Copper(II)-mediated resolution of α-halo carboxylic acids with chiral O,O′-dibenzoyltartaric acid: Spontaneous racemization and crystallization-induced dynamic resolution

Article abstract of DOI:10.1039/b510170k

Herein we present a new example of coordination-mediated resolution of racemic acids by a chiral acid. The reaction of copper(II) acetate monohydrate, optically pure O,O′-dibenzoyltartaric acid (DBTA) and racemic α-bromo-2-chlorophenylacetic acid (HL

Full text of DOI:10.1039/b510170k

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Copper(II)-mediated resolution of a-halo carboxylic acids with chiral
O,Oꢀ-dibenzoyltartaric acid: spontaneous racemization and
crystallization-induced dynamic resolution†
Hong-wu Xu,a,c Qi-wei Wang,a Jin Zhu,a,c Jin-gen Deng,*a,c Lin-feng Cun,a Xin Cui,a
Jun Wu,a Xin-liang Xub and Yu-liang Wub
a Key Laboratory of Asymmetric Synthesis & Chirotechnology of Sichuan Province and Union
Laboratory of Asymmetric Synthesis, Chengdu Institute of Organic Chemistry, Chinese
Academy of Sciences, Chengdu, 610041, China. E-mail: jgdeng@cioc.ac.cn; Fax: 86 28 8522
3978
b Zhejiang APELOA Medical Technology Co. Ltd., Dongyang, 322118, China
c Graduate School of the Chinese Academy of Sciences, Beijing, 100039, China
Received 18th July 2005, Accepted 10th October 2005
First published as an Advance Article on the web 31st October 2005
Herein we present a new example of coordination-mediated resolution of racemic acids by a chiral acid. The reaction
of copper(II) acetate monohydrate, optically pure O,Oꢀ-dibenzoyltartaric acid (DBTA) and racemic
a-bromo-2-chlorophenylacetic acid (HL1) in acetonitrile solution afforded a binuclear copper(II) complex with
D-DBTA dianion, a-bromo-2-chlorophenylacetate and acetate as ligands. After decomposition of the complex with
acid, the optically active acid ((R)–HL1) was obtained. Similarly, a-bromo-2-fluorophenylacetic acid (HL2),
a-bromo-2-bromophenylacetic acid (HL3), a-chloro-2-chlorophenylacetic acid (HL4), a-chloro-2-fluorophenylacetic
acid (HL5), a-bromophenylacetic acid (HL6), a-bromo-4-chlorophenylacetic acid (HL7), 2-bromopropionic acid
(HL8) and 2-chloropropionic acid (HL9) were resolved by the same method. Satisfactory results were obtained for
HL2 to HL5. For HL6 and HL7, only racemic acids were obtained. For the two a-halo aliphatic acids (HL8 and HL9),
poor enantioselectivity was obtained. It is more interesting that three acids (HL1, HL2 and HL3) could spontaneously
racemize in acetonitrile solution, which resulted in crystallization-induced dynamic resolution (CIDR) with greater
than 50% yield.
We report here a new example in which a series of racemic
Introduction
a-halo acids (Scheme 1) were resolved with a chiral acid. The re-
Optically active molecules play important roles in both indus-
trial and academic sectors. Methods of obtaining the enan-
tiomerically enriched substances can be classified into two broad
actions of Cu(OAc)2·H2O, optically pure O,Oꢀ-dibenzoyltartaric
acid monohydrate and the corresponding racemic halo acid in
acetonitrilesolutionaffordedmixedligandcopper(II)complexes.
categories: optical resolution of racemates, and asymmetrization
After their decomposition with hydrogen chloride solution,
of prochiral or meso compounds. Although many traditional
optically active acids were obtained. Moreover, three acids
resolutions have already been well-established methods for
(HL1, HL2, and HL3) can spontaneously racemize in acetonitrile
the preparation of optically active compounds, they have an
solution, and thus were resolved by crystallization-induced
economic disadvantage in that the maximum yield for each
dynamic resolution.
of the two pure enantiomers is theoretically limited to 50%.1
Crystallization-induced dynamic resolution (CIDR), which is
a method utilizing the epimerization of the substrate and the
solubility difference of the diastereomeric salts, is one of the most
important recent developments in the preparation of optically
pure compounds due to the greater than 50% yield of the
obtained enantiomer.2,3
Racemic carboxylic acids are generally resolved by optically
active bases. However, these bases are often extremely toxic and
expensive.1 For example, a-bromo-2-chlorophenylacetic acid,
which is an important intermediate for some medicines,4–6
was resolved by such bases in low yield (<25%).6 Recently,
the Mravik group applied optically active DBTA, which is
Scheme 1 The structures of a-halo acids.
usually used for the resolution of bases, as a new resolving
reagent for the racemic carboxylic acids. Some new metal–mixed
ligand complexes of DBTA can form with the corresponding
acids and then optically active acids can be obtained by their
decomposition.7–9 In contrast to the bases, optically pure DBTA
is a relatively cheap, nontoxic reagent. Therefore, this method is
promising for the resolution of acids.
