Migratory dynamic kinetic resolution of carbocyclic allylic alcohols

Article abstract of DOI:10.1021/ol502979g

A novel migratory dynamic kinetic resolution based on the interplay between an enzyme acylation catalyst and a heterogeneous Bronsted acid as an isomerization/racemization catalyst gives rise to carbocyclic allylic esters with excellent stereoselectivity

Full text of DOI:10.1021/ol502979g

Letter
pubs.acs.org/OrgLett
Migratory Dynamic Kinetic Resolution of Carbocyclic Allylic Alcohols
Chicco Manzuna Sapu,† Tamas Gorbe,‡ Richard Lihammar,‡ Jan-E. Backvall,*,‡ and Jan Deska*,†
́
̈
̈
†Department fur Chemie, Universitat zu Koln, 50939 Cologne, Germany
̈
̈
̈
‡Department of Organic Chemistry, Stockholms Universitet, 10691 Stockholm, Sweden
S
* Supporting Information
ABSTRACT: A novel migratory dynamic kinetic resolution based on
the interplay between an enzyme acylation catalyst and a heterogeneous
Brønsted acid as an isomerization/racemization catalyst gives rise to
carbocyclic allylic esters with excellent stereoselectivity from readily
available tertiary carbinols. An easy-to-use teabag setup combining resin-
bound catalysts, a biphasic isooctane−water solvent system, and a highly
lipophilic acyl donor efficiently suppresses side reactions and allows for
the preparation of functionalized carbocyclic building blocks in high yields and optical purity.
Scheme 1. Inherent Lability of Tertiary Allylic Carbinols:
Synthetic Limitation or Opportunity?
he use of lipases as inexpensive and reliable biocatalysts in
kinetic resolutions and desymmetrization reactions for the
T
preparation of enantiomerically enriched alcohols has evolved
into a fundamental and well accepted approach within the
synthetic-organic community.1 Both primary and secondary
hydroxy-functionalized compounds or their corresponding ester
derivatives can be engaged in enantiodiscriminating trans-
formations giving rise to optically active products in excellent
optical purities.2 Tertiary alcohols, on the other hand, proved to
be much more problematic, as the majority of hydrolases lack
both activity and selectivity for these kinds of structural
entities.3 However, some exceptions to this rule have been
reported for tertiary propargylic carbinols with a few serine
hydrolases, where the enzyme was able to productively
accommodate the substrates’ terminal alkyne moiety.4 As part
of our studies toward the total synthesis of taylofuran A, we
pursued investigations employing 1a as starting material en
route to the sesquiterpene natural product.5 With an ethynyl
substituent at the tertiary carbinol, we envisioned a lipase-
catalyzed hydrolytic kinetic resolution of the corresponding
acetate as entry to an enantioselective path. To our surprise,
Candida antarctica lipase B (CALB) yielded a highly
enantioenriched carbinol product beyond selectivities previ-
ously described for these substrate structures. Unfortunately,
spectroscopic analysis disclosed formation of a secondary
alcohol product (2a) rather than the desired 1a (Scheme 1,
top). The formation of 2a could be traced back to a 1,3-
transposition during the preceding, mildly acidic, acetylation
step in acetic anhydride/pyridine. Acid-catalyzed isomerizations
of allylic alcohols have recently been reported by Mulzer and
co-workers using trifluoroacetic acid,6 but obviously this
isomerization is also promoted by weaker acids. Our finding
that less acidic conditions can be used brought us to reflect
upon a way to exploit this reactivity in a more productive
fashion. The inherent lability of these allylic alcohols should
make them attractive targets for stereoselective transformations
taking advantage of this kind of dynamic behavior (Scheme 1,
bottom).7,8 The preparation of cyclic tertiary allylic carbinols is
often a very easy task (addition of C-nucleophiles to enones)
while the synthesis of their secondary isomers involves in most
cases multistep procedures. Hence, from a synthetic point of
view, the development of a catalytic process combining the
isomerization of tertiary to secondary allylic alcohols with a
subsequent dynamic kinetic resolution based on the interplay
between an acid-catalyzed racemization and a lipase-mediated
enantiodiscriminating acylation appeared to be a worthwhile
challenge, providing a rapid access to optically enriched,
synthetically valuable allylic carbinol derivatives.
