2970
Z. Wang et al. / Tetrahedron Letters 55 (2014) 2969–2972
Table 1
Table 2
Silver mediated epoxidation of phenyldiazoacetate 1a with anisaldehyde 2aa
NHC-Ag+ mediated epoxidation reaction of 1a with various aldehydes.a,
b,
c
N2
CHO
5 mol% IPrAgCl
/AgOTf
O
N2
O
O
+
Silver catalyst
& Additive
solvent, rt
COOMe
+
R
COOMe
MeOOC
DCM, rt
R
MeOOC
3a-m
1a
2a-m
OMe
2a
OMe
1a
3a
O
O
O
Ag catalyst (mol %)
Additive
(mol %)
Solvent/
time
(h)
Yieldb
(%)
O
MeOOC
OBn
MeOOC
3a,
MeOOC
N
OMe
H
3b,
3c,
88% yield
O
88% yield
O
54% yield
O
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20c
21
AgOTf (5)
No
No
No
No
DCM/24
DCM/8
DCM/8
DCM/24
DCM/8
DCM/8
DCM/8
DCM/8
DCM/8
DCM/8
DCM/8
DCM
DCM/8
CHCl3/8
DCE/8
Toluene
Et2O/8
DCM/24
DCM/24
DCM/8
DCM
<5
52
57
<5
69
41
42
39
12
26
54
NR
20
58
52
NR
40
42
30
88
NR
AgOTf (5)/Ph3P (5)
AgOTf(5)/Cy3P (5)
AgOTf (5)/(C6F5)3P (5)
IPrAgCl (5)
SIPrAgCl (5)
ICyAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
IPrAgCl (5)
MeOOC
MeOOC
AgOTf (5)
AgOTf (5)
AgOTf (5)
AgBF4 (5)
AgSbF6 (5)
AgPF6 (5)
AgNTf2 (5)
Ag2CO3 (5)
AgClO4 (5)
AgOTf (5)
AgOTf (5)
AgOTf (5)
AgOTf (5)
AgOTf (2.5)
AgOTf (10)
AgOTf (5)
—
MeOOC
Br
Cl
3d,
3e,
3f,
62% yield
O
52% yield
O
58% yield
O
MeOOC
MeOOC
MeOOC
F
3g,
3h,
3i,
56% yield
91% yield
O
81% yield
O
O
MeOOC
3j,
Ph
MeOOC
MeOOC
Ph
3k,
3m,
95% yield
51% yield
72% yield
a
Unless noted, all reactions were carried out on 0.1 mmol scale in 2 mL CH2Cl2 at
rt with the addition of 50 mg 4 Å molecular sieves.
IPrAgCl (5)
b
Isolated yields.
a
c
In Table 1 scheme, 3a’s structure only shows relative configuration. Unless
Product 3a–m’s structure only show their relative configuration
noted, all reactions were carried out on 0.1 mmol scale in 2 mL solvent at rt with the
addition of 50 mg 4 Å molecular sieves.
The reaction yields were determined by 1H NMR spectral data.
b
c
2 mol equiv of 1a was used. NR = No Reaction.
the formation of a mixture of carbene dimerization and decompo-
sition products (Table 1, entry 1). When 4 Å molecular sieves were
removed, no desired product was obtained. We envisioned that
suitable ligand binding would enhance silver carbenoid stability
and therefore enhance their reactivity. To our delight, when 5 mo-
l % equiv of Ph3P was added, 3a’s yield could be improved to 52%
(Table 1, entry 2). Electron rich phosphine ligand Cy3P performed
better than Ph3P (Table 1, entry 3), while a combination of silver
salt with electron deficient (C6F5)3P showed poor reactivity. We
carbene source (Table 2). It has been reported that epoxidation
reaction started from the nucleophilic attack of an aldehyde onto
metal carbenoid. Therefore, the aldehyde substrate’s oxygen nucle-
ophilicity would affect the reaction yield. As shown in Table 1, sev-
eral para-, ortho-, and meta-substituted benzaldehydes were
scrutinized. Both electron-donating and electron-withdrawing
benzaldehydes worked very well in this silver carbene [2+1] cyclo-
addition reaction. Electron rich p-alkoxy benzaldehydes afforded
3a and 3b in high reaction yields (Table 2, 3a–b), while p-acetoa-
minobenzaldehyde gave only a moderate yield (Table 2, 3c). The
electron-poor p-halo substituted benzaldehydes provided the de-
sired oxiranes in relatively low yields (Table 2, 3e–g). The m- and
o-substituted benzaldehyde and 2-naphthaldehyde were then
considered that
r donating NHC ligand (N-heterocyclic carbene)
would further improve silver carbenoid reactivity. As shown in
Table 1, treating the reaction mixture with 5 mol % equiv of
IPrAgCl/AgOTf could provide the desired product 3a in 69% yield
(Table 1, entry 5). Other NHC ligands, such as: SIPrAgCl and ICyAgCl,
were also tested, which gave much less yields (Table 1, entry 6–7).
Then, different silver co-catalysts (Table 1, entry 8–13) and sol-
vents (Table 1, entry 14–17) were screened. It was found that
AgBF4, AgSbF6, AgPF6, and AgNTf2 gave the desired product in mod-
erate yields. AgClO4 afforded 3a in a relatively low yield, while Ag2-
CO3 gave no reaction. Further exploration on various solvents was
also fruitless. Thus, catalyst IPrAgCl/AgOTf in CH2Cl2 was proved to
be the best combination (Table 1, entry 5). Improving or lowering
the co-catalyst AgOTf’s equivalence reduced 3a’s reaction yield
(Table 1, entry 18–19). When 1a’s amount was improved to
2 equiv, 3a could be obtained in 88% yield (Table 1, entry 20). In
the control experiment, the sole silver catalyst IPrAgCl without sil-
ver co-catalyst gave no desired product (Table 1, entry 21).10
With the optimal reaction conditions in hand, the substrate
scope was explored and a series of oxirane derivatives were syn-
tested, which gave 3h, 3i, and 3j in moderate to good yields.
a,
b-Unsaturated aldehydes were also evaluated. Both vinyl and
alkynyl substrates worked very smoothly in this reaction (Table 2,
3k–m).
Various diazocarbonyl substrates were then explored (Table 3).
Both phenyl and vinyl diazocarbonyl substrates reacted smoothly
with 2a, which provided a series of desired oxirane products
(Table 3, 4b–g). Alkyl diazocarbonyl substrate was also tested,
but it did not give the desired oxirane product.
In order to elucidate the exact active reaction mediator, we
examined the effect of the ligand loading. As compared with the
effect of the sole silver salt (5 mol % of AgOTf, Table 1, entry 1)
and the combination of AgOTf/Cy3P (5 mol %, Table 1, entry 3),
improving Cy3P’s equivalence to 7 mol % lowered down 3a’s yield
(Scheme 2). Furthermore, only trace amounts of 3a could be de-
tected in the condition of 5 mol % AgOTf and 10 mol % of Cy3P. This
result, combined with the control experiment (Table 1, entry 21),
indicated that IPrAg+ might be the real active reaction mediator.
thesized. At first, various aromatic and
a, b-unsaturated aldehydes
were tested by using methyl phenyldiazoacetate 1a as the reacting