Efficient synthesis of an α-trifluoromethyl-α-tosyloxymethyl epoxide enabling stepwise double functionalisation to afford CF3-substituted tertiary alcohols

Article abstract of DOI:10.1016/j.tetlet.2008.06.011

The efficient synthesis of an α-trifluoromethyl-α-tosyloxymethyl epoxide is reported. This highly versatile building block may be reacted sequentially with two different nucleophiles to furnish α-trifluoromethyl tertiary alcohols. Furthermore, the two enantiomers of this key intermediate have been separated using chiral HPLC and the stereochemistry shown to be conserved during subsequent chemical manipulations. Finally, an enzyme-driven desymmetrisation approach has been successfully employed to confer chirality on an intermediate in the sequence.

Full text of DOI:10.1016/j.tetlet.2008.06.011

Tetrahedron Letters 49 (2008) 5101–5104
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Efficient synthesis of an a-trifluoromethyl-a-tosyloxymethyl epoxide
enabling stepwise double functionalisation to afford CF3-substituted tertiary
alcohols
*
Steven P. Keeling , Ian B. Campbell, Diane M. Coe, Tony W. J. Cooper, George W. Hardy, Torquil I. Jack,
Haydn T. Jones, Deborah Needham, Tracy J. Shipley, Philip A. Skone, Peter W. Sutton, Gordon A. Weingarten,
*
Simon J. F. Macdonald
Respiratory CEDD, GlaxoSmithKline, Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The efficient synthesis of an
tile building block may be reacted sequentially with two different nucleophiles to furnish
methyl tertiary alcohols. Furthermore, the two enantiomers of this key intermediate have been
separated using chiral HPLC and the stereochemistry shown to be conserved during subsequent chemical
manipulations. Finally, an enzyme-driven desymmetrisation approach has been successfully employed to
confer chirality on an intermediate in the sequence.
a
-trifluoromethyl-
a
-tosyloxymethyl epoxide is reported. This highly versa-
-trifluoro-
Received 19 March 2008
Revised 23 May 2008
Accepted 2 June 2008
Available online 6 June 2008
a
Keywords:
Ó 2008 Elsevier Ltd. All rights reserved.
a
-Trifluoromethyl epoxide
Doubly electrophilic
Desymmetrisation
Glucocorticoid receptor agonists
R1
N
It has long been known that fluorinated ketones and, to a les-
ser extent fluorinated alcohols, can act as inhibitors of hydrolytic
enzymes by mimicking the tetrahedral transition state of peptide
N
OH
F3C
R
O
H
N
CF3
OTs
R2
N
bond hydrolysis.1,2 Thus, due to the potential application of
a-
NH2
O
O
trifluoromethylated
a-amino acids and their derivatives in the
1
2
biochemical and pharmacological fields, there is widespread
interest in the construction of stereogenic quaternary centres
featuring trifluoromethyl groups under mild conditions.3 For
Figure 1. Structures of epoxytosylate 1 and non-steroidal GR agonist 2.
example,
a-trifluoromethyl tertiary alcohols have featured prom-
misation. The work described in this Letter may also prove of va-
lue in providing a different approach to enantiomeric derivatives
featuring trifluoromethylated quaternary centres, ultimately
inantly in a new generation of glucocorticoid receptor (GR) ago-
nists from several pharmaceutical companies.4,5 The tertiary
alcohol is believed to be involved in a critical H-bond interaction
with the receptor. At least one of these agonists is being ex-
plored in development, with elegant work by Song et al. describ-
allowing access to analogues related to
-amino acids.
a-trifluoromethylated
a
The preparation of optically active polyfluoroalkyl epoxide syn-
ing the formation of enantiomerically pure
a-trifluoromethyl
thons was reported in the literature9 some years ago. Whilst this
strategy appeared attractive and was considered for our work, it
was ultimately discounted due to the necessity for the use of
diazomethane in the epoxide formation and the inherent safety is-
sues this raises for a large-scale synthesis.
