing only L1 and the other L1 plus hydroxide anion
(see Figure 8). The two bands can be attributed to the
emission of the neutral and deprotonated forms of
the TFP–urea group, respectively, making L1 sensitive
to the presence of OH2, although in large excess when
compared to L1. L1 responds to the presence of OH2 by
visible fluorescent emission in the same [OH2] range as in
the UV–vis experiments, suggesting that it could be a
possible application for L1; a fluorescent sensor for OH2
or another strong base able to deprotonate the TFP–ureido

374
M. Formica et al.
Figure 7. (a) UV–vis [L1] ¼ 2.5 £ 1025 M and (b) fluorescence titration [L1] ¼ 2.5 £ 1026 M of L1 with OH2 in CH3CN solution at
298 K.
unit in non-aqueous solvents. As mentioned above and in
agreement with the UV–vis experiments, the fluorescence
spectrum is affected by the presence of AcO2, F2, Cl2 and
Br2 anions while I2 does not affect the fluorescence of L1,
suggesting the lack of interaction between L1 and iodide.
Figure 5(b) reports the titration of L1 with acetate; as
can be seen, the lem at 334 nm decreases when adding the
anion while a new band centred at 457 nm appears in the
spectrum. This band lies in the visible range, thus making
L1 suitable to detect the anion also with the naked eyes and
the solution shows a colour similar to that reported in
Figure 8. In the inset of Figure 5(b), it is possible to see
that the band centred at 334 nm stops to vary after the
addition of about 4 equivalents of AcO2.
This spectral behaviour is attributed to the interaction
via H-bonding of L1 with AcO2 taking place at the ureido
groups. Even when adding acetate, as a solid compound,
no deprotonation process was observed in agreement with
the NMR and UV–vis experiments. Examining the
fluorescence of L1 in the presence of F2 (Figure 6(b)), a
different trend with respect to that observed with AcO2
can be highlighted. While the presence of acetate gives rise

Supramolecular Chemistry
375
often reported in the literature in similar cases (15).
However, it is to remark that the presence of AcO2 in
solution can be signalled by fluorescence emission in the
visible range (lem ¼ 457 nm) which the other anions do
not yield.
Association constants
All the G–L1 titrations, carried out by UV–vis,
fluorescence and when it was possible by 1H NMR
experiments, were processed and the stability constants of
the species formed are summarised in Table 1. The
constants evaluated by NMR titrations were obtained by
processing all shifting resonances together or separately;
no significant changes in the constant values were found.
As reported above, in DMSO, only NMR is sensitive to the
presence of the guests, while, in CH3CN, the constants
were obtained only from UV–vis and fluorescence data.
The NMR experiments are not suitable, since the
association constants with halides and acetate are too
high in this medium (16). The processing of the 1H NMR
titration gave rise to the stability constants due to the
formation of L1–AcO2 adducts in DMSO. The speciation
of the adducts, as well as the values of the association
constants, resulted similar to those obtained with the
previously studied receptor L2 (6); in fact, L1 forms two
adducts, [L1AcO]2 and [L1(AcO)2]22, with association
constant values of log K ¼ 4.5 and 2.1, a bit lower but with
a similar trend to those obtained with L2 (log K ¼ 5.6 and
2.9 for the addition of the first and second acetate,
respectively). The addition of the first acetate being higher
than the second and higher with respect to a V-shape
stabilisation by only one aromatic-ureido group in DMSO
(11), a coordination of AcO2 in the [L1AcO]2 species
similar to that suggested in the [L2AcO]2 species can be
supposed (6). In other words, both side arms with both
ureido groups participate in stabilising one AcO2 via H-
bonding. As previously demonstrated (14), the key to
preorganisation is given by the two NZCvO groups
linking each side arm to the 12-membered macrocyclic
ring, thus preorganising them so that both urea groups
cooperate in stabilising the acetate via H-bond.
