M. Yasumoto et al. / Journal of Fluorine Chemistry 131 (2010) 535–539
537
knowledge of SDE via sublimation, which is clearly not as simple as
one may assume [9–12,14–19]. For instance, besides the basic
thermodynamic factors as energy/stability of crystals, kinetic
features, such a minute nuance as shape of crystals [13], and
therefore the sublimation surface, should be expected to have a
dramatic effect on the difference in the sublimation rates of
racemic and enantiomerically pure forms. We hope this paper will
stimulate discussions and raise an additional interest in this new,
fascinating and practically important area of research.
Acknowledgments
Fig. 5. Sublimation of samples of 1 of various enantiomeric composition (Petri dish).
The authors gratefully acknowledge generous financial support
from Central Glass Company (Tokyo, Japan) and Ajinomoto
Company (Tokyo, Japan).
The results or sublimation (% ee vs. time) of samples of
different enantiomeric composition are presented in Fig. 5. Thus,
regardless of the starting enantiomeric composition more racemic
fractions sublimed faster, eventually leaving behind enantio-
merically pure remainder. As one may see these results disagree
with the theoretical predictions made in the book [9]. For instance,
if, according to the predictions, the melting eutectic composition
of the compound of ca. 40% should sublime preferentially, then in
the case of samples of 36.8 and 58.7% ee (most close to 40% ee) the
SDE should be barely observed. Furthermore, in the case of
samples of 20.8 and 79.4% ee the order of SDE should be opposite,
leaving behind the racemic remainder in the former (20.8% ee)
case.
Experimental part
A.1. Large-scale preparation of 3,3,3-trifluoro-2-hydroxypropionic
acid
To the solution of 30 wt% aqueous sodium hydroxide (227 g,
From the practical stand point, these data demonstrate the
greater potential of simple sublimation technique for preparation of
optically pure compounds. For instance, purification of an enantio-
merically enriched sample of 1 of 75% ee to optically pure (S)-1 form
via conventional crystallization technique, requires several conse-
cutive re-crystallizations furnishing an enantiomerically pure
compound in about 60% yield [7]. On the other hand sublimation
ofsampleof1of79.4%eeover25 minresultedinopticallypure(S)-1,
which was collected with 81.7% yield. In this case no solvents, labor,
heating or cooling devises, filtration apparatus were used, as the
optical purification via sublimation was accomplished simply by
spreading the startingmixture over a Petri dish in the open air on the
laboratory bench. Sublimation of the samples of lower % ees were
less efficient in terms of the yield of enantiomerically pure
compound 1, however again, overall simplicity of the sublimation
approach over conventional crystallization was obvious. Further-
more, one may assume that for practical reasons, the sublimed
material can be collected and recycled via several sublimation
experiments furnishing the racemic and enantiomerically pure
fractions in virtually quantitative yield.
It should be emphasized that while the presence of fluoroalkyl
[24] or fluoroaryl [25] groups in organic compounds significantly
influences their physical properties, the difference in sublimation
rates between racemic and optically pure compounds is irrelevant
to the presence of fluorine-containing groups. Thus, in the same
order as well-known differences in melting points and solubility
between racemic and enantiomerically pure compounds, the
differences in sublimation rates is a physico-chemical conse-
quence of differences in the crystallographic structures of racemic
and optically pure crystals. Therefore separation of racemate from
the excess enantiomer via sublimation is ultimately general
phenomenon and can be used as an alternative approach for
optical purifications of any chiral organic or inorganic compound
which forms racemic and enantiomerically pure crystals.
1.70 mol),
1,1-dichloro-3,3,3-trifluoroacetone
hydrate
(100 g,
0.43 mol) was added drop-wise, keeping the solution temperature
under 30 8C. After the addition, the mixture was stirred for 12 h at room
temperature. 18% aqueous hydrochloric acid (172 g, 0.85 mol) and
ethyl acetate (200 mL) were added to the solution. After separation, the
organic layer was washed with brine (100 mL). The organic layer was
evaporate to obtain the 3,3,3-trifluoro-2-hydroxypropionic acid (55 g).
The crude product was used for the next step without any purification.
A.2. Preparation of isopropyl 3,3,3-trifluoro-2-hydroxypropanoate
To the solution of 2-propanol (69 mL), the crude 3,3,3-trifluoro-2-
hydroxypropionic acid (10 g, 69 mmol) and sulfuric acid (0.07 g,
0.7 mmol) were added. The mixture was refluxed for 24 h and the
solution was distilled. The main fraction was purified by sublimation
to obtain the racemic isopropyl 3,3,3-trifluoro-2-hydroxypropanoate.
A.3. Preparation of (S)-3,3,3-trifluoro-2-hydroxypropionic acid
To the solution of the crude 3,3,3-trifluoro-2-hydroxypropionic
acid (40 g, 0.28 mol) in ethyl acetate (120 mL) and n-heptane (12 mL),
(S)-a-(methyl)benzylamine (33 g, 0.28 mol) was added. The mixture
was heated at 60 8C for 30 min, then cooled to room temperature to
crystallize 3,3,3-trifluoro-2-hydroxypropionic acid salt (26 g, 68%ee).
The crystal was purified by further recrystallization (total three
3. Conclusion
In summary, we believe that the data presented here and the
recent literature results [20] clearly suggest a glaring gap in our