Full Papers
Next, the effect of the starting concentration of the secon-
dary alcohol was studied by varying its concentration to 0.25,
0.50, and 0.75m. A slow resolution was obtained with the
lowest starting concentration, with a difference in conversion
of almost 15% compared with the other concentrations. The
conversions achieved using concentrations of 0.50 and 0.75m
were basically the same. Nevertheless, 0.50m was chosen in-
stead of 0.75m because it provided a higher E (901 vs. 554,
Table 1, entry 2 vs. entry 5). Additionally, the use of 0.75m gen-
erates a thick and opaque reaction medium, which could be
problematic for the solubilization of other substrates.
1 in a modest yield of 60%, most likely owing to increased
degradation of the ionic acid 1 during the concentration step.
Substrate scope and enzyme reuse
The next step was to apply the optimized method to various
secondary alcohols in which both enantiomers are valuable
(Figure 4). 1-Phenylethanol (3), 4-phenyl-2-butanol (4), 1-cyclo-
hexylethanol (5), and 2-octanol (6) are important chiral build-
ing blocks for agrochemicals, pharmaceuticals, and natural pro-
ducts.[1b,8a] Sulcatol (6-methylhept-5-en-2-ol; 7) is a pheromone
used for pest control of Gnathotrichus retusus and Gnathotri-
chus sulcatus as the pure (S)-enantiomer and as a controlled
To investigate the effect of the temperature, the kinetic reso-
lutions were performed at 30, 35, and 408C. As expected, the
conversion improved with higher reaction temperatures. The
enantioselectivity of the enzyme was not disturbed by the in-
crease in energy available in the system. It was possible to ach-
ieve a conversion of almost 50% with a very high E value (698,
Table 1, entry 7) at 408C.
Finally, a survey with other similar coupling agents, such as
dicyclohexylcarbodiimide (DCC) and N-(3-dimethylaminoprop-
yl)-N’-ethylcarbodiimide hydrochloride (EDC), was performed to
assess if they could be applied in this method. Unfortunately,
both coupling agents were not as efficient and practical as DIC
(Table 1, entries 8 and 9 vs. entry 7). EDC, a salt, was mostly in-
soluble in acetone and consequently the yield of the coupling
reaction was low, which led to a poor conversion (11%,
Table 1, entry 8). Conversely, DCC is almost as soluble as DIC in
acetone so the anhydride formation was successful and
a good conversion was achieved, although it was not as high
as the conversion with DIC (43% vs. 49%, Table 1, entries 9
and 7). In addition, DCC was difficult to transfer to the reaction
medium because it was a waxy solid, and the dicyclohexylurea
that is formed during the coupling reaction is not soluble in
acetone, unlike the diisopropylurea byproduct, which is partial-
ly soluble. The high amount of dicyclohexylurea suspension in-
creases the viscosity of the reaction medium, probably limiting
the mass transfer mechanisms and consequently slowing
down the enzymatic resolution rate.
Figure 4. Secondary alcohols employed in the EKRs using in situ ionic anhy-
dride 2 as acylating agent.
mixture of both enantiomers in a (S)/(R) ratio of 65:35, respec-
tively.[13] 2-Hydroxycyclohexanecarbonitrile (8) is a key precur-
sor to an androgen receptor antagonist, which is in develop-
ment for the treatment of alopecia and excess amounts of
sebum (oily skin).[14] For these substrates, the reaction scale
was increased from 0.41 mmol to 0.82 mmol of alcohol
(ꢀ100 mg of substrate). For all alcohols, the resolution was
first followed by GC to determine the specific time to achieve
a conversion of approximately 50% (14 h for 3, 7 h for 4, 21 h
for 5, 2 h for 6 and 7, and 10 h for 8; see Supporting Informa-
tion for more details). This information was used for the EKRs
presented in Table 2.
Because the diisopropylurea formed during the coupling re-
action was partially soluble in acetone and ether, further purifi-
cation of the recovered enantiomers was required to isolate
the products. Fortunately, the byproduct was insoluble in alka-
nes, so after each organic extraction it was only necessary to
evaporate the solvent and extract the obtained mixture with
n-hexane to obtain pure alcohols.
The resolutions resulted in very high enantiomeric ratios
(321ꢁEꢁ711) for almost all substrates studied. The only ex-
ception was 8 (E=170), although it was still one of the best
enantiomeric ratios achieved so far for this substrate.[5c,8b,14]
Very high conversions, between 46% and 48%, were achieved
for all substrates. The enantiomeric excess of the (S)-enantio-
mers, recovered after the first extraction with ether, were good
(87% on average). The best enantiomeric excess (92%) was
obtained for sulcatol. The enantiomeric excess of the (R)-enan-
tiomers, recovered after hydrolysis and a second organic ex-
traction, was all above 97%. The isolated yields, which were
obtained without the need for a chromatographic step, were
very good (between 48% and 53%) for the (S)-enantiomer,
and moderate to good (31–40%) for the (R)-enantiomer. It was
Another important feature of the method described herein
is the possibility to recover the ionic acid 1 (after hydrolysis
and extraction of the (R)-enriched enantiomer) and to reuse it
immediately. To maximize the amount of 1 recovered, different
protocols for the removal of 1 from the aqueous acidic
medium were explored: i) Extraction with dichloromethane,
which provided a yield of 73%; ii) saturation of the aqueous
medium with NaCl followed by extraction with dichlorome-
thane, which gave 82% of 1; and iii) evaporation of aqueous
medium under reduced pressure followed by solid–liquid ex-
traction with dichloromethane, which allowed the isolation of
&
ChemSusChem 2016, 9, 1 – 8
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Rocha, ?ngelo
