Accessible sugars as asymmetric olefin epoxidation organocatalysts: Glucosaminide ketones in the synthesis of terminal epoxides

Article abstract of DOI:10.1039/b911675c

A systematically varied series of conformationally restricted ketones, readily prepared from N-acetyl-d-glucosamine, were tested against representative olefins as asymmetric epoxidation catalysts showing useful selectivities against terminal olefins and, in particular, typically difficult 2,2-disubstituted terminal olefins.

Full text of DOI:10.1039/b911675c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Accessible sugars as asymmetric olefin epoxidation organocatalysts:
glucosaminide ketones in the synthesis of terminal epoxides†
Omar Boutureira, Joanna F. McGouran, Robert L. Stafford, Daniel P. G. Emmerson and Benjamin G. Davis*
Received 15th June 2009, Accepted 15th July 2009
First published as an Advance Article on the web 14th August 2009
DOI: 10.1039/b911675c
A systematically varied series of conformationally restricted ketones, readily prepared from
N-acetyl-D-glucosamine, were tested against representative olefins as asymmetric epoxidation catalysts
showing useful selectivities against terminal olefins and, in particular, typically difficult
2,2-disubstituted terminal olefins.
provides excellent selectivities, this specialist biocatalytic method
is most successful for 2,2-disubstituted aliphatic alkenes and
only some 2-aryl olefins such as a-methylstyrene derivatives
have been employed to date. During the preparation of this
Introduction
Chiral epoxides are found in many natural products1 and are highly
versatile electrophilic intermediates which can be manipulated
manuscript, Shibasaki and co-workers10 revealed the addition of
dimethyloxosulfonium methylide to ketones catalysed by chiral
multimetallic complexes (up to 97% ee) and a procedure reported
by Shi6 employed a new lactam ketone system which enhances the
enantioselectivity for the epoxidation of 2,2-disubstituted terminal
olefins up to 88% ee. Since 2,2-disubstituted terminal epoxides are
implicated in the synthesis of various a-arylpropionic acids (non-
steroidal antiinflammatory drugs–NSAIDs and household pain
killers) and these analogues are commonly prepared via enzymatic
resolution procedures,11 the development of a new catalytic system
able to synthesise 2,2-disubstituted terminal epoxides via chemical
asymmetric epoxidation of the corresponding alkenes would be
highly desirable.
into a large variety of related functional groups. Asymmetric
epoxidations2 have long been performed using metal-mediated
methodology such as the powerful transformations developed
by Sharpless3 and Jacobsen.4 Other important methods in-
volve the use of chiral dioxiranes and oxaziridines, typically
derived from precursor ketones, aldehydes and imines as ver-
satile organocatalysts5 with increasingly broad olefin substrate
specificities allowing normally high conversions and selectivities.
Among these methods, heterocyclic ketones, including fructose-
and oxazolidinone-bearing ketones 1–4, have been successfully
employed for the epoxidation of a certain range of olefins such
as unfunctionalised trans-, trisubstituted and terminal olefins5a
(Fig. 1).
With a view to addressing this goal and some prior limitations,12
we present here a novel class of chiral ketones readily derived from
an abundant, chiral pool carbohydrate, N-acetyl-D-glucosamine.
Carbohydrates are a naturally occurring, inexpensive, renewable,
readily available source of chirality which contain a high density
of stereogenic centres and may provide advantageously rigid
frameworks.13 Due to these favourable properties, they have
already proven to be effective scaffolds for organocatalysts in
asymmetric epoxidations.5a In our choice of scaffold I (Fig. 2),
we considered key advantages to be the often powerful directing
effect of a-substituents,14 and conformational rigidity afforded
by the trans-decalin-like benzylidene-4,6-O-acetal. Indeed, it has
been noted previously that conformational flexibility in such
epoxidising organocatalysts may have dramatically detrimental
effects on stereoselectivity.5a Moreover, we considered that the
use of an a-substituent with the potential to act as Lewis base
(C-2 carboxamide) would allow scope for wide-ranging tuning of
Fig. 1 Selected examples of carbohydrate-based asymmetric epoxidation
organocatalysts.
However, certain motifs and especially 2,2-disubstituted ter-
minal olefinic systems remain challenging substrates for epox-
idation with high stereoselectivity6 (e.g. 30% and 58% ee for
a-methylstyrene and a-isopropylstyrene, respectively, with cat-
alyst 3).7 As a solution to these problems, alternative methods
have been reported for the synthesis of 2,2-disubstituted terminal
epoxides such as the use of stoichiometric amounts of keto bile
acids8 as asymmetric inducers (30–76% ee for a-methylstyrene)
and chloroperoxidases9 (20–89% ee). Although the last method
Department of Chemistry, University of Oxford, Chemistry Research
Laboratory, 12 Mansfield Road, Oxford, UK OX1 3TA. E-mail: Ben.
