Communications
tion was made on testing the milling stability of acetate rac-3a
in the presence of the enzyme. Under these conditions, a par-
tial hydrolysis occurred, which proved to be highly enantiose-
lective, providing alcohol (R)-1a with 98% ee (at 14% conver-
sion; the remaining acetate (S)-3a had 16% ee).[10,11] This result
was very important because it revealed a substantial degree of
stability of the biocatalyst under the milling conditions.
In the subsequent set of reactions (Scheme 1, bottom reac-
tion branch), various amounts of immobilized CALB were
added to a combination of rac-1a (0.409 mmol) and acyl
donor 2, and the mixture was then milled in a 10 mL ZrO2 jar
in the presence of one ZrO2 ball of 10 mm in diameter at
25 Hz for a given period of time (Table 1). As assumed, this
change of the reaction conditions resulted in a major improve-
ment, and now, acetate 3a was indeed formed. Applying 5 mg
of the enzyme led to a 37% conversion of rac-1a, affording
acetate (R)-3a with >99% ee (Table 1; entry 2).[12,14] Gradually
increasing the amount of CALB (from the initial 5 mg to a maxi-
mum of 60 mg) had a significant effect on both the conversion
of rac-1a and the stereocontrol in the kinetic resolution pro-
cess. Whereas in each case the formation of (R)-3a proceeded
with excellent enantioselectivity (>99% ee), both the conver-
sion of rac-1a as well as the ee of the remaining alcohol (S)-1a
first rose and then decreased (see Supporting Information).
Usage of 30 mg of CALB proved optimal under these condi-
tions and led to a 47% conversion of rac-1a providing (R)-3a
and (S)-1a with >99% and 90% ee, respectively, which corre-
sponded to an E value of >200 (Table 1, entry 3). With less (0.5
or 0.75 equiv.) or more (1.5 or 2.0 equiv.) of the acyl donor the
ee of (S)-1a was affected negatively (see ESI). Whereas milling
times of 30 min or 1 h (instead of the standard 3 h) led to
lower conversions of rac-1a, a shortening to 2 h or an exten-
sion to 4 h proved possible without significantly altering the
overall results (Table 1, entries 3–7).[15]
Figure 1. Recyclability of immobilized CALB in the mechanochemical kinetic
resolution of rac-1a in the ball mill at 25 Hz for 3 h (according to the reac-
tion conditions shown in Scheme 1, bottom).
with >99% ee. In the case of heterocyclic secondary alcohols,
such as 1-(2-furyl)ethanol (1h) and 1-(2-thienyl)ethanol (1i), the
conversion was high, but the lipase was less selective leading
to acetate (R)-3h and (R)-3i, with only 93% ee (E=83) and
94% ee (E=110), respectively. 1-(4-Pyridyl)ethanol (1j) proved
to be compatible with the mechanochemical enzymatic resolu-
tion conditions as well, yielding (R)-3j with excellent enantiose-
lectivity (99% ee). CALB was also selective to catalyze the acy-
lation of the racemic aliphatic secondary alcohol (1k) affording
(R)-3k with 98% ee. The racemates of tetrahydro-1-naphthol
(1l) and 2-naphthyl ethanol (1m) reacted with very good ster-
eopreference, too. In contrast, the kinetic resolution of the
sterically more hindered 1-naphthyl ethanol (rac-1n) was very
slow, and after 4 h of milling only traces of acetate (R)-3n were
observed.[19] Likewise, the reaction of 1-phenylpropanol (rac-
1o) proved difficult. After 4 h of milling in the presence of
1.5 equiv. of 2 the conversion reached only 22%, but delight-
fully, the product (R)-3o was formed with very high enantio-
selectivity (97% ee).
Immobilized biocatalysts have been recovered, retaining
their activity and specificity.[17] For CALB adsorbed on a granular
acrylic resin, however, mechanical erosion has been ob-
served.[18] To test the recyclability here, the immobilized
enzyme was recovered after a standard mechanochemical ki-
netic resolution of rac-1a (for 3 h at 25 Hz) by centrifugation
and reapplied in the same process. Although the beads were
pulverized under the mechanical stress, the enzyme remained
active. As shown in Figure 1, the conversion of rac-1a de-
creased in the four mechanochemical cycles, and, thus, the ee
of (S)-1a dropped (to 27% ee in the 4th cycle) accordingly. In
contrast, the enantioselectivity in the formation of (R)-3a
remained excellent (>99 ee) in each run.
The scalability of the mechanochemical kinetic resolution
was demonstrated in a gram-scale experiment (20-fold scaled
up), with rac-1a performed in a planetary ball mill, using
a 45 mL ZrO2 milling vessel and three ZrO2 milling balls of
10 mm in diameter.[20] After 3 h of milling at 600 rpm, the con-
version of 1a was 49% and both acetate (R)-3a as well as alco-
hol (S)-1a were isolated in 44% yield. The enantioselectivity
was very high (eeacetate >99%; eealcohol 98%), corresponding to
an E value of >200 (Scheme 2).
Next, the substrate scope of the mechanochemical enzymat-
ic kinetic resolution was evaluated. Again, isopropenyl acetate
(2) served as acyl donor (Table 2). Aromatic alcohols rac-1b
and rac-1c with electron-donating groups on the arene react-
ed well and afforded the corresponding acetates [(R)-3b and
(R)-3c] with excellent enantioselectivities after 3 h of milling
(>99% ee). Also substrates with electron-withdrawing sub-
stituents (rac-1d-f) led to highly enantioenriched acetates [(R)-
3d–f]. CALB was also effective in catalyzing the acylation of
the racemic propargylic alcohol 1g, yielding acetate (R)-3g
Scheme 2. Gram-scale mechanochemical enzymatic resolution of rac-1a.
ChemCatChem 2016, 8, 1769 – 1772
1770
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim