Enantioselective fluorescence sensing of amino acids by modified cyclodextrins: Role of the cavity and sensing mechanism

Article abstract of DOI:10.1002/chem.200305448

Two selectors based on modified cyclodextrins containing a metal binding site and a dansyl fluorophore-6-deoxy-6-N-(N^α-[(5- dimethylamino-1-naphthalenesulfonyl)aminoethyl]phenylalanylamino-β- cyclodextrin-containing D-Phe (3) and L-Phe (4) moie

Full text of DOI:10.1002/chem.200305448

FULL PAPER
Enantioselective Fluorescence Sensing of Amino Acids by Modified
Cyclodextrins: Role of the Cavity and Sensing Mechanism
Sara Pagliari,[a, c] Roberto Corradini,*[a] Gianni Galaverna,[a] Stefano Sforza,[a]
Arnaldo Dossena,[a] Marco Montalti,[b] Luca Prodi,[b] Nelsi Zaccheroni,[b] and
Rosangela Marchelli[a]
Abstract: Two selectors based on modi-
fied cyclodextrins containing a metal
clodextrins were found to bind cop-
per(ii) with different affinities, as re-
vealed by fluorescence quenching in
competitive binding measurements.
Addition of d- or l-amino acids in-
duced increases in fluorescence intensi-
ty, which were dependent on the amino
acid used and in some cases on its ab-
solute configuration. The cyclodextrin
4 was found to be more enantioselec-
tive than 3, suggesting that the self-in-
clusion in the cyclodextrin cavity
strongly increases the chiral discrimina-
tion ability of the copper(ii) complex.
Accordingly, a linear fluorescent ligand
Na-[(5-dimethylamino-1-naphthalene-
sulfonyl)aminoethyl]-N1-propyl-phenyl-
alaninamide, which has the same bind-
ing site and absolute configuration as
4, showed very low chiral discrimina-
tion ability. The enantioselectivity in
fluorescence response was found to be
due to the formation of diastereomeric
ternary complexes, which were detect-
ed by ESI-MS and by circular dichro-
ism. Time-resolved fluorescence studies
showed that the fluorescence of the
dansyl group was completely quenched
in the ternary complexes formed, and
that the residual fluorescence was due
to uncomplexed ligand.
binding site and
a dansyl fluoro-
phore–6-deoxy-6-N-(Na-[(5-dimethyl-
amino-1-naphthalenesulfonyl)amino-
ethyl]phenylalanylamino-b-cyclodex-
trin–containing d-Phe (3) and l-Phe
(4) moieties were synthesized. The con-
formations of the two selectors were
studied by circular dichroism, two-di-
mensional NMR spectroscopy and
time-resolved fluorescence spectrosco-
py. Cyclodextrin 4 was found to have a
predominant conformation in which
the dansyl group is self-included in the
cyclodextrin cavity, while 3 showed a
larger proportion of the conformation
with the dansyl group outside the
cavity. As a consequence, the two cy-
Keywords: amino acids ¥ copper ¥
cyclodextrins ¥ enantioselectivity ¥
sensors
Introduction
Optical sensing by molecular recognition principles has
become increasingly important in recent years.[1] Optical
sensing molecules have been designed for ions and organic
compounds, by use of supramolecular chemistry in combina-
tion with optical and photophysical properties.[2] Fluorescent
sensors are of particular interest, due to the high sensitivity
and selectivity of the detection method,[3] and to their prop-
erties as photoionic gates and switches.[4] Fluorescent sens-
ing molecules capable of detecting metal ions have been re-
ported, based on changes in fluorescence intensity due to
photoinduced electron transfer (PET)[5] or proton transfer
processes,[6] or shifts in the fluorescence maxima induced by
complexation,[7] by excimer formation[8] or by irreversible
reactions.[9] Sensors for anions[10] and bioactive organic mol-
ecules[11] have also been synthesized, some of them showing
remarkable selectivity.
[a] Dr. S. Pagliari, Prof. R. Corradini, Dr. G. Galaverna, Dr. S. Sforza,
Prof. A. Dossena, Prof. R. Marchelli
Dipartimento di Chimica Organica e Industriale
Universit‡ di Parma, Parco Area delle Scienze 17/A
43100 Parma (Italy)
Fax : (+39)0521-905472
E-mail: roberto.corradini@unipr.it
[b] Dr. M. Montalti, Dr. L. Prodi, Dr. N. Zaccheroni
Dipartimento di Chimica ™G. Ciamician∫
Universit‡ di Bologna, Via Selmi 2
40126 Bologna (Italy)
[c] Dr. S. Pagliari
Present address:
Cyanagen S.r.l. via Stradelli Guelfi 40/c
40138 Bologna (Italy)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author. ESI-MS spectra
of the ternary mixtures containing cyclodextrin 4, copper(ii) and d- or
l-proline.
Application of chemosensors as components of an ™artifi-
cial tongue∫ for the parallel detection of organic molecules
2749
Chem. Eur. J. 2004, 10, 2749 2758
DOI: 10.1002/chem.200305448
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

FULL PAPER
in solution has recently been
proposed.[12] One of the essen-
tial features of natural taste sen-
sory systems is the ability to dis-
criminate between enantiomers;
for example, l-amino acids are
bitter, while d-amino acids are
sweet. Moreover, enantioselec-
tivity is also one of the most im-
portant goals for the sensing of
organic substances,[13] since bio-
logical activity is strictly corre-
lated with stereochemistry. Al-
though some enantioselective
sensors based on enzymes,[14] on
piezoelectric effects,[15] or on
colorimetric methods[16] have
been proposed in recent years,
very few examples of enantiose-
lective fluorescence sensors
have been reported.[17,18]
Modified cyclodextrins bearing fluorescent moieties have
been used by Ueno and co-workers as sensors for a wide va-
riety of organic guests, including chiral molecules.[19] The ad-
vantages of the use of cyclodextrin derivatives are: 1) solu-
bility in water, which circumvents problems with highly lipo-
philic fluorescent groups that can limit application in biolog-
ical systems, 2) fluorescence enhancement of some fluoro-
phores through inclusion within the cyclodextrin cavity,[20]
3) the presence of a lipophilic cavity which can act as a rec-
ognition motif for apolar groups in water,[21] and 4) the use
of intramolecular interactions to obtain rigidly preorganized
structures from flexible synthons.
acid used,[27] as a result of the formation of ternary com-
plexes.[28] However, no enantioselectivity was observed in
this case.
In a preliminary communication, we reported the use of
this approach to obtain enantioselective fluorescence sen-
sors for unmodified amino acids. We synthesized the two
modified cyclodextrins 3 and 4, each bearing an Na-(N2-dan-
sylaminoethyl)-d-(or l)-phenylalanine unit as a copper(ii)-
binding moiety.[29] Our design being based on our previous
experience in enantiomer separation by copper(ii) com-
plexes of amino acid amides (ligand-exchange chromatogra-
31]
phy),[30
the fluorescent group was chosen to have an
The use of cyclodextrins modified with chiral fluorescent
side arms has been described. Cyclodextrin mono-deriva-
tives containing either d- or l-dansylleucine were shown to
recognize organic guests with both chemo- and diastereose-
lectivity, the recognition properties being affected by the
configuration of the leucine residue.[22] An l-tryptophan-
modified b-cyclodextrin was recently described, and showed
enantioselective binding of organic guests such as borneol
or menthol.[23] However, this type of approach is not suitable
for the enantioselective binding of bifunctional polar mole-
cules such as unmodified amino acids.
amino, an amide, and a sulfonamide group as binding sites
for copper(ii), and an additional chiral centre to enhance
enantioselectivity. b-Cyclodextrin was chosen as platform,
since the size of the cavity is most suited for the complexa-
tion of the dansyl residue.[26]
Herein we report a systematic study of the sensing mecha-
nism of the two cyclodextrins 3 and 4 by spectroscopic
methods. Comparison between the two cyclodextrins and
the analogous compound 5, without the cyclodextrin part,
provided precious insights into the role of the cavity in the
recognition process.
