Amphiphilic chiral receptor as efficient chiral solvating agent for both lipophilic and hydrophilic carboxylic acids

Article abstract of DOI:10.1016/j.tetlet.2008.03.113

Two amphiphilic chiral receptors 2a and 2b were designed and synthesized. Both are efficient chiral solvating agents for chiral carboxylic acids. In particular, 2a is an excellent CSA not only for lipophilic guests, but also for some hydrophilic guests. I

Full text of DOI:10.1016/j.tetlet.2008.03.113

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 3385–3390
Amphiphilic chiral receptor as efficient chiral solvating
agent for both lipophilic and hydrophilic carboxylic acids
Zengwei Luo, Cheng Zhong, Xiaojun Wu, Enqin Fu *
Hubei Key Laboratory on Organic and Polymeric Optoelectronic Materials, Department of Chemistry, Wuhan University, Wuhan 430072, PR China
Received 3 March 2008; accepted 25 March 2008
Available online 28 March 2008
Abstract
Two amphiphilic chiral receptors 2a and 2b were designed and synthesized. Both are efficient chiral solvating agents for chiral
carboxylic acids. In particular, 2a is an excellent CSA not only for lipophilic guests, but also for some hydrophilic guests. It is the first
CSA for the direct determination of the enantiomeric composition of hydrophilic chiral hydroxylated acid in protic polar solvent.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: Chiral solvating agent; Chiral recognition; High enantioselectivity; Amphiphilic receptor
Chiral carboxylic acids are basic building blocks of nat-
ural products and drug molecules as well as versatile func-
tional synthons.1–3 Due to their importance, there is a
growing demand for the development of fast and accurate
methods for determining enantiomeric excess (ee) and
assigning the absolute configuration (AC) of these chiral
compounds. Apart from other analytical methods such as
chromatography4 and capillary electrophoresis,5 using chi-
ral solvating agents (CSAs) on NMR spectroscopy is a sat-
isfactory and convenient method to meet this demand,6–8
and the method can also provide with direct structural
and dynamic information of host–guest complexes in solu-
tion. Although the development of new CSAs continues to
be an active area of research, and various synthetic recep-
tors have been reported in the previous literatures,9–18 the
CSAs that can lead to clear baseline separation of the
multiplet of the probe group in two enantiomers are rare.
In addition, it has been known that almost all reported
CSAs for carboxylic acids could only be used in the less
polar solvents, such as deuterated chloroform and benzene
so far. This can be ascribed to two reasons. First, most
CSAs are only soluble in the less polar solvents due to
the existence of the large aromatic groups in these mole-
cules. Second, the noncovalent interactions between host
and guest molecules for molecular recognition would be
weakened remarkably by the competitive influence of prot-
ic solvents. So the discrimination toward hydrophilic guests
could not be carried out by using these CSAs. That is why
the hydrophilic compounds have to be changed to their
lipophilic derivatives prior to discrimination. In fact, most
of the bioactive compounds are chiral and hydrophilic and
all of the recognition events in nature take place in aqueous
medium. So the development of new CSAs which can be
used in protic solvents, especially in water, is very impor-
tant not only for the analysis of bioactive compounds but
also for a better understanding and control of the major
processes in nature. Therefore the design of the synthetic
receptors that can be used as CSAs in protic solvents pre-
sents special challenge.
In our previous work,18 the CSAs (1a, 1b, and 1c)
(Fig. 1) have shown excellent enantioselective recognition
ability for some chiral carboxylic acids. However, the acids
suitable as the guests for discrimination are limited to lipo-
philic, and the enantioselective discrimination can be
achieved only in the less polar solvents.
*
In this Letter, we report the design, synthesis, and
property of new amphiphilic chiral solvating agents 2a
Corresponding author. Tel.: +86 27 87219044.
E-mail address: fueq@whu.edu.cn (E. Fu).
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.03.113