The 1,3-transposition and the racemization of the secondary
alcohol both rely on the acid-catalyzed formation of transient
allylic cationic intermediates. Thus, on the basis of the model
substrates 1b and 2b, the efficiency of the 1,3-transposition
from carbinol 1 to the racemic secondary alcohol 2 as well as
the racemization of enantioenriched 2 (ee =99%) was
investigated employing different acidic ion-exchange resins.
Received: October 9, 2014
Published: October 30, 2014
© 2014 American Chemical Society
5952
dx.doi.org/10.1021/ol502979g | Org. Lett. 2014, 16, 5952−5955

Organic Letters
Letter
Pentane was initially chosen as the reaction medium, due to its
compatibility with the lipase-catalyzed acylation reaction, later
to be added. To our delight, carboxylic acid loaded Amberlite
120 was active in the isomerization of 1b (Table 1, entry 1) as
presence of water lipases can catalyze the reverse reaction, that
is, ester hydrolysis. As described by Jacobs and co-workers in
their work on the zeolite/lipase-catalyzed dynamic kinetic
resolution (DKR) of arylethanols,10 this problem can be solved
for instance by fixation of the enzyme in a rotating steel basket
inside the reaction vessel.11 However, we imagined that a
simple nylon fabric might serve as a practical alternative to the
steel basket. Once filled with the bead of the immobilized
enzyme, the nylon basket could be immersed into the organic
top layer like a catalytic teabag (Figure 1).12 Standard fine-
Table 1. Screening Catalyst Abilities for the Selective
a
Transposition and Racemization
2b
ee 2b
4b
entry alcohol
acid resin
solvent
pentane
(%)
(%)
(%)
1
rac-1b Amberlite
31
−
34
120
2
(S)-2b Amberlite
pentane
24
31
36
120
3
4
5
rac-1b Amberlyst 15 pentane
(S)-2b Amberlyst 15 pentane
39
9
−
5
26
35
31
rac-1b Dowex
50Wx4
pentane
26
−
6
(S)-2b Dowex
50Wx4
pentane
21
68
99
74
99
25
−
5
36
5
Figure 1. Schematic drawing and practical implementation of the
teabag approach.
7
rac-1b Amberlyst 15 isooctane/
H2O
8
(S)-2b Amberlyst 15 isooctane/
0
woven nylon mesh from the local department store (available
by the meter) proved to be a well-suited and cheap material for
this purpose that combined perfect immobilization of the
acrylic beads of Novozym 435 (CALB) with high flexibility to
adjust reaction scale and enzyme loading without being limited
to a certain size of a reaction vessel’s components.
H2O
9
rac-1b Dowex
50Wx4
isooctane/
H2O
−
22
0
10
(S)-2b Dowex
50Wx4
isooctane/
H2O
0
a
Reaction conditions: rac-1b (10 mg) or (S)-2b (10 mg, ee = 99), ion-
exchange resin (10 mg) in 4 mL of pentane or isooctane/water (1:1)
at room temperature. Product ratios were determined by H NMR;
enantiomeric excesses were determined by chiral HPLC.
The success of the envisaged migratory dynamic kinetic
resolution should apparently be strongly dependent on the
efficacy of separating the individual reaction steps. With
increased water solubility of the reaction components both
undesired background decomposition of the vinyl carboxylate
as well as acid-mediated (de)esterification ought to become
significant factors. Hence, lipophilicity of the acyl donor and
consequently lipophilicity of the resulting ester product proved
to be a key parameter in this regard. In a simple kinetic
resolution of rac-2b by CALB in anhydrous isooctane the chain
length of the donor had only marginal influence on the
selectivity with relative rates (k(R)/k(S)) greater than 200 in all
cases. However, initial attempts to combine the Dowex-
catalyzed isomerizations with a lipase-mediated acylation
employing the teabag setup revealed a strong dependence of
the products’ optical purity from the nature of the vinyl ester.