Modification with groups containing metal binding moiet-
ies allows the use of metal complexes of cyclodextrins as
binding sites for organic molecules with donor groups.[24] For
example, enantiomeric recognition of unmodified amino
acids was performed with the copper complex of a hista-
mine-modified b-cyclodextrin and related molecules.[25]
We have recently described the synthesis and binding
properties of modified b-cyclodextrins 1 and 2, containing a
binding site for copper(ii) and a dansyl fluorophore.[26,27]
The cyclodextrin CD-dien-DNS 2 has been found to un-
dergo fluorescence quenching in the presence of copper(ii),
thus acting as a fluorescent chemosensor with good metal
ion selectivity. One interesting property of the complex Cu-
2 was that the fluorescence was ™switched on∫ in the pres-
ence of amino acids, with responses depending on the amino
The mechanism of fluorescence sensing was studied by
steady-state and time-resolved fluorescence and circular di-
chroism, and the enantioselectivity was evaluated as a func-
tion of the analyte amino acid and of the cyclodextrin struc-
ture.
Results and Discussion
Synthesis: Na-(N2-Dansylaminoethyl)-d-(or l)-phenylalanine
d-8 or l-8 were synthesized from d- and l-phenylalanine
benzyl ester with a simple two-step strategy (Scheme 1).
tert-Butoxycarbonyl(Boc)-protected aminoethylphenylala-
nine benzyl ester 6 was first synthesized by a general strat-
egy for the synthesis of chiral peptide nucleic acids
2750
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2749 2758

Enantioselective Sensing of Amino Acids by Modified Cyclodextrins
2749 2758
Scheme 1. Synthesis of the cyclodextrins 3 and 4 and of the amide 5:
i) NaBH3CN, CH3COOH, in MeOH, RT, 2 h; ii) AlCl3, anisole, CH3NO2,
CH2Cl2, 08C, 15 min; iii) dansyl chloride, Li2CO3 in H2O/CH3CN, 08C,
4h; iv) HBTU/DIEA in DMF dry, 3 h.
Figure 1. Circular dichroism spectra of: a) 4, b) 3 in aqueous solution
(0.1m borate buffer, pH 7.3), and c) 5 in MeOH/H2O 9:1 (0.02m borate
buffer, pH 8.0).
(PNAs).[32] Simultaneous removal of Boc and benzyl groups
was carried out with AlCl3. The resulting mixture containing
aminoethylphenylalanine (7) was treated with dansyl chlo-
ride at low temperature (08C, in order to avoid sulfonyla-
tion of the secondary amine) to yield d-8 or l-8. The two
modified cyclodextrins 3 and 4 were synthesized by treat-
ment of 6-deoxy-6-amino-b-cyclodextrin 9 with d-8 or l-8,
respectively, in the presence of a coupling reagent (HBTU).
Good enantiomeric purities (ee = 95% for 3 and 94% for
4) were obtained, as measured by hydrolysis of the ligands
with HCl followed by chiral GC-MS analysis of the resulting
N-2-aminoethylphenylalanine by a method previously devel-
oped by us.[33]
dichroism spectrum of 5, with the same absolute configura-
tion of the side arm of 4, showed a weak negative nature of
the 350 nm band, indicating that the spectrum of the latter
compound is not due to the asymmetry of the side arm, but
rather to its interaction with the cyclodextrin cavity.
Two-dimensional NMR spectra (ROESY and TOCSY)
were used to better characterize the different conformations.
Figure 2 reports the portions of the ROESY spectra showing
the cross-peaks connecting the aromatic with the cyclodex-
trin protons.
In the case of the l-Phe-containing cyclodextrin 4, corre-
lations between the protons of the dansyl group– protons 3’
and 7’ in particular–with those of the cyclodextrin were de-
tected, suggesting equatorial inclusion within the CD cavity.
In the case of 3, only correlations between the phenylala-
nine aromatic protons and the cyclodextrin cavity were de-
tected, thus suggesting a predominant conformation with
the dansyl group outside and the phenyl ring inside or on
top of the cavity.
The linear ligand 5 was obtained in a similar way, by
treatment of l-8 with n-propylamine.
Conformations of the ligands: The two diastereomeric cyclo-
dextrins 3 and 4 showed different conformational properties
in solution. Their circular dichroism spectra, reported in
Figure 1, showed 350 nm bands of opposite signs.
In the inclusion complexes of substituted naphthalenes
with cyclodextrins, the signs of the induced circular dichro-
ism (ICD) bands are dependent on the type of complex
formed (axial or equatorial) and on the position of the chro-
mophore with respect to the cavity (inside or outside).[34] In
our previous work we had been able to assign the differen-
ces in the ICDs of cyclodextrins 1 and 2 to different orienta-
tions of the dansyl moiety within the cavity, as confirmed by
NMR data; in particular, an intense negative band in the
350 nm region was associated with deep axial complexation,
while a positive one was interpreted as due to equatorial
complexation.
In the present case, the positive band of 4 can be inter-
preted as due to equatorial inclusion of the dansyl group,
while the weak negative band of 3 could be due either to
axial complexation or to the dansyl group being positioned
outside the cavity in an equatorial orientation. The circular
Since the NMR timescale did not allow the contributions
of the different conformations to be evaluated, steady-state
and time-resolved fluorescence measurements were also per-
formed.
The absorption spectra of 4, 3 and 5 in water solutions are
similar but not superimposible, showing a small red shift
(from 333 to 335 nm) and a lowering of the molar absorptiv-
ity (3900, 3700 and 3600mꢀ1 cmꢀ1, respectively) in that order.
Figure 3 shows the fluorescence emission spectra of the
two cyclodextrins 3 and 4, measured under the same condi-
tions, and that of the linear analogue 5. Table 1 lists the cor-
responding excited state lifetimes and pre-exponential
terms.
The fluorescence spectra of 3, 4 and 5 in water solutions
(Figure 3, normalized to the same relative intensity) show a
trend very similar to that observed for the absorption spec-
tra, but quantitatively more pronounced. A red shift in the
2751
Chem. Eur. J. 2004, 10, 2749 2758
www.chemeurj.org
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

R. Corradini et al.
FULL PAPER
Figure 2. Portions of the ROESY 400 MHz spectra (D2O, pH 7.0) of: a) 3, b) 4, showing cross-peaks between aromatic and aliphatic protons. A G refer
to different glucose rings.
Table 1. Excited state lifetimes (t) and pre-exponential terms (a) ob-
served for samples of 3, 4 and 5 solutions in aqueous borate buffer (0.1m,
pH 7.3), and of 5 in MeOH/H2O 9:1 borate buffer (0.2m, pH 7.9).
Sample
tA [ns]
aA [%]
tB [ns]
aB [%]
3
15
16
16
16
16
16
17
18
-
38
26
30
31
69
59
51
53
-
462
474
470
469
431
44
Cu-3
Cu-3/l-Pro[a]
Cu-3/d-Pro[a]
4
Cu-4
1
Cu-4/l-Pro[a]
Cu-4/d-Pro[a]
5
5
5
49
47
4100
13.2
5[b]
-
-
100
Figure 3. Fluorescence emission spectra. Left: a) 3 (6î10ꢀ5 m), b) 3 (6î
10ꢀ5 m) after addition of a 50-fold excess of adamantane-1-carboxylic acid
(ACA). Right: a) 4 (6î10ꢀ5 m), b) 4 (6î10ꢀ5 m) after addition of a 50-fold
excess of ACA, c) 5 (6î10ꢀ5 m). Solvent: borate buffer 0.1m, pH 7.3. lex
= 345 nm for 3 and 4, 335 for 5. Intensities are expressed as the percent-
age of the fluorescence of cyclodextrin 4 (corrected as described in the
Experimental Section).
[a] Copper(ii)/cyclodextrin/amino acid 1:1:1.[b] Measured in MeOH/H2O
(9:1) borate buffer (0.2m, pH 7.9).
two different lifetimes are taken into account (t1 = 4ns;
t2 = 15 ns).