3386
Z. Luo et al. / Tetrahedron Letters 49 (2008) 3385–3390
to overcome the competitive influence of protic solvents
to molecular recognition.
The investigation reveals that chiral receptors 2a and 2b
are amphiphilic, and they are effective chiral solvating
agents. In particular, 2a is an excellent CSA not only for
lipophilic guests, but also for some hydrophilic guests, such
as tartaric acid and lactic acid. To the best of our knowl-
edge, 2a is the first receptor which can be used as CSA
for the direct determination of the enantiomeric composi-
tion of hydrophilic chiral hydroxylated acid in protic polar
solvent. Therefore it may be applicable to bioanalytical
problems.
The synthetic route for compounds 2a and 2b is outlined
in Scheme 1. (S)-2-(Boc-amino)-3-phenylpropyl bromide or
(S)-2-(Boc-amino)-4-methylpentyl bromide19 was reacted,
respectively, with 1 equiv of prolinol in the presence of
K2CO3 in refluxing acetonitrile affording 4a and 4b. The
treatment of 4a or 4b with TFA in CH2Cl2 furnished chiral
amino alcohols 3a and 3b in 43.2% and 35.7% yield, respec-
tively. Then 2a and 2b were obtained by condensation of 3a
or 3b and 1,8-naphthalic anhydride in refluxing toluene
with Dean–Stark trap, the yield was 21.3% and 31.5%,
respectively. The structures of new chiral receptors 2a
Fig. 1. (1R,2R)-1-(10,80-Naphthalimide)-2-aminocyclohexane and its
40-derivatives.
and 2b. NMR studies demonstrated that they were effective
chiral solvating agents. In particular, 2a not only leads to
clear baseline separation of the multiplet of the probe
groups in two enantiomers, but also is versatile for a wide
range of chiral carboxylic acids including both lipophilic
and hydrophilic carboxylic acids.
In our approach, chiral amino alcohols (1S,2S)-N-(1-(2-
amino-3-phenylpropyl) pyrrolidin-2-yl) methanol (3a) and
1
and 2b were characterized by H NMR, 13C NMR, IR,
(1S,2S)-N-(1-(2-amino-4-methylpentyl)
pyrrolidin-2-yl)
MS, and EA, respectively.
1H NMR spectroscopy was utilized to investigate the
enantiomeric discriminating ability of 2a and 2b. We chose
a broad variety of racemic chiral carboxylic acids as the
guests, including lipophilic and hydrophilic carboxylic
acids, which were mandelic acid 5 and some of its deriva-
tives 6–9, dibenzoyltartaric acid 10, lactic acid 11, tartaric
acid 12, and some N-protected L-amino acids, such as
p-tolysulfonyl alanine (Ts-alanine) 13, p-tolylsulfonyl valine
(Ts-valine) 14, p-nitrobenzenesulfonyl alanine (p-NBS-ala-
nine) 15, and 3,5-dinitrobenzoyl valine (DNB-valine) 16.
The structures of all guests were shown in Figure 2.
Table 1 summarized the chemical shift nonequivalences
(DDd) of CH3, CH, and NHTs of these guests in the
methanol (3b) instead of (1R,2R)-1,2-diaminocyclohexane
were used to react with 1,8-naphthalic anhydride to get
new chiral receptors 2a and 2b. Comparing with 1a, 1b,
and 1c, the new chiral receptors 2a and 2b have one more
functional group (primary hydroxyl group) and better flex-
ibility. It was expected that the increased hydrophilic group
might make them amphiphilic, which is important for them
to be used as the CSAs in many test solvents, such as deu-
terated methanol, ethanol, acetonitrile, acetone, chloro-
form, and ethyl acetate. The existence of primary
hydroxyl group and high flexibility would help 2a and 2b
form more stable diastereoisomeric complexes through
multiple weak, noncovalent interactions with the guests
Scheme 1. The synthesis route of compounds 2a and 2b.

