Nucleation inhibition in attrition-enhanced Pope-Peachey type of diastereomeric resolutions

Article abstract of DOI:10.1016/j.tetasy.2009.05.025

The resolution of racemic 1-(3-methoxyphenyl)ethylamine with (S)-mandelic acid by diastereomeric salt formation can be made more economical by application of the Pope-Peachey method. This resolution was further improved upon by the addition of small amoun

Full text of DOI:10.1016/j.tetasy.2009.05.025

Tetrahedron: Asymmetry 20 (2009) 1363–1364
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Nucleation inhibition in attrition-enhanced Pope-Peachey type
of diastereomeric resolutions
a
b
a,*
Michel Leeman , Bernard Kaptein , Richard M. Kellogg
a
Syncom BV, Kadijk 3, 9747 AT Groningen, The Netherlands
DSM Pharmaceutical Products, Innovative Synthesis & Catalysis, PO Box 18, 6160 MD Geleen, The Netherlands
b
a r t i c l e i n f o
a b s t r a c t
Article history:
The resolution of racemic 1-(3-methoxyphenyl)ethylamine with (S)-mandelic acid by diastereomeric salt
formation can be made more economical by application of the Pope-Peachey method. This resolution was
further improved upon by the addition of small amounts of a nucleation inhibitor, which prevents the
formation of crystals of the more soluble diastereomer. The mandelic acid left in the solution is then used
for further crystal growth of the less soluble diastereomer, furnishing high yields and diastereomeric
excesses. Slow cooling and grinding of the crystals formed lead to increased secondary nucleation and
thus to more consumption of the less soluble diastereomer and even better results.
Received 19 March 2009
Accepted 26 May 2009
Available online 21 June 2009
Ó 2009 Elsevier Ltd. All rights reserved.
1
. Introduction
A relatively new insight into diastereomeric resolutions is that
inhibitor can be used and/or even higher concentrations can be
applied.
A variation of a classical resolution frequently applied on an
industrial scale is the use of half of the resolving agent (in practice
usually 0.7 equiv) or optionally the Pope-Peachey approach in
which half of the resolving agent is replaced by a (low cost) achiral
nucleation inhibitors that resemble structurally either the resolv-
ing agent or racemate can effectively inhibit the crystallization of
the more soluble diastereomer in a resolution. Application of this
principle in classical resolutions has been described for several sys-
tems.1 Recently we described this principle in detail for the res-
olution of racemic 1-(3-methoxyphenyl)ethylamine (3MeOPEA)
with (S)-mandelic acid (MA) as depicted in Figure 1. The standard
resolution at high concentration and using 1 equiv of MA resulted
in a solid in 72% yield and 10% de (thus to 40% less soluble diaste-
reomer and 32% of the more soluble diastereomer). The mother li-
quor (28%) was then saturated in both diastereomers with a
eutectic composition of 29% de. However, on addition of a small
amount (as little as 1 mol %) of (S)-2-acetoxy-2-phenylacetic acid
6
,7
acid or base. The role of the achiral acid or base is to form highly
soluble salts with the enantiomer remaining in the solution. Note
that on crystallization of the less soluble diastereomer the resolv-
ing agent is depleted in the solution and the solution equilibria
shift to the salts of the achiral acid or base. The crystallization yield
will never exceed 50%—the maximum yield of a perfect resolution
without racemization. Since only half of an equivalent of the
resolving agent is available the solubility of the diastereomeric salt
will also be increased and therefore the resolution conditions need
to be changed by using higher concentrations or less polar (poorer)
solvents. A search for a solvent that would dissolve the less soluble
salt, (S)-3MeOPEA-(S)-MA at a reasonable concentration at reflux
temperature but only a small part of this salt at room temperature,
led us to toluene as the most suitable solvent.
–3
(
AcMA), one of the best nucleation inhibitors for this resolution,
the crystallization of the more soluble diastereomer was inhibited
for more than three weeks. The result was that the diastereomeric
excess (de) of the isolated salts increased from 10% to 97%.
4
2
. Results and discussion
OH
(
S) CO H
Inhibition of the crystallization of the more soluble diastereo-
NH2
2
less soluble salt: (S)·(S)
mer under these conditions by addition of only 1 mol % of (S)-
MeO
MA
5
+
AcMA results in a supersaturation of 278% , the result being that
the less soluble diastereomer is obtained in 42% yield and 97% de.
The resolution might be further improved if less of the resolving
agent remained in the solution. In this way, less of the nucleation
OAc
more soluble salt (R)·(S)
3
MeOPEA
(
S) CO H
2
AcMA
*
Corresponding author. Tel.: +31 0 50 5757272; fax: +31 0 50 57577399.
Figure 1. Resolution of (±)-3MeOPEA with (S)-MA, using (S)-AcMA as a chiral
E-mail address: r.m.kellogg@syncom.nl (R.M. Kellogg).
nucleation inhibitor.
0
957-4166/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2009.05.025