Results and discussion
Resolution of HL1
To a solution of D-DBTA·H2O (H2L0·H2O, 1 equiv.) and
racemic a-bromo-2-chlorophenylacetic acid (HL1, 2 equiv.) in
acetonitrile, was added copper(II) acetate monohydrate (1 equiv.)
and the resulting mixture was stirred for several days at 25 ◦C. A
green precipitate was collected by filtration and characterized as
[Cu2(L0)(L1)(OAc)(H2O)2]·1.5 MeCN (complex 1) by elemental
† Electronic supplementary information (ESI) available: Experimen-
tal procedures, IR spectra, UV-vis spectra, thermograms. See DOI:
10.1039/b510170k
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 2 2 7 – 4 2 3 2
4 2 2 7
©

View Article Online
Table 1 Resolution of the a-halo acids
Entry
HA
Solvent
Ratioa
T (◦C)
Time/d
Obtained complex
Ee of the acid (%)b
Yield of the acid (%)c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
HL1
HL1
HL1
HL1
HL1
HL1
HL1
HL1
HL2
HL2
HL2
HL3
HL3
HL3
HL4
HL4
HL4
HL5
HL5
HL5
HL6
HL6
HL7
HL7
HL8
HL8
HL8
HL9
HL9
HL9
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN–H2O (16 : 1)
EtOH–H2O (3 : 1)
MeCN
MeCN
MeCN–H2O (16 : 1)
MeCN
MeCN–H2O (16 : 1)
MeCN–H2O (16 : 1)
MeCN
MeCN
MeCN–H2O (16 : 1)
MeCN
MeCN
MeCN–H2O (16 : 1)
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN–H2O (16 : 1)
MeCN
MeCN
MeCN–H2O (16 : 1)
2 : 1 : 2
1 : 1 : 2
1 : 1 : 2
1 : 1 : 2
1 : 1 : 2
1 : 1 : 2
1 : 1 : 2
2 : 1 : 2
2 : 1 : 2
1 : 1 : 2
1 : 1 : 2
2 : 1 : 2
2 : 1 : 2
1 : 1 : 2
2 : 1 : 2
1 : 1 : 2
1 : 1 : 2
2 : 1 : 2
1 : 1 : 2
1 : 1 : 2
2 : 1 : 2
1 : 1 : 1
2 : 1 : 2
1 : 1 : 1
2 : 1 : 2
1 : 1 : 2
1 : 1 : 2
2 : 1 : 2
1 : 1 : 2
1 : 1 : 2
25
25
30
35
40
50
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
7
7
7
7
7
7
1
1
1
1
1
1
1
10
2
2
2
85.0 (R)
85.0 (R)
90.5 (R)
91.8 (R)
94.7 (R)
91.3 (R)
94.1 (R)
25.7 (S)
92.0 (−)
91.5 (−)
89.9 (−)
—
93.7 (+)
93.6 (+)
87.6 (−)
94.8 (−)
—
87.9 (−)
80.5 (−)
—
0
0
3.2
0
42.2
84.0
84.3
84.3
82.4
83.5
84.1
20.2
43.3
83.5
83.3
—
43.7
85.5
30.6
37.2
—
5
1d
7
7
5
7
5
e
—
3
3
4
4
5
7
14f
5
e
—
7
5
5
—
34.0
38.6
—
14f
5
e
7
7
7
7
7
7
5
7
6
6
7
7
8
8
8
9
9
9
45.7
89.5
43.8
90.0
45.3
74.2
74.1
42.6
72.5
72.4
20.9 (R)
14.8 (R)
17.3 (R)
26.8 (R)
27.9 (R)
30.8 (R)
7
5
a The molar ratio of the reactants, HA, D-DBTA·H2O and Cu(OAc)2·H2O. b Enantiomeric excesses of HL1 to HL5 were determined by HPLC analysis
on Chiralcel AS Column with heptane-2-propanol-trifluoroacetic acid (90 : 10 : 0.1) as elute at 1 mL min−1 flow rate. The enantiomeric excess of HL6
to HL9 were determined by GC analysis on Chiralcel Dex 225 column after conversion to the corresponding ethyl esters. The configuration of HL1,
HL8 and HL9 were determined by comparing the sign of rotation to that of the literature. c Based on the racemic acid. d The complex was obtained
by cooling the resulting solution to −10 ◦C. e Only the Cu(II) complex of DBTA was obtained. f The precipitate formed 8 d later.
and ICP analyses. The (R)-(+)-form6 of HL1 was obtained in
good enantioselectivity (85.0% ee, Table 1, entry 1) by treatment
of the solid complex with 10% hydrogen chloride.
be indicative of a dimeric complex,12 suggesting that the complex
is binuclear, which also confirms the result of elemental analyses.
The IR spectrum of the complex exhibits strong absorptions
at 1721 and 1271 cm−1 which is assigned to the DBTA ligands.
It also shows a broad band at 3438 cm−1 which is assigned to
water molecules.
The thermogram of the complex (see the electronic supple-
mentary information) displays a two-step decomposition profile
from ambient temperature to 300 ◦C. The first continuous
weight loss step was observed from ambient temperature to
140 ◦C. The observed weight loss corresponds to the loss of
loosely bound acetonitrile and water molecules. This confirmed
the result of the elemental analyses. The second step was
observed from 148 to 225 ◦C, which should be attributed to the
decomposition of the other organic ligands. After the complex
was dried in vacuum for 5 h at 50 ◦C, the loss of acetonitrile
was observed by microanalysis10 and the result was identical
with the thermogram of the dried solid (see the electronic
supplementary information), in which only the loss of water
molecules was observed. Moreover, after the sample was exposed
to air for a long period, the partial loss of acetonitrile molecules
in the complex was confirmed by microanalysis. Both the results
indicate that the acetonitrile molecules only act as lattice solvent
molecules, rather than the ligands of the complex.
