Enantioselective olefin epoxidation using homologous amine and iminium catalysts-a direct comparison

Article abstract of DOI:10.1016/j.tetlet.2006.05.132

Homologous biphenyl and (diastereomeric) binaphthyl tertiary azepines and quaternary iminium salts were prepared from (+)-(S,S)-l-acetonamine. Both the amines and iminium ions behave as effective catalysts for the enantioselective epoxidation of unfunctio

Full text of DOI:10.1016/j.tetlet.2006.05.132

Tetrahedron Letters 47 (2006) 5297–5301
Enantioselective olefin epoxidation using homologous amine
and iminium catalysts—a direct comparison
a
a
a
´ ´
´
Maria-Helena Gonc¸alves, Alexandre Martinez, Stephane Grass,
b
a,
*
´ ˆ
Philip C. Bulman Page and Jerome Lacour
a
`
`
De´partement de Chimie Organique, Universite´ de Geneve, quai Ernest Ansermet 30, CH-1211 Geneve-4, Switzerland
bDepartment of Chemistry, Loughborough University, Loughborough, Leicestershire, LE11 3TU, UK
Received 13 April 2006; revised 18 May 2006; accepted 22 May 2006
Available online 12 June 2006
Abstract—Homologous biphenyl and (diastereomeric) binaphthyl tertiary azepines and quaternary iminium salts were prepared
from (+)-(S,S)-L-acetonamine. Both the amines and iminium ions behave as effective catalysts for the enantioselective epoxidation
of unfunctionalized olefins (ee up to 83%).
Ó 2006 Elsevier Ltd. All rights reserved.
Chiral non-racemic epoxides are not only useful precur-
sors for organic chemists, but also frequently met struc-
tures in natural products, often related to their
biological activities (Eq. 1).1 A number of efficient meth-
ods exist for their preparation from olefins and many of
them use transition metal catalysts such as the Katsuki–
Sharpless or Katsuki–Jacobsen protocols.2 In the recent
years, much effort has been devoted to the development
of organocatalyzed epoxidation conditions that afford
metal-free procedures; the catalysts being perhydrate,
dioxirane, oxaziridine, or oxoammonium moieties as
well as ammonium or oxaziridinium salts.3
processes possible.5 The first example of an enantioselec-
tive iminium catalyzed reaction was reported in 1993.6
Since this pioneering work, several successful enantiose-
lective variants of the reaction have been reported,7–9
among which are studies using biphenyl 1i10 and binaph-
thyl 2i and 3i iminium salts;11 these compounds were
derived from (+)-L-acetonamine used as an exocyclic
chiral auxiliary (Fig. 1).12
In the case of 1i, the twisted [7]-membered ring is
conformationally labile and single enantiomers are read-
ily prepared (vide infra). Two different types of salts,
namely compounds [1i][BPh4] and [1i][TRISPHAT],
catalyst
NaHCO3, oxone
CH3CN / Water 10:1
0 C (conditions A)
˚
R1
R2
H
O
X
R1
R2
H
R3
ð1Þ
O
R3
(L)-[1i][X]
or
N
NaHCO3, oxone
O
CH2Cl2 / Water 3:2
Ph
0
C (conditions B)
˚
Oxaziridinium ions are interesting alternatives to the
commonly used dioxiranes.4 Such organic salts are effec-
tive oxygen transfer reagents towards nucleophilic sub-
strates and electron-rich unfunctionalized olefins in
particular. Moreover, the propensity of iminium ions
to react with OxoneÒ triple salt to generate the oxazirid-
inium species renders the development of catalytic
X
X
O
O
O
O
N
N
Ph
Ph
(
Ra,L)-[2i][X]
(Sa,L)-[3i][X]
Figure 1. Selected non-racemic iminium salts and their absolute
configuration, Xꢀ being a lipophilic non-coordinating anion (BPh4
or TRISPHAT).