The charge-transfer excited state responsible for the
dansyl fluorescence is very sensitive to the polarity of its en-
vironment; in particular, its maximum experiences a shift to-
wards higher energy in apolar media, as shown by fluores-
cent measurements on other dansyl derivatives.[35] At the
same time, increases in the fluorescence quantum yield and
fluorescence maximum is again observed on going from 4 to
3 and on to 5, together with a decrease in the relative fluo-
rescence intensity in the same order. In addition, while the
excited state decay of 5 was strictly mono-exponential (t =
4ns), the decay of 3 and 4 can be conveniently fitted only if
2752
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2749 2758

Enantioselective Sensing of Amino Acids by Modified Cyclodextrins
2749 2758
in the excited state lifetime are observed, according to the
shielding of the cyclodextrin cavity[19] and in agreement with
the energy gap rule.[36]
From the presence of two different lifetimes, the overall
fluorescence bands observed for 3 and 4 can be interpreted
as a linear combination of two different bands: one with
higher energy and longer lifetime, originating from dansyl
units lying in a relatively apolar environment, and the other
at lower energy with shorter lifetime, originating from units
lying in a more polar environment.
The observed behaviour is thus consistent with a two-
state model for the dansylated cyclodextrins, as previously
described, with equilibrium between one conformation (out)
in which the dansyl group is outside the cavity (more polar
Figure 4. a) Fluorescence intensity decrease of 3 and 4 upon addition of
environment) and has a short fluorescence lifetime (t1), and
one (in) with self-inclusion of the dansyl moiety (t2, more
polar environment). The pre-exponential terms (a1 and a2),
corresponding to the molar fractions of these two conforma-
tions, were found to be different for the two cyclodextrins: 3
showed only 38% of the ™in∫ conformation, while 4 showed
69%. The difference in the normalized fluorescence intensi-
ties between the two cyclodextrins is thus due solely to the
different extent of self-inclusion of the dansyl group, which
is larger for 4. This conclusion finds further support in anal-
ysis of the excited state lifetime of 5, which increases from 4
to 16 ns in the presence of an excess of b-cyclodextrin, as a
result of its inclusion in the cavity of the host.
Moreover, the addition of a 50:1 excess of 1-adamantane-
carboxylic acid (ACA), which is known to form inclusion
complexes with b-cyclodextrin, induced decreases in the
fluorescence intensity for both 3 and 4 (Figure 3); after the
addition, both cyclodextrins showed the same fluorescence
maxima, suggesting that the inclusion of the guest had shift-
ed the equilibrium towards the ™out∫ for both 3 and 4.
Similar results had previously been described by Ueno
and co-workers for diastereomeric mono-functionalized b-
cyclodextrins containing dansyl-d- or -l-leucine:[22] the latter
was found to have a larger fraction (a = 0.77) of the isomer
showing a long excited state lifetime than the former (a =
0.67), indicating that the chirality of the leucine residue
could bias the conformational–and hence the molecular
recognition and sensing–properties of these molecules. In
the present case the difference in the distributions of the
two conformations is much larger, suggesting that the self-
inclusion of the dansyl moiety is more affected by the con-
figuration of the phenylalanine linker.
CuSO4 in 1:1 ratio, b) titration of 3 (&) and 4 ( ) with Cu(EDTA)2ꢀ com-
~
plex at pH 7.0. Concentration: 6î10ꢀ5 m, lex = 345 nm, lem = 516 nm.
Vertical bars indicate standard deviations.
as for the free 3 and 4, indicating that they are due to un-
complexed ligand (Table 1). Since direct titration with
excess of Cu2+ ion was unsuccessful, due to precipitation,
competition experiments were performed with Cu(EDTA)2ꢀ
as titrant (Figure 4b).
From these data the conditional equilibrium constants (K)
for the equilibrium (1) at pH 7.0 could be calculated (K(3)
= 0.020, s.d. = 0.002 and K(4) = 0.013, s.d. = 0.001). If it
is assumed that no tertiary complexes are formed in these
experiments, the stability of Cu-3 can be gauged higher than
that of Cu-4. This effect could be due to a lower degree of
flexibility in 4 due to the inclusion of the dansyl group
within the cavity.
CuðEDTAÞ2ꢀ þ 3 ðor 4Þ Ð Cu-3 ðor Cu-4Þ þ EDTA þ n Hþ
ð1Þ
Enantioselective recognition of amino acids: The addition of
amino acids to the copper(ii) complexes of 3 or 4 in a 1:1
ratio caused increases in the fluorescence. These were de-
pendent on the type of amino acid used and, for some
amino acids, on the absolute configuration.[29]
The enantioselectivity displayed by the copper(ii) com-
plexes of cyclodextrins 3 and 4 was evaluated in terms of re-
producibility, in order to ascertain whether and under which
conditions the enantioselective effect might be useful for an-
alytical applications. Three samples were analysed for each
amino acid, to assess the significance of enantioselectivity
observed at three different amino acid/copper(ii) complex
molar ratios (1:1, 2:1 and 10:1). The results obtained were
compared by use of Student×s t-test to verify the significance
of the differences observed in fluorescence intensity
(Table 2 and Table 3).
The best enantioselectivity was observed for proline with
both cyclodextrins, and cyclodextrin 4 showed better enan-
tiomeric recognition than 3.
With 3, good enantioselectivity and high significance were
found only in the case of Pro for all the molar ratios utiliz-
ed; Val and Ser gave significant differences at low amino
The fluorescence properties of 5 in methanol/water (9:1)
solutions are very similar to those typical of dansyl deriva-
tives in methanol, and, as expected, its excited state decay
was again strictly mono-exponential (t = 13 ns).
Copper(ii) binding: Both cyclodextrins were shown to be
quenched by copper(ii) (Figure 4a); the quenching effects
upon addition of Cu2+ in a 1:1 ratio were found to be in the
same range for the two cyclodextrins.
It is important to note that in both cases a residual fluo-
rescence band can still be observed even upon addition of
one equivalent of metal ion. These bands have the same
shapes and the same lifetimes (with the same distribution)
2753
Chem. Eur. J. 2004, 10, 2749 2758
www.chemeurj.org
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

R. Corradini et al.
FULL PAPER
Table 2. Enantioselectivity factors (a = DFD/DFL) of Cu-3 (60 mm) to-
The good enantioselectivity obtained with proline for 4 al-
lowed us to perform quantitative measurements of the enan-
tiomeric purities of samples of intermediate composition by
fast fluorescence measurements, with an accuracy within
5%.[29]
The Cu-3 and Cu-4 complexes might therefore be usable
as model systems for the development of enantioselective
fluorescence chemosensors for unmodified amino acids for
use in analytical applications. It was thus of great interest to
understand the origin of the enantioselectivity and the sens-
ing mechanism involved.
wards amino acids (0.1m borate buffer, pH 7.3; lex = 345 nm, lem
=
516 nm).
a.a.
[a.a.]:[Cu-3] = 1
[a.a.]:[Cu-3] = 2
[a.a.]:[Cu-3] = 10
a. Signific. [%] a.
Signific. [%] a.
Signific. [%]
Pro
Ala
Val
Leu
His
Asp
Lys
Ser
Phe
Phgly 0.78
Trp
Tyr
0.77
0.98
1.31
1.46
1.05
1.11
1.39
0.57
1.01
99 0.7499 0.87
99
<20 0.85
99 1.20
60 0.97
60 1.00
80 1.02
80 1.13
99 0.91
<20 1.02
80 0.96
60 1.07
60 0.92
99 0.9480
99 1.0495
<20 1.39
<20 0.99
20 0.9450
99 1.03
99
20
20
95
<20
80
<20
20
90 1.02
<20 1.00
20 0.92
90 0.99
Mechanism of chiral discrimination: The observed enantio-
selectivity in the ™switching on∫ process of the fluorescence
of the two complexes can only be explained in terms of the
formation of ternary cyclodextrin/copper(ii)/amino acid
complexes. The formation of these species was confirmed by
ESI-MS measurements (see Supporting Information) per-
formed under pH conditions comparable to those used for
fluorescence experiments (pH 7.0). In the presence of 4,
copper(ii) ions, and proline enantiomers, we detected signals
for the ligand alone (as potassium adduct [LK]+), for the
copper(ii) complex ([CuLHꢀ2K2]2+ and [CuLHꢀ2K]+), and
of the ternary complex ([CuLHꢀ1(Pro)K]+ and [CuLHꢀ1-
(Pro)K2]2+). Therefore, both the amide and sulfonamide are
deprotonated at this pH in the binary complex, while the
ternary complex is mono-deprotonated, suggesting that the
bidentate amino acid has displaced the sulfonamide group
from the copper ion. The binary copper(ii)/amino acid com-
plex [Cu(Pro)2K]+ was also detected. Although the concen-
trations of the species were higher in these experiments
than in the fluorescence measurements, due to the lower
sensitivity of this technique, the formation of ternary com-
plexes can be clearly demonstrated by this experiments.