Figure 8. Photo of the sample of L1 in CH3CN solution: alone
(right) and in [OH2] ¼ 1 £ 1022 M (left).
to the decrease in the emission at higher energy with the
simultaneous appearance of a new band at lower energy,
the presence of fluoride does not produce the appearance
of the latter band (lem ¼ 457 nm) but a marked red shift of
the band at higher energy which moves from lem ¼ 334 to
lem ¼ 355 nm (Figure 6(b)) with a slight increase in the
emission; the spectrum obtained after the addition of 3
equivalents of fluoride is not further modified. A similar
behaviour was also observed by adding Cl2 and Br2,
although with a lower shift of the band at lem ¼ 334 nm
(the final lem is 343 and 341 nm for Cl2 and Br2,
respectively). These data highlight that the emission of L1
is affected in a different way by the presence of AcO2 vs.
F2 and the other halides in solution; in particular, while the
halides shift the lem of the neutral form of L1, and this
shift can be correlated with the strength of H-bonding
occurring, the highest shift for F2 and the lowest for Br2,
the acetate quenches the emission of this band and
produces the appearance of a new band, with a wavelength
similar to that of the deprotonated ones but of lower
intensity. In other words, while the spectra of L1 in the
presence of halides show only one emission band, the
presence of acetate gives rise to two distinct emission
bands, one being of the neutral and another of the
deprotonated species of L1. The hypotheses to explain this
aspect, as already suggested in the NMR assays and also
taking into account that no L1 deprotonation was observed
in NMR or UV–vis experiments in CH3CN in the presence
of AcO2, could be a different way of coordination and thus
a different H-bonding network provided by L1 in
stabilising the bidentate AcO2 vs. the spherical halides,
in addition to the higher basicity of AcO2 that favours the
deprotonation of the fluorophore in the excited state, as
The speciation of adducts formed in CH3CN is similar
to that found in DMSO with acetate; in fact, L1 forms
adducts with the anion G of both [L1G] and [L1G2]
stoichiometries. The value of the association constants
results higher for the addition of AcO2 in CH3CN than in
DMSO, as usually expected when using these solvents
(11). In particular, the formation of both [L1AcO]2 and
[L1(AcO)2]22 species resulted higher by about 2
logarithmic units in CH3CN than in DMSO; moreover,
the addition of the first acetate is higher than the second
and higher with respect to a V-shape stabilisation of AcO2
by only one aromatic-ureido fragment. This again suggests

376
M. Formica et al.
the cooperation of both side arms in stabilising the acetate
in this solvent as well. The second acetate is stabilised by
lower association constant; this can be attributed to the fact
that both urea units are involved in the stabilisation of the
first AcO2, hindering the binding of the second acetate,
once again indicating the cooperation of both side arms in
binding the first AcO2.
CH3CN. The host–guest interaction occurs via H-bond at
the TFP–ureido moieties; in DMSO, the receptor is
sensitive to the presence of the anion only when employing
1H and 19F NMR, while, in CH3CN, all the techniques
explored are sensitive. The fluorescence in CH3CN shows
different spectral behaviours in the presence of AcO2 or
halide anions; this can be attributed to a different way of
stabilising the two guests by the side arms, in agreement
with the stability constant values observed for the
formation of the host–guest adducts. L1 forms species
of 1:1 ([L1G]) and 1:2 ([L1G2]) stoichiometry with all the
interacting anions. The higher value of the constant for
the formation of the [L1AcO]2 species as compared to the
[L1(AcO)2]22 one suggests that both side arms cooperate
in binding this guest at least in the first complex but this
does not occur with the spherical halide anion since the two
association constants are similar. In other words, the
cooperative effects of the two side arms can only be
obtained with guests able to interact via H-bond in a bridge
disposition with the two side arms, as in the case of the
carboxylate group, whereas spherical anions, such as the
halides, are more readily stabilised by only one side arm.
The stability constants resulted higher in CH3CN than
in DMSO, with AcO2 being the most strongly bound
among all the anions investigated.
