Entrainment resolution of carnitinamide chloride

Article abstract of DOI:10.1016/j.tetasy.2008.06.011

The ready availability of (R)-carnitinamide, an immediate synthetic precursor of (R)-carnitine, is an ambitious goal and resolutions, due to the very low cost of racemic carnitinamide, can be the most convenient stereotechnology to reach it. We have efficiently resolved carnitinamide chloride by preferential crystallization from methanol according to simple entrainment procedure. The previous transformation of the chloride into other salts or the use of specific solvent systems was not required for successful resolution, as in the case of the reported entrainment of carnitinamide precursor, carnitine nitrile.

Full text of DOI:10.1016/j.tetasy.2008.06.011

Tetrahedron: Asymmetry 19 (2008) 1637–1640
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Entrainment resolution of carnitinamide chloride q
a
a
a
a
Marco Pallavicini a, , Cristiano Bolchi , Matteo Binda , Rossana Ferrara , Laura Fumagalli ,
*
Oreste Piccolo b, Ermanno Valoti a
a Istituto di Chimica Farmaceutica e Tossicologica ‘‘Pietro Pratesi”, Università di Milano, via Mangiagalli 25, I-20133 Milano, Italy
b Studio di Consulenza Scientifica, via Bornò 5, I-23896 Sirtori (LC), Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
The ready availability of (R)-carnitinamide, an immediate synthetic precursor of (R)-carnitine, is an ambi-
tious goal and resolutions, due to the very low cost of racemic carnitinamide, can be the most convenient
stereotechnology to reach it. We have efficiently resolved carnitinamide chloride by preferential crystal-
lization from methanol according to simple entrainment procedure. The previous transformation of the
chloride into other salts or the use of specific solvent systems was not required for successful resolution,
as in the case of the reported entrainment of carnitinamide precursor, carnitine nitrile.
Ó 2008 Elsevier Ltd. All rights reserved.
Received 27 May 2008
Accepted 11 June 2008
Available online 12 July 2008
1. Introduction
ing agent, thanks to its conglomerate nature.8 This is the case for
carnitine nitrile 3, a carnitine precursor, whose chloride is a con-
Over the last 30 years, many significant nutritional and medical
applications have been found for (R)-carnitine (R)-1, thus increas-
ing its commercial demand and increasing the efforts of both fun-
damental and applied research to improve its production.
Asymmetric chemical syntheses, biotransformation of achiral pre-
cursors, diastereomeric salts resolutions, enzymatic resolutions
and transformations of chiral pool substances have been investi-
gated to develop advantageous methods for the industrial scale
production of (R)-1.2 Crotonobetaine and 4-butyrobetaine bio-
glomerate, readily available by a one-pot conversion of inexpen-
sive racemic epichlorohydrin.9 A 1968 patent claims the direct
resolution of ( )-3 chloride by crystallization from methanol or
water or aqueous methanol.10 However, the efficiencies of such
procedures are poor, giving the pure enantiomer recovered at each
crystallization in 4–11% of the total enantiomer initially present in
the racemate supersaturated solution. Fourteen years later, Jacques
and Collet reported that the perchlorate and the hydrogen oxalate
of ( )-3 are also conglomerates and patented the entrainment res-
olution of these salts by crystallization from water or 70/30
methoxyethanol–water mixture.11 The efficiencies of the new pro-
cedures are higher, in particular those of the mono-oxalate from
water, which allows more than 20% of the total enantiomer to be
separated at each crystallization. Furthermore, the same authors
attempted the entrainment resolution of chloride ( )-3 again,
accomplishing a series of alternate preferential crystallizations of
the two enantiomers from 70/30 methoxyethanol–water, but, as
reported in the previous patent, the precipitates have a low enan-
tiomeric purity (ꢀ50%) and the pure enantiomer recovered at each
crystallization averages to a modest 9% of the total enantiomer ini-
tially present in the racemate supersaturated solution. Therefore, it
is not surprising that the direct resolution of ( )-3 has not found
industrial application in spite of the general convenience of
entrainment procedures. In fact, chloride ( )-3 is readily available,
but not efficiently resolved by this method10,11 and, conversely,
mono-oxalate ( )-3 better resolved, but is not readily available,
since its preparation requires the intermediary conversion of chlo-
ride into hydroxide by ion-exchange chromatography, treatment
with oxalic acid, complete water removal and crystallization of
the racemate from methoxyethanol.11
transformation3 and racemic carnitinamide 2
D-camphorate reso-
lution via crystallization4 may be cited as examples of methods
scaled up and currently applied in industry. These successes have
not, however, relaxed the interest in new preparative procedures
for (R)-1, which still remains a challenging target as demonstrated
by the recently patented stereoselective enzymatic hydrolyses of
( )-4-chloro-3-hydroxybutyric acid alkyl ester, synthesized from
racemic epichlorohydrin,5 and of ( )-3-acyloxy-
c-butyrolactone,
synthesized from racemic 3-hydroxy-
c
-butyrolactone,6 or the li-
pase-catalyzed stereoselctive acetylation of 3-hydroxy-4-tos-
yloxybutanenitrile.7 Compared with other stereotechnologies,
racemate resolutions, in particular the chemical ones, suffer from
two major drawbacks: the cost of the non-catalytic chiral auxiliary,
the resolving agent, which often needs to be recovered, and the re-
use of the undesired enantiomer. Such problems can be overcome
or minimized when the racemate is an inexpensive material, di-
rectly resolvable by preferential crystallization, without the resolv-
q
See Ref. 1.
* Corresponding author. Tel.: +39 2 50319336; fax: +39 2 50319335.
E-mail address: marco.pallavicini@unimi.it (M. Pallavicini).
0957-4166/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2008.06.011