tertiary alcohols.6 The recent explosion of interest7 in GR ago-
nists has prompted us to describe our work in this field.5,8 This
has been greatly enabled by a key ‘doubly electrophilic’
fluoromethyl- -tosyloxymethyl epoxide synthon 1 (Fig. 1) which,
a-tri-
a
due to its versatility, has allowed entry into highly tractable
non-steroidal GR agonists8 such as 2 (Fig. 1) for rapid lead opti-
Several strategies currently exist in the literature for the syn-
thesis of
a-trifluoromethyl alcohols. It is more than twenty years
since Ruppert et al. disclosed the first synthesis of trimethyl(tri-
fluoromethyl)silane.10 Since then this has become widely estab-
lished as a standard means of introducing a trifluoromethyl
group into molecules via addition to carbonyl compounds (e.g.,
* Corresponding authors. Tel.: +44 143 876 4018; fax: +44 143 876 3616 (S.P.K.);
tel.: +44 143 876 8249; fax: +44 143 876 3616 (S.J.F.M.).
E-mail addresses: steve.p.keeling@gsk.com (S. P. Keeling), simon.jf.macdonald@
gsk.com (S. J. F. Macdonald).
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.06.011

5102
S. P. Keeling et al. / Tetrahedron Letters 49 (2008) 5101–5104
ketones for the preparation of
a
-trifluoromethyl tertiary alcohols)
This route is robust and has been scaled to produce kilogram
quantities of 1 without issue; on this scale, the TEMPO oxidation
and potassium carbonate epoxidation reaction conditions were
used.
In another approach used in recent years, trifluoromethyl epoxides
have been employed6,11 as intermediates permitting more diver-
sity in the structure of the tertiary alcohol products.
A different approach involves generation of an
a
-trifluoro-
The epoxytosylate 1 is essentially doubly electrophilic and
whilst both electrophilic centres are reasonably reactive, one elec-
trophile is appreciably more reactive than the other. For example
(Scheme 2), on treatment of 1 with a slight excess of benzylamine
in dioxane, the epoxide is smoothly opened in preference to dis-
placement of the tosylate. The tosylate can then be displaced with
a different nucleophile such as hydroxide to give the diol 9 in 57–
64% overall yield. This is a robust sequence and has been carried
out on >80 g scale.
methyl oxiranyl anion by the deprotonation of commercially avail-
able 2-(trifluoromethyl)oxirane using n-butyllithium.12 This has
been shown to react with electrophiles to give a-substituted-a-tri-
fluoromethyl epoxides which can further react with nucleophiles
to furnish tertiary alcohols, hence offering a further level of diver-
sification relative to other methods. One practical limitation of this
methodology is that very cold temperatures (ca. À100 °C) are re-
quired to avoid possible side reactions of the anion.
We sought a complimentary approach to these methods where-
by a suitably functionalised trifluoromethyl epoxide might lend it-
self to sequential reactions with two structurally diverse
nucleophiles. A prerequisite was that the intermediate would be
a stable robust entity which could be used under a range of reac-
tion conditions. As such, an efficient five-step synthesis of the 2-
Alternatively, the epoxytosylate 1 can be treated first with the
anion of ethylbenzamide generated with sodium hydride in DMF
(Scheme 2). After 18 h, benzylamine may be added to react with
the second electrophile to give 11 in 32% overall yield. This exam-
ple illustrates the one-pot reaction of the epoxytosylate with two
different nucleophiles.
trifluoromethyl-2-tosyloxymethyl epoxide
(Scheme 1).
1
was developed
In a third example, treatment of epoxytosylate 1 with dibenzyl-
amine in the presence of excess lithium perchlorate in ether
(Scheme 2) leads to isolation of the epoxide 12. It is unclear
whether this occurs by preferential displacement of the tosylate,
or epoxide opening followed by ring closure of the resulting hydro-
xy tosylate. Epoxide 12 can then be reacted with ethylamine to
provide the diamine 13 in reasonable yield.
Thus, the reactivity of the epoxytosylate 1 may be manipulated
by the judicious selection of nucleophiles (and reagents) to provide
the product of choice.