*
Corresponding author. E-mail: jerome.lacour@chiorg.unige.ch
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.05.132

5298
M.-H. Gonc¸alves et al. / Tetrahedron Letters 47 (2006) 5297–5301
have been utilized previously in epoxidation reactions—
no major differences being observed between the two ion
pairing systems.10,13 In the case of 2i and 3i, the presence
of the stereogenic configurationally rigid binaphthyl
core creates a diastereomeric relationship. Both salts
(ꢀ)-[2i][BPh4] and (+)-[3i][BPh4] of (Ra,L) and (Sa,L)
configuration, respectively, were prepared. An interest-
ing matched/mismatched behaviour was characterized;
salt [2i][BPh4] leading to quite higher conversions than
its diastereomer. On the whole, compound (ꢀ)-
[2i][BPh4] is one of the most effective iminium salt cata-
lysts to date (ee up to 95%).11
O
O
O
O
N
N
Ph
Ph
(R
a,L)-2a
(Sa,L)-3a
Figure 2. Tertiary amines 2a and 3a directly related to iminium cations
2i and 3i.
the iminium salt; and (iii) an ion pair metathesis with
an ammonium TRISPHAT salt to afford the final prod-
uct (60%, two steps).19 With both compounds 1a and
[1i][TRISPHAT] available, there was thus a unique
opportunity to perform an amine/ammonium versus
iminium comparison—tertiary amines of type 1a being
undocumented prior to this study as catalysts in (enan-
tioselective) olefin epoxidation reactions.
Whereas the epoxidation of olefins catalyzed by iminium
salts has been known for quite some time, the mediation
of the reaction by amines and/or ammonium salts is still
a new topic.14 It was only in 2000 that the catalyzed
enantioselective epoxidation of olefins by secondary
amines was reported (ee up to 66%), the involvement
of ammonium species in the key oxidation transfer step
only being described in 2003.15,16 Recently, various sec-
ondary amines were studied in this context and a bene-
ficial influence of electron-withdrawing atoms (such as
fluorine) at the b-position relative to the amino group
was demonstrated. In that report, the influence of the
reaction medium was also examined and different out-
puts resulted from the reactions that were performed
in slightly acidic conditions: type A: CH3CN/NaH-
CO3/H2O and type B: CH2Cl2/NaHCO3/18-crown-6/
H2O.17
Me
4
5
6
Two different sets of epoxidation conditions (A and B,
vide supra) and three different prochiral trisubstituted
unfunctionalized alkenes (4–6) were selected for the
study. The results are reported in Table 1. Significantly,
both reagents 1a and [1i][TRISPHAT] behaved as effec-
tive catalysts under the two sets of experimental condi-
tions.20 Non-racemic epoxides of analogous absolute
configurations were isolated from the reactions with 1a
and [1i][TRISPHAT]. Whereas amine 1a performed bet-
ter in terms of conversions and enantiomeric excesses in
CH3CN/H2O (conditions A), iminium salt [1i][TRIS-
PHAT] gave better (overall) results in biphasic
CH2Cl2/H2O medium (conditions B). Enantiomeric
excesses up to 51% and 68% (alkene 5) were obtained
with 1a and [1i][TRISPHAT], respectively, the 51%
value being in fair comparison with that previously
obtained with secondary amine/ammonium salts.15,17
So far, the most selective amine/ammonium catalysts
have been a-substituted pyrrolidine moieties for which
no stable iminium analogues can be found.18 As such,
it has been difficult to compare the catalytic activity
and selectivity of ammonium moieties with that of
related iminium species. It was therefore debatable as to
which of these two classes of related catalysts is the most
effective—if either. Herein, we report a study in which
tertiary amines 1a, 2a and 3a (Scheme 1 and Fig. 2), di-
rectly related to iminium cations 1i, 2i and 3i, have been
synthesized, and all these derivatives were tested as cat-
alysts for the enantioselective epoxidation of olefins.