A carboxylate, RCOO−, can coordinate to metals in a number
of ways, viz. as a unidentate ligand, as a chelating ligand,
as a bridging bidentate ligand, or as a monatomic bridging
ligand. The mode of coordination to the center copper ion can
be determined from the value of D (D = masym − msym) in the
IR spectrum.13 This criterion can also be used in this copper
complex. The moderately strong band at 1401 cm−1 is assigned
to the symmetric carboxyl stretching frequency (msym). The
analogous asymmetric stretching frequency (masym) is 1602 cm−1.
The value of D (201 cm−1) and the positions of these bands
are comparable to those reported for other binuclear copper(II)
complexes,11–14 in which a bridging bidentate coordination mode
of carboxylate groups exists.
TGA, UV-vis, IR spectra, elemental and ICP analyses of the
complex suggest that it is a binuclear with the D-DBTA dianion,
a-bromo-2-chlorophenylacetate and acetate as ligands. On the
other hand, that DBTA often acts as bridging ligand because
of the two carboxylate groups,8,15 together with the fact that
the complex is soluble only in DMF and DMSO, indicate that
the complex is polymeric (Scheme 2). Unfortunately, the single
The solid state UV-vis spectrum of the complex showed a
broad absorption band (band I) in the visible region around
672 nm, which is assigned to a metal–ligand interaction.11 More-
over, it also displays a shoulder around the 372 nm (band II).
Although the origin of band II is controversial, it is believed to
4 2 2 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 2 2 7 – 4 2 3 2

View Article Online
unaffected and the best result was obtained at 40 ◦C (entries
2–5). However, when the reaction was conducted at more
◦
higher temperature (50 C), the enantioselectivity and yield of
acid obtained by decomposition of the solid complex declined
(entry 6).
(b) Solvents. Interestingly, it was found that adding water
to the reaction mixture can increase the reaction rate greatly.
When the ratio of acetonitrile and water was 16 : 1, the best
result was obtained. After 30 min, a green precipitate formed
and after 5 d, the yield and the diastereomeric excess of the
complex were close to those obtained from the same reaction in
absolute acetonitrile (entry 7 vs. entry 5). The water molecules
are auxiliary ligands of the resulting compound, and must be
regarded as a kind of reactant. The addition of the water meant
that the concentration of water was increased. Therefore, the
reaction was speeded up.
When the experiment was performed in mixed solvents
(EtOH : H2O = 3 : 1), a green precipitate was obtained by
cooling the resulting mixture to −10 ◦C. Microanalysis and ICP
and TGA analyses showed that the complex was formulated as
Cu2(L0)(L1)(OAc)(H2O)2 (complex 10) and the diastereomeric
excess (25.7%) of the resulting complex was lower than that
obtained from acetonitrile. In addition, the predominant form,
(S)-HL1, was different to that from acetonitrile (entry 8).19 Both
the results indicated the importance of solvents to the resolution
of HL1.
Scheme 2 The two possible structures for the binuclear copper(II)
complex of a-bromo-2-chlorophenylacetic acid (HL1).
crystal of the complex was not obtained although many methods
were tried. The definite structure can not be clarified by the
present results.
Crystallization-induced dynamic resolution of HL1
Initially, it was thought that the ratio of the three reactants (rac-
HL1, H2L0 and Cu(OAc)2·H2O) should be 2 : 1 : 2 because
the resulting complex was formulated as [Cu2(L0)(L1)(OAc)
(H2O)2]·1.5 MeCN. However, after the mother liquid was
concentrated to dryness, only racemic a-bromo-2-chloro-
phenylacetic acid was obtained by following the same procedure
for the decomposition of the copper(II) complex. It showed that
there was a process in which (S)-HL1 was converted to racemic
HL1. Subsequent study showed that optically active HL1 can
racemize spontaneously in acetonitrile,16 which indicated that
the acid could be resolved by crystallization-induced dynamic
resolution.
Furthermore, crystallization-induced dynamic resolution was
not observed when mixed solvents (EtOH : H2O = 3 : 1) were
used, owing to the slow spontaneous racemization rate.20,21
When dichloromethane, methanol, anhydrous ethanol or
dimethylformamide was used as the solvent, no precipitate was
obtained even though the reaction was conducted over 7 d.
Thus, the experiment, in which the ratio of rac-HL1, D-
DBTA·H2O and Cu(OAc)2·H2O was 1 : 1 : 2, was performed at
room temperature. The resulting complex was also fomulated as
[Cu2(L0)((R)-L1)(OAc)(H2O)2]·1.5 MeCN and the yield (84.0%)
of the acid was almost doubled with a comparable diastere-
omeric excess (entry 2 vs. entry 1). So crystallization-induced
dynamic resolution occurred.
As is shown in Scheme 3, after the mixtures of Cu(OAc)2·H2O,
optically pure D-DBTA·H2O and racemic HL1 in acetoni-
trile had been stirred for 2 d at 40 ◦C, the less soluble
complex [Cu2(L0)((R)-L1)(OAc)(H2O)2]·1.5 MeCN deposited
preferentially17 with 79.8% diastereomeric excess.18 The result
was that the concentration of (S)-HL1 was higher than that
of (R)-HL1 in the solution and the inversion of (S)-HL1 to
(R)-HL1 occurred.16,17 Therefore, more and more [Cu2(L0)((R)-
L1)(OAc)(H2O)2]·1.5 MeCN accumulated as time went by.18 At
the same time, the fact that the more soluble complex dissolves
more easily than the less soluble one can also cause a concen-
tration difference between the two diastereomers in the mother
liquid. At last, the diastereomeric excess was increased to 94.7%.
