COMMUNICATION
Table 1. Screening of catalysts and reaction parameters in the epoxida-
improvement of ee value up to 97% (Table 1, entries 6–10).
Cumyl hydroperoxide was shown to be a slightly less effec-
tive oxidant than TBHP (Table 1, entry 11), whereas the ee
value significantly dropped to 54% when using hydrogen
peroxide as terminal oxidant (Table 1, entry 12). Pleasingly,
the eQNT loading could be reduced as low as 5 mol% with-
out affecting the process when employing TBHP as the oxi-
dant (Table 1, entries 13 and 14). A brief screening of low
polarity solvents, such as p-xylene, chloroform, and diethyl
ether, confirmed toluene to be the optimum medium for the
epoxidation (Table 1, entries 15–17).
tion of a-benzoyl-N-phenylacrylamide 1a.[a]
The scope and limitation of the asymmetric epoxidation
was next investigated with respect to the nature of substitu-
ents and carbonyl residue present in the starting alkene
1 using eQNT at 5 mol% loading with TBHP at room tem-
perature (Table 2).
Entry
Cat.
t [h]
Yield [%][b]
ee [%][c]
1
2
3
4
5
6
7
8
CN
CD
QN
QD
2
1
1
1
93
97
92
98
98
98
98
98
98
98
96
90
98
98
98
98
98
61
À59
À33
À12
À13
À97
À96
95
Electron-donating groups at different position in the a-
aroyl moiety and a larger naphthoyl residue in alkenes
1 were well tolerated leading to epoxides in excellent yield
and ee value (Table 2, entries 2–6). Compounds 1 with elec-
tron-withdrawing groups at the the a-aroyl moiety proved
to be very reactive and suffered easy decomposition. Never-
theless, when the epoxidation was carried out on compound
1 f, containing a methoxy group at the meta position, the
product was recovered in good yield and 98% ee (Table 2,
entry 7). Alkenes with electron-withdrawing groups in the
aromatic residue of the amide group were smoothly convert-
ed into the product in very high enantioselectivity (Table 2,
entry 8). Aliphatic substituted alkenes 1i–l, either at the
amido, ketone, or both moieties, gave the epoxide with high
to excellent ee values (Table 2, entries 9–12). The corre-
sponding a-benzoyl acrylates 1m–n were converted into the
product after prolonged reaction times using 10 mol% load-
ing of eHQNT as the catalyst (see the Supporting Informa-
tion). The epoxides were obtained in good yield and 74%,
77% ee, respectively (Table 2, entries 13 and 14). The enan-
tioselectivity could be improved to a good level by perform-
ing the reaction at À208C (Table 2, entry 15). Mixed sub-
strates, such as carbamoyl acrylates 1o–p smoothly reacted,
affording the product in good yield and enantioselectivity
(Table 2, entries 16 and 17). Interestingly, alkene 1q contain-
ing a tertiary amide group, treated under otherwise identical
conditions, even after prolonged reaction times, proved to
be unreactive (Table 2, entry 18). This result clearly indi-
(DHQD)2PHAL
2
eQNT
eQDT
eCDT
eCNT
eHQNT
eQNT
eQNT
eQNT
eQNT
eQNT
eQNT
eQNT
0.5
0.5
0.5
0.5
0.5
1
1
1
1
1
9
À95
10
11[d]
12[e]
13[f]
14[g]
15[f,h]
16[f,i]
17[f,j]
96
93
54
96
84
96
90
55
1
1
[a] Reactions performed on 0.1 mmol scale, with 1a/TBHP 1/1.2 ratio at
[1a]0 =0.1m. [b] Isolated yield after silica gel chromatography. [c] Deter-
mined by HPLC analysis using a chiral column. [d] Cumyl hydroperoxide
was used. [e] 50% H2O2 was used. [f] 5 mol% catalyst was used.
[g] 2 mol% catalyst was used. [h] p-Xylene as solvent. [i] CHCl3 as sol-
vent. [j] Et2O as solvent.
Interestingly, when cinchonine (CN) was employed at
10 mol% loading, the epoxide was isolated in high yield and
61% enantiomeric excess (ee) after a short reaction time
(Table 1, entry 1). In our opinion this is a remarkable result
for different reasons: 1) it represents the first example of
a natural cinchona alkaloid catalyzed asymmetric epoxida-
tion; 2) the reaction was clean. Indeed, the epoxide was ob-
tained in high yield, considering the reactive nature of com-
pound 1a as a Michael acceptor and the competing polymer-
ization that could have occurred as a side reaction. The
pseudoenantiomeric cinchonidine (CD) afforded the oppo-
site enantiomer of 2a in comparable yield and asymmetric
induction (Table 1, entry 2). Quinine (QN) and quinidine
(QD) proved to be equally effective in terms of product
yield, although significantly less enantioselective catalysts
(Table 1, entries 3 and 4). The dimeric derivative
(DHQD)2PHAL gave almost quantitatively the epoxide,
albeit with only 13% ee (Table 1, entry 5). To our great de-
light, the epoxidation mediated by commonly used cinchona
thioureas, proceeded rapidly to give epoxide 2a, of both ab-
solute configurations, in excellent yield and with a notable
À
cates the crucial role of the amide N H proton in the pro-
cess. Finally, 2-alkylidene-1,3-dione 1r reacted sluggishly to
give the epoxide in low yield and poor ee value (Table 2,
entry 19).
To better understand the role of the promoter in the cat-
alysis, the epoxidation was further investigated (Table 3).
Model compound 1a was treated with 10 mol% of Schrein-
erꢁs thiourea[11a] and TBHP in toluene at room temperature
(Table 3, entry 1). After 1 hour, the conversion to epoxide
was very low in comparison to the data reported in Table 1
(entries 6–10). The reaction catalyzed by 20 mol% of 1,8-di-
azabicycloundec-7-ene (DBU) likewise proceeded to
a lesser extent when compared with the use of cinchona thi-
Chem. Eur. J. 2012, 18, 6152 – 6157
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6153