As indicated above, iminium salt [1i][TRISPHAT] is an
effective catalyst for the asymmetric epoxidation of pro-
chiral alkenes. This compound can be prepared in three
steps from 2,20-bis(bromomethyl)biphenyl using
standard reactions (Scheme 1): (i) an alkylation with
(+)-L-acetonamine to afford amine 1a (88%); (ii) a sub-
sequent elimination with N-bromosuccinimide to form
To extend the scope of the study, and potentially
increase the selectivity of the amine/ammonium cata-
lyzed reactions, compounds 2a and 3a were prepared
following the protocol detailed above (Scheme 1) with
(R)- and (S)-2,20-bis(bromomethyl)-1,10-binaphthyl as
substrates, respectively, these compounds being further
derived into the diastereomeric iminium salts [2i][TRIS-
PHAT] and [3i][TRISPHAT].
Br
Olefins 4–6 were then treated under conditions A and B
with substoichiometric amounts (5 mol %) of 2a, 3a,
[2i][TRISPHAT] and [3i][TRISPHAT]. The results are
reported in Tables 2 and 3; all four derivatives behave
as catalysts. Careful analysis of the data reveals a num-
ber of subtleties, but some general trends can be found.
O
i)
ii)
N
[1i][TRISPHAT]
iii)
O
Br
Ph
1a
Scheme 1. Reagents and conditions: (i) (+)-L-acetonamine (1.0 equiv),
CH3CN, reflux, 88%; (ii) NBS (1.1 equiv), CH2Cl2, 20 °C; (iii) [R3NH]-
[TRISPHAT] (1.2 equiv), chromatography (basic Al2O3, CH2Cl2),
60% (two steps).
As far as solvent effects are concerned, CH3CN/H2O
conditions (A) were better overall than biphasic

M.-H. Gonc¸alves et al. / Tetrahedron Letters 47 (2006) 5297–5301
5299
Table 1. Enantioselective epoxidation of olefins 4–6 using 1a and [1i][TRISPHAT] as catalysts
Alkenec
Amine 1a
Iminium [1i][TRISPHAT]
Conditions Aa Conditions Bb
Conditions Aa
Conditions Bb
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
4
5
6
90e,f
50e,f
97d
53
51
36
(ꢀ)-(S,S)
(+)-(1R,2S)
(ꢀ)-(S,S)
78d
66d
73d
26
23
21
(ꢀ)-(S,S)
(+)-(1R,2S)
(ꢀ)-(S,S)
75e,f
36e,f
95d
54
57
33
(ꢀ)-(S,S)
(+)-(1R,2S)
(ꢀ)-(S,S)
81d
85d
88d
54
68
36
(ꢀ)-(S,S)
(+)-(1R,2S)
(ꢀ)-(S,S)
a Conditions A: 5 mol % of catalyst, 2.0 equiv OxoneÒ, 5.0 equiv NaHCO3, CH3CN/H2O (10:1), 0 °C. Average of at least two runs.
b Conditions B: 5 mol % of catalyst, 2.5 mol % 18-C-6, 1.1 equiv OxoneÒ, 4.0 equiv NaHCO3, CH2Cl2/H2O (3:2), 0 °C. Average of at least two runs.
c The enantiomeric excesses were determined by CSP-GC (4, Chiraldex Hydrodex b-3P) or CSP-HPLC (5 and 6, Chiralcel OD-H); the conversions
using an internal standard (naphthalene).
d 2 h reaction time.
e 15 min reaction time.
f Complete conversion was observed in 2 h along with some product decomposition. Care was thus taken to select a shorter reaction time.
Table 2. Enantioselective epoxidation of olefins 4–6 using 2a and [2i][TRISPHAT] as catalysts
Alkenec
Amine 2a
Iminium [2i][TRISPHAT]
Conditions Aa Conditions Bb
Conditions Aa
Conditions Bb
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
4
5
6
100e,f
99e,f
94d
78
80
48
(ꢀ)-(S,S)
(+)-(1R,2S)
(ꢀ)-(S,S)
90d
87d
58d
65
45
48
(ꢀ)-(S,S)
(+)-(1R,2S)
(ꢀ)-(S,S)
64e,f
34e,f
88d
79
71
47
(ꢀ)-(S,S)
(+)-(1R,2S)
(ꢀ)-(S,S)
99d
90d
80d
77
78
46
(ꢀ)-(S,S)
(+)-(1R,2S)
(ꢀ)-(S,S)
a Conditions A: 5 mol % of catalyst, 2.0 equiv OxoneÒ, 5.0 equiv NaHCO3, CH3CN/H2O (10:1), 0 °C. Average of at least two runs.