Z. Luo et al. / Tetrahedron Letters 49 (2008) 3385–3390
3387
Fig. 2. The structure of guests used herein.
presence of compounds 2a and 2b, respectively. As shown
in Table 1, in the presence of receptors 2a and 2b, there
were large enough chemical shift nonequivalences to give
baseline resolution of appropriate proton signals for
almost all the chosen carboxylic acids measured on a
300 MHz NMR instrument. In the presence of compound
2a, the DDd of the methine proton of 1-methoxy-phenyl-
acetic acid (MPAA) (6) was up to 84.3 Hz (see Fig. 3),
and the DDd of the methine proton of DNB-valine 16
was up to 74.1 Hz. The clear resolution of the multiplet
peaks of the methine proton could be observed (see
Fig. 4). Furthermore, as what we expected, the recognition
of water soluble lactic acid 11 can also be achieved per-
fectly in CDCl3 when 2a is used as CSA. It means that
the formation of stable host–guest diastereoisomeric com-
plexes has facilitated the guest compound resolving in the
test solvent. It is remarkable that even the enantiomers of
racemic tartaric acid 12 can also be discerned clearly in
the presence of receptor 2a, the DDd is 18.9 Hz in
CD3COCD3 and 6.9 Hz in CD3OD, respectively, (shown
in Fig. 4), while in the same condition, 2b did not show
any discriminating ability toward the enantiomers of 12.
As far as we know, 2a is the first receptor, which can be
used as CSA for the direct determination of the enantio-
meric composition of chiral hydroxylated acid in protic
polar solvent. All the results reveal that chiral receptors
2a and 2b are excellent chiral solvating agents. Compared
with 2b, 2a shows applicability toward a wider range of
guests and in a broad variety of solvents.
Table 1
Chemical shift nonequivalences (DDd, 300 MHz) for 1:1 diastereoisomeric
complexes of 2a and 2b with racemic guests in CDCl3 at 25 °C
Guest
Probe group
Dd (Hz)
DDd (Hz)
2a
2b
2a
2b
5
6
CH
ꢁ0.575
ꢁ0.706
ꢁ0.235
ꢁ0.516
ꢁ0.105
ꢁ0.183
ꢁ1.010
ꢁ1.062
ꢁ0.639
ꢁ0.817
ꢁ0.681
ꢁ0.747
ꢁ0.461
ꢁ0.524
ꢁ0.256
ꢁ0.421
0.119
ꢁ0.830
ꢁ0.888
ꢁ0.351
ꢁ0.417
ꢁ0.200
ꢁ0.245
ꢁ1.047
ꢁ1.119
ꢁ0.972
ꢁ1.037
ꢁ0.889
ꢁ0.975
ꢁ0.527
ꢁ0.616
39.3
84.3
23.4
15.6
53.4
45.6
18.9a
49.5
6.9
17.4
19.8
13.5
21.6
19.5
25.8
CH
CH3
CH
7
8
CH
9
CH
10
11
12c
12d
13
CH
26.7a
b
b
CH3
CH
—
—
0.124
0.215
0
0
0.096
0.344
0.152
ꢁ0.280
ꢁ0.410
0.466
CH
18.9
39.0
72.0
44.7
69.6
33.0
74.1
The 1H NMR spectra of hosts 2a and 2b with some opti-
cal pure guests in a variety of ratios in CDCl3 at a constant
total concentration of 3.0 ꢀ 10ꢁ3 M were obtained. The
stoichiometric ratio of the host–guest complexes was
determined according to the Job’s method of continuous
variations.20 The Job plots of 2a with (R)-mandelic acid
and (S)-mandelic acid were illustrated in Figure 5, showing
a minimum of DdX at X = 0.5, which indicated that a 1:1
instantaneous complex was formed. The Job plots we have
got indicated that hosts 2a and 2b form 1:1 instantaneous
complexes with all these enantiomerically pure guests,
respectively.
CH3
NHTs
CH3
NHTs
CH3
CH(CH3)2
ꢁ0.643
ꢁ0.711
0.08
ꢁ0.286
ꢁ0.413
ꢁ0.602
0.135
20.4
109.8
62.7
97.8
0.226
14
ꢁ0.25
ꢁ0.399
0.243
ꢁ0.011
ꢁ0.349
ꢁ0.596
ꢁ0.28
ꢁ0.191
b
b
15
16
a
—
—
b
b
—
—
ꢁ0.39
The molar ratio of CSA and guest is 2:1.
b
The peaks of hosts overlapped with the peaks of the probe groups of
guests.
We also obtained the titration curves of hosts 2a and 2b
with these enantiomerically pure guests, respectively.
The association constants of the above complexes were
c
The deuterated solvent is CD3OD.
The deuterated solvent is CD3OCD3.
d

3388
Z. Luo et al. / Tetrahedron Letters 49 (2008) 3385–3390
Fig. 3. 1H NMR spectra of equimolar mixtures (20 mM each) of MPAA/compound 2a. (a) (R)-MPAA and compound 2a; (b) (S)-MPAA and compound
2a; (c) (R)/(S)-MPAA and compound 2a; (d) (R)/(S)-MPAA without compound 2a.
Fig. 4. 1H NMR (300 Hz, CDCl3, 25 °C, ppm referred to TMS as external standard) spectral regions corresponding to CH proton absorptions of racemic
guest (20 mM): (a) equimolar mixture 2a/DNB-valine, (b) free racemic DNB-valine, (c) mixture 2a/tartaric (2:1) in CD3OD, (d) mixture 2a/tartaric (2:1) in
CD3OCD3 ꢂProbe signal.
determined from the titration curves by a nonlinear least-
squares fitting method (Table 2). It showed that host 2b
bound guests stronger than host 2a did for its flexibility,
and the (S)/(D)-enantiomer was more strongly bound to
2a or 2b than the (R)/(L)-enantiomer. Particularly, the ratio
between the association constant of the complex formed
from host 2a and (S)-mandelic acid and that from host
2a and (R)-mandelic acid (Ka(S)/Ka(R)) exceeds 10.
important role for the formation of host–guest complex.
Host 2a may bind in situ to each enantiomer of mandelic
acid through noncovalent intermolecular forces as the
major driving force of molecular recognition. The possible
interaction models (Fig. 6) between host 2a and each enan-
tiomer of mandelic acid were established through theoreti-
cal calculation based on PM3 of MOPAC 2007 software. The
most stable forms are shown in Figure 6.
From the data in Table 2, it is clear that mandelic acid
exhibited much stronger bonding ability toward chiral
hosts 2a and 2b than the other guests did. The results imply
that the hydroxyl group in mandelic acid may play an
From Figure 6 we can see that the ionpairing interaction
formed from the carboxylic group in (R) or (S)-mandelic
acid and the tertiary amine in host 2a is a major force
for the formation of host–guest complexes. This is well