Davis@chem.ox.ac.uk; Fax: +44 (0)1865 285002; Tel: +44 (0) 1865 275
652
† Electronic supplementary information (ESI) available: Synthesis and
characterization of ketones 7a–e and 8, epoxidation procedures, op-
timization studies, proposed model for the stereochemical outcome,
characterization of epoxides 10, 15–18 and 24–28 along with the data for
the determination of the enantiomeric excess. See DOI: 10.1039/b911675c
Fig. 2 Putative interactions between dioxirane at C-3 and carboxamide
at C-2.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4285–4288 | 4285
©

View Article Online
Table 1 Screening of ketones 7a–e and 8 in trans-stilbene 9 epoxidationa
a putative dioxirane at C-3 II (Fig. 2), through substituent (R)
alteration.
Results and discussion
Putative dioxirane precursors, ketonic amides 7a–e and 8 contain-
ing just such a vicinal functional group pattern, were synthesised
over 5–6 steps from commercially available N-acetylglucosamine
in overall yields up to 59% giving simple and robust access to these
potential catalysts on gram scales (Scheme 1). Thus, anomeric
protection as methyl pyranoside using acetyl chloride in methanol,
regioselective 4,6-O-protection as benzylidene acetal and base-
mediated deamidation of N-2 gave versatile aminoalcohol in-
termediate 6. Amidation of the alpha anomer of 6 with either
acetic anhydride or appropriate para-substituted benzoyl chloride
afforded the corresponding amide which was then oxidised under
Swern conditions to give 7a–e and 8, displaying a representative
range of electronic properties (varied sp).
Oxone
Conv.
Entry Catalyst
R
(equiv) (%)b
ee (%)c
1
2
3
4
5
6
7
8
a-Me-GlcNAc
n/a
1.4
1.4
1.4
1.4
1.4
1.4
1.4
3
3
3
3
6
0
3
19
26
39
7
43
10
51
64
10
57
nd
7a
7b
7c
7d
7e
8
7a
7b
7d
7e
7d
p-CF3C6H4
p-ClC6H4
C6H5
p-MeOC6H4
p-Me2NC6H4
Me
p-CF3C6H4
p-ClC6H4
p-MeOC6H4
p-Me2NC6H4
p-MeOC6H4
56 (R,R)
65 (R,R)
69 (R,R)
73 (R,R)
74 (R,R)
61 (R,R)
nd
67 (R,R)
66 (R,R)
76 (R,R)
69 (R,R)
9
10
11
12
a General conditions: 3 : 2 CH3CN/diglyme, OxoneꢀR added over 30 min,
TBAHS (0.1 equiv), K2CO3 (5.8 equiv), buffer pH 10.6, catalyst (1 equiv),
2.5 h, rt (see ESI† for full details). b Determined by 1H NMR analysis.
c Determined by chiral HPLC (Chiralcel OD column).
observed yielding a relative reaction coefficient (log(kR/kS) vs.
sp) of r = -0.28 consistent with differential development of
greater partial positive charge within the amide (Fig. 2) during
the more diastereoselective epoxidation modes.15 Encouraged by
these initial results, we screened several other representative olefin
substrates such as bis-substituted 11, terminal 12 and 13, and
tri-substituted 14 (Table 2).
Scheme 1 Synthesis of carbohydrate ketone catalyst precursors based on
scaffold I. Reagents and conditions: (a) AcCl, MeOH; (b) PhCH(OMe)2,
p-TsOH, DMF, 70 ◦C, 80% from 5; (c) KOH, EtOH, 80 ◦C, 76%; (d)
p-R-BzCl, 0 ◦C, pyridine, DCM; (e) NaOH, MeOH; (f) DMSO, (COCl)2,
Et3N, DCM/THF, -78 ◦C to rt, 45–97% (yields quoted over 3 steps, see
ESI† for full details); (g) Ac2O, NaOMe, MeOH, 81%.