Further evidence of the formation of ternary complexes
and of their structures was provided by circular dichroism
spectra (Figure 5).
The formation of the copper complex is accompanied by
an increase in the signal intensity of the band at 350 nm,
which indicated a more rigid structure, with the dansyl
group equatorially oriented. A second band appeared at
280 nm, and was attributed to the deprotonated sulfonamide
group, on the basis of circular dichroism spectra of the cy-
clodextrin 2 copper complex. This band appears at the same
pH at which the deprotonation of the sulfonamide takes
place (results not shown). Upon addition of one molar
equivalent of d- or l-Pro, or of d- or l-Phe, the CD spectra
showed a positive band at 350 nm, more intense than that of
the free ligand and, in the case of l-Pro and d-Phe, of the
copper(ii) binary complex. At the same time, the band at
280 nm disappeared, indicating decomplexation of the sul-
fonamide group. The CD spectra therefore strongly support
the hypothesis of formation of ternary complexes.
1.19
1.49
40 1.06
Table 3. Enantioselectivity factors (a = DFD/DFL) of Cu-4 and Cu-5 to-
wards amino acids.
a.a.
[a.a.]/[Cu-4(or 5)]
= 1
[a.a.]/[Cu-4(or 5)]
= 2
[a.a.]/[Cu-4(or 5)]
= 10
a
Signific. [%]
a
Signific. [%]
a
Signific. [%]
Cu-4[a]
Pro
Ala
Val
Leu
His
Asp
Lys
Ser
Phe
Phgly
Trp
3.93
99 2.90
99 1.55
<20 0.99
99
40
1.0450 1.01
0.3499 0.50
0.31
0.9495 1.00
0.89
0.79
0.68
0.19
0.43
99 0.65
50 0.99
99
99 0.41
99 0.78
99
20
95 0.92
95 0.7499 0.98
99 0.77
99 0.58
99 0.70
95 0.9490
99
99 0.87
99 0.89
95 0.87
99 0.98
99 0.91
99
99
99
90
95
0.62
0.44
95 0.77
99 0.71
Tyr
Cu-5[b]
Pro
1.20
1.00
95 1.1499 0.97
<20 0.95
95
Phe
50 1.00
40
[a] 60 mm in 0.1m aqueous borate buffer, pH 7.3; lex = 345 nm, lem
516 nm. [b] 60 mm in MeOH/H2O 9:1, 0.02m borate buffer, pH 8.0; lex
335 nm, lem = 535 nm.
=
=
acid excess, while Leu gave a significant difference only at
10:1 excess. When cyclodextrin 4 was used instead, the
copper complex gave good enantioselectivities for Pro, Val,
Leu, Ser, Phe, and PhGly at all ratios, while Lys and Tyr
enantiomers were discriminated only at low molar excess.
It is worth noting that alanine, with a small aliphatic side
chain, produced negligible differences between the two
enantiomers in fluorescence response, while the larger ring
of proline and the aromatic moiety of phenylalanine in-
duced enantiomeric differentiation. Large differences were
also found in the case of tryptophan, but the large standard
deviations obtained for this amino acid did not allow enan-
tioselectivity to be assessed.
In order to understand the mechanism involved in the
enantioselective sensing, we investigated the excited state
lifetimes of the dansyl chromophore in the Cu/3/amino acid
(1:1:1) and Cu/4/amino acid (1:1:1) mixtures (Table 1). In
each case we found that the same two lifetimes with the
same pre-exponential terms as the free ligand were also
The linear ligand 5 was tested with Pro and Phe enan-
tiomers, and the results showed remarkably lower enantiose-
lectivity, thus indicating that the cyclodextrin cavity in 4
plays a role in determining the discriminating ability of the
ligand (Table 3).
2754
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2749 2758

Enantioselective Sensing of Amino Acids by Modified Cyclodextrins
2749 2758
is still effective in the ternary
complex, though the dansyl
group is displaced by the metal
ion, as indicated by its neutral
charge (observed in the ESI
spectrum) and by circular di-
chroism studies.
The observed sensing process
can therefore be interpreted in
terms of a simple model as
shown in Scheme 2, Equili-
bria (2) (4).
Equilibrium (2) is copper(ii)
complexation, and is dependent
on the conformation of the
ligand, being more effective for
the cyclodextrin 4, as shown by
the quenching experiments re-
ported in Figure 3. The ™switch
Figure 5. Circular dichroism spectra of: a) 4, b) Cu/4 (1:1), c) Cu/4/d-Pro (1:1:1), d) Cu/4/l-Pro (1:1:1), e) Cu/4/
d-Phe (1:1:1), f) Cu/4/l-Phe (1:1:1) in aqueous borate buffer, pH 7.0 at concentrations of 6î10ꢀ5 m.
present after the addition of the amino acid. From these re-
sults, we believe that the dansyl fluorescence is almost com-
pletely quenched in the ternary complex. It seems very un-
likely that each ternary complex could have the same photo-
physical properties and the same conformer distribution as
the parent free ligand. In addition, the copper ion in the ter-
nary complex lies very close to the chromophore, and for
this reason, it can still reasonably produce efficient energy-
or electron-transfer processes. In a previous study we dem-
onstrated that simple dansylated polyamine ligands could
bind copper(ii) over a wide pH range.[35] The fluorescence of
the dansyl group was quenched when copper(ii) complexes
were formed, as well as in species in which the dansyl group
was not coordinated to the metal ion. Therefore, a quench-
ing process of the dansyl group involving the copper(ii) ion
on∫ process is then composed of two equilibria: Equili-
brium (3) the formation of a non-fluorescent ternary com-
plex, and Equilibrium (4) the displacement of copper(ii) by
the amino acid to form a 1:2 complex with displacement of
the free fluorescent ligand. The enantioselectivity in the
fluorescence ™switching on∫ is therefore based on competi-
tion between the non-fluorescent ternary complex and the
binary copper(ii):amino acid complex, since K2(l)¼K2(d)
and K3(l)¼K3(d).
From this scheme, enantioselectivity could also be ob-
served in the reverse process reported in Equilibrium (4), by
titration of the cyclodextrin 4 with the 2:1 copper(ii) com-
plex of the amino acid. Figure 6 reports the quenching ob-
served during the titration experiment, which allows the dif-
ference in the stability constants of the ternary diastereo-
meric complexes formed to be evaluated.
According to this model, the high ratio of the association
constants for the two enantiomers (KSV(l)/KSV(d) = 2.36)
corresponds to the ratio between the overall stability con-
stants of the diastereomeric ternary complexes, since the
Figure 6. Stern Volmer plot obtained by quenching the fluorescence of 4
(6î10ꢀ5 m) with Cu(l-Pro)2 (upper line) or Cu(d-Pro)2 (lower line) in
0.1m borate buffer, pH 7.3.
Scheme 2. Equilibria involved in the sensing process and proposed cop-
per(ii) coordination in the various species.
2755
Chem. Eur. J. 2004, 10, 2749 2758
www.chemeurj.org
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

R. Corradini et al.
FULL PAPER
mined on a Gallenkamp melting point apparatus and are uncorrected.
TLC chromatography was performed on precoated silica gel SIL G/
UV254 aluminium support (0.20 mm layer; Macherey Nagel). HPLC chro-
matography was performed on a Waters HPLC Chromatograph, equip-
ped with a model 4000 pump and a model 2487 UV detector, set at
330 nm, with a C18 Spherisorb column (25î200 mm). IR spectra were re-
corded on a Nicolet 5PC FT-IR spectrometer, and 1H NMR and 13C
NMR were recorded on Bruker AC 300 or AMX 400 spectrometers. The
optical purities of 3, 4 and 5 were checked by GC/MS analysis on a Hew-
binary species have the same stability for the two enantiom-
ers.