The stability constants with the halides, evaluated only
in CH3CN for shortness, show that L1 is able to bind 2
equivalents of all the anions (Table 1), with similar
association constants for the formation of the [L1G]2 and
[L1G2]22 adducts. The similar values for the addition of
the first and second anion suggest that the side arms
probably behave independently in binding spherical guests
such as the halide series. However, a small decrease in the
association constant of the second anion can be observed
with respect to the first one; these data can be ascribed to
several contributions, even if difficult to determine. For
example, the statistic effect, the second addition which
takes place on a negatively charged species as well as the
conformational changes of the ligand due to electrostatic
repulsions in the negative [L1G2]22 adduct, could explain
the lower value of the second constant. Nevertheless, the
larger difference between the two stability constants
observed in the case of chloride cannot completely exclude
the participation of both side arms in the binding of the
first chloride.
In conclusion, the introduction of the TFP group in this
ligand topology enhances the handling and solubility in
common organic solvents preserving the binding proper-
ties already found for this class of receptors. Moreover, the
ability of L1 to signal the presence of the acetate by
significant fluorescence changes also in the visible range
will stimulate us to develop the use of L1 as a sensor for
more complicated carboxylate guests.
The values were higher for F2 than for Cl2 and higher
for Cl2 than for Br2 in both steps, supporting the
statement that the interaction occurs via H-bonding; in
particular, fluoride exhibits very high association constants
almost comparable to the addition of two acetates,
although the latter shows the highest first association
constant.
Experimental section
Conclusions
General methods
In this study, we have reported the synthesis and the
binding properties of L1, a new versatile host for anions,
both in DMSO and CH3CN. Different synthetic
approaches were investigated and the acyl-fluoride
strategy was found to be the most reliable. An interesting
one-pot procedure for the removal of the Cbz-protecting
groups and formation of urea has also been performed. The
present synthetic approach allowed to obtain the new
receptor L1 with a 54% overall yield, on a 10–15 mmol
scale. Furthermore, the N-3,5-bis(trifluoromethyl)phenyl
moiety was found to be fluorescent, extending the
possibility to the determination of the host–guest
interactions.
IR spectra were recorded on a Shimadzu FTIR-8300
spectrometer. Melting points were determined on a Buchi
melting point B 540 apparatus and are uncorrected. EI-MS
spectra (70 eV) were recorded with a Fisons Trio 1000
spectrometer. The HPLC system (Waters Alliance 2795)
was coupled with a photodiode array detector (Waters
2996 PDA), followed by an electrospray mass spec-
trometer detector (ESI-MS) (Waters-Micromass ZQ)
worked by Mass Lynx 4.1 SP4 software. The ESI-MS
analyses were performed in a positive mode under the
following conditions: source and desolvation temperature,
100 and 2508C; capillary and cone voltage, 2.5 kV and
25 V; cone and desolvation flow (nitrogen gas), 40 and
The anion coordination properties of the new receptor
were investigated in solution by NMR, UV–vis and
fluorescence spectroscopy; L1 can bind AcO2, F2 and
Cl2 anions in DMSO solution and, in addition, Br2 in
400 l/h. HPLC analyses were performed on
a
25 cm £ 4.6 mm Discovery C-18, 5 cm column (Supelco,
Bellefonte, PA, USA) equipped with a Supelguard
Discovery C-18 guard column (2 cm £ 4 mm, 5 cm).

Supramolecular Chemistry
An analytical sample
377
Solvents A (MeOH) and C (5 nM ammonium acetate)
were run at a flow rate of 1 ml/min. The running gradient
was adjusted to 10% A (2 min), increasing to 90% A
(18 min), and then 100% C (2 min), followed by a re-
equilibration at 10% A (15 min). All solvents were HPLC
grade (Aldrich-Sigma, Milwaukee, WI, USA) and water
was purified via a Millex Q-plus system (Millipore,
Bedford, MA, USA).