Table 2 Screening of ketones 7a–d and 8 against other olefin substratesa
With these putative ketone organocatalysts in hand, suitable
conditions for the epoxidation of a testing model olefin, trans-
stilbene 9, using p-Cl benzamide 7b were investigated with respect
to: solvent system; temperature; volume of organic solvent; pH;
R
equivalents of co-oxidant Oxoneꢀ (2KHSO5·KHSO4·K2SO4); re-
R
action time; and duration of Oxoneꢀ addition. Promising reaction
conditions were found: 3 : 2 CH3CN/diglyme at pH 10.6 using 1
equivalent of 7b at 25 ◦C over 2.5 h reaction in the presence of 3
R
equivalents of Oxoneꢀ added over 30 min. These conditions were
then screened against trans-stilbene as a substrate using the full
range of ketones 7a–e and 8 to test both potential activity and
selectivity of such ketones (Table 1).
Entry Ketone
R
Substrate
Conv. (%)b ee (%)c
1
2
7a
7b
8
p-CF3C6H4
p-ClC6H4
Me
trans-anethole 11
trans-anethole 11
trans-anethole 11
trans-anethole 11
vinylnaphthalene 12
vinylnaphthalene 12
styrene 13
95
87
> 98
> 98
33
41
45
50
49
25 (R)
26 (R)
67 (R)
66 (R)
81 (R)
74 (R)
58 (R)
Although some conversions were found to be low under these
conditions, for trans-stilbene 9 as a difficult substrate (Table 1,
entries 2–7) the across-the-board activity with all ketones and
broadly reasonable stereoselectivities (56–76% ee) showed the
initial potential of the keto glucosaminide scaffold. Conversions
could be improved to usable levels without too severe a loss
of selectivity (Table 1, entries 8–12) using a greater excess of
3
4d
5
8
Me
7a
7b
7a
7c
7d
8
p-CF3C6H4
p-ClC6H4
p-CF3C6H4
C6H5
p-MeOC6H4 styrene 13
Me
6
7
30
67
95
76
8
9
styrene 13
10d
11
styrene 13
triphenylethene 14
85
40
R
Oxoneꢀ. The desired ability to tune reactivity through substituent
7c
C6H5
(R) was also confirmed and gave insight into mechanism. For
benzamide ketones 7a–e, a clear trend relating ee to Lewis basicity
was also observed. Thus, variation of the para-substituent tuned
ee from 56% for para-CF3 benzamide (sp = +0.53) up to 76%
for para-NMe2 (sp = -0.63). A correlation (R2 = 0.88) was
a General conditions: 3 : 2 CH3CN/diglyme, OxoneꢀR (1.4 equiv) added over
30 min, TBAHS (0.1 equiv), K2CO3 (5.8 equiv), buffer pH 10.6, ketone
1
(1 equiv), 2.5 h, rt (see ESI† for full details). b Determined by H NMR
analysis. c Determined by chiral HPLC (Chiralcel OD column). d 0.3 equiv
of ketone used.
4286 | Org. Biomol. Chem., 2009, 7, 4285–4288
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 3 Asymmetric epoxidation of 2,2-disubstituted terminal olefins
with ketones 7a–d and 8a
substrates for the first time (e.g. NSAID precursor 22, 70% conv.,
42% ee) in a manner that outstrips other strategies19 may prove
usefully complementary. To this end, we continue to focus on the
development of these readily created and tuned catalysts under
the mild conditions investigated here with other putative substrate
types.
R1 R2 Substrate Conv. (%)b ee (%)c
Experimental section
Entry Ketone
R
1d
2d
3d
4d
5d
6e
7
8
9
10
7a
7b
7c
7d
8
p-CF3C6H4 Me
H
H
H
H
H
H
19
19
19
19
19
19
92
33
60
53
98
4 (S)
General epoxidation procedure
p-ClC6H4
Me
Me
23 (S)
30 (S)
19 (S)
40 (S)
40 (S)
42f (S)
36f (R)g
42f (S)
24f
Trans-stilbene 9 (54 mg, 0.3 mmol) and the chosen ketone
catalyst (0.3 mmol) were dissolved in the chosen solvent system
(4.5 mL). The phase transfer agent, NBu4HSO4 (TBAHS),
(5 mg, 0.1 mmol) and chosen buffer solution (3 mL in 4 ¥
C6H5
p-MeOC6H4 Me
Me
Me
Me
Me
Me
Me
Me
Me
8
98
8
Me Me 20
Me Cl 21
Me i-Bu 22
> 99
> 99
70
R
10-4 M Na2EDTA) were added to the solution. Oxoneꢀ (254 mg,
8
0.41 mmol) in aqueous Na2EDTA (4 ¥ 10-4 M, 2 mL) and a
solution of K2CO3 in water20 (2 mL) were added dropwise over a
period of 30 min using two separate plastic syringes with teflon
needles. The mixture was stirred at room temperature for 2.5 hours.