The higher enantioselectivity observed in the case of 4
can be explained by taking the conformational properties of
the ligand into account: the higher degree of self-inclusion
of the dansyl group within the cyclodextrin cavity did not
affect the copper(ii) binding ability, suggesting that in the
copper(ii) complex the dansyl group retained the same con-
formation. A more rigid ternary complex could therefore be
formed, and the different orientation of the amino acid side
chain with respect to the cavity could give rise to different
interactions. The same effect could not be obtained with the
cyclodextrin 3, which has the dansyl group outside the
cavity, or by the linear ligand 5, which has the same configu-
ration as 4 but lacks the pre-organization induced by the cy-
clodextrin cavity.
lett Packard HP 6890 chromatograph fitted with
a Hewlett Packard
HP 5973 detector and a Chirasil-Val column. ESI-MS experiments were
performed on a Micromass ZMD mass spectrometer (Micromass, Man-
chester, UK), fitted with an electrospray source and a single quadrupole
mass analyser. Data were acquired by use of Masslynx 3.4software (Mi-
cromass, Manchester, UK). UV spectra were recorded on a Perkin
Elmer Lambda 40 spectrophotometer. CD spectra were obtained on a
Jasco 715A spectropolarimeter. Fluorescence spectra were recorded on a
Perkin Elmer Ls50B instrument in a 1î0.2 cm quartz cell thermostatted
at 258C. The fluorescence lifetimes (uncertainty Æ5%) were obtained
with an Edinburgh single-photon counting apparatus (D2-filled flash
lamp). To allow comparison of emission intensities, corrections were per-
formed for instrumental response, inner filter effects, and phototube sen-
sitivity. Elemental analysis were performed with a Carlo Erba CHNS-O
EA1108 elemental analyzer.
Conclusion
Na-[N2-(tert-Butoxycarbonyl)aminoethyl]-d(or l)-phenylalanine (d-6 or
Rationally designed enantioselective fluorescent sensors are
of great importance in the development of new analytical
tools for assessment of optical purity. Herein we have dem-
onstrated that the combination of a fluorophore, a metal
binding site and a lipophilic cavity gives rise to high enantio-
selectivity for some amino acids. The mechanism of sensing
can be attributed to competition between the ternary cop-
per(ii)/ligand/amino acid complex and the binary copper(ii)/
amino acid complex.
Our results indicated that a highly pre-organized structure
can be achieved by allowing self-inclusion of the fluoro-
phore in the cyclodextrin cavity, thus giving rise to more
rigid structures. The stereochemistry of the side arm used
was found to be important, in the present case, for determi-
nation of the conformational properties of ligands 3 and 4,
since they affect the ability to complex copper(ii) and to dis-
criminate between enantiomers of amino acids. Other selec-
tors designed with the same structural features are being
synthesized in order to obtain a pool of selectors for parallel
analysis.
The proposed method can be used for analytical purposes,
at least when fast screening procedures are required. Unlike
many other proposed methods for parallel analysis of enan-
tiomers, the present one is based not on kinetic enantiodis-
crimination, but on enantioselectivity in complex formation.
Therefore, only a simple mixing procedure, which can easily
be automated, is required. Parallel read-out devices can be
used for this purpose. Future development could be envis-
aged for chiral analysis of unmodified natural or synthetic
amino acids, providing new and efficient tools for screening
of synthetic processes or of enantioselective catalysts both
in traditional organic synthesis and in combinatorial chemis-
try.
l-6): These compounds were synthesized as described previously.[37]
Na-[N2-(5-Dimethylamino-1-naphthalenesulfonyl)aminoethyl]-d (or l)-
phenylalanine (d-8 and l-8): Anisole (0.49 mL, 4.52 mmol) and d-6 (or l-
6) (0.3 g, 0.75 mmol) were dissolved in CH2Cl2 (30 mL) and cooled to
08C. AlCl3 (0.6 g, 4.52 mmol) in CH3NO2 (15 mL) was then added with
stirring. The mixture was stirred at room temperature for 15 min, and the
reaction was stopped by addition of few mL of water. The solution was
extracted with CH2Cl2 to remove anisole, the pH of the aqueous phase
was adjusted to 8, and the formed Al(OH)3 precipitate was removed by
filtration. The filtrate was evaporated to dryness to yield a solid residue,
which was dissolved in a solution of Li2CO3 (0.17 g, 2.25 mmol) in water
(10 mL, pH 9.5), prepared immediately before use. Dansyl chloride
(0.20 g, 0.75 mmol) in CH3CN (10 mL) was added dropwise at 08C
(0.25 equiv per hour). Unreacted dansyl chloride was removed by extrac-
tion with diethyl ether. The product was precipitated after acidification
(pH 8) of the aqueous layer. Yield 69%. Rf = 0.55 (BuOH/H2O/acetic
acid 4:1:1).
Compound d-8: M.p. 200 2158C (decomp); [a]D20
=
ꢀ9.7 (c
= 1 in
water, pH 12); compound l-8: m.p. 200 2158C (decomp); [a]2D0 = +10.5
1
(c = 1 in water, pH 12); H NMR (300 MHz, D2O, 25 8C): d = 2.18 2.26
(m, 3H), 2.41 2.50 (m, 1H), 2.89 (s, 6H; N(CH3)2), 2.78 3.01 (m, 3H;
ꢀ
ꢀ
N CH, SO2N CH2), 6.97 7.01 (m, 2H; Ph), 7.24 7.30 (m, 3H; Ph), 7.43
(d, J(H,H) = 7.4Hz, 1H; H 6’), 7.68 (t, J(H,H) = 8.4Hz, 1H; H 3’), 7.75
3
3
3
(t, J(H,H) = 7.8 Hz, 1H; H7’), 8.14(d, 3J(H,H) = 7.1 Hz, 1H; H8’), 8.39
(d, 3J(H,H)
= = 8.6 Hz, 1H;
8.4Hz, 1H; H 2’), 8.69 (d, 3J(H,H)
H4’) ppm; 13C NMR (75 MHz, D2O, 25 8C): d = 42.08, 47.92, 48.29, 51.13,
68.51, 118.83, 124.61, 127.44, 129.69, 130.70, 130.93, 131.31, 131.68, 132.29,
132.35, 133.18, 141.41, 142.11, 153.55, 184.31 ppm; IR (KBr): n˜ =
3308 cmꢀ1 (N H), 1574cm
(C=O), 1323 cmꢀ1 (S=O); MS (C.I.): m/z
ꢀ1
ꢀ
(%): 442 (12) [M+H]+, 171 (57) [dimethylaminonaphthalene+H]+, 122
(100) [C8H12N]+ ; elemental analysis calcd (%) for C23H27N3O4S (441.54):
C 62.57, H 6.16, N 9.42; found: C 62.20, H 6.41, N 9.42.
6-Deoxy-6-N-(Na-[(5-dimethylamino-1-naphthalenesulfonyl)aminoethyl]-
phenylalanylamino-b-cyclodextrin (3, 4): HBTU (0.024g, 0.062 mmol)
and mono-6-deoxy-6-amino-b-cyclodextrin hydrochloride (0.066 g,
0.056 mmol) were dissolved in dry DMF (4mL) at 0 8C. Compound d-8
(or l-8; 0.027 g, 0.062 mmol) and DIEA (0.031 mL, 0.180 mmol) were
added in small portions every five minutes. The reaction was stopped
after three hours by addition of few mL of water. Water and DMF were
evaporated under vacuum at 408C. The obtained yellow solid was dis-
solved in water, and the insoluble residue, unreacted 8, was removed by
filtration. The product was precipitated in acetone and purified by HPLC
chromatography; the mobile phase was water (eluent A) and CH3CN/
TFA (0.1%) (eluent B), with a linear gradient from 100% of A to 100%
of B in 10 minutes with a flow rate of 20 mLminꢀ1; retention times of 3
Experimental Section
General: Starting materials were obtained from commercial suppliers
and were used without further purification. Melting points were deter-
2756
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2749 2758

Enantioselective Sensing of Amino Acids by Modified Cyclodextrins
2749 2758
or 4 were 8 8.5 minutes. The products were thus isolated and character-
ized as trifluoroacetate salts.