1
364
M. Leeman et al. / Tetrahedron: Asymmetry 20 (2009) 1363–1364
Achiral acids were screened to deliver a salt with (R)-3MeOPEA
tions in a test tube without slow cooling or grinding, will most
likely result in far from optimal (thermodynamic) results as a result
of higher supersaturation. Faster cooling might be possible if the
grinding action on the first formed crystals is more intense, for
example, by the addition of solid glass beads and vigorous stirring.9
Entries 2 and 3 in Table 2 show that addition of a small amount
of (S)-AcMA seems to result in more reproducible results and in
higher de’s as a result of the widening of the metastable zone
width of the more soluble diastereomeric salt.
and which would dissolve in toluene at high concentrations. Acetic
acid was chosen as the achiral acid since it is cheap, non-toxic,
readily available, and produced oils that did not crystallize, at least
within a period of 3 weeks.
In the first experiments, a half equivalent of AcOH, a half equiv-
alent of (S)-MA, varying amounts of (S)-AcMA (replacing the AcOH)
as a nucleation inhibitor, and 1 equiv of racemic 3MeOPEA were
mixed with toluene in stirred reactor tubes, heated at reflux
(
7
ꢀ111 °C) to dissolution, and the tubes were stirred for 30 min at
0 °C during which time crystallization started. The tubes were
3. Conclusion
À1
subsequently stirred and further cooled to 20 °C at 0.1 °C min
.
After eight more hours of stirring at 20 °C, the solids were col-
The results in Table 1, entry 1 represent the thermodynamic
end-state of the resolution. The resolutions that give better results
are kinetically steered and after prolonged stirring (days–years),
these will produce the same thermodynamical end-state. Thus, in
large-scale Pope-Peachey resolutions, a precise cooling profile,
grinding of the less soluble crystals, seeding, and/or addition of a
nucleation inhibitor might be essential to obtain reproducibly good
results. The addition of a small amount of a proper nucleation
inhibitor can dramatically increase the success rate even without
precise cooling.
lected, washed, and the yields and de’s were determined as shown
8
in Table 1.
Table 1
The effect of (S)-AcMA on the resolution efficiency
Yielda (%)
dea (%)
S-factorb
Entry
(S)-AcMA (%)
1
2
3
4
None
0.01
0.10
1.00
47
46
47
43
54
56
61
97
0.51
0.50
0.56
0.83
Acknowledgments
a
The results are averages of duplicates.
S-factor = 2  yield  de.
Partial financial support for this work was provided by the
Samenwerkingsverband Noord- Nederland (Cooperation Northern
Netherlands) and the European Fund for Regional Development
b
(
EFRO).
For entry 1 in Table 1, no additive was added and, as explained
in the introduction, the crystalline salt cannot be isolated in more
than 50% yield. As expected, the isolated salts only have moderate
de’s. However, as seen in entry 4, only 1% of (S)-AcMA blocks the
nucleation of the more soluble diastereomer (R)-3MeOPEA-(S)-
MA and almost all of (S)-MA in the solution is used for the precip-
itation of the less soluble (S)-3MeOPEA-(S)-MA yielding nearly
enantiomerically pure material. Lower concentrations of (S)-AcMA
References
1. Nieuwenhuijzen, J. W.; Grimbergen, R. F. P.; Koopman, C.; Kellogg, R. M.; Vries,
T. R.; Pouwer, K.; van Echten, E.; Kaptein, B.; Hulshof, L. A.; Broxterman, Q. B.
Angew. Chem., Int. Ed. 2002, 41, 4281.
2.
Dalmolen, J.; Tiemersma-Wegman, T. D.; Nieuwenhuijzen, J. W.; van der Sluis,
M.; van Echten, E.; Vries, T. R.; Kaptein, B.; Broxterman, Q. B.; Kellogg, R. M.
Chem. Eur. J. 2005, 11, 5619.
3.
Addadi, L.; Weinstein, S.; Gati, E.; Weissbuch, I.; Lahav, M. J. Am. Chem. Soc.
(
entries 2 and 3) were not effective with this setup.
1982, 104, 4610.
On the basis of a recent discovery9 of attrition-enhanced
4. Leeman, M.; Brasile, G.; Gelens, E.; Vries, T.; Kaptein, B.; Kellogg, R. Angew.
Chem., Int. Ed. 2008, 47, 1287.
deracemization, we turned our attention to a different protocol.