While the use of partially water-miscible vinyl acetate (log P =
0.7) led to acetyl ester 3b in only a moderate enantiomeric
excess of 81%, more lipophilic donors such as vinyl butyrate
(log P = 1.6) and in particular vinyl caprate (log P = 4.6)
helped to overcome this shortcoming and the latter one
allowed for an efficient dynamic kinetic resolution with good
yield and excellent enantiopurity of the caprate (R)-6b (ee =
99%) (Scheme 2). Estimation of the partition coefficients of the
carbocyclic reaction constituents further underlined the
beneficial effect of long-chain donors in this system by
efficiently raising the affinity of the reaction products for the
1
well as in the racemization of (S)-2b (Table 1, entry 2);
however, analysis of the reaction mixture after 5 h revealed
substantial contamination with several side products with ether
4b as the major constituent. Similar results were obtained using
the sulfonic acid bearing resins Amberlyst 15 and Dowex
50Wx4, respectively. Both showed good activity in the
isomerization reactions but likewise suffered from pronounced
ether formation (Table 1, entries 3−6). Providing water as the
nucleophile to quickly trap potential cationic species and thus
avoid dehydration product 4b, a two-phase system based on
isooctane and water as nonmiscible liquid layers was tested.10
In this setting, isomerization to the secondary alcohol 2b by
Amberlyst 15 proceeded with reasonable rates detecting only a
few percent of the Etheral side product (Table 1, entry 7). With
enantiopure (S)-2b as starting material, almost complete
racemization was achieved after 5 h and as opposed to the
purely organic solvent system perfect recovery of the secondary
alcohol was accomplished (Table 1, entry 8). Also Dowex
50Wx4 proved to be a suitable catalyst under the biphasic
reaction conditions, and while rates of racemization were
slightly reduced, the absence of any detectable 4b in either 1,3-
transposition or racemization prompted us to continue the
method development with this acid resin (Table 1, entries 9−
10).
nonaqueous layer as a safe haven (log P(2b) = 2.6; log P(4b)
=
The necessity of an aqueous cosolvent in the optimal
transposition/racemization system raised concerns about the
interference with the desired biocatalytic esterification, as in the
3.3; log P(6b) > 6.0).13
In order to survey the general applicability of this method, a
series of tertiary carbinols were synthesized by addition of
5953
dx.doi.org/10.1021/ol502979g | Org. Lett. 2014, 16, 5952−5955

Organic Letters
Letter
enantiopure conjugated enyne 6a was isolated in 82% yield.
As for 1a also cyclopentenol 1c required a lower acid content to
suppress side reactions. After these marginal adjustments, 6c
could be isolated in high yield and decent enantioselectivity.14
By contrast, the one-carbon ring expanded cycloheptenol 1d
isomerized much slower than 1b and extended reaction times
were required to achieve acceptable conversion. Nonetheless,
after 4 to 5 days not only cycloheptenyl caprate 6d but also
various cyclohexenyl derivatives carrying substituted aryl
moieties (6e−6g) were obtained in good to excellent yields
with optical purities beyond 99%. For the alkyl-substituted
substrate 1h the migratory DKR gave a low yield due to a very
slow isomerization to the secondary alcohol 2h. Finally, as
exemplified for styrene 6i the teabag protocol is also well
applicable for acyclic allylic carbinols and migratory DKR led to
(R)-6i in 82% yield and 99% enantiomeric excess.
Scheme 2. Influence of Acyl Donor Chain Length on Optical
Purity in the Migratory Dynamic Kinetic Resolution
organolithium or organomagnesium nucleophiles (RLi or
RMgX; see Supporting Information for experimental details)
to the corresponding enones. The racemic allylic alcohols 1a−i
thus obtained were subsequently subjected to an acid/lipase-
catalyzed migratory dynamic kinetic resolution (Scheme 3).
