Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
A. M. Foley et al. / Tetrahedron: Asymmetry xxx (2017) xxx–xxx
3
resolution (a phenyl group at b-position is favourable),19 the length
of the alkyl chains (longer chains tend to give better selectiv-
ity),15,42 and the electronic effects of the acyl chain (with isomeric
pentenoic acids as acyl donors it was shown that the position of
the double bond is important).15
Enantioenriched 2-aryl propanols resolved using lipases have
been used in the synthesis of sesquiterpenes28 and fragrance
molecules.33,34
The resolution of 2-phenylpropyl benzoate 2a by enzymatic
hydrolysis with just four lipases has previously been reported.18
Conversion of up to 39% was achieved with the highest ee for the
alcohol being 88%.
lipase was subjected to a limited solvent screen, including 1-octa-
nol (Table 1, entry 16), which furnished ees = 27%, eep = 20%,
c = 42%. While this is poorly selective overall, it shows a significant
increase in selectivity compared to the outcome in the absence of
the organic cosolvent. The resolution was also attempted using
the transesterification reaction, since the selectivity can be chan-
ged by changing the mode of reaction.48 The transesterification
reaction with vinyl benzoate gave either no conversion or full con-
version, depending on the lipase used. The results can be found in
the ESI.
Herein, some loss of 2-phenyl-1-propanol 1a was observed on
prolonged evaporation of the solvents. Accordingly, Ecalc [Ecalc
=
Investigations into the impact of smaller acyl groups, aside from
acetate,28 are uncommon, with very few reports in the resolution
of 2-phenyl-1-propanol 1a and with poor selectivity
(E < 10).25,43,44 Herein we examined lipase mediated resolution
with small acyl groups as a means to resolve 2-phenyl-1-propanol
1a, which are more atom economical than use of larger acyl groups,
and hold a clear advantage in product isolation, as the acid by-pro-
duct can be removed in aqueous work up.
The preparation of derivatives of 2-phenyl-1-propanol 1a with
larger alkyl substituents 1b and 1c by lipase catalysed reaction
has rarely been investigated. 2-Phenyl-1-butanol 1b has previously
been prepared by hydrolysis of the corresponding acetate, with
poor E-value (E = 2).25 There are a few reports of the transesterifi-
cation of 2-phenyl-1-butanol 1b with ee values of up to 69% for the
alcohol 1b.19,45–48 E-Values of up to 142 have been achieved, how-
ever, these resolutions require the use of acyl donors with large
acyl groups.19 3-Methyl-2-phenyl-1-butanol 1c resolution by
lipase-catalysed transesterification has been reported, usually with
much longer reaction times than for 1b, and with ee values of up to
74% for the alcohol.45–47
ees/(ees + eep)] is a more reliable indicator of the extent of conver-
sion than the 1H NMR in this instance.
In order to achieve high enantioselectivity with good conver-
sion, alternative ester groups were employed; firstly, it was
decided to test the commercial enzymes against 2-phenylpropyl
acetate 2b. Resolution by hydrolysis has previously been reported
but is poorly selective (E = 4) for 2-phenylpropyl acetate 2b.54
More common however, is the resolution of the acetate 2b by
transesterification, again with limited enantioselectivity under
standard conditions.28,40,45,46
The resolution of 2-phenyl-1-propanol 1a was attempted using
a smaller acyl group, acetate 2b. Before carrying out the transester-
ification reactions, the effect of molecular sieves was investigated
(data in ESI). The presence or absence of molecular sieves had little
effect on the transesterification reactions with vinyl acetate, with
molecular sieves increasing the conversion, but having very little
effect on the E value, which is consistent with previously reported
findings.22 The resolution of 1a was attempted using both transes-
terification and hydrolysis (Table 2) reactions. The transesterifica-
tion reactions were carried out in the absence of a lipase, as a
control, which showed no chemical acylation over 24 h despite
the high loading of vinyl acetate. Similarly, it was demonstrated
that hydrolysis did not occur without a lipase in the hydrolytic
reactions.
The extent of acylation is very sensitive to reaction time, with
reactions proceeding essentially to completion at longer reaction
times with some lipases. By careful control of the reaction time,
the extent of the transformation was optimised to ꢀ50%, although
with modest enantioselectivity in each case. The five lipases tested
gave >50% hydrolysis product after 65 h, showing that the reaction
was time sensitive. The times were reduced but the resolutions
were not selective enough when the acetate 2b was the substrate.
This is not unexpected, as acetate 2b provides limited selectivity at
30 °C, however, this screen provided an important reference
point.18,28,40,45,54,55
In the context of the high selectivity but poor conversion using
the pivalate ester 2c and the fast reaction but poor selectivity using
the acetate ester 2b, isobutyrate ester 2d was next explored.
In order to explore the impact of a more sterically demanding
ester, the pivaloyl ester, 2-phenylpropyl pivalate 2c, was prepared
2. Results and discussion
The esters 2a–f were prepared from the corresponding alcohols
and acid chlorides using literature methods49 as shown in
Scheme 1, where 2a and 2b are previously reported com-
pounds,49,50 and 2c–f are novel compounds. Alcohols 1a and 1b
are commercially available, while 1c was prepared by literature
methods.51
The resolution was first attempted using hydrolytic reactions
with 2-phenylpropyl benzoate 2a as substrate, which was screened
using a targeted panel of lipases,52 as previous studies were limited
to only a few lipases.18 Of the 52 enzymes tested, 15 gave no con-
version, 22 gave conversion under 10%, while 4 gave 100% conver-
sion with no selectivity. A selection of the results are shown below
(Table 1). The lipase from Pseudomonas cepacia (Table 1, entry 4)
furnished the product alcohol with excellent enantioselectivity,
which surpassed previously reported results (E = 42, cf. 28 highest
to date).18 The lipase from Candida cylindracea (Table 1, entry 15)
resulted in good conversion but without selectivity. The use of this
R
∗
t
i
OMe
OMe
KO Bu, PrBr
LiAlH4
Ph
Ph
Et2O
6h
100%
O
O
DMF, 0°C→RT, 1h
i
4
5
R = Pr
68%
R = Me, R'= Ph 2a 66%
R
R
*
Me 2b 55%
t
*
1a
1b
Cl
R'
R = Me
Et
2c
68%
NEt3 or pyridine and DMAP
Bu
OH
O
R'
iPr 2d 90%
Ph
Ph
iPr 1c
DCM, 0°→RT, 16-24h
i
O
E = Et, R' = Pr 2e 50%
O
i
E = Pr, R' = iPr
3a-d
2f
35%
35-90%
2a-f
Scheme 1. Preparation of ester substrates 2a–f.
Foley, Aoife M.