1638
M. Pallavicini et al. / Tetrahedron: Asymmetry 19 (2008) 1637–1640
and the sulfate is a solid solution.12 The chloride of 2 is an imme-
OH
OH O
N+
CN
N+
N+
O
diate carnitine precursor, easily prepared from 3 chloride by
hydrolysis13 and thus, similar to chloride 3, an inexpensive racemic
material. This makes ( )-2 chloride a good candidate to resolution
by preferential crystallization making the entrainment resolution
of ( )-2 an attractive method for preparing (R)-1. Herein, we report
a simple and efficient procedure to obtain chlorides (R)-2 and (S)-2
by direct crystallization from methanolic supersaturated solutions
of the racemate.
NH2
3
2
OH
O-
(R)-1
Recently, we have characterized the enantiomer systems
formed by some inorganic salts of carnitine amide 2 proving that
the chloride is a conglomerate, the nitrate is a racemic compound
2. Results and discussion
On the basis of the previously demonstrated conglomerate nat-
ure of chloride 2, we decided to attempt its resolution according to
an entrainment procedure, which involves the preferential precip-
itation of one enantiomer from a supersaturated solution of the
racemate enriched with a slight excess of the same enantiomer. Be-
fore performing the crystallization experiments, we measured the
solubility of ( )-2 chloride and of (R)-2 chloride in water, in meth-
anol and in ethanol. The results, reported in Table 1, show that
methanol, thanks to the moderate dissolving capability of 2, is a
crystallization solvent preferable to water and ethanol, where the
Table 1
Experimental solubility (grams of solute per 100 g of solution) of the chlorides of ( )-2
and of (R)-2
Solvent
T (°C)
Chloride ( )-2
Chloride (R)-2
Water
22
22
30
22
56
3.6
4.6
Methanol
Methanol
Ethanol
6.8
8.5
<0.3
a
% (S)-2
% (R)-2
% MeOH
chloride chloride
87.71
88.81
87.72
88.78
87.72
88.76
88.40
89.79
88.40
89.71
88.40
89.90
88.40
89.80
88.40
89.68
88.40
89.76
88.40
89.76
A
B
C
D
E
F
5.85
5.79
6.34
5.36
5.90
5.75
5.92
4.72
5.42
5.32
5.97
4.67
5.42
5.32
6.02
4.72
5.37
5.30
5.98
4.68
6.44
5.40
5.94
5.85
6.38
5.49
5.68
5.49
6.18
4.97
5.63
5.43
6.18
4.87
5.58
5.60
6.23
4.94
5.62
5.56
e
l
c
y
c
t
s
1
e
l
c
b
c
y
c
d
G
H
I
n
2
unsaturation
area
e
l
J
c
y
c
K
L
M
N
O
P
d
r
3
entrainment
area
e
l
c
y
c
h
t
4
Q
R
S
e
l
c
y
c
h
t
5
T
Figure 1. (a) Solubility diagram of chloride 2 in methanol at 30 °C. The concentrations of the components are expressed as weight percentages. (b) Unsaturation area (black
quadrilateral) and area usable for the entrainment (blue parallelogram) in the solubility diagram of 2 chloride. (c) Entrainment area with the ternary compositions describing
the five cycles of resolution of ( )-2 chloride (see Table 2). The points A-E-I-M-Q are the ternary compositions of the (R)-enriched supersaturated solutions of ( )-2 chloride,
the points C-G-K-O-S those of the (S)-enriched supersaturated solutions of ( )-2 chloride, the points B-F-J-N-R those of the mother liquors remaining from the crystallization
of (R)-2 chloride and the points D-H-L-P-T those of the mother liquors remaining from the crystallization of (S)-2 chloride. The respective terns of values (weight percentages)
are listed in the side table.