Having demonstrated an efficient and robust racemic route, we
then began to address the tractability of a chiral synthesis. Two
questions can be raised. Firstly, can enantiomerically pure epoxy-
tosylate be reacted sequentially with two different nucleophiles
without racemisation? Secondly, can enantiomerically enriched
or pure epoxytosylate 1 be prepared from the symmetrical triol 6
(or other precursor) using a desymmetrisation strategy? The latter
would have the major theoretical benefit that 100% of the triol 6
could be converted into the single desired enantiomer. This stands
in contrast to a resolution of the epoxytosylate (or some chiral pre-
cursor or derivative) where theoretically only 50% of the desired
enantiomer can be obtained from the racemate.
Two options were considered to synthesise the ketone interme-
diate 4. Attempts to introduce alcohol protecting groups by
O-alkylating 1,3-dihydroxyacetone met with limited success, so
1,3-dibenzyloxy-2-propanol 3 was oxidised using Ley’s method13
in 83% yield. The oxidation could also be effected using sulfur tri-
oxide pyridine complex in DMSO, or 2,2,6,6-tetramethyl-1-piperid-
inyloxy free radical (TEMPO) along with sodium hypochlorite (13%
w/v). Reaction of 4 with trimethyl(trifluoromethyl)silane employ-
ing cesium fluoride or cesium carbonate/cat. tetrabutylammonium
fluoride (TBAF) in DMF gave moderate yields (25–38%) of the ter-
tiary alcohol 5. However, changing the solvent to THF and employ-
ing TBAF stoichiometrically14 increased the yield markedly to over
80%. The benzyl groups were exchanged for p-toluenesulfonyl
groups by firstly subjecting 5 to hydrogenation conditions employ-
ing 5% palladium on carbon in ethanol. The resulting triol 6 was
treated with p-toluenesulfonyl chloride in pyridine to give the ter-
tiary alcohol 7. Despite using 3 equiv of the sulfonyl chloride, only
the primary alcohols were functionalised. Due to the acidic nature
of the alcohol, mild basic conditions effected cyclisation to afford
the epoxide 1. On a small scale, polymer-supported carbonate15
was used for convenience but on a larger scale this became cost
prohibitive, so potassium carbonate was employed.
To address the first question, the epoxytosylate 1 was separated
by preparative chiral HPLC16 into its two enantiomers 1a and 1b
(Scheme 3). Epoxytosylate 1a (of ca. 95% enantiomeric purity)
OH
(a)
O
HO
HO
Ph
O
O
Ph
Ph
Ph
O
O
O
Ph
Ph
CF3
OH
CF3
OTs
(b)
(a)
4
3
N
H
N
H
Ph
Ph
(b)
8
9
HO
CF3
HO
CF3
OTs
O
OH
(c)
H
OH
CF3
Ph
(c)
(d)
Ph
N
N
O
N
HO
OH
Ph
OTs
CF3
O
CF3
1
O
6
5
11
10
(d)
HO
CF3
O
CF3
Ph
Ph
O
(e)
Ph
Ph
OH
(f)
CF3
(e)
TsO
OTs
N
NHEt
N
OTs
CF3
1
13
7
12
Scheme 1. Reagents and conditions: (a) TPAP, NMO, 4 Å molecular sieves, DCM, rt,
18 h (83%); (b) TMSCF3, TBAF, THF, 0 °C ? rt, 18 h (81%); (c) H2, 5% Pd.C, EtOH, rt,
24 h (98%); (d) TsCl, pyridine, 0 °C ? rt, 20 h (91%); (e) PS-carbonate, DCM,
0 °C ? rt, 18 h (94%).
Scheme 2. Reagents and conditions: (a) BnNH2 (1.05 equiv), dioxane, rt, 18 h; (b)
2 M NaOH, dioxane, reflux, 18 h (57–64% for 2 steps); (c) PhCONHEt, NaH, DMF then
1 (1 equiv), rt, 18 h; (d) BnNH2 (1 equiv), rt, 24 h (32% for two steps); (e) Bn2NH
(1 equiv), LiClO4 (10 equiv), Et2O rt, 18 h (59%); (f) EtNH2 (2 equiv), THF (62%).