Fluorescence measurements: Concentrated stock solutions of 3 and 4
(1î10ꢀ3 m in water) and 5 (1î10ꢀ3 m in methanol) were prepared; solu-
tions of 3 and 4 were diluted to final concentrations of 6î10ꢀ5 m at
pH 7.3 in 0.1m borate buffer, the solution of 5 was diluted to a final con-
centration of 6î10ꢀ5 m at pH 8.0 methanol/water (9:1 v/v, 0.02m borate
buffer). In the titration experiments with copper(ii), 0.5 mL aliquots of
these solutions were titrated in the cell by addition of 6î10ꢀ3 m CuSO4 in
water with a 10 mL syringe. Three measurements were performed. The
Compound 3: Yield 22%. Rf = 0.56 (PrOH/H2O/NH3 5:3:1). M.p. 2068C
(decomp); ee = 95%; 1H NMR (300 MHz, D2O, 25 8C): d = 2.78 (m,
1H), 2.96 (m, 1H), 3.04(s, 6H; N(CH 3)2), 3.23 3.42 (m, 6H), 3.55 3.88
(m, 26H), 3.93 4.16 (m, 14H), 4.93 (d, 3J(H,H) = 3.4Hz, 1H; H 1), 5.05
3
3
(d, J(H,H) = 3.4Hz, 1H; H 1), 5.10 (d, J(H,H) = 3.4Hz, 1H; H 1), 5.13
3
3
(d, J(H,H) = 3.7 Hz, 1H; H1), 5.19 (d, J(H,H) = 3.4Hz, 2H; H 1), 5.23
same procedure was followed in the titrations of
3 and 4 with
3
(d, J(H,H) = 3.5 Hz, 1H; H1), 7.15 7.25 (m, 2H; Ph), 7.40 7.50 (m, 3H;
Cu(EDTA)2ꢀ (6î10ꢀ2 m in water), adamantanecarboxylic acid (ACA, 6î
10ꢀ2 m in methanol) and in the titration of 4 with Cu(Pro)2 (6î10ꢀ2 m in
water). Fluorescence intensities were corrected by a literature proce-
dure;[38] a futher correction of the fluorescence intensity of all the sam-
ples was made according to In = I/Ist, where I is the observed fluores-
cence intensity and Ist is the intensity of a reference solution of 4, both
measured at the same excitation and emission wavelength.
Ph), 7.57 (d, J(H,H) = 7.6 Hz, 1H; H6’), 7.84(t, 3J(H,H) = 8.6 Hz, 1H;
3
3
3
H3’), 7.87 (t, J(H,H) = 8.7 Hz, 1H; H7’), 8.40 (d, J(H,H) = 7.5 Hz, 1H;
3
3
H2’), 8.44 (d, J(H,H) = 8.7 Hz, 1H; H8’), 8.65 (d, J(H,H) = 8.6 Hz, 1H;
H4’) ppm; 13C NMR (75 MHz, D2O, 25 8C): d = 40.77, 43.56, 48.56, 49.10,
62.88, 63.23, 63.37, 63.53, 65.44, 74.49, 74.90, 74.98, 75.09, 75.56, 75.91,
76.20, 76.46, 76.63, 83.10, 83.53, 83.79, 84.29, 84.38, 85.97, 103.76, 104.76,
105.07, 105.31, 119.17, 122.35, 127.43, 130.82, 131.93, 132.27, 132.44,
Fluorescence intensities from the titration with Cu(EDTA)2ꢀ were used
according to the following model. By considering that [EDTA] = [Cu-3
(or Cu-4)] in equilibrium (1) (i.e., by neglecting the dissociation of the
Cu(EDTA) complex, due to its high stability) and by keeping the pH
buffered, Equation (5) could be obtained, where I is the corrected fluo-
rescence intensity, I0 is the fluorescence of each cyclodextrin alone, CCD
and CCu(EDTA) are the analytical concentrations of the cyclodextrin and of
133.11, 133.61, 137.08, 138.42, 154.53 ppm; IR (KBr): n˜ = 3650 3050 cmꢀ1
ꢀ1
(O H, N H), 1650 cm (C=O), 1150 cmꢀ1 (S=O); MS (ESI): m/z (%):
1558 (100) [M+H]+, 1579 (20) [M+Na]+, 1597 (26) [M+K]+ ; elemental
analysis calcd for C69H98N4O41SF6¥1H2O (1804): C 45.95, H 5.59, N 3.11;
found: C 45.87, H 5.28, N 3.23.
ꢀ
ꢀ
Compound 4: Yield 29%. Rf = 0.56 (PrOH/H2O/NH3 5:3:1); m.p. 2008C
(decomp); ee = 94%; 1H NMR (300 MHz, D2O, 25 8C): d = 3.10 4.05
(m, 53H), 4.15 (m, 1H), 4.33 (m, 1H), 5.00 5.20 (m, 7H; H1), 7.25 7.34
(m, 2H; Ph), 7.45 7.58 (m, 3H; Ph), 8.07 8.12 (m, 2H; H3’, H7’), 8.27 (d,
Cu(EDTA)2ꢀ respectively, and A=KCCu(EDTA)ꢀKCCD +2CCD
.
1=2
ꢀA þ ½A2 þ 4 ðKCCDꢀCCDÞCCD
ꢁ
3
3J(H,H) = 7.8 Hz, 1H; H6’), 8.46 (d, J(H,H) = 7.5 Hz, 1H; H8’), 8.64(d,
ð5Þ
I=I0
¼
2ðKCCDꢀCCD
Þ
3J(H,H) = 8.4Hz, 1H; H 2’), 8.93 ppm (d, 3J(H,H) = 8.7 Hz, 1H; H4’);
13C NMR (75 MHz, D2O, 25 8C): d = 39.70, 42.25, 45.22, 50.17, 63.51,
63.66, 63.83, 64.45, 74.79, 74.99, 75.16, 75.21, 75.35, 75.53, 75.77, 75.90,
76.31, 76.43, 76.48, 76.60, 84.15, 84.51, 84.57, 85.06, 85.47, 88.15, 104.57,
Nonlinear regressions of I/I0 values were performed with the aid of the
Sigmaplot program (18 data points for each cyclodextrin, from three in-
dependent measurements).
105.00, 105.37, 105.46, 105.51, 105.63, 119.84 (q, 1J(C,F)
=
289.7 Hz;
Titrations of Cu-3 and Cu-4 with amino acids were carried out with 6î
10ꢀ5 m solutions of the complexes in 0.1m borate buffer at pH 7. To
0.5 mL of these solutions were added aliquots of d- or l-amino acids (6î
10ꢀ2 m in water); each titration was repeated three times: mean values
and standard deviations were calculated. During each titration, the pH
was constant, as verified by a combined glass electrode.
CF3), 123.12, 129.17, 129.25, 129.86, 130.32, 131.69, 131.80, 132.06, 132.40,
132.47, 136.48, 141.97, 166.29 (q, 3J(C,F)
=
34.6 Hz; COCF3),
ꢀ
170.70 ppm; IR (KBr): n˜ = 3650 3050 cmꢀ1 (O H, N H), 1683 cm (C=
O), 1150 cmꢀ1 (S=O); MS (ESI): m/z (%): 1558 (100) [M+H]+, 798 (65)
ꢀ1
ꢀ
[M+H+K]2+
, ; elemental analysis calcd for
779 (83) [M+2H]2+
C67H97N4O39SF3¥15H2O (1941): C 41.46, H 6.54, N 2.89; found: C 41.42,
Samples analyzed by time-resolved fluorescence were prepared in the
same way; amino acids were added in tenfold excess to the solutions of
Cu-3 and Cu-4.
H 6.82, N 3.03.
N1-Propyl-Na-[N2-(5-dimethylamino-1-naphthalenesulfonyl)aminoethyl]-
l-phenylalaninamide (5): HBTU (0.171 g, 0.451 mmol) and n-propyl-
amine hydrochloride (0.058 g, 0.607 mmol) were dissolved in dry DMF.
Compound l-8 (0.132 g, 0.300 mmol) and DIEA (0.235 mL, 1.350 mmol)
were added in small portions every five minutes over one hour. The reac-
tion was stopped by addition of few mL of water. Water and DMF were
evaporated under vacuum at 408C. The yellow solid obtained was dis-
solved in water at pH 2 and extracted with chloroform; the organic layer
was washed with basic water (pH 8), dried over MgSO4 and evaporated
to dryness. The product was purified by HPLC: the mobile phase was
MeOH/H2O 95:5 (v/v) with TFA (0.035%), flow rate = 18 mLminꢀ1: re-
tention time of 5 was 4.5 min. The product was isolated as a trifluoroace-
tate salt, dissolved in water at pH 8 and extracted in chloroform; it was
then treated with a HCl/methanol solution and precipitated with metha-
nol/diethyl ether.