M. Pallavicini et al. / Tetrahedron: Asymmetry 19 (2008) 1637–1640
1639
Table 2
Resolution of ( )-2 chloride by entrainment
Cycle
Precipitation timea
Chloride 2 added (mg)
(R)
Recovery of resolved chloride 2b (mg)
a3D0 of the mother liquors
( )
(R)
(S)
1
2
3
4
5
120 (A?B)
140 (C?D)
150 (E?F)
460 (G?H)
310 (I?J)
750 (K?L)
420 (M?N)
800 (O?P)
560 (Q?R)
1050 (S?T)
4000
420
410
400
560
530
605
570
520
541
200
420 (79.6% ee)
+0.086
ꢁ0.074
+0.090
ꢁ0.125
+0.100
ꢁ0.108
+0.120
ꢁ0.150
+0.102
ꢁ0.131
410 (73.2% ee)
560 (64.8% ee)
605 (65.3% ee)
520 (91.2% ee)
400 (63.8% ee)
530 (75.1% ee)
570 (76.3% ee)
541 (80.3% ee)
510 (79.2% ee)
2605d (74.3% ee)
Total
8556
3510
200
2461c (75.5% ee)
Residuee
a
After the third crystallization (E?F), methanol was increased from 30 to 32 g. Between brackets, the ternary composition of the solution before and after the precipitation
(see Table in Fig. 1).
Between brackets, the enantiomeric purity resulting from the percent ratio of the observed specific rotation to the specific rotation {½a D20
¼ ꢁ24:5 (c 1, methanol)} of pure
ꢂ
b
chloride (R)-2.
c
d
e
1800 mg (100% ee) after recrystallization from methanol.
1874 mg (98.1% ee) after recrystallization from methanol.
Recovered, with 9.6% ee [(R) enriched], by concentration of the mother liquors remaining from the last crystallization.
solubility of the chloride is too high and too low, respectively. In
methanol, we found a 1.8 ratio between the solubilities, expressed
as a solute/solution weight percent, of ( )-2 and (R)-2. Such a va-
lue, lower than 2, is typical for conglomerates, whose molecules
are dissociable in solution as is the case for those of chloride 2.14
The ternary phase diagram of (R)-2 chloride/(S)-2 chloride/
methanol was constructed using values of 8.5% and 4.6% solubility
for chloride ( )-2 and chloride (R)-2, respectively, in methanol at
30 °C (see Fig. 1a) At this temperature, the weight percentage of
methanol for a saturated solution of chloride ( )-2 was 91.5, while
it was 95.4 for a saturated solution of chloride (R)-2. The resultant
unsaturation area is not represented by a parallelogram, but by a
quadrilateral whose two lower sides, constituting the solubility
curve, are not parallel to the sides of the triangle and intersect to
Fig. 1c); the procedure was repeated for the (S)-enantiomer. Again,
we isolated a crystalline, but now destrorotatory precipitate,
which contained an excess of chloride (S)-2 more than double
the (S)-enrichment of the filtered mother liquors of the previous
crystallization, while the filtered (R)-enriched mother liquors of
this second crystallization had now assumed composition D (see
Fig. 1c) and showed a negative value for the rotation. After the
third crystallization, the amount of methanol was increased from
30 to 32 g raising the crystallization time from 12–15 min to about
60 min for the subsequent seven resolutions. After five cycles, we
obtained 2.46 g of (R)-2 chloride with 75.5% enantiomeric purity
and 2.61 g of (S)-2 chloride with 74.3% enantiomeric purity (Table
2). Finally, these two quantities were recrystallized from methanol
to yield about 1.8 g of each enantiomer, 9 times the initial invest-
ment in chloride (R)-2, with >98% enantiomeric purity. The mother
liquors remaining from the last crystallization were laevorotatory
and, after concentration, afforded 3.51 g of chloride 2 enriched in
the levo enantiomer with 9.6% enantiomeric purity. These values
are consistent with those of the mixture initially subjected to res-
olution, that is, 4.2 g and 4.8% enantiomeric purity [4.0 g of chlo-
ride ( )-2 + 0.2 g of chloride (R)-2].
form an angle little wider than 60° (a <2) (see Fig. 1b).
In this triangular phase diagram, we defined the region in which
resolution by entrainment is favourable, that is, suitable conditions
of supersaturation for an efficient resolution by preferential crys-
tallization. A 14 wt % of chloride ( )-2, that is 1.6 times its solubil-
ity in methanol at 30 °C, was empirically established as
a
concentration corresponding to a sufficiently high, but neverthe-
less even metastable supersaturation, since without seeding, no
crystallization took place in a methanolic solution with such a con-
centration over several days at this temperature. Therefore, the
point corresponding to 7–7–86% ternary composition of (R)-2 chlo-
ride-(S)-2 chloride–methanol was fixed as a supersaturation limit
and used to graphically delimitate the area of the ternary diagram
useful for entrainment (see the blue parallelogram in Fig. 1b).
Within this area, we drew five consecutive crystallization cycles.
Figure 1c shows the ternary compositions (points A-T) assumed
by the system over the course of the cycles.
In the first cycle, chloride ( )-2 (4 g), enriched with a slight ex-
cess of chloride (R)-2 (0.2 g), was dissolved in boiling methanol
(30 g) (see ternary composition A in Fig. 1c) and slowly cooled
to 33 °C while stirring. The resulting supersaturated solution was
seeded with crystals of (R)-2 chloride, cooled further to 30–32 °C
and, after 12 min, filtered to isolate a laevorotatory precipitate,
containing an excess of (R)-2 chloride, which was 1.7 times the
initial (R)-enrichment. The filtered mother liquors, now having a
composition B (see Fig. 1c) and rotation changed to a positive va-
lue because of the (S)-enrichment, were added with chloride ( )-2
to restore the initial supersaturation (see ternary composition C in
3. Conclusion
The unreported resolution of chloride ( )-2 by preferential crys-
tallization according to the entrainment method was tried and
developed successfully. The main advantages of such a new proce-
dure, in addition to the typical ones of entrainment (very simple
operations and no chiral auxiliaries), can be summarized as fol-
lows: (i) the only necessary materials are methanol and the easily
available, inexpensive and very stable chloride of ( )-2; (ii) trans-
formation of this latter into other salts, such as mono-oxalate or
perchlorate, or specific solvent systems is not required for success-
ful resolution, as in the case of the carnitine nitrile entrainment;
(iii) the resolution efficiency (chemical yield x enantiomeric ex-
cess) is absolutely good (approx. 20% medium recovery of total
enantiomer at each crystallization) and highly reproducible.
Acknowledgements
This work was supported by the Ministero dell’Istruzione,
dell’Università e della Ricerca of Italy. Thanks are due to Sigma-