Our originally tested carbinol 1a proved to be substantially
more sensitive to the acid catalyst than cyclohexenol 1b, and
extensive substrate decomposition was observed under standard
conditions. By reduction of the Dowex loading by a factor of
10, however, an efficient DKR was obtained and the
The simple product isolation from the two-phase system in
combination with the easy recovery of the two heterogeneous
catalysts prompted us to perform recycling studies on this
dynamic resolution system (Figure 2). It was possible to reuse
Scheme 3. Scope of the Migratory Dynamic Kinetic
a
Resolution
Figure 2. Recyclability of the resin-bound catalysts in the migratory
dynamic kinetic resolution of alcohol rac-1b.
both Dowex 50Wx4 and CALB for more than six rounds of
catalysis in the migratory DKR of substrate 1b without
substantial loss of reactivity or selectivity. Here, at the end of
each reaction cycle the organic layer was removed and the
product (R)-6b was extracted from the aqueous layer
containing the heterogeneous acid. A new batch of substrate
and acyl donor in isooctane was added on top of the water layer
and the lipase-containing teabag was reimmersed into the
organic layer. Throughout the recyclability study, the
conversion rates were close to identical within the margins of
error with final conversions after 22 h between 80% and 90% in
all runs. In all cases, perfect optical purity for the caprate 6b (ee
>99%) was achieved.
In summary, by exploiting the lability of tertiary allylic
alcohols under mildly acidic conditions, a preparatively
convenient protocol for the synthesis of enantiopure secondary
carbocyclic esters was developed based on a migratory dynamic
kinetic resolution mechanism. Featuring a resin-bound acidic
catalyst in combination with immobilized lipases kept in a
handy teabag-like immobilization system, this chemoenzymatic
strategy represents an easy-to-use and straightforward approach
toward synthetically valuable carbocyclic building blocks.
a
Reaction conditions: rac-1 (0.5 mmol), vinyl caprate (2.5 mmol),
Dowex 50Wx4 (100 wt %), lipase (100 wt %), isooctane (20 mL),
water (20 mL), room temperature, 300 rpm. Isolated yields after flash
chromatography; optical purity determined by chiral HPLC. Capr =
caprate. b Reduced Dowex loading (10 wt %) in the case of 1a and 1c.
5954
dx.doi.org/10.1021/ol502979g | Org. Lett. 2014, 16, 5952−5955

Organic Letters
Letter
(13) Lipophilicity was estimated using the ALOGPS 2.1 online
analysis predictor ( www.vcclab.org): Tetko, I. V.; Gasteiger, J.;
Todeschini, R.; Mauri, A.; Livingstone, D.; Ertl, P.; Palyulin, V. A.;
Radchenko, E. V.; Zefirov, N. S.; Makarenko, A. S.; Tanchuk, V. Y.;
Prokopenko, V. V. J. Comput. Aid. Mol. Des. 2005, 19, 453−63.
(14) Lipase from Pseudomonas cepacia (Amano PS-D I, diatomite as
support) was used for this substrate due to a substantially higher
enantioselectivity (ECALB = 1.8, EPSL = 13.1, in the acylative kinetic
resolution of alcohol 2c using vinyl acetate). However, to achieve very
high optical purity for 6c, an in-depth enzyme screening would be
necessary.
ASSOCIATED CONTENT
■
S
* Supporting Information
Full experimental details for the preparation of the compounds,
characterization data sets, and copies of H and 13C NMR are
1
provided. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: jan.deska@uni-koeln.de.
*E-mail: jeb@organ.su.se.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge financial support from the Fonds
der Chemischen Industrie (Liebig fellowship of J.D.), the
Deutscher Akademischer Austauschdienst (DAAD, fellowship
of C.M.S.), the Deutsche Forschungsgemeinschaft (DE1599/4-
1), the Swedish Research Council, and the Knut and Alice
Wallenberg Foundation.