S. P. Keeling et al. / Tetrahedron Letters 49 (2008) 5101–5104
5103
format, the monobutyrate 15 was furnished in 85% solution yield
and 92% ee from a 10 g/L solution of triol 6 in vinyl butyrate/TBME
(1:9). As the diester does not participate in the subsequent chem-
ical step and can be removed readily at a later stage, the monoes-
ter/diester mixture resulting from the continuous reaction can be
used without the need for purification steps.
As in
Scheme 2
F3C
O
O
OH
Ref 16
H
N
CF3
OTs
CF3
OTs
Ph
Ph
OH
1a
+
O
1
9a
(a)
Ph
CF3
OTs
A batch of desymmetrised butyrate 15 of 84% ee was converted
selectively into the tosylate 16 in excellent yield (Scheme 4). The
epoxy-butyrate 17 could then be prepared readily by using poly-
mer-supported carbonate. This was smoothly reacted with benzyl-
amine to give the aminoester 18 in excellent yield over two steps.
Base hydrolysis of 18 caused O- to N-migration of the butyryl
group. Although the resultant amide can be hydrolysed in acid to
the diol 9a, acid hydrolysis of the ester 18 gives the diol 9a directly
in quantitative yield in one step. Chiral HPLC analysis of 9a showed
it still to be enantiomerically enriched in the (S) enantiomer (ca.
83% ee) demonstrating that the route does not cause racemisation,
and also assigning the absolute stereochemistry of the enzyme-
mediated desymmetrisation. This sequence opened-up a route to
enantiomeric analogues for lead optimisation (such as 2) and
potentially for large-scale preparation of final drug substances.
In summary, we have described the synthesis and reactivity of a
versatile intermediate 1. We have also demonstrated the prepara-
tion of enantiomerically pure 1 (and related analogues), and fur-
ther shown that these can be used without racemisation.
F3C
OH
N
OH
1b
14
Scheme 3. Reagents and conditions: (a) PhCHO (1 equiv), NaBH(OAc)3, AcOH, DCM,
rt, 72 h (10%).
has been processed through to 9a which is 95% enantiomerically
pure by chiral HPLC when compared with 9, demonstrating that
the described sequence may be carried out without racemisation.17
Compound 9a was then reductively alkylated with benzaldehyde
to give 14, the measured optical rotation of which is the antipode
of that quoted in the literature for the (R)-enantiomer9 indicating
the stereochemistry of 14 is therefore (S).
One problem with the preparative separation of the enantio-
mers by chiral HPLC was that only low loadings were possible.
Thus, a desymmetrisation approach was of even greater interest.
The chirality was introduced early in the synthesis by a lipase-
catalysed desymmetrisation of the triol 6 (Scheme 4). A screen
for triol acylation in neat vinyl butyrate using kits of commer-
cially available lipases gave a number of hits.18 Of these hits, the
lipase from Burkholderia cepacia (AE012) was selected for further
investigation as it produced the desired monobutyrate 15 with
the best enantioselectivity (86% ee) and with little dibutyrate
formation.19,20
In an unoptimised reaction using lipase Amano PS, the monobu-
tyrate 15 was produced in 81% solution yield and 84% ee. Due to
incomplete conversion, the unreacted triol 6 needed to be removed
by column chromatography which resulted in significant loss of
the monoester (42% isolated yield). In contrast, complete consump-
tion of triol could be attained by passing a solution of triol 6 in vi-
nyl butyrate and TBME through a plug flow reactor containing
Burkholderia cepacia lipase immobilised on sepabeads EC-EP.21 By
allowing some conversion of monoester 15 to diester, even after
the disappearance of triol 6, an enantioenrichment of the monobu-
tyrate 15 could be attained that resulted from further enzyme-
catalysed acylation of the minor monobutyrate enantiomer to
dibutyrate. In preliminary experiments using this continuous flow
1. Experimental for preparation of epoxytosylate 1
1.1. 1,3-Bis[(benzyl)oxy]-2-propanone 4
To a suspension of powdered 4 Å molecular sieves (30 g) in
DCM (500 mL) was added 4-methylmorpholine N-oxide (21 g)
and 1,3-dibenzyloxy-2-propanol (24.7 mL). The mixture was stir-
red vigorously under a nitrogen atmosphere for 1.5 h then treated
with tetrapropylammonium perruthenate (1.75 g) (note: exo-
therm). Stirring was continued for 18 h then the mixture was fil-
tered through a pad of Celite. The filtrate was evaporated in
vacuo and the residue was purified by flash chromatography using
a 0–25% ethyl acetate/cyclohexane gradient to obtain the title com-
pound as a white solid (22.45 g, 83%). 1H NMR (400 MHz, CDCl3) d
ppm 7.31–7.40 (m, 10H); 4.59 (s, 4H); 4.26 (s, 4H).