1640
M. Pallavicini et al. / Tetrahedron: Asymmetry 19 (2008) 1637–1640
4. de Witt, P.; Diamanti, E. US Patent 4,254,053, 1979.; . Chem. Abstr. 1980, 93,
114973.
5. Hwang, S. O.; Ryu, H. Y.; Chung, S. H. WO 139238, 2007.
6. Hwang, S. O.; Chung, S. H. WO 108572, 2007.
Tau S.p.A. (Pomezia, Italy) for providing a sample of (R)-carnitina-
mide chloride.
7. Kamal, A.; Khanna, G. B. R.; Krishnaji, T. Helv. Chim. Acta 2007, 90, 1723–1730.
8. Lorenz, I. J. Prakt. Chem. (Leipzig) 1980, 322, 785–792.
9. Kasai, N.; Sakaguchi, K. Tetrahedron Lett. 1992, 33, 1211–1212.
10. Shiba-Saki, M.; Ohnoki, J.; Hongo, C.; Shibata, K. JP 19680008713, 1968.
11. Jacques, J.; Brienne, M. J.; Collet, A.; Cier, A. FR 2536391, 1984.
12. Pallavicini, M.; Bolchi, C.; Fumagalli, L.; Piccolo, O.; Valoti, E. Tetrahedron:
Asymmetry 2007, 18, 906–909.
13. Xu, B.; Cheng, G. Huaxue Shijie 2000, 41, 459–461.
14. Jacques, J.; Collet, A.; Wilen, S. Enantiomers, Racemates and Resolutions; Wiley:
New York, 1981. p 191.
References
1. Pallavicini, M.; Valoti, E.; Bolchi, C.; Fumagalli, L.; Piccolo, O. EP 6325304.0,
2006.
2. Bernal, V.; Sevilla, A.; Canovas, M.; Iborra, J. L.. Microbial Cell Factories 2007, 6,
31.
3. Naidu, G. S. N.; Lee, I. Y.; Lee, E. G.; Kang, G. H.; Park, Y. H. Bioprocess Eng. 2000,
23, 627–635.