(c) Reagents. When Ca(OH)2 was used, only calcium salt
of DBTA was obtained. And when Zn(OAc)2 was used, although
a mixed ligand Zn(II) complex was obtained, only the racemic
acid was obtained by decomposition of the solid complex.
On the other hand, when O,Oꢀ-di-p-toluoyltartaric acid
(DTTA) as well as tartaric acid (TA) were used as the resolving
agent, only the copper(II) salts of DTTA or TA were obtained.15
There are two kinds of free acids (DBTA and a-halo acid) in
the reactants. However, the ligands in the resulting complex are
D-DBTA dianion, a-bromo-2-chlorophenylacetate and acetate.
It seems that using the salts in place of the corresponding acids
(NaL1 for example) might increase the reaction.11 However, such
an attempt was unsuccessful and only the copper(II) salt of O,Oꢀ-
dibenzoyltartaric acid was obtained.
Resolution of the other a-halo acids
(a) HL2 and HL3. When racemic a-bromo-2-fluorophenyl-
acetic acid (HL2) was resolved in acetonitrile, satisfactory results
(the acid with 91.5% ee and 83.5% yield) were obtained (entry
10). Similarly to the resolution of HL1, water can accelerate the
resolution rate (entry 11).
When racemic a-bromo-2-bromophenylacetic acid (HL3) was
resolved in absolute acetonitrile, only the copper(II) salt of
DBTA was obtained (entry 12). However, when it was resolved in
mixed solvents (MeCN : H2O = 16 : 1), a mixed ligand complex
(complex 3), which was similar to complex 1, was obtained in
93.6% ee and 85.5% yield (entry 14). The results showed the
important roles of water and the substituents of the benzene
ring in the formation of the complex.
(b) HL4 and HL5. Racemic a-chloro-2-chlorophenylacetic
acid (HL4) and a-chloro-2-fluorophenylacetic acid (HL5) can be
resolved by the same method, while the yields and selectivities
were slightly lower than those in the resolution of HL1 (entries
15 and 18). Moreover, when they were resolved in the mixed
solvents (MeCN : H2O = 16 : 1), only the copper(II) salt of
DBTA was obtained (entries 17 and 20).
Scheme 3 Proposed mechanism for the crystallization-induced dy-
namic resolution of a-bromo-2-chlorophenylacetic acid (HL1).
Investigation of the resolution conditions
(a) Temperatures. When the reaction was conducted at
0
◦C, no precipitate was obtained even though it had been
conducted for 7 d. With increasing reaction temperature, the
diastereomeric excess increased, with the yield being almost
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 2 2 7 – 4 2 3 2
4 2 2 9

View Article Online
Crystallization-induced dynamic resolution was not observed
in the resolution process of these two acids (entries 16 and 19)
and these results may originate from that a-chloroarylacetic acid
are more configurationally stable than a-bromoarylacetic acid
in acetonitrile solution.
IR spectrophotometer over the range 4000–400 cm−1, using KBr
pellets. 1H NMR spectra were measured on a Bruker (300 MHz)
spectrometer in CDCl3 or D2O solutions with tetramethylsilane
as the internal standard. The elemental (C, H, N) analyses
of the powdered samples were performed on a Perkin-Elmer
model 240 C automatic instrument. The concentrations of Cu
in the complexes were determined using an◦IRIS Advantage
ICP spectrophotometer after heating to 400 C and treatment
with acid. The thermograms of the complexes were conducted
with a Perkin-Elmer TGA instrument. The UV-vis spectra of the
solid complexes were recorded from 200–800 nm, using an UV-
1000 spectrophotometer. Liquid chromatographic analyses were
conducted on a Beckman 110 instrument equipped with a model
168 detector. Gas chromatographic analyses were conducted on
a Varian 3380 instrument.
(c) HL6 and HL7. Reaction of D-DBTA·H2O, racemic a-
bromophenylacetic acid (HL6) and Cu(OAc)2·H2O in acetoni-
trile only afforded a mononuclear compound formulated as
Cu(HL0)(L6)·1.5H2O. The result can be confirmed from that
the solid state UV-vis spectrum of the complex did not display
the band II of the dimeric complexes. After decomposition of
the solid complex, only racemic HL6 was obtained (entries 21
and 22).
In the resolution of a-bromo-4-chlorophenylacetic acid (HL7),
although a binuclear mixed ligand complex, which is formulated
as Cu2(L0)(L7)(OAc)(H2O)2, was obtained by same method, the
acid obtained by decomposition was nearly racemic (entries 23
and 24).
Resolution of the acids: syntheses and decomposition of the
complexes
[Cu2(L0)(L1)(OAc)(H2O)2]·1.5 MeCN (complex 1). To a
solution of D-DBTA·H2O (760 mg, 0.2 mmol) and racemic a-
bromo-2-chlorophenylacetic acid (500 mg, 0.2 mmol) in 10 mL
acetonitrile at 40 ◦C was added Cu(OAc)2·H2O (800 mg,
0.4 mmol). After stirring for about 5 min, Cu(OAc)2 was
dissolved and the mixture became clear. 2 d later, a green
precipitate formed, and the mixture was stirred for an additional
5 d. The complex was filtered, washed with 10 mL acetonitrile
and air-dried overnight at room temperature. 1.581 g green
solid were obtained. Anal. (C31H28.5N1.5O14BrClCu2) C, H, N,
Cu: calcd, 41.91%, 3.23%, 2.36%, 14.3%; found, 42.12%, 3.16%,
2.40%, 14.0%. IR absorption bands (cm−1): 3438 (br), 3069 (w),
1721 (s), 1602 (vs), 1450 (w), 1401 (s), 1271 (s), 1178 (w), 1115
(m), 1070 (w), 1027 (w), 741 (m), 709 (m). UV-vis of the solid
sample (k/nm): 372, 672.