5.
A resolution without nucleation inhibitor yielded 18% (total 50%) of the more
soluble diastereomer in the saturated mother liquor. The remaining 32% of the
more soluble diastereomer had thus crystallized. This resolution with 1% (S)-
AcMA as nucleation inhibitor where the more soluble diastereomer did not
crystallize therefore produced a supersaturation of: 50%/18% Â 100% = 278%,
relative to 100% for a saturated solution.
When the mixtures were heated to dissolution at 100 °C and then
À1
cooled to 20 °C at 0.1 °C min , instead of fast cooling from 111 °C
À1
to 70 °C and subsequently cooled to 20 °C at 0.1 °C min , the res-
olution efficiency increased dramatically as shown in Table 2.
6.
Jacques, J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates and Resolutions;
Krieger: Malabar, Florida, 1994.
Table 2
The effect of (S)-AcMA with slow crystallization
7. Pope, W. J.; Peachey, S. J. J.Chem. Soc. 1899, 75, 1066.
8.
A typical Pope-Peachey resolution experiment was performed by charging a
Kimble reactor tube (Ø 25 Â 150 mm) with a PTFE-coated egg-shaped magnetic
stirring bar (19 mm  10 mm), (±)-3MeOPEA (322 mg, 2.13 mmol, 1.0 equiv),
Entry
(S)-AcMA (%)
Yielda (%)
dea (%)
S-factorb
1
2
3
None
0.05
0.10
42–43
41–42
41–42
92–99
98–99
99
0.79–0.83
0.81–0.83
0.81–0.83
(
S)-MA (162 mg, 1.06 mmol, 0.5 equiv), and AcOH (63 mg, 1.06 mmol,
0.5 equiv) in toluene (10 mL). This mixture was stirred and after some
minutes, crystals started to form. When additives were used, an equimolar
a
amount of AcOH and toluene were replaced by
a
solution of (S)-AcMA
The results are extremes of triplicates.
S-factor = 2  yield  de.
À1
b
(41.35 mgÁ10 mL toluene) ensuring that the whole system remained neutral
and of equal volume. The suspension was heated at reflux (dissolution) and
placed in
a Reactiv8 computer-controlled reactor station and stirred
As can be concluded from the differences between Table 1, en-
magnetically at 600 rpm and at 70 °C (Table 1) or 100 °C (Table 2) for
30 min. The tubes were then cooled to 20 °C at 0.1°/min and kept at 20 °C for an
additional 8 h. The crystals formed were collected on pre-weighed disposable
try 1 and Table 2, entry 1 the temperature profile used in the res-
9
olution is essential. This can be explained by the crystallization of
filters and washed with 2 Â 2.0 mL Et
2
O. The solids were dried in vacuo,
the less soluble diastereomer together with the grinding action of
weighed, and subsequently the de of the salts was determined. Chiral HPLC
1
0,11
the magnetic stirrer,
which delivers more crystal surface (sec-
analysis of 3MeOPEA salts was carried out on a Crownpak CR(À) column with
À1
an aqueous solution of HClO
4
(pH 2) as eluent at 20 °C and 0.6 mL min . UV–
ondary nucleation) and promotes the consumption of the (S)-MA in
the solution, therefore reducing (but not eliminating) the supersat-
uration of the more soluble diastereomer. The results seen here are
thus strictly kinetic in nature.
During the cooling, the crystallization was observed. In Table 2,
entry 1, the reaction mixture with the lowest de (92%), started to
crystallize at 84 °C whereas both reaction mixtures with >98% de
started to crystallize above 85 °C. Screening Pope-Peachey resolu-
vis detection was performed at 192 nm. The salts were dissolved in eluent and
injected as such. (R)-3MeOPEA R : 39.59 min, (S)-3MeOPEA R : 42.98 min.
f
f
9. (a) Noorduin, W. L.; Izumi, T.; Millimaggi, A.; Leeman, M.; Meekes, H.; van
Enckevort, W. J. P.; Kellogg, R. M.; Kaptein, B.; Vlieg, E.; Blackmond, D. B. J. Am.
Chem. Soc. 2008, 130, 1158; (b) Leeman, M.; Noorduin, W.L.; Kaptein, B.; Vlieg,
E.; Meekes, H.; van Enckevort, W.J.P.; Zwaagstra, K.; de Gooijer, J.M.; Boer, K.;
Kellogg, R.M.; submitted for publication.
1
1
0. Kondepudi, D. K.; Kaufman, R. J.; Singh, N. Science 1990, 250, 975.
1. McBride, J. M.; Carter, R. L. Angew. Chem., Int. Ed. 1991, 30, 293.