b Conditions B: 5 mol % of catalyst, 2.5 mol % 18-C-6, 1.1 eq OxoneÒ, 4.0 equiv NaHCO3, CH2Cl2/H2O (3:2), 0 °C. Average of at least two runs.
c The enantiomeric excesses were determined by CSP-GC (4, Chiraldex Hydrodex b-3P) or CSP-HPLC (5 and 6, Chiralcel OD-H); the conversions
using an internal standard (naphthalene).
d 2 h reaction time.
e 15 min reaction time.
f Complete conversion was observed in 2 h along with some product decomposition. Care was thus taken to select a shorter reaction time.
Table 3. Enantioselective epoxidation of olefins 4–6 using 3a and [3i][TRISPHAT] as catalysts
Alkenec
Amine 3a
Iminium [3i][TRISPHAT]
Conditions Aa Conditions Bb
Conditions Aa
Conditions Bb
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
Conv.
(%)
ee
(%)
Conf.
4
5
6
99
97
97
76
78
52
(+)-(R,R)
(ꢀ)-(1S,2R)
(+)-(R,R)
70
<5
23
77
57
53
(+)-(R,R)
(ꢀ)-(1S,2R)
(+)-(R,R)
98
99
85
81
83
52
(+)-(R,R)
(ꢀ)-(1S,2R)
(+)-(R,R)
54
33
15
78
69
54
(+)-(R,R)
(ꢀ)-(1S,2R)
(+)-(R,R)
a Conditions A: 5 mol % of catalyst, 2.0 equiv OxoneÒ, 5.0 equiv NaHCO3, CH3CN/H2O (10:1), 0 °C, 2 h. Average of at least two runs.
b Conditions B: 5 mol % of catalyst, 2.5 mol % 18-C-6, 1.1 equiv OxoneÒ, 4.0 equiv NaHCO3, CH2Cl2/H2O (3:2) , 0 °C, 2 h. Average of at least two
runs.
c The enantiomeric excesses were determined by CSP-GC (4, Chiraldex Hydrodex b-3P) or CSP-HPLC (5 and 6, Chiralcel OD-H); the conversions
using an internal standard (naphthalene).
CH2Cl2/H2O (B), better conversions occurred in the
more polar conditions. In several instances, the reac-
tions were complete in 15 min using conditions A,
whereas a time of 2 h was necessary with the haloge-
nated solvent mixture. This is true for all catalysts and
compound 2a in particular (e.g., olefin 4, A: 15 min,
100% versus B: 2 h, 90%). This trend also holds true
for the enantiomeric excesses, which were higher in the
more polar conditions (olefin 5, catalyst 2a, A: ee 80%
versus B: ee 45%).21
one generally observes analogous levels of stereoinduc-
tion in the (Ra,L) and (Sa,L) series, the only major
difference being reversal of the sense of induction for
the non-racemic epoxides. It indicates that the binaph-
thyl framework is a more effective chiral auxiliary than
L-acetonamine, since the configuration of the epoxides
changes with the inversion of the absolute configuration
of the biaryl moiety.
This general lack of ‘matched’/‘mismatched’ distinction,
as far as enantiomeric excesses are concerned, does not
apply to conversions. Catalyst 2i performed better than
3i—as previously reported.11 Amine 2a also catalyzed
If one now compares the selectivity of the diastereomeric
catalysts together—that is 2a with 3a, and 2i with 3i—

5300
M.-H. Gonc¸alves et al. / Tetrahedron Letters 47 (2006) 5297–5301
Rassias, G. A.; Bethell, D.; Schilling, M. B. J. Org. Chem.