1
Rotamer mixture: H NMR (CDCl3) d ¼ 2.24–2.38 (6H,
m), 2.63–2.71 (8H, m), 3.28–3.46 (8H, m), 4.05–4.07
(4H, m), 5.10–5.11 (4H, m), 5.82–5.89 (2H, m), 7.30–
7.40 (10H, m); 13C NMR (CDCl3) d ¼ 42.8, 43.1, 46.6,
55.0, 65.9, 127.2, 129.0, 141.2, 156.3, 164.3 ppm; FTIR
(KBr): 3268, 3066, 2986, 2940, 1726, 1642, 1550, 1454,
1255, 1167 cm21; MS (þESI) 583.7 (M þ 1); Anal. for
C30H42N6O6 (582.70): Calcd: C 61.84, H 7.27, N 14.42;
Found: C 62.0, H 7.2, N 14.3.
All chemicals were purchased from Aldrich, Fluka and
Lancaster in the highest quality commercially available.
All the solvents were dried prior to use. Chromatographic
separations were performed on silica gel columns by flash
chromatography (Kieselgel 60, 0.040–0.063 mm;
Merck, Dr Whitehouse Station, NJ, USA). TLC analyses
were performed on precoated silica gel on aluminium
sheets (Kieselgel 60 F254, Merck). 1,7-Dimethyl-
1,4,7,10-tetraazacyclododecane 3 was prepared as pre-
viously described (17).
1,10-(2,20-(4,10-Dimethyl-1,4,7,10-
tetraazacyclododecane-1,7-diyl)bis(2-oxoethane-2,1-
diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (L1)
Pd/C (10%, 2.1 g) was added to a solution of the above bis-
carbamate 4 (10.3 g, 17.7 mmol) in EtOH abs (150 ml).
The mixture was then hydrogenated at 1 atm with
hydrogen balloon for 4 h at room temperature. When
TLC showed disappearance of the starting material, 3,5-
bis(trifluoromethyl)phenyl isocyanate 5 (40 mmol, 6.9 ml)
was added at room temperature. After 2 h, the suspension
was filtered through a short pad of Celite and the solvent
was removed under vacuum. The solid thus obtained was
purified on alumina using (CH2Cl2–MeOH 92:8) and then
by silica gel column chromatography (CH2Cl2–MeOH–
NH3 conc. 91:8:1) to give the product L1 as a white-off solid
in 61% yield (9.7 g).
Synthesis
4(N),10(N)-Bis-[N-(benzyloxycarbonyl)glycyl]-
1,7dimethyl-1,4,7,10-tetraazacyclododecane (4)
To a well-stirred solution of N-Cbz-glycine 1 (56 mmol,
11.7 g) in dry CH2Cl2 (150 ml) and pyridine (56 mmol,
4.44 ml) kept under a N2 atmosphere was added cyanuric
fluoride (0.12 mol, 9.7 ml) at 220 to 2108C. A precipitate
or emulsion was formed and gradually increased in
amount. After stirring at 2108C for 2 h, crushed ice was
added along with 300 ml of CH2Cl2. The organic layer was
separated and the aqueous layer was extracted twice with
300 ml of CH2Cl2. The combined organic extracts were
washed with 150 ml of ice-cold water, brine and dried
(Na2SO4). The organic solution was filtered and
concentrated under reduced pressure at room temperature
to give the pure acid fluoride 2 (infrared carbonyl
absorption 1847 cm21).