1.2. 1,1,1-Trifluoro-3-[(benzyl)oxy]-2-{[(benzyl)oxy]methyl}-2-
propanol 5
To a solution of 4 (21.94 g) in anhydrous THF (300 mL) under a
nitrogen atmosphere was added trimethyl (trifluoromethyl)silane
(16.8 mL). After cooling to 0–5 °C tetrabutylammonium fluoride
(90 mL of a 1 M solution in THF) was added slowly maintaining
the reaction temperature below 10 °C. The cooling bath was re-
moved and stirring was continued for 18 h. The reaction mixture
was partitioned between 1 M HCl (800 mL) and diethyl ether
(300 mL). The layers were separated, and the aqueous phase was
further extracted with diethyl ether (300 mL). The combined or-
ganic extracts were dried over sodium sulfate and evaporated in
vacuo. The residue was purified by flash chromatography using a
0–25% ethyl acetate/cyclohexane gradient to obtain the title com-
pound as a pale yellow oil (22.37 g, 81%). 1H NMR (400 MHz,
CDCl3) d ppm 7.29–7.39 (m, 10H); 4.60 (s, 4H); 3.72 (s, 4H); 3.38
(s, 1H).
F3C OH
HO
(a)
F3C OH
HO
O
O
Pr
Pr
OH
15
84% ee
O
O
6
(b)
O
CF3
(c)
F3C OH
TsO
O
Pr
O
17
(d)
16
F3C
(e)
F3C
OH
OH
H
H
N
Ph
N
O
Pr
Ph
OH
1.3. 2-(Trifluoromethyl)-1,2,3-propanetriol 6
O
9a
18
A solution of 5 (21.87 g) in ethanol (400 mL) containing 5% pal-
ladium on carbon (2.2 g, wet, Degussa type E101 NO/W) was stir-
red under an atmosphere of hydrogen for 24 h. The reaction
Scheme 4. Reagents and conditions: (a) lipase vinyl butyrate (42%); (b) TsCl
(1 equiv), pyridine (1 equiv), DCM (70–100%); (c) PS-carbonate, THF; (d) BnNH2
(1 equiv) (80–100% for two steps); (e) 5 M HCl, EtOH (100%).

5104
S. P. Keeling et al. / Tetrahedron Letters 49 (2008) 5101–5104
mixture was filtered through a pad of Celite and the filtrate was
evaporated in vacuo. The residue was evaporated twice from
DCM to obtain the title compound as a white solid (10.1 g, 98%).
1H NMR (400 MHz, DMSO-d6) d ppm 5.65 (s, 1H); 4.89 (t,
J = 6.0 Hz, 2H); 3.54 (d, J = 5.8 Hz, 4H).