The poor diastereoselectivities for the two acids further con-
firmed that the ortho-substitute of the benzene ring of a-bromo-
arylacetic acid play an important role in chiral recognition of
the copper(II) complexes.
(d) HL8 and HL9. It was reported by Mravik group that
2-bromopropionic acid and 2-chloropropionic acid can be
resolved by forming two mixed ligand Cu(II) complexes in EtOH
solution, which were formulated as [Cu(A)n1(H2O)nw](HL0)(L)
(L0 is DBTA dianion, L is a a-halo acid anion and A stands for an
alcohol molecule).9 However, when we did the same experiments
in acetonitrile solution, the obtained complexes were formulated
as [Cu2(L0)(L)(OAc) (H2O)2]·1.5 MeCN. After decomposition of
them with HCl solution, optically active acids (20.9 and 26.8%
ee for HL8 and HL9, respectively) were obtained (entries 25 and
28). The poor enantioselectivities for the two acids, HL8 and
HL9, indicate the important role of substitutes at a-position in
the chiral recognition of the copper(II) complexes.
Interestingly, when the molar ratio of reactants changed to 1 :
1 : 2 (entries 26 and 27, and entries 29 and 30), high yield of
the acids was obtained and resulted in that the optical purities
(39.3% ee and 68.2% ee for (S)-HL8 and (S)-HL9, respectively)
in the mother liquid were higher than those (14.8 and 27.9% ee
for (R)-HL8 and (R)-HL9, respectively) in the crystals (entries
26 and 29). Moreover, these results showed that no racemization
occurred during the resolution process and thus crystallization-
induced dynamic resolution was not observed for these two
acids.
The complex 1 was suspended in 25 mL 10% hydrochloric
◦
acid and 5 mL toluene. The mixture was warmed to 50 C for
a short period with vigorous stirring, and then cooled to the
room temperature. The precipitate (DBTA·H2O) was filtered.
The toluene layer was separated and the aqueous layer was
extracted with ethyl acetate (3 × 10 mL). The organic phases
were combined, dried over sodium sulfate and concentrated in
vacuum. To the residue was added 20 mL of toluene to dissolve
the product and then the undissolved DBTA was filtered off. The
filtrate was concentrated to give optically active (R)-a-bromo-
2-chlorophenylacetic acid (412 mg). Yield: 82.4% (based on the
racemic acid). Ee: 94.7%; [a]2D8 = +102.7 (c 1.0, EtOH), [a]2D5
+6.3 in MeOH. 1H NMR (300 MHz, CDCl3), d: 10.38 (br, 1H,
=
+8.9 (c 2.4, MeOH); the (R)-isomer reported in literature:6 [a]D25
=
Conclusions
COOH), 7.40–7.93 (m, 4H, Ar–H), 6.09 (s, 1H, CH) ppm.
[Cu2(L0)(L2)(OAc)(H2O)2]·1.5 MeCN (complex 2). This
compound was prepared with a similar method to the prepa-
ration of complex 1. Anal. (C31H28.5N1.5O14BrFCu2) C, H, N,
Cu: calcd, 42.70%, 3.29%, 2.41%, 14.6%; found, 43.01%, 3.22%,
2.44%, 14.9%. IR absorption bands (cm−1): 3443 (br), 3069 (w),
1719 (s), 1646 (vs), 1602 (vs), 1490 (m), 1451 (m), 1406 (s), 1316
(w), 1267 (s), 1178 (w), 1114 (m), 1096 (m), 1070 (m), 1026 (w),
750 (m), 712 (m). UV-vis of the solid sample (k/nm): 364, 692.
Optically active a-bromo-2-fluorophenylacetic acid was ob-
tained by the similar method for HL1. Yield: 83.5% (based on
We have described a successful example of coordination-
mediated resolution of a series of a-halo acids using O,Oꢀ-
dibenzoyltartaric acid and copper(II) acetate monohydrate as
resolving reagents. Moreover, we also found that crystallization-
induced dynamic resolution occurred in the resolution processes
of HL1, HL2 and HL3, because they can spontaneously racemize
in acetonitrile solution. Finally, substituents at the a-position of
the a-halo acids play important roles in the chiral recognition of
the copper(II) complexes.
1
the racemic acid). Ee: 91.5%; [a]2D8 = −50.4 (c 1.0, EtOH). H
Experimental section
NMR (300 MHz, CDCl3), d: 10.43 (br, 1H, COOH), 7.05–7.70
(m, 4H, Ar–H), 5.74 (s, 1H, CH) ppm.
Material and reagents
HL2 to HL7 were prepared by the reported methods.5,22 All the
other reagents were purchased from commercial sources and
were used without purification.
[Cu2(L0)(L3)(OAc)(H2O)2]·2 MeCN (complex 3). This com-
pound was prepared with a similar method to the preparation
of complex 1 in a solution of acetonitrile and H2O (16 : 1). Anal.