The reaction progress was followed by TLC (20 : 1 petrol : diethyl
ether) showing consumption of the starting material (Rf 0.7) and
formation of product (Rf 0.5). The reaction was quenched with
water (10 mL) and the resulting precipitate was filtered off and
washed with a 5 : 1 petrol : diethyl ether solution (30 mL). The
remaining precipitate was found to be the pure ketone (40–95%).
The epoxide and corresponding olefin were found to be entirely
within the petrol solution. The epoxide was separated from the
olefin using column chromatography (97 : 2 : 1, petrol : diethyl
ether : triethylamine) after conversion had been determined by
8
8
i-Pr
H
23
98
a General conditions: 3 : 2 CH3CN/diglyme, OxoneꢀR (1.4 equiv) added over
30 min, TBAHS (0.1 equiv), K2CO3 (5.8 equiv), buffer pH 10.6, ketone
(0.06 equiv), 2.5 h, 0 ◦C. b Determined by 1H NMR analysis. c Determined
by chiral HPLC (Chiralcel OD column). d 1 equiv of ketone used at rt. e 0.3
equiv of ketone used at rt. f Determined by chiral GC (Cydex B column).
g Opposite enantiomer obtained (see ESI† for full details).
Ready alteration of reactivity and selectivity of the organocat-
alysts in the substituent (R) allowed tuning towards substrate. In
all cases, good to excellent conversions (67–98%) were possible
except with more difficult substrates, terminal 2-vinylnaphthalene
12 (33% with ketone 7a) and the most hindered, triphenylethene 14
(40% with ketone 7c).16 Ketone 8 was an all-round performer and
consistently gave high conversions with comparable selectivities
across all substrates tested, with little loss of activity even down
to 0.06 equiv. Representative epoxidations of notoriously difficult
substrate styrene 13 (Table 2, entry 9) complement existing meth-
ods (81% ee with catalyst 3).17 These epoxidation organocatalysts
were then applied to key target 2,2-disubstituted terminal olefins
(Table 3). Such substrates might allow direct asymmetric access to
NSAID scaffolds, an approach not previously explored. Thus, a
variety of 2-alkyl-2-aryl olefins 19–23 were effectively epoxidised
in fair to excellent conversions (up to >99%) albeit in moderate
selectivity (Table 3, entries 5–8 and 10). In particular, the 70%
conversion observed in alkene 22 with ketone 8 (Table 3, entry 9)
is, to the best of our knowledge, the highest conversion reported
for this substrate, (S)-ibuprofen18 precursor, using an asymmetric
organocatalytic epoxidation strategy. Finally, catalyst loading
was successfully reduced to 0.3 and 0.06 equiv, with the same
enantioselectivity being maintained (Table 3, entries 6–10).
1
crude H NMR (calculated from the integration of the olefinic
and oxiranic protons of the crude reaction mixture). In the cases
where a low conversion was observed, a small portion was purified
using semi-prep TLC.
Acknowledgements
We gratefully acknowledge the European Comission (Marie
Curie Intra European Fellowship, O.B.) and BBSRC (J.F.M.) for
financial support and Dr Carole Bataille for technical support
(chiral GC analysis).
Notes and references
1 J. Marco-Contelles, M. T. Molina and S. Anjum, Chem. Rev., 2004,
104, 2857.
2 For a general review, see: Q.-H. Xia, H.-Q. Ge, C.-P. Ye, Z.-M. Liu and
K.-X. Su., Chem. Rev., 2005, 105, 1603.
3 (a) For general reviews, see: A. U. Barlan, W. Zhang and H. Yamamoto,
Tetrahedron, 2007, 63, 6075; (b) T. Katsuky, in Catalytic Asymmetric
Synthesis, 2nd ed., Ed. I. Ojima, Wiley–VCH, New York, 2000.
4 (a) For general reviews, see: K. Matsumoto, B. Saito and T. Katsuki,
Chem. Commun., 2007, 3619; (b) E. M. McGarrible and D. G. Gilheany,
Chem. Rev., 2005, 105, 1563; (c) T. Katsuki, Synlett, 2003, 281.
5 (a) For general reviews, see: O. A. Wong and Y. Shi, Chem. Rev., 2008,
108, 3958 and references therein; (b) D. Yang, Acc. Chem. Res., 2004,
37, 497; (c) Y. Shi, Acc. Chem. Res., 2004, 37, 488.