REFERENCES
■
(1) For reviews, see: (a) Drauz, K.; Groger, H.; May, O. Enzyme
̈
Catalysis in Organic Synthesis, 3rd ed.; Wiley-VCH: Weinheim, 2012.
(b) García-Urdiales, E.; Alfonso, I.; Gotor, V. Chem. Rev. 2011, 111,
PR110−PR180. (c) Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in
Organic Synthesis: Regio- and Stereoselective Biotransformations, 2nd ed.;
Wiley-VCH: Weinheim, 2005.
(2) Recent examples from our lab: (a) Manzuna Sapu, C.; Backvall,
̈
J.-E.; Deska, J. Angew. Chem. 2011, 123, 9905−9908; Angew. Chem.,
Int. Ed. 2011, 50, 9731−9734. (b) Hammel, M.; Deska, J. Synthesis
2012, 44, 3789−3796. (c) Manzuna Sapu, C.; Deska, J. Org. Biomol.
Chem. 2013, 11, 1376−1382. (d) Thiel, D.; Deska, J. Synlett 2013, 24,
̈
1529−1532. (e) Kamiyama, T.; Ozer, M. S.; Otth, E.; Deska, J.;
Cvengros, J. ChemPlusChem 2013, 78, 1510−1516. (f) Thiel, D.;
́
Doknic, D.; Deska, J. Nat. Commun. 2014, 5, 5278.
(3) Henke, E.; Pleiss, J.; Bornscheuer, U. T. Angew. Chem. 2002, 114,
3338−3341; Angew. Chem., Int. Ed. 2002, 41, 3211−3213.
(4) For a review, see: Kourist, R.; Domínguez de María, P.;
Bornscheuer, U. T. ChemBioChem 2008, 9, 491−498.
(5) Von Reuß, S. H.; Wu, C.-L.; Muhle, H.; Konig, W. A.
̈
Phytochemistry 2004, 65, 2277−2291.
(6) Ramharter, J.; Mulzer, J. Org. Lett. 2011, 13, 5310−5313.
(7) For metal-based isomerization catalysts, see: (a) Morrill, C.;
Grubbs, R. H. J. Am. Chem. Soc. 2005, 127, 2842−2843.
(b) Mitsudome, T.; Sueoka, S.; Ideka, S.; Mizugaki, T.; Jitsukawa,
K.; Kaneda, K. ChemCatChem 2013, 5, 2879−2882. (c) Li, J.; Tan, C.;
Gong, J.; Yang, Z. Org. Lett. 2014, 16, 5370−5373.
(8) For a lipase-vanadium combo approach, see: Akai, S.; Hanada, R.;
Fujiwara, N.; Kita, Y.; Egi, M. Org. Lett. 2010, 12, 4900−4903.
(9) For reviews on chemoenzymatic DKR, see: (a) Pam
̀
ies, O.;
Backvall, J.-E. Chem. Rev. 2003, 103, 3247−3261. (b) Pellisier, H.
̈
Tetrahedron 2011, 67, 3769−3802.
(10) Wuyts, S.; De Temmerman, K.; De Vos, D. E.; Jacobs, P. A.
Chem.Eur. J. 2005, 11, 386−397.
(11) For other examples on compartmented catalysts in dynamic
̈
kinetic resolutions, see: (a) Odman, P.; Wessjohann, L. A.;
Bornscheuer, U. T. J. Org. Chem. 2005, 70, 9551−9555.
(b) Roengpithya, C.; Patterson, D. A.; Livingston, A. G.; Taylor, P.
C.; Irwin, J. L.; Parrett, M. R. Chem. Commun. 2007, 3462−3463.
(12) Recent example using a membrane teabag for the immobiliza-
tion of whole cells: Wachtmeister, J.; Jakoblinnert, A.; Kulig, J.;
Offermann, H.; Rother, D. ChemCatChem 2014, 6, 1051−1058.
5955
dx.doi.org/10.1021/ol502979g | Org. Lett. 2014, 16, 5952−5955