References and notes
1. Gelb, M. H.; Svaren, J. P.; Abeles, R. H. Biochemistry 1985, 24, 1813.
2. Patel, D. V.; Rielly-Gauvin, K.; Ryono, D. E. Tetrahedron Lett. 1988, 29, 4665.
3. (a) Huguenot, F.; Brigaud, T. J. Org. Chem. 2006, 71, 7075; (b) Wang, H.; Zhao, X.;
Li, Y.; Lu, L. Org. Lett. 2006, 8, 1379; (c) Ogu, K.; Matsumoto, S.; Akazome, M.;
Ogura, K. Org. Lett. 2005, 7, 589; (d) Sinisi, R.; Sani, M.; Candiani, G.; Parente, R.;
Pecker, F.; Bellosta, S.; Zanda, M. Tetrahedron Lett. 2005, 46, 6515; (e) Lauzon,
C.; Charette, A. B. Org. Lett. 2006, 8, 2743.
4. Schacke, H.; Schottelius, A.; Docke, W.-D.; Strehlke, P.; Jaroch, S.; Schmees, N.;
Rehwinkel, H.; Hennekes, H.; Asadullah, K. Proc. Nat. Acad. Sci. 2004, 101, 227.
5. Clackers, M.; Coe, D. M.; Demaine, D. A.; Hardy, G. W.; Humphreys, D.; Inglis, G.
G. A.; Johnston, M. J.; Jones, H. T.; House, D.; Loiseau, R.; Minick, D. J.; Skone, P.
A.; Uings, I.; McLay, I. M.; Macdonald, S. J. F. Bioorg. Med. Chem. Lett. 2007, 17,
4737. and references therein.
6. Song, J. J.; Tan, Z.; Xu, J.; Reeves, J. T.; Yee, N. K.; Ramdas, R.; Gallou, F.; Kuzmich,
K.; DeLattre, L.; Lee, H.; Feng, X.; Senanayake, C. H. J. Org. Chem. 2007, 72, 292.
7. Mohler, M. L.; He, Y.; Wu, Z.; Hong, S.-S.; Miller, D. D. Expert Opin. Ther. Patents
2007, 17, 37. This paper provides a succinct review of the non-steroidal GR
agonist area and contains many leading references.
8. (a) Barnett, H. A.; Campbell, I. B.; Coe, D. M.; Cooper, A. W. J.; Inglis, G. G. A.;
Jones, H. T.; Keeling, S. P.; Macdonald, S. J. F.; McLay, I. M.; Skone, P. A.;
Weingarten, G. G.; Woolven, J. M. PCT Int. Appl. 2008, WO2008000777. CAN
148:100602.; (b) Barnett, H. A.; Campbell, I. B.; Coe, D. M.; Cooper, A. W. J.;
Inglis, G. G. A.; Jones, H. T.; Keeling, S. P.; Macdonald, S. J. F.; McLay, I. M.; Skone,
P. A.; Weingarten, G. G.; Woolven, J. M. PCT Int. Appl. 2007 WO2007144327.
CAN 148:79023.
1.4. 3,3,3-Trifluoro-2-hydroxy-2({[(4-
methylphenyl)sulfonyl]oxy}methyl)propyl
4-methylbenzenesulfonate 7
A stirred solution of 6 (9.42 g) in pyridine (100 mL) under a
nitrogen atmosphere was cooled to 0–5 °C. To this was added p-
toluenesulfonyl chloride (33.67 g) to give an orange solution. The
cooling bath was removed and stirring was continued for 20 h,
then the reaction mixture was evaporated in vacuo. The residue
was partitioned between 1 M HCl (700 mL) and ethyl acetate
(300 mL). The layers were separated, and the aqueous phase was
further extracted with ethyl acetate (300 mL). The combined or-
ganic extracts were washed with saturated aqueous sodium bicar-
bonate and brine, dried over sodium sulfate and evaporated in
vacuo. The residue was purified by flash chromatography using a
0–50% ethyl acetate/ cyclohexane gradient to obtain the title com-
pound as a very pale yellow oil, which crystallised on standing
9. Bravo, P.; Farina, A.; Frigerio, M.; Meille, S. V.; Viani, F.; Soloshonok, V.
Tetrahedron: Asymmetry 1994, 5, 987–1004.