Rotamer mixture: mp ¼ 226–2288C; FTIR (film):
3390, 2924, 1680, 1648, 1559, 1325, 1269, 1174,
1110 cm21; H NMR (DMSO) d ¼ 2.31–2.47 (6H, m),
1
2.71–2.85 (8H, m), 3.48–3.52 (8H, m), 4.12–4.17 (4H,
m), 6.36–6.50 (2H, m), 7.34–7.38 (2H, m), 7.92–7.98
(4H, m), 9.11–9.18 (2H, m); 13C NMR (DMSO): 41.5,
41.8, 46.6, 57.4, 118.6, 124.1(q), 129.1, 131.7, 140.3,
154.7, 167.8 ppm; MS (þESI) 825.3 (MHþ); Anal. for
C32H36F12N8O4 (824.67): Calcd: C 46.61, H 4.40, F 27.65,
N 13.59, O 7.76; Found: C 46.65, H 4.45, N 13.69.
A solution of the above acid fluoride in dry CH2Cl2
(80 ml) was added dropwise over a period of 120 s at room
temperature to a stirred solution of 1,7-dimethyl-1,4,7,10-
tetraazacyclododecane 3 (4 g, 20 mmol) in 120 ml of H2O
containing 6.8 g (80 mmol) of NaHCO3.
Spectroscopic experiments
1H, 13C and 19F NMR spectra were recorded on a Bruker
Avance 200 instrument, operating at 200.13, 50.33 and
188.31 MHz, respectively, and equipped with a variable
temperature controller. The temperature of the NMR probe
was calibrated using 1,2-ethanediol as a calibration
sample. For the spectra recorded in CDCl3, CD3CN and
DMSO-d6, the 1H and 13C peak positions are reported with
respect to TMS while CFCl3 was used as a standard
reference for the 19F NMR spectroscopy. 1H–1H and
1H–13C correlation experiments were performed to assign
the signals.
The mixture was stirred at room temperature for
60 min (after 50 min, IR examination showed the acid
fluoride band at 1847 cm21 to have essentially disap-
peared), followed by the addition of 350 ml of CH2Cl2.
The organic yellow solution was separated and the
aqueous layer was extracted twice with 200 ml of CH2Cl2.
The combined organic extracts were washed with
200 ml of brine, dried (Na2SO4) and concentrated under
reduced pressure at room temperature to give compound 4
(10.3 g, 89% yield), which was used in the following step
without further purification.
NMR titrations were carried out in CD3CN or in a
DMSO-d6–0.5% water mixture; 0.5% of water was added

378
M. Formica et al.
Tuntulani, T. Chem. Soc. Rev. 2003, 32, 192–202. (f) Beer,
P.D.; Gale, P.A. Angew. Chem. Int. Ed. 2001, 40, 486–516.
(g) Schmidtchen, F.P.; Berger, M. Chem. Rev. 1997, 97,
1609–1646.
to DMSO to avoid the non-uniform absorption of water
from the atmosphere by anhydrous DMSO during the
titration. In a typical experiment, a 5 £ 1022 mol dm23
solution of the anion was added in 0.1 eq at a time to a
1 £ 1022 mol dm23 solution of the ligand directly in the
NMR tube; the tube was then kept for 5 min at a
temperature of 298 K before starting the acquisition of the
spectrum. The anions tested were added as their
tetrabutylammonium salts. The monitoring of the shift of
the signals in the ligand spectra (see Results and
discussion) permitted evaluation of the association
constants of ligand–anion using the HYPNMR computer
program (18).
(2) Amendola, V.; Fabbrizzi, L. Chem. Commun. 2009,
513–531.
(3) (a) Miranda, C.; Escarti, F.; Lamarque, L.; Yunta, M.J.R.;
Navarro, P.; Garcia-Espan˜a, E.; Jimeno, M.L. J. Am. Chem.
Soc. 2004, 126, 823–833. (b) Lamarque, L.; Navarro, P.;
Miranda, C.; Aran, V.J.; Ochoa, C.; Escarti, F.; Garcia-
Espan˜a, E.; Latorre, J.; Luis, S.V.; Miravet, J.F. J. Am.
Chem. Soc. 2001, 123, 10560–10570.
(4) (a) Kang, S.O.; Begum, R.A.; Bowman-James, K. Angew.