Z. Luo et al. / Tetrahedron Letters 49 (2008) 3385–3390
3389
strong ionpairing interaction between host and guest mol-
ecules. In addition, the hydrogen-bonding between the car-
boxyl group in the guest molecule and the primary
hydroxyl group in host 2a, as well as the p–p interaction
between host and guest molecules, is also important forces
for achieving cooperative binding. From Figure 6 we found
that there was another hydrogen-bonding between the car-
boxyl and a-hydroxyl groups in mandelic acid, which may
be helpful to facilitate the formation of ionpairing interac-
tion. The interaction models may explain why hosts 2a and
2b can facilitate lactic acid resolving in CDCl3 for the test.
It is known that the stability of the host–guest complexes
could be weakened remarkably in protic polar solvent.
The clear discerning of the enantiomers of tartaric acid
12 can be achieved even in protic polar solvent (CD3OD),
which means the formation of the stable host–guest com-
plexes from the enantiomers of 12 and 2a in the test condi-
tion. The possible reason is that the more functional groups
(one more hydroxyl and one more carboxyl groups) and
the flexibility of tartaric acid may be suitable for it to form
more stable host–guest complex with 2a through multiple
noncovalent interactions, decreasing efficiently the influ-
ence of protic polar solvent to molecular recognition.
Although both 2a and 2b are amphiphilic chiral receptors
and all can form stable host–guest complexes with tartaric
acid, but only 2a can discriminate the enantiomers of it effi-
ciently. It may be due to the presence and the suitable posi-
tion of the phenyl group in 2a, in the two complexes
formed, respectively, from the two enantiomers of 12 and
2a, the two corresponding probe groups of the guests
may locate at different points in space around the phenyl
group of the host molecule, and the anisotropic effect of
the phenyl group would play an important role for chemi-
cal shift nonequivalence. As to 2b as the CSA, the corre-
sponding probe groups in the two complexes formed,
respectively, from the two enantiomers of 12 and 2b may
locate far from the naphthalene ring and carbonyl groups
of host 2b, so that the anisotropic effect of the naphthalene
ring and carbonyl groups could not play any role for chem-
ical shift nonequivalence. When mandelic acid was used as
the guest, which brings an phenyl group, the corresponding
probe groups of the two enantiomers may locate at differ-
ent points in space near the naphthalene ring or carbonyl
(R)-mandelic acid (-CH)
(S)-mandelic acid (-CH)
0.00
-0.05
-0.10
-0.15
-0.20
-0.25
-0.30
-0.35
-0.40
0.0
0.2
0.4
0.6
0.8
1.0
X
Fig. 5. Job plots of 2a with (R)- and (S)-mandelic acid [X = molar
fraction of mandelic acid, Dd = chemical shift change of the methine of
(R)- and (S)-mandelic acid].
Table 2
Association constants Ka (mol/L)ꢁ1 of 2a and 2b with chiral carboxylic
acids
Entry
CSA
Guests
Ka
Ka (S or D)/
Ka (R or L)
1
2
3
4
5
6
7
8
9
2a
2a
2b
2b
2a
2a
2b
2b
2a
2a
2b
2b
(R)-Mandelic acid (1.1 0.07) ꢀ 104 >10
(S)-Mandelic acid >105
(R)-Mandelic acid (1.6 0.4) ꢀ 104
(S)-Mandelic acid >105
>6
(R)-MPAA
(S)-MPAA
(R)-MPAA
(S)-MPAA
Ts-(D)-valine
Ts-(L)-valine
Ts-(D)-valine
Ts-(L)-valine
(2.2 0.03) ꢀ 102 1.72
(3.8 0.01) ꢀ 102
(2.3 0.04) ꢀ 102 1.74
(4.0 0.04) ꢀ 102
(3.5 0.2) ꢀ 103
(1.2 0.01) ꢀ 103
2.92
10
11
12
(4.9 0.01) ꢀ 103 1.01
(4.9 0.01) ꢀ 103
1
consistent with the results of NMR experiment. The H
NMR spectra of equimolar host 2a and (R) or (S)-mandelic
acid showed that the methine proton signals of (R) and (S)-
mandelic acids were shifted upfield by 0.55 and 0.68 ppm,
respectively, while the proton signals of methine and meth-
ylene adjacent to the nitrogen atom of the tertiary amine in
host 2a were shifted downfield by about 0.4 ppm due to the
Fig. 6. The possible most stable models for the interaction between (a) 2a and (R)-mandelic acid, (b) 2a and (S)-mandelic acid based on PM3 of MOPAC
2007 software.