(C32H30N2O14Br2Cu2) C, H, N, Cu: calcd, 40.31%, 3.17%, 2.94%,
13.3%; found, 40.59%, 3.20%, 2.90%, 13.7%. IR absorption
bands (cm−1): 3443 (br), 3068 (w), 2942 (sh), 1720 (s), 1648 (vs),
1602 (vs), 1585 (sh), 1467 (w), 1450 (m), 1401 (s), 1316 (w), 1268
Physical measurements
Optical rotations were measured on a Perkin-Elmer 341 po-
larimeter. IR spectra were recorded on a Nicolet 200SXV FT-
4 2 3 0
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 2 2 7 – 4 2 3 2

View Article Online
(s), 1177 (w), 1115 (m), 1099 (m), 1071 (m), 1026 (m), 738 (m),
713 (m), 686 (m), 623 (w). UV-vis of the solid sample (k/nm):
373, 669.
(w), 1451 (m), 1412 (s), 1319 (w), 1265 (s), 1177 (w), 1117 (m),
1071 (m), 711 (m), 686 (m). UV-vis of the solid sample (k/nm):
361, 676.
Optically active a-bromo-2-bromophenylacetic acid was ob-
Optically active (R)-2-bromopropionic acid was obtained by
the decomposition of the complex and distillation of the residue.
tained by the similar method for HL1. Yield: 85.5% (based on
1
the racemic acid). Ee: 93.6%; [a]2D8 = +12.3 (c 1.0, EtOH). H
Yield: 45.3% (based on the racemic acid). Ee: 20.9%; [a]D22
=
NMR (300 MHz, CDCl3), d: 11.40 (br, 1H, COOH), 7.22–7.80
+6.2 (c 1.0, MeOH); the (S)-isomer reported in literature:23 [a]D22
1
(m, 4H, Ar–H), 5.96 (s, 1H, CH) ppm.
= −29.8 in methanol. H NMR (300 MHz, CDCl3), d: 11.78
(br, 1H, COOH), 4.40 (q, J = 6.96 Hz, 1H, CH), 1.83 (d, J =
[Cu2(L0)(L4)(OAc)(H2O)2]·1.5MeCN (complex 4). This
compound was prepared with a similar method to the
preparation of complex 1 except that 2 equiv. racemic HL4 was
used. Anal. (C31H28.5N1.5O14Cl2Cu2) C, H, N, Cu: calcd, 44.11%,
3.40%, 2.49%, 15.1%; found, 44.39%, 3.31%, 2.43%, 14.7%. IR
absorption bands (cm−1): 3440 (br), 3070 (w), 1720 (s), 1648
(vs), 1602 (vs), 1491 (w), 1451 (m), 1404 (s), 1317 (w), 1268 (s),
1177 (w), 1114 (m), 1098 (m), 1026 (w), 754 (m), 711 (m), 686
(m). UV-vis of the solid sample (k/nm): 362, 692.
6.96 Hz, 3H, CH3) ppm.
[Cu2(L0)(L9)(OAc)(H2O)2]·1.5 MeCN (complex 9). This
compound was prepared with a similar method to the prepa-
ration of complex 4. Anal. (C26H28.5N1.5O14ClCu2) C, H, N, Cu:
calcd, 41.72%, 3.84%, 2.81%, 17.0%; found, 41.65%, 3.74%,
2.80%, 17.4%. Major IR absorption bands (cm−1): 3435 (br),
3075 (w), 2929 (w), 2275 (vw), 1723 (s), 1641 (vs), 1602 (vs),
1585 (sh), 1491 (w), 1451 (m), 1413 (s), 1328 (w), 1267 (s), 1177
(w), 1117 (m), 1071 (m), 1026 (w), 711 (m), 686 (m). UV-vis of
the solid sample (k/nm): 360, 685.
Optically active (R)-2-chloropropionic acid was obtained by
the similar method for HL8. Yield: 42.6% (based on the racemic
acid). Ee: 26.8%; [a]2D5 = +4.3 (c 1.0, CHCl3); the (S)-isomer
reported in literature:24 [a]D25 = −15.1 in CHCl3. 1H NMR
(300 MHz, CDCl3), d: 11.54 (br, 1H, COOH), 4.46 (q, J =
9.78 Hz, 1H, CH), 1.72 (d, J = 9.78 Hz, 3H, CH3) ppm.
Optically active a-chloro-2-chlorophenylacetic acid was ob-
tained by the similar method for HL1. Yield: 30.6% (based on
1
the racemic acid). Ee: 87.6%; [a]2D8 = −87.6 (c 1.0, EtOH). H
NMR (300 MHz, CDCl3), d: 10.58 (br, 1H, COOH), 7.32–7.67
(m, 4H, Ar–H), 5.86 (s, 1H, CH) ppm.
[Cu2(L0)(L5)(OAc)(H2O)2]·1.5 MeCN (complex 5). This
compound was prepared with a similar method to the prepa-
ration of complex 4. Anal. (C31H28.5N1.5O14ClFCu2) C, H, N,
Cu: calcd, 44.99%, 3.47%, 2.54%, 15.6%; found, 44.85%, 3.41%,
2.63%, 16.0%. IR absorption bands (cm−1): 3425 (br), 3069 (w),
1719 (s), 1642 (vs), 1602 (vs), 1492 (w), 1451 (m), 1404 (s), 1317
(w), 1268 (s), 1177 (w), 1115 (m), 1096 (m), 1071 (m), 1026 (w),
754 (m), 711 (m), 686 (m). UV-vis of the solid sample (k/nm):
362, 677.
Acknowledgements
Financial support from the National Natural Science Foun-
dation of China (No. 203900507 and 20025205) is greatly
acknowledged.