1998, 63, 2774–2777; Page, P. C. B.; Rassias, G. A.;
Barros, D.; Bethell, D.; Schilling, M. B. J. Chem. Soc.,
Perkin Trans. 1 2000, 3325–3334; Page, P. C. B.; Rassias,
G. A.; Barros, D.; Ardakani, A.; Buckley, B.; Bethell, D.;
Smith, T. A. D.; Slawin, A. M. Z. J. Org. Chem. 2001, 66,
6926–6931; Page, P. C. B.; Barros, D.; Buckley, B. R.;
Ardakani, A.; Marples, B. A. J. Org. Chem. 2004, 69,
3595–3597; Page, P. C. B.; Buckley, B. R.; Heaney, H.;
Blacker, A. J. Org. Lett. 2005, 7, 375–377; Page, P. C. B.;
Buckley, B. R.; Rassias, G. A.; Blacker, A. J. Eur. J. Org.
Chem. 2006, 803–813.
the reaction better than 3a, in biphasic CH2Cl2/water
conditions in particular (e.g., olefin 5, conditions B,
2a: 87% versus 3a: <5%).
If one now compares the selectivity of the homologous
amine and iminium salts—that is 2a with 2i, and 3a
with 3i—one notices that the amines and iminium salts
(i) induce the same sense of stereoselective induction
into the non-racemic epoxides, and (ii) lead to
comparable levels of enantiomeric excesses (with the
‘exception’ of olefin 5).22 A subtle solvent effect is
observed for compounds 2a and 2i, the amine perform-
ing slightly better in conditions A and the iminium in
CH2Cl2/water (conditions B). For derivatives 3a and
3i, the iminium cation leads to slightly better results
in both solvent conditions.
8. For acyclic iminium catalysts, see: Armstrong, A.; Ahmed,
G.; Garnett, I.; Goacolou, K.; Wailes, J. S. Tetrahedron
1999, 55, 2341–2352; Minakata, S.; Takemiya, A.;
Nakamura, K.; Ryu, I.; Komatsu, M. Synlett 2000,
1810–1812; Wong, M.-K.; Ho, L.-M.; Zheng, Y.-S.; Ho,
C.-Y.; Yang, D. Org. Lett. 2001, 3, 2587–2590.
9. For mechanistic studies, see: Washington, I.; Houk, K. N.
J. Am. Chem. Soc. 2000, 122, 2948–2949; Page, P. C. B.;
Barros, D.; Buckley, B. R.; Marples, B. A. Tetrahedron:
Asymmetry 2005, 16, 3488–3491.
10. (a) Page, P. C. B.; Rassias, G. A.; Barros, D.; Ardakani,
A.; Bethell, D.; Merifield, E. Synlett 2002, 580–582; (b)
Lacour, J.; Monchaud, D.; Marsol, C. Tetrahedron Lett.
2002, 43, 8257–8260.
To conclude, amines 1a–3a perform essentially as well as
their iminium salts 1i–3i as catalysts for the enantioselec-
tive epoxidation of some prochiral olefins—in particular
in the acetonitrile/water conditions. As making the
amines requires less synthetic steps than the preparation
of the iminium salts, it is therefore advantageous to use
these ‘simpler’ reagents for synthetic applications.
11. Page, P. C. B.; Buckley Benjamin, R.; Blacker, A. J. Org.
Lett. 2004, 6, 1543–1546.
12. (+)-L-Acetonamine is (+)-5-amino-2,2-dimethyl-4-phenyl-
1,3-dioxane.
Acknowledgements
13. TRISPHAT stands for Tris(tetrachlorobenzenediolato)-
phosphate(v) anion: Lacour, J.; Ginglinger, C.; Grivet, C.;
Bernardinelli, G. Angew. Chem., Int. Ed. 1997, 36, 608–
609; Favarger, F.; Goujon-Ginglinger, C.; Monchaud, D.;
Lacour, J. J. Org. Chem. 2004, 69, 8521–8524.
We are grateful for financial support of this work by the
University of Geneva, the Swiss National Science Foun-
dation and the State Secretariat for Research and
Science.
14. Armstrong, A. Angew. Chem., Int. Ed. 2004, 43, 1460–
1462.