3390
Z. Luo et al. / Tetrahedron Letters 49 (2008) 3385–3390
3. Taran, F.; Gauchet, C.; Mohar, B.; Meunier, S.; Valleix, A.; Renard,
P. Y.; Ger’minon, C.; Grassi, J.; Wagner, A.; Mioskowski, C. Angew.
Chem., Int. Ed. 2002, 41, 124–127.
4. Schurig, V.; Nowotny, H. P. Angew. Chem. 1990, 102, 969–1108.
5. Allenmark, S. G. Chromatographic Enantioseparation: Methods and
Applications; Elliss Horwood: Chichester, 1988.
6. Pirkle, W. H.; Hoover, D. J. Top. Stereochem. 1982, 13, 263–331.
7. Parker, D. Chem. Rev. 1991, 91, 1441–1457.
8. Wenzel, T. J.; Wilcox, J. D. Chirality 2003, 15, 256–270.
9. Fulwood, R.; Parker, D. Tetrahedron: Asymmetry 1992, 3, 25–28.
groups of host 2b, which may be caused by the p–p inter-
action between host and guest compounds. In this case,
the discerning of the enantiomers of the guests can be
1
achieved in H NMR spectra.
In conclusion, we have designed and synthesized a new
class of chiral solvating agents 2a and 2b derived from chi-
ral amino alcohol and 1,8-naphthalic anhydride. They are
amphiphilic and efficient CSAs. In particular, 2a shows
excellent ability to discriminate the enantiomers of a broad
range of carboxylic acids, including hydrophilic carboxylic
acid even in protic polar solvent. This implies that it may
be used in biological and medical systems.
10. Kuhn, M.; Buddrus, J. Tetrahedron: Asymmetry 1993, 4, 207–210.
¨
11. Yang, D.; Li, X.; Fan, Y.; Zhang, D. J. Am. Chem. Soc. 2005, 127,
7996–7997.
12. Ma, F.-N.; Ai, L.; Shen, X.-M.; Zhang, C. Org. Lett. 2007, 9, 125–
127.
13. Wang, W.-G.; Ma, F.-N.; Shen, X.-M.; Zhang, C. Tetrahedron:
Asymmetry 2007, 18, 832–837.
Acknowledgment
14. Enders, D.; Thomas, C. R.; Runsink, J. Tetrahedron: Asymmetry
1999, 10, 323–326.
15. Bailey, D. J.; O’Hagan, D.; Tavasli, M. Tetrahedron: Asymmetry
1997, 8, 149–153.
16. Bilz, A.; Stork, T.; Helmchen, G. Tetrahedron: Asymmetry 1997, 8,
3999–4002.
The authors are grateful to the National Natural Science
Foundation of China (Nos. 20272045 and 20672085) for
financial support.
17. Port, A.; Virgili, A.; Alvarez-Larena, A.; Piniella, J. F. Tetrahedron:
Asymmetry 2000, 11, 3747–3757.
References and notes
18. Yang, X.; Wang, G.; Zhong, C.; Wu, X.; Fu, E. Tetrahedron:
Asymmetry 2006, 17, 916–921.
19. Brouwer, A. J.; Mulders, S. J. E.; Liskamp, R. M. J. Eur. J. Org.
Chem. 2001, 1903–1915.
1. Coppola, G. M.; Schuster, H. F. a-Hydroxyl Acids in Enantioselective
Synthesis; VCH: Weinheim, 1997.
2. Hanessian, S. Total synthesis of Natural Products: The Chiron
Approach; Pergamon: Oxford, 1983.
20. Fielding, L. Tetrahedron 2000, 56, 6151–6170.