Chem. Int. Ed. 2006, 45, 7882–7894. (b) Esteban-Gomez,
D.; Fabbrizzi, L.; Licchelli, M.; Monzani, E. Org. Biomol.
Chem. 2005, 3, 1495–1500. (c) Lowe, A.J.; Dyson, G.A.;
Pfeffer, F.M. Org. Biomol. Chem. 2007, 5, 1343–1346. (d)
Fluorescence spectra were recorded at 298 K with a
Varian Cary Eclipse spectrofluorimeter. UV–vis absorption
spectra were recorded at 298 K with a Varian Cary-100
spectrophotometer equipped with a temperature control unit.
Theinteraction ofanionswith ligandsL1and L2 was studied
in similar conditions of the NMR titration experiments using
CH3CN or DMSO-d6–0.5% water mixture as the solvent;
the solution containing the guest anion (F2, Cl2, Br2, I2 or
AcO2) up to 5 equivalents with respect to the ligand was
added to the solution containing L1 ([L1] ¼ 1.5 £ 105 M).
At least three sets of spectrophotometric titration curves for
each G/L system were performed. All sets of curves were
treated either as single sets or as separate entities, for each
system; no significant variations were found in the values of
the determined constants. The HYPERQUAD computer
program was used to process the spectrophotometric data
(18b).
¨
Suhs, T.; Konig, B. Chem. Eur. J. 2006, 12, 8150–8157. (e)
`
Andres, A.; Arago, J.; Bencini, A.; Bianchi, A.; Domenech,
˜a, E.; Paoletti, P.; Ramirez, J.A.
Inorg. Chem. 1993, 32, 3418–3424.
A.; Fusi, V.; Garcia-Espan
(5) (a) Kang, S.O.; Hossain, M.A.; Bowman-James, K. Coord.
Chem. Rev. 2006, 250, 3038–3052. (b) Bowman-James, K.
Acc. Chem. Res. 2005, 38, 671–678. (c) Amendola, V.;
Esteban-Gomez, D.; Fabbrizzi, L.; Licchelli, M. Acc. Chem.
Res. 2006, 39, 343–353. (d) Gale, P.; Quesada, R. Coord.
Chem. Rev. 2006, 250, 3219–3244. (e)Kim, T.H.;Choi, M.S.;
Sohn, B.-H.; Park, S.-Y. Chem. Commun. 2008, 2364–2366.
(f) Lowe, A.J.; Dyson, G.A.; Pfeffer, F.M. Eur. J. Org. Chem.
2008, 1559–1567. (g) Amendola, V.; Baiocchi, D.M.;
Fabbrizzi, L.; Mosca, L. Chem. Eur. J. 2008, 14, 9683–9696.
(6) Formica, M.; Fusi, V.; Macedi, E.; Paoli, P.; Piersanti, G.;
Rossi, P.; Zappia, G.; Orlando, P. New J. Chem. 2008, 32,
1204–1214.
(7) For selected reviews of (thio)urea-based organocatalysis,
see: (a) Zhang, Z.; Schreiner, P.R. Chem. Soc. Rev. 2009,
38, 1187–1198. (b) Connon, S.J. Chem. Commun. 2008,
2499–2510. (c) Doyle, A.G.; Jacobsen, E.N. Chem. Rev.
2007, 107, 5713–5743. (d) Taylor, M.S.; Jacobsen, E.N.
Angew. Chem., Int. Ed. 2006, 45, 1520–1543. (e)
Takemoto, Y. Org. Biomol. Chem. 2005, 3, 4299–4306.
(f) Schreiner, P.R. Chem. Soc. Rev. 2003, 32, 289–296.
(8) Valeur, E.; Bradley, M. Chem. Soc. Rev. 2009, 38, 606–631.
(9) (a) Carpino, L.A.; Beyermann, M.; Wenschuh, H.; Bienert,
M. Acc. Chem. Res. 1996, 29, 268–274 and reference
therein. (b) Jedrzejczak, M.; Motie, R.E.; Satchell, D.P.N.;
Satchell, R.S.; Wassef, W.N. J. Chem. Soc., Perkin Trans. 2
1994, 1471–1479.