15. Adamo, M. F. A.; Aggarwal, V. K.; Sage, M. A. J. Am.
Chem. Soc. 2000, 122, 8317–8318; Adamo, M. F. A.;
Aggarwal, V. K.; Sage, M. A. J. Am. Chem. Soc. 2002,
124, 11223; Aggarwal, V. K.; Lopin, C.; Sandrinelli, F.
J. Am. Chem. Soc. 2003, 125, 7596–7601.
16. Diastereoselective epoxidation reactions of allylic amines
with OxoneÒ have also been reported: Aggarwal, V. K.;
Fang, G. Y. Chem. Commun. 2005, 3448–3450.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.
2006.05.132.
17. Ho, C. Y.; Chen, Y. C.; Wong, M. K.; Yang, D. J. Org.
Chem. 2005, 70, 898–906.
References and notes
18. Appropriate pyrrolidine and carbonyl moieties can react
together to generate iminium species. Unfortunately,
most of these species, in particular the most hindered
ones, are prone to solvolysis in the reaction conditions:
see Ref. 8.
19. TRISPHAT is chiral. However, it was shown (see Ref.
10b) that its configuration plays no role in the enantio-
selective epoxidation reaction. For ease of synthesis and
better chemical yields in ion pair exchange metathesis,
enantiopure [cinchonidinium][D-TRISPHAT] was used as
a source of hexacoordinated phosphate anion.
20. In view of the previous results by the group of Yang (Ref.
17), one can reason that the presence of two electron-
withdrawing oxygen atoms at b-position relative to the
amino group in L-acetonamine activates the catalytic
activity of the amine/ammonium salt.
21. These results confirm the previous observations that
higher values are obtained for both conversions and
enantiomeric excesses using more polar solvent conditions
and amines as catalysts, see Refs. 15 and 17.
1. Contelles, J. M.; Moilna, M. T.; Anjum, S. Chem. Rev.
2004, 104, 2857–2900.
2. Comprehensive Asymmetric Catalysis I–III; Jacobsen, E.
N., Pfaltz, A., Yamamoto, H., Eds., 1999; Vol. 2.
3. Adam, W.; Saha-Moller, C. R.; Ganeshpure, P. A. Chem.
Rev. 2001, 101, 3499–3548.
4. Barbaro, P.; Bianchini, C. Chemtracts 2001, 14, 274–277;
Adam, W.; Saha-Moeller, C. R.; Zhao, C.-G. Org. React.
2002, 61, 219–516; Wu, X.-Y.; She, X.; Shi, Y. J. Am.
Chem. Soc. 2002, 124, 8792–8793; Curci, R.; D’Accolti, L.;
Fusco, C. Acc. Chem. Res. 2006, 39, 1–9; Goeddel, D.;
Shu, L.; Yuan, Y.; Wong, O. A.; Wang, B.; Shi, Y. J. Org.
Chem. 2006, 71, 1715–1717.
5. Hanquet, G.; Lusinchi, X.; Milliet, P. C. R. Acad. Sci.,
Ser. II: Mec., Phys., Chim., Sci. Terre Univers 1991, 313,
625–628; Lusinchi, X.; Hanquet, G. Tetrahedron 1997, 53,
13727–13738.
6. Bohe, L.; Hanquet, G.; Lusinchi, M.; Lusinchi, X.
Tetrahedron Lett. 1993, 34, 7271–7274.
7. For cyclic iminium catalysts, see: Aggarwal, V. K.; Wang,
M. F. Chem. Commun. 1996, 191–192; Page, P. C. B.;
22. It is remarkable that the amines and iminium salts leads to
such similar enantiomeric excess values. Although it is

M.-H. Gonc¸alves et al. / Tetrahedron Letters 47 (2006) 5297–5301
5301
reasonable to consider a single mechanistic pathway for
the two processes, we have no evidence for it. Under
conditions A or B, no products of oxidation of amines
1a–3a could be characterized or, for that matter, could
iminium cations 1i–3i be recovered at the end of the
reactions.