10. Ruppert, I.; Schlich, K.; Volbach, W. Tetrahedron Lett. 1984, 25, 2195.
11. Lee, T. W.; Proudfoot, J. R.; Thomson, D. S. Bioorg. Med. Chem. Lett. 2006, 16, 654.
12. Yamauchi, Y.; Katagiri, T.; Uneyama, K. Org. Lett. 2002, 4, 173.
13. Griffith, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D. J. Chem. Soc., Chem.
Commun. 1987, 1625.
(25.08 g, 91%). 1H NMR (400 MHz, CDCl3)
d ppm 7.78 (d,
J = 8.3 Hz, 4H); 7.38 (d, J = 8.3 Hz, 4H); 4.18 (s, 4H); 3.66 (s, 1H);
2.48 (s, 6H).
14. Canney, D. J.; Lu, H.-F.; McKeon, A. C.; Yoon, K.-W.; Xu, K.; Holland, K. D.;
Rothman, S. M.; Ferrendelli, J. A.; Covey, D. F. Bioorg. Med. Chem. 1998, 6, 43.
15. Carbonate on polymer support supplied by Fluka (capacity ꢀ3.5 mmol/g).
16. Preparative separation of the enantiomers of 1. Column Chiralpak AD,
2 inch  20 cm eluting with heptane/ethanol 98:2 with a flow rate of 75 mL/
min. 1 b elutes at around 17 min and 1a at around 21 min. Racemate loadings
<200 mg are required to achieve acceptable separation of the enantiomers.
17. Analytical separation of the enantiomers of 9. Column Chiralcel OD-H, 25 cm,
1.5. [2-(Trifluoromethyl)-2-oxiranyl]methyl
4-methylbenzenesulfonate 1
A stirred solution of 7 (27 g) in DCM (400 mL) under a nitro-
gen atmosphere was cooled to 0–5 °C. To this was added poly-
mer-supported carbonate15 (32.9 g). The cooling bath was
removed and stirring was continued for 18 h. The reaction mix-
ture was filtered and the filtrate was evaporated in vacuo to ob-
eluting with 5% ethanol in heptane with
a flow rate of 1 mL/min. The
enantiomers appear at 8.61 and 9.98 min. The material processed to 9a from 1a
is composed of 5% of the first eluting enantiomer and 95% of the second eluting
enantiomer.
tain the title compound as
a very pale yellow oil, which
18. Using Alphamerix Screening Kits 1 and 2 (Mann Associates) hits were obtained
from C. rugosa lipase (AE02), Achromobacter spp. lipase (AE04), Alcaligenes spp.
lipase (AE05), Burkholderia cepacia lipase P1 (AE06), Pseudomonas stutzeri lipase
(AE07), Burkholderia cepacia lipase P2 (AE012) and Mucor javanicus lipase
(AE013).
crystallised on standing (16.04 g, 94%). 1H NMR (400 MHz, CDCl3)
d ppm 7.80 (d, J = 8.3 Hz, 2H); 7.38 (d, J = 8.3 Hz, 2H); 4.26–4.45
(ab, J = 12.0 Hz, 2H); 3.14 (d, J = 4.8 Hz, 1H); 3.01 (dd, 1H); 2.47
(s, 3H).
19. The screen was monitored by TLC and product ee determined by 31P NMR after
derivatisation of the monobutyrate, present in crude reaction mixtures, as the
mono Mosher’s ester. Previous Mosher’s ester derivatisation of authentic
racemic monobutyrate and triol had shown that their respective 31P NMR
signals do not overlap. The authentic diester was inert to Mosher’s ester
derivatisation. All enzyme hits preferentially produced the same enantiomer.
20. Parallel screening for diacetate or dibutyrate hydrolysis activity also gave a
number of hits, all of which resulted in complete hydrolysis to the triol under a
variety of conditions.
Acknowledgements
We thank Eric Hortense and Steve Jackson for analytical
and preparative chiral HPLC support, and Otman Benali for assis-
tance with the flow chemistry set-up for the desymmetrisation
work.
21. Sepabeads EC-EP is a porous epoxy-resin support produced by Resindion–
Mitzubishi.