UV and CD measurements: Samples for UV and CD measurements were
prepared as described for fluorescence measurements.
2D NMR spectroscopy: 2D spectra were performed at 400 MHz on a
Bruker AMX 400 spectrometer. Compounds 3 and 4 were dissolved in
D2O at concentrations of 10 mm and the pH was adjusted to 7 with
NaOD. The corresponding 1H chemical shifts observed in the 2D experi-
ments were different from those reported above in the characterization
part, due to different degrees of protonation; the signal assignments are
based on TOCSY and ROESY spectra. The mlevtp pulse program was
used for the TOCSY experiments, with different mixing times (30, 60 and
90 ms). The roesysh pulse program was used for the ROESY experi-
ments, with 250 ms mixing time.
ESI-MS measurements: Solutions for MS experiments were prepared by
diluting standard solutions of 3 or 4, CuSO4, l- and d-proline to final
concentrations of 5î10ꢀ4 m in water: solutions were basified to pH 7 by
addition of 0.1m KOH.
Compound 5: Yield 48%. Rf
= 0.30 (ethyl acetate/hexane 1:1, TFA
0.1%); m.p. 1688C; ee = 91%; 1H NMR (300 MHz, CD3OD, 258C): d
3
3
= 0.72 (t, J(H,H) = 7.4Hz, 3H; CH 3 propyl), 1.24 1.34 (m, J(H,H) =
7.4Hz, 3J(H,H)
7.2 Hz, 2H; CH2 propyl), 2.94 3.28 (m, 8H; CH2
propyl, CH2Ph, CH2N), 3.11 (s, 6H; N(CH3)2), 4.06 (dd, 3J(H,H)
ESI-MS spectra were recorded in positive ion mode by perfusion of the
=
solutions directly into the mass spectrometer at
a flow rate =
=
10 mLminꢀ1 (conditions: capillary voltage 3000 V, cone voltage 50 V, de-
solvation flow (N2) = 500 Lhꢀ1, nebulizer flow (N2) = 100 Lhꢀ1, desol-
vation temperature = 808C, source block temperature = 1508C, scan
range m/z 200 2000 Da, scan time 4s).
9.8 Hz, 3J(H,H) = 5.3 Hz, 1H; CH), 7.19 7.37 (m, 5H; Ph), 7.59 (d,
3J(H,H)
3J(H,H)
=
7.7 Hz, 1H; H6’), 7.69 7.75 (m, 2H; H3’, H7’), 8.30 (dd,
=
7.3 Hz, 4J(H,H) 1.1 Hz, 1H; H8’), 8.53 (d, 3J(H,H)
=
=
8.6 Hz, 1H; H2’), 8.60 (d, 3J(H,H) = 8.6 Hz, 1H; H4’) ppm; 13C NMR
(75 MHz, CD3OD, 258C): d = 11.88, 23.40, 37.99, 40.20, 42.68, 46.70,
47.71, 63.43, 118.31, 123.17, 125.83, 129.06, 129.63, 130.25, 130.41, 130.64,
130.82, 130.99, 131.19, 135.53, 136.68, 149.41, 169.03 ppm; IR (KBr): n˜ =
GC-MS measurements: The enantiomeric excesses of the ligands 3, 4 and
5 were checked by GC/MS analysis of the corresponding 4-N-trifluoroa-
cetyl-3-benzylpiperazine-2-ones.The ligands (1 2 mg) were hydrolysed at
1008C with HCl (6n, 2 mL) for 6 h to the corresponding N-2-aminoethyl-
phenylalanine, then suspended in CH2Cl2 (2 mL) and treated with tri-
fluoroacetic anhydride (0.3 mL) for 1 hour at 608C. After removal of
excess reagent by evaporation under vacuum, the samples were dissolved
in CH2Cl2 and injected into the GC (1 mL). Analysis were performed
3266 cmꢀ1 (N H), 3066 cm (C H), 2964cm (C H), 1683 cm (C=O),
ꢀ1
ꢀ1
ꢀ1
ꢀ
ꢀ
ꢀ
1566 cmꢀ1 (N H), 1337 cm (S=O), 1147 cmꢀ1 (S=O); MS (CI): m/z
ꢀ1
ꢀ
(%): 483 (100) [M]+, 396 (67) [Mꢀ(CONHPr)]+, 170 (35) [dimethylami-
nonaphthalene]+ ; elemental analysis calcd for C26H34N4O3S¥2HCl¥3= H2O
4
(569.07): C 54.87, H 6.64, N 9.85; found: C 54.59, H 6.27, N 9.53.
2757
Chem. Eur. J. 2004, 10, 2749 2758
www.chemeurj.org
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

R. Corradini et al.
FULL PAPER
with a chiral capillary column: Chirasil-Val, 10 m, I.D. = 0.25 mm, film
thickness: 0.12 mm. carrier: He. Flow: 1.1 mLminꢀ1. injector temperature:
2308C. detector temperature: 2308C.; isotherm 1808C, detection: MS,
SIM mode (m/z 91, 286, 167).
[13] F. Vogtle, P. Knops, Angew. Chem. 1991, 103, 972 974; Angew.
Chem. Int. Ed. Engl. 1991, 30, 958 959.
[14] L. Labadini, R. Marchelli, A. Dossena, G. Palla, G. Mori, C. Bocchi,
M. Pilone, S. Butto in Current Status andFuture Trends in Analytical
FoodChemistry, Volume 1 (Eds: G. Somitag, W. Pfannhauser), Aus-
trian Chem. Soc., Wien, 1995, pp. 127 130 .
[15] K. Bodenhoefer, A. Hierlemann, J. Seemann, G. Gauglitz, B. Kop-
penhoefer, W. Gopel, Nature 1997, 387, 577 580.
Acknowledgments
[16] Y. Kubo, S. Maeda, S. Tokita, M. Kubo, Nature 1996, 382, 522 524.
[17] a) T. D. James, K. R. A. S. Sandanayake, S. Shinkai, Nature 1995,
374, 345 347; b) L. Prodi, F. Bolletta, M. Montalti, N. Zaccheroni,
P. Huszthy, E. Samu, B. Vermes, New J. Chem. 2000, 24, 781 785;
c) G. Beer, K. Rurack, J. Daub, Chem. Commun. 2001, 1138 1139.
[18] J. Lin, Q. -S. Hu, M. -H. Hua, L. Pu, J. Am Chem. Soc. 2002, 124,
2088 2089.
This work was supported by the Ministero dell’Istruzione, dell’Universit‡
e della Ricerca (MIUR) (™Separation Engineering Based on Molecular
and Chiral Recognition∫, ™Supramolecular Devices∫ and ™Solid Super-
molecules∫ projects). NMR and Mass Spectra (C.I.) were recorded at the
Centro Interfacolt‡ Misure of the Universit‡ di Parma. We thank Alessia
Franciosi for her technical assistance.
[19] A. Ueno, Adv. Mater. 1993, 5, 132 134.
[20] a) K. Hamasaki, S. Usui, H. Ikeda, T. Ikeda, A. Ueno, Supramol.
Chem. 1997, 8, 125 135; b) A. Ueno, I. Suzuki, T. Osa, Anal. Chem.
1990, 62, 2461 2466; c) A. Ueno, F. Toda, I. Suzuki, T. Osa, J. Am.
Chem. Soc. 1993, 115, 5035 5040; U. Narang, R. A. Dunbar, F. V.
Bright, P. N. Prasad, Appl. Spectrosc. 1993, 47, 1700 1703.
[21] a) G. Wenz, Angew. Chem. 1994, 106, 851; Angew. Chem. Int. Ed.
Engl. 1994, 33, 803 822; b) M. V. Rekharsky, Y. Inoue, Chem. Rev.
1998, 98, 1875 1917.
[22] H. Ikeda, M. Nakamura, N. Ise, N. Oguma, A. Nakamura, T. Ikeda,
F. Toda, A. Ueno, J. Am. Chem. Soc. 1996, 118, 10980 10988.
[23] Y. Liu, B.-H. Han, S.-X. Sun, T. Wada, Y. Inoue, J. Org. Chem. 1999,
64, 1487 1493.