Optically active a-chloro-2-fluorophenylacetic acid was ob-
tained by the similar method for HL1. Yield: 34.0% (based on
References
1
the racemic acid). Ee: 87.9%; [a]2D8 = −60.2 (c 1.0, EtOH). H
1 J. Jaques, A. Collet, and S. H. Wilen, Enantiomers, Racemates and
Resolutions, Wiley, New York, 1981.
NMR (300 MHz, CDCl3), d: 8.81 (br, 1H, COOH), 7.08–7.60
(m, 4H, Ar–H), 5.76 (s, 1H, CH) ppm.
2 For recent reports on crystallization-induced dynamic resolution
(CIDR), see: J. G. Chen, J. Zhu, P. M. Skonezny, V. Rosso and
J. J. Venit, Org. Lett., 2004, 6, 3233; S. Kiau, R. P. Discordia, G.
Madding, F. J. Okuniewicz, V. Rosso and J. J. Venit, J. Org. Chem.,
2004, 69, 4256; J. Kosˇmrlj, L. O. Weigel, D. A. Evans, C. W. Downey
and J. Wu, J. Am. Chem. Soc., 2003, 125, 3208; H. Kaku, S. Takaoka
and T. Tsunoda, Tetrahedron, 2002, 58, 3401 and references therein;
W. H. J. Boesten, J.-P. G. Seerden, B. de Lange, H. J. A. Dielemans,
H. L. M. Elsenberg, B. Kaptein, H. M. Moody, R. M. Kellogg and
Q. B. Broxterman, Org. Lett., 2001, 3, 1121; W. Aelterman, Y. Lang,
B. Willemsens, I. Vervest, S. Leurs and F. D. Knaep, Org. Process
Res. Dev., 2001, 5, 467; A. Kolarovic, D. Berkesˇ, P. Baran and F.
Povazanec, Tetrahedron Lett., 2001, 42, 2579.
3 For related reviews on crystallization-induced dynamic resolution
(CIDR), see: M. C. Hillier and P. J. Reider, Drug Discovery Today,
2002, 7, 303; E. J. Ebbers, G. J. A. Ariaans, J. P. M. Houbiers,
A. Bruggink and B. Zwanenburg, Tetrahedron, 1997, 53, 9417; S.
Caddick and K. Jenkins, Chem. Soc. Rev., 1996, 447; R. S. Ward,
Tetrahedron: Asymmetry, 1995, 6, 1475.
Cu(HL0)(L6)·1.5H2O (complex 6). This compound was pre-
pared with a similar method to the preparation of complex 4.
Anal. (C26H22O11.5BrCu) C, H, N, Cu: calcd, 47.18%, 3.35%,
0, 9.6%; found, 47.19%, 3.29%, 0, 9.4%. IR absorption bands
(cm−1): 3439 (br), 3244 (sh), 3158 (sh), 3064 (sh), 1718 (s), 1640
(vs), 1603 (vs), 1493 (w), 1451 (m), 1400 (m), 1337 (m), 1318 (w),
1268 (s), 1178 (w), 1115 (m), 1097 (m), 1070 (m), 1026 (m), 997
(w), 943 (w), 799 (w), 754 (m), 712 (m), 695 (m). UV-vis of the
solid sample (k/nm): 683.
a-Bromophenylacetic acid was obtained by the similar
method for HL1. Yield: 45.7% (based on the racemic acid). Ee:
0. 1H NMR (300 MHz, D2O), d: 7.42 (brs, 5H, Ar–H), 5.26 (s,
1H, CH) ppm.
Cu2(L0)(L7)(OAc)(H2O)2 (complex 7). This compound was
prepared with a similar method to the preparation of complex 4.
Anal. (C28H24O14ClBrCu2) C, H, N, Cu: calcd, 40.67%, 2.93%,
0, 15.4%; found, 40.40%, 3.10%, 0, 15.7%. IR absorption bands
(cm−1): 3436 (br), 1717 (s), 1638 (vs), 1601 (vs), 1492 (w), 1451
(m), 1404 (m), 1268 (s), 1178 (w), 1115 (m), 1096 (m), 1070 (w),
1027 (w), 792 (m), 710 (m), 685 (w). UV-vis of the solid sample
(k/nm): 352, 683.
4 M. Valeriano, P. Daverio, and S. Bianchi, US Pat. 6 737 411, 2004.
5 A. Badore, D. Frehel, J. P. Maffrand, and E. Vallee, US Pat. 4 740 510,
1988 (CA, 107: 89900); M. Bouisset, and J. Radisson, US Pat.
5 036 156, 1991 (CA, 115: 114486).
6 O. Ko, H. Yukihiro, and K. Isao, JP Pat. 200 034 256, 2000 (CA, 132:
122386).
7 A. Mravik, Z. Bo¨cskei, Z. Katona, I. Markovits and E. Fogassy,
Angew. Chem., Int. Ed., 1997, 36, 1534.
a-Bromo-4-chlorophenylacetic acid was obtained by the sim-
ilar method for HL1. Yield: 43.8% (based on the racemic acid).
Ee: 3.2%. 1H NMR (300 MHz, D2O), d: 7.42 (brs, 4H, Ar–H),
5.27 (s, 1H, CH) ppm.
8 A. Mravik, Z. Bo¨cskei, Z. Katona, I. Markovits, G. Pokol, D. K.
Menyha´rd and E. Fogassy, Chem. Commun., 1996, 1983.
9 A. Mravik, Z. Bo¨cskei, K. Simon, F. Elekes and Z. Izsa´ki, Chem.
Eur. J., 1998, 4, 1621.