The fluorescence quantum yields (Ff) were determined
by comparing the integrated fluorescence spectra of
the sample with 2,20-biphenol in acetonitrile (Ff ¼ 0.29)
(17).
Supporting Information
19F NMR titrations of L1 with AcO2 or F2 in DMSO-d6,
1H NMR titration of L1 with F2 in CD3CN and 13C NMR
spectrum of L1 are available online.
(10) White, J.M.; Rao Tunoori, A.; Turunen, B.J.; Georg, G.I.
J. Org. Chem. 2004, 69, 2573–2576.
Acknowledgement
´
(11) Perez-Casa, C.; Yatsimirsky, A.K. J. Org. Chem. 2008, 73,
2275–2284.
The authors thank the Italian Ministero dell’Istruzione dell’U-
(12) Shenderovich, I.G.; Tolstoy, P.M.; Golubev, N.S.; Smirnov,
S.N.; Denisov, G.S.; Limbach, H.H. J. Am. Chem. Soc.
2003, 125, 11710–11720.
`
niversita e della Ricerca (MIUR) PRIN2007 for the financial
support.
(13) Boiocchi, M.; Del Boca, L.; Esteban-Gomez, D.; Fabbrizzi,
L.; Licchelli, M.; Monzani, E. Chem. Eur. J. 2005, 11,
3097–3104.
References
(14) (a) Formica, M.; Fusi, V.; Giorgi, L.; Guerri, A.; Lucarini,
S.; Micheloni, M.; Paoli, P.; Pontellini, R.; Rossi, P.; Tarzia,
G.; Zappia, G. New J. Chem. 2003, 27, 1575–1583. (b)
Ambrosi, G.; Dapporto, P.; Formica, M.; Fusi, V.; Giorgi,
L.; Guerri, A.; Lucarini, S.; Micheloni, M.; Paoli, P.;
Pontellini, R.; Rossi, P.; Zappia, G. New J. Chem. 2004, 28,
1359–1367. (c) Ambrosi, G.; Formica, M.; Fusi, V.; Giorgi,
(1) (a) Bianchi, A.; Bowman-James, K.; Garcia-Espan˜a, E.
Supramolecular Chemistry of Anions; Wiley-VCH:
New York, 1997. (b) Sessler, J.L.; Gale, P.A.; Cho, W.S.
Anion Receptor Chemistry; Royal Society of Chemistry:
Cambridge, 2006. For recent reviews, see: (c) Caltagirone,
C.; Gale, P.A. Chem. Soc. Rev. 2009, 38, 520–563. (d) Gale,
P.A. Acc. Chem. Res. 2006, 39, 465–475. (e) Suksai, C.;

Supramolecular Chemistry
379
L.; Guerri, A.; Lucarini, S.; Micheloni, M.; Paoli, P.; Rossi,
P.; Zappia, G. Inorg. Chem. 2005, 44, 3249–3260.
(15) Sharma, A.; Shulman, S.G. Introduction to Fluorescence
Spectroscopy; Wiley: New York, 1999.
(16) Jyotirmayee, M.; Pal, H.; Sapre, A.V. Bull. Chem. Soc. Jpn
1999, 72, 2193–2202.
(17) Ciampolini, M.; Dapporto, P.; Micheloni, M.; Nardi, N.;
Paoletti, P.; Zanobini, F. J. Chem. Soc., Dalton Trans. 1984,
1357–1362.
(18) (a) Frassinetti, C.; Gelli, S.; Sabatini, A.; Moruzzi, M.S.;
Vacca, A. Anal. Biochem. 1995, 231, 374–382. (b) Gans, P.;
Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739–1753.

Copyright of Supramolecular Chemistry is the property of Taylor & Francis Ltd and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.