[1] Fluorescent Chemosensors for Ion andMolecule Recognition (Ed.:
A. W. Czarnik), ACS Symposium Series 538, American Chemical
Society, Washington, DC, 1992.
[2] a) A. P. de Silva, H. Q. N. Gunaratne, T. Gunnlaugsson, A. J. M.
Huxley, C. P. McCoy, J. T. Rademacher, T. E. Rice, Chem. Rev. 1997,
97, 1515 1566; b) L. Prodi, F. Bolletta, M. Montalti, N. Zaccheroni,
Coord. Chem. Rev. 2000, 205, 59 83.
[3] J. Janata, Anal. Chem. 1992, 64, 33R-44R.
[4] A. P. de Silva, H. Q. N. Gunaratne, C. P. McCoy, Nature 1993, 364,
42 44.
[5] a) M. E. Huston, K. W. Haider, A. W. Czarnik, J. Am. Chem. Soc.
1988, 110, 4460 4462; b) A. J. Bryan, A. P. de Silva, S. A. de Silva,
D. D. Rupasinghe, S. Sandanayake, Biosensors 1989, 4, 169 179;
c) L. Fabrizzi, A. Poggi, Chem. Soc. Rev. 1995, 24, 197 202; d) E.
Kimura, T. Koike, Chem. Soc. Rev. 1998, 27, 179 184; e) R.
Kramer, Angew Chem. 1998, 110, 804 806; Angew. Chem. Int. Ed.
Engl. 1998, 37, 772 773.
[6] a) L. Prodi, F. Bolletta, M. Montalti, N. Zaccheroni, J. S. Bradshaw,
P. B. Savage, R. M. Izatt, Tetrahedron Lett. 1998, 39, 5451 5454;
b) L. Prodi, C. Bargossi, M. Montalti, N. Zaccheroni, N. Su, J. S.
Bradshaw, R. M. Izatt, P. B. Savage, J. Am. Chem. Soc. 2000, 122,
6769 6770; c) L. Prodi, M. Montalti, N. Zaccheroni, J. S. Bradshaw,
R. M. Izatt, P. B. Savage, Tetrahedron Lett. 2001, 42, 2941 2944.
[7] a) G. Grynkiewicz, M. Poenie, R. Y. Tsien, J. Biol. Chem. 1985, 260,
3440 3450; b) R. Yuste, A. Peinado, L. C. Katz, Science 1992, 257,
665 669; c) K. Kahn, R. De Biaso, D. L. Taylor, Nature 1992, 359,
736 738.
[24] G. Impellizzeri, G. Maccarrone, E. Rizzarelli, G. Vecchio, R. Corra-
dini, R. Marchelli, Angew Chem. 1991, 103, 1363 1364; Angew.
Chem. Int. Ed. Engl. 1991, 30, 1348 1349.
[25] R. Corradini, A. Dossena, G. Impellizzeri, G. Maccarrone, R. Mar-
chelli, E. Rizzarelli, G. Sartor, G. Vecchio, J. Am. Chem. Soc. 1994,
116, 10267 10274.
[26] R. Corradini, A. Dossena, R. Marchelli, A. Panagia, G. Sartor, M.
Saviano, A. Lombardi, V. Pavone, Chem. Eur. J. 1996, 2, 373 381.
[27] R. Corradini, A. Dossena, G. Galaverna, R. Marchelli, A. Panagia,
G. Sartor, J. Org. Chem. 1997, 62, 6283 6289.
[28] MS/ESI experiments showed the presence of the 2/Cu2+/Trp ternary
complex.
[29] S. Pagliari, R. Corradini, G. Galaverna, S. Sforza, A. Dossena, R.
Marchelli, Tetrahedron Lett. 2000, 41, 3691 3695.
[30] V. A. Davankov, J. D. Navratil, H. F. Walton LigandExchange Chro-
matography, Boca Raton (Florida), CRC Press, 1988.
[31] a) R. Marchelli, R. Corradini, T. Bertuzzi, G. Galaverna, A. Dosse-
na, F. Gasparrini, B. Galli, C. Villani, D. Misiti, Chirality 1996, 8,
452 461; b) G. Galaverna, R. Corradini, E. De Munari, A. Dossena,
R. Marchelli, J. Chromatogr. 1993, 657, 43 54.
[32] G. Haaima, A. Lohse, O. Buchardt, P. E. Nielsen, Angew. Chem.
1996, 108, 2068 2070; Angew. Chem. Int. Ed. Engl. 1996, 35, 1939
1942.
[33] R. Corradini, G. Di Silvestro, S. Sforza, G. Palla, A. Dossena, P. E.
Nielsen, R. Marchelli, Tetrahedron: Asymmetry 1999, 10, 2063
2066.
[34] a) K. Harata, H. Uedaira, Bull. Chem. Soc. Jpn. 1975, 48, 375 378;
b) M. Kodaka, J. Am. Chem. Soc. 1993, 115, 3702 3705.
[35] Y.-H. Li, L.-M. Chan, L. Tyer, R. T. Moody, C. M. Himel, D. M.
Hercules, J. Am. Chem. Soc. 1975, 97, 3118 3126.
[36] L. Prodi, M. Montalti, N. Zaccheroni, F. Dallavalle, G. Folesani, M.
Lanfranchi, R. Corradini, S. Pagliari, R. Marchelli, Helv. Chim. Acta
2001, 84, 690 706.
[37] A. P¸schl, S. Sforza, G. Haaima, O. Dahl, P. E. Nielsen, Tetrahedron
Lett. 1998, 39, 4707 4710.
[8] D. Y. Sasaki, D. R. Shnek, D. W. Pack, F. H. Arnold, Angew Chem.
1995, 107, 994 996; Angew Chem. Int. Ed. Engl. 1995, 34, 905 907.
[9] V. Dujols, F. Ford, A. W. Czarnik, J. Am. Chem. Soc. 1997, 119,
7386 7387.
[10] a) A. W. Czarnik, Acc. Chem. Res. 1994, 27, 302 308, and references
therein.; b) G. De Santis, L. Fabrizzi, M. Licchelli, A. Poggi, A. Ta-
glietti, Angew. Chem. 1996, 108, 224 227; Angew. Chem. Int. Ed.
Engl. 1996, 35, 202 204; c) A. Metzger, E. V. Anslyn, Angew.
Chem. 1998, 110, 682 684; Angew Chem. Int. Ed. Engl. 1998, 37,
649 652; d) M. Montalti, L. Prodi, N. Zaccheroni, L. J. Charbon-
niõre, L. Douce, R. Ziessel, J. Am. Chem. Soc. 2001, 123, 12694
12695; e) L. J. Charbonniõre, R. Ziessel, M. Montalti, L. Prodi, N.
Zaccheroni, C. Boehme, G. Wipff, J. Am. Chem. Soc. 2002, 124,
7779 7788.
[11] a) S. R. Adams, A. T. Harootunian, Y. J. Buechler, S. S. Taylor, R. Y.
Tsien, Nature 1991, 349, 694 697; b) T. D. James, P. Linnane, S.
Shinkai, Chem. Commun. 1996, 281 288; c) A. P. de Silva, H. Q. N.
Gunaratne, C. McVeigh, G. E. M. Maguire, P. R. S. Maxwell, E.
O’Hanlon, Chem. Commun. 1996, 2191; d) C. Chen, H. Wagner,
W. C. Still, Science 1998, 279, 851 853.
[38] A. Credi, L. Prodi, Spectrochim. Acta 1998, 54, 159 170.
[12] a) J. J. Lavigne, S. Savoy, M. B. Clevenger, J. E. Ritchie, B. McDo-
niel, S.-J. Yoo, E. V. Anslyn, J. T. McDevitt, J. B. Shear, D. Neikirk,
J. Am. Chem. Soc. 1998, 120, 6429 6430; b) Z. Zhong, E. V. Anslyn,
J. Am. Chem. Soc. 2002, 124, 9014 9015; c) L. Prodi, F. Bolletta, N.
Zaccheroni, C. I. F. Watt, N. J. Mooney, Chem. Eur. J. 1998, 4, 1090
1094.
Received: August 13, 2003
Revised: January 13, 2004
Published online: April 16, 2004
2758
¹ 2004Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2749 2758