10 Microanalysis and ICP analysis further confirmed the result. Anal.
(C28H24O14ClBrCu2) C, H, N, Cu: calcd, 40.67%, 2.93%, 0, 15.4%;
found, 40.54%, 3.03%, 0, 15.2%.
11 C. Dendrinou-Samara, D. P. Kessissoglou, G. E. Manoussakis, D.
Mentzafos and A. Terzis, J. Chem. Soc., Dalton Trans., 1990, 959
and references therein;; C. Dendrinou-Samara, P. D. Jannakoudakis,
D. P. Kessissoglou, G. E. Manoussakis, D. Mentzafos and A. Terzis,
J. Chem. Soc., Dalton Trans., 1992, 3259.
[Cu2(L0)(L8)(OAc)(H2O)2]·1.5 MeCN (complex 8). This
compound was prepared with a similar method to the prepa-
ration of complex 4. Anal. (C26H28.5N1.5O14BrCu2) C, H, N, Cu:
calcd, 39.38%, 3.62%, 2.65%, 16.0%; found, 39.45%, 3.55%,
2.64%, 16.3%. Major IR absorption bands (cm−1): 3444 (br),
3051 (w), 2929 (w), 2277 (w), 1723 (s), 1641 (vs), 1601 (vs), 1492
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 2 2 7 – 4 2 3 2
4 2 3 1

View Article Online
12 J. E. Weder, T. W. Hambely, B. J. Kennedy, P. A. Lay, D. MacLachlan,
R. Bramley, C. D. Delfs, K. S. Murray, B. Moubaraki, B. Warwick,
J. R. Biffin and H. L. Regtop, Inorg. Chem., 1999, 38, 1736 and
references therein.
13 G. B. Deacon and R. J. Phillips, Coord. Chem. Rev., 1980, 33,
227; K. Nakamoto, Infrared Spectra of Inorganic and Coordination
Compounds, Wiley, New York, 1963.
18 This can be confirmed by the following experiment. The complex
with 79.8% diastereomeric excess, which was obtained by filtration
when the reaction had been conducted for 2 d, was suspended in
acetonitrile, and stirred at 40 ◦C for 5 d. HPLC analysis showed that
the diastereomic excess was increased to 85.2%.
19 The phenomenon that different enantiomer was obtained when
different solvents were used can also be observed in the coordination-
mediated resolution of mandelic acid ester (see reference 8).
20 After (R)-HL1 (93.6% ee) was stirred in ethanol for 2 d at 40 ◦C,
HPLC analysis showed that there was only slight racemization (88.7%
ee) and also no racemization was observed in ethyl acetate and
toluene.
14 Y. Mathey, D. R. Greig and D. F. Shriver, Inorg. Chem., 1982, 21,
3409.
15 J. Liao, X. Peng, J. Zhang, K. Yu, X. Cui, J. Zhu and J. Deng, Org.
Biomol. Chem., 2003, 1080; H. Tomori, H. Yoshihara and K. Ogura,
Bull. Chem. Soc. Jpn., 1996, 69, 3581.
16 This was confirmed by the following experiment. After (R)-HL1
(93.6% ee) was stirred for 20 h in acetonitrile at 40 ◦C, HPLC analysis
showed that racemic HL1 was obtained.
21 Reaction of (R)-HL1 (93.3% ee), L-DBTA·H2O and Cu(OAc)2·H2O
(1 : 1 : 2 molar ratio) in EtOH–H2O (3 : 1) afforded Cu2(L-DBTA)((R)-
L1)(OAc) (H2O)2 and (R)-HL1 with 92.6% ee was obtained after the
treatment of the complex with HCl. When D-DBTA was used in
place of L-DBTA, Cu2(D-DBTA)((R)-L1)(OAc)(H2O)2 was obtained
and (R)-HL1 with 87.7% ee was obtained after the treatment of the
complex with HCl.
22 H. L. Haller, P. D. Bartlett, N. L. Drake, M. S. Newman, S. J. Cristol,
C. M. Eaker, R. A. Hayes, G. W. Kilmer, B. Magerlein, G. P. Mueller,
A. Schneider and W. Wheatley, J. Am. Chem. Soc., 1945, 67, 1591.
23 A. S. Dutta, M. B. Giles, J. J. Gormley, J. C. Williams and E. J.
Kusner, J. Chem. Soc., Perkin Trans. 1, 1987, 111.
17 To verify our hypothesis further, we did the following experiments.
Reaction of (R)-HL1 (93.6% ee), D-DBTA·H2O and Cu(OAc)2·H2O
(1 : 1 : 2 molar ratio) in acetonitrile solution afforded [Cu2(D-
DBTA)((R)-L1)(OAc) (H2O)2]·1.5 MeCN in 30 min and (R)-HL1
with 90.1% ee was obtained after the treatment of the complex with
HCl. While L-DBTA was used in place of D-DBTA, green precipitate
formed only after 2 d. Moreover, microanalysis and ICP analysis
of the complex and HPLC analysis of the acid (85.6% ee) obtained
by decomposition of the complex showed that it was formulated as
[Cu2(L-DBTA)((S)-L1)(OAc)(H2O)2]·1.5 MeCN, rather than [Cu2(L-
DBTA)((R)-L1)(OAc)(H2O)2]·1.5 MeCN.
24 G. Kirchner, M. P. Scollar and A. M. Klibanov, J. Am. Chem. Soc.,
1985, 107, 7072.
4 2 3 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 2 2 7 – 4 2 3 2