6 B. Wang, O. A. Wong, M.-X. Zhao and Y. Shi, J. Org. Chem., 2008, 73,
9539, and references therein.
7 H. Tian, X. She, H. Yu, L. Shu and Y. Shi, J. Org. Chem., 2002, 67,
2435.
Conclusions
In summary, the keto glucosaminide scaffold I made in expedi-
tious, high yielding syntheses from renewable, low cost starting
materials gives access to tuneable robust epoxidation organocata-
lysts which were readily recovered. Until now, single examples of
such catalysts have emerged based on single scaffolds with little
intended design or scope for substituent tuning of reactivity or
selectivity. Although some stereoselectivities were modest, their
ability to handle typically challenging terminal olefin substrates
(e.g. styrene 13, 76% conv., 81% ee) and including relevant
8 O. Bortolini, G. Fantin, M. Fogagnolo and L. Mari, Tetrahedron, 2006,
62, 4482.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4285–4288 | 4287
©

View Article Online
9 (a) For selected reviews, see: F. van Rantwijk and R. A. Sheldon, Curr.
Opin. Biotechnol., 2000, 11, 554; (b) L. P. Hager, F. J. Lakner and A.
Basavapathruni, J. Mol. Catal. B: Enzym., 1998, 5, 95.
10 (a) T. Sone, A. Yamaguchi, S. Matsunaga and M. Shibasaki, J. Am.
Chem. Soc., 2008, 130, 10078; (b) T. Sone, G. Lu, S. Matsunaga and M.
Shibasaki, Angew. Chem., 2009, 121, 1705 and references therein.
11 (a) For general reviews, see: A. Steinreiber and K. Faber, Curr. Opin.
Biotechnol., 2001, 12, 552; (b) A. Archelas and R. Furtoss, Curr. Opin.
Chem. Biol., 2001, 5, 112.
12 The procedure reported by Shibasaki requires the preparation of a
complex chiral multimetallic complex and its applicability is limited
by the presence of basic-labile groups and/or epimerisable positions,
whereas the current disadvantage with regard to catalysts 1–4 is the
lengthy synthetic sequence reported for their synthesis giving low
overall preparative yields.
13 (a) D. P. G. Emmerson, W. P. Hems and B. G. Davis, Org. Lett.,
2006, 8, 207; (b) D. P. G. Emmerson, W. P. Hems and B. G. Davis,
Tetrahedron: Asymmetry, 2005, 16, 213; (c) D. P. G. Emmerson, R.
Villard, C. Mugnaini, A. Batsanov, J. A. K. Howard, W. P. Hems, R. P.
Tooze and B. G. Davis, Org. Biomol. Chem., 2003, 1, 3826.
14 A. Armstrong, I. Washington and K. N. Houk, J. Am. Chem. Soc.,
2000, 122, 6297.
15 No conversion was observed using methyl N-acetyl-a-D-
glucosaminide, the hydroxyamide analogue of ketone 8 (Table 1,
entry 1), and indicated that the amide group alone was not acting as
an epoxidation catalyst and that the uncatalysed background reaction
between OxoneꢀR and olefin occurs at a negligible rate.
16 It should be noted that good conversions (up to 91%) for 2-
vinylnaphthalene 12 could be reached although with complete
loss of stereoselectivity by using ketone
details).
8 (see ESI† for full
17 A. Goeddel, L. Shu, Y. Yuan, O. A. Wong, B. Wang and Y. Shi, J. Org.
Chem., 2006, 71, 1715, and references therein.
18 R. A. Sheldon, J. Chem. Technol. Biotechnol., 1996, 67, 1. Despite the
fact that the (S)-enantiomer is responsible for the desired therapeutic
effect, ibuprofen is currently administered as the racemate.
19 (a) For some reviews on the stereoselective chemical synthesis of
arylpropionic acids, see: H. R. Sonawane, N. S. Bellur, J. R. Ahuja
and D. G. Kulkarni, Tetrahedron: Asymmetry, 1992, 3, 163; (b) J.-P.
Rieu, A. Boucherle, H. Cousse and G. Mouzin, Tetrahedron, 1986, 42,
4095.
20 Reactions at pH 10.6 were carried out using buffer 0.05
M
Na2B4O7·10H2O (3 mL in 4 ¥ 10-4 M Na2EDTA) and K2CO3 (240 mg,
1.74 mmol) in water (2 mL) as base solution.
4288 | Org. Biomol. Chem., 2009, 7, 4285–4288
This journal is
The Royal Society of Chemistry 2009
©