Angewandte
Chemie
alkene and catalyst. In it the olefin is the limiting reagent,
albeit with a large excess of the diazoalkane (5–10 equiv). The
research groups of Charette and Braddock have reported the
generation of ethyldiazoacetate in situ from the correspond-
ing amine and its use in the rhodium-catalyzed cyclopropa-
nation of styrene to give moderate yields (45–62%) and no
turned to examining a cobalt–porphyrin catalysts [Co(TPP)],
because they have been shown by Chen and Zhang to be more
active than iron–porphyrins in the cyclopropanation reaction
[
15]
of olefins with ethyl diazoacetate. Under otherwise iden-
tical reaction conditions, only partial conversion was
observed, and the product was formed as a single diastereo-
mer (Table 1, entry 3). As shown in Table 1, the correspond-
ing [Ru(TPP)CO] catalyst was less efficient than [Fe(TPP)Cl]
and furnished the product with a lower d.r. (86:14; Table 1,
entry 4).
[9]
diastereoselectivity (1.5–1:1). Inspired by this work, we
decided to explore the generation of F CCHN in situ in the
3
2
presence of olefins with catalysts in aqueous media. With
respect to the latter, we were particularly interested in
investigating simple complexes of iron. The recent interest in
this metal stems from the obvious advantages, namely its
ready availability along with its relatively low price and
toxicity. Of additional interest is that its use in the synthesis of
fine chemicals is still relatively unexplored when compared to
Nonporphyrin-based complexes were then examined
[
16]
(Scheme 2). (Æ )-Salen–cobalt complex 5
afforded the
product in 51% yield and > 95:5 d.r (Table 1, entry 5).
[10]
late-transition metals.
We commenced our investigations using the commercially
available iron–porphyrin complex [Fe(TPP)Cl] (TPP =
5
,10,15,20-tetraphenyl-21H,23H-porphine), which is known
[
11]
to be a robust catalyst in carbene-transfer processes. In the
experimental set-up, a mixture of test substrate (4-methox-
ystyrene), F CCH NH ·HCl (1.5 equiv), H SO (10 mol%),
3
2
2
2
4
and NaOAc (20 mol%) at ambient temperature was treated
with an aqueous solution of NaNO (0.8m, 1.8 equiv) that was
2
delivered by a syringe pump. The trial run that employed
[
12]
[
Fe(TPP)Cl] and catalytic DMAP afforded the trifluoro-
methyl-substituted cyclopropane as a single trans diaster-
eomer in 89% yield (Table 1, entry 1). Importantly, as little as
1
.5 equivalents of trifluoroethyl amine hydrochloride was
Scheme 2. Structures of nonporphyrin catalysts tested. Tf=trifluoro-
methanesulfonyl.
needed for full consumption of the olefin, which bodes well
for the use of valuable substrates. Notably, the same reaction
performed in a water/toluene (5:1) mixture afforded product
in only 50% yield. Thus, the reaction worked better in the
absence of an organic solvent; it is possible that it proceeds
Complex 6 developed by Nishiyama and co-workers is a
[
13]
“
on” water and that this leads to preferential reaction of the
well-known catalyst for cyclopropanation and has been shown
[
14]
[17]
metal–carbene with the organic substrate.
Lowering the
to be catalytically active in water. Unfortunately, under the
catalyst loading to 1 mol% led to a diminished 70% yield of
the isolated product (Table 1, entry 2). We subsequently
reaction conditions necessary for the formation of the
diazoalkane in situ, none of the desired product was formed
(Table 1, entry 6). Two of the most extensively used metals for
olefin cyclopropanation were examined, namely, rhodium and
[18]
[
a]
copper. Du Bois’s [Rh (esp) ] furnished product in 74%
2 2
Table 1: Scope with diverse catalysts.
yield, albeit with 1:1 d.r (Table 1, entry 7). Interestingly, the
[
19]
Cu–BOX complex 7 failed to give any traces of adduct
Table 1, entry 8). On the basis of these screening results, we
(
decided to focus on the iron catalyst.
The styrene derivatives studied to date have proven to be
excellent substrates for the transformation, leading to product
in high yield as a single trans diastereomer, as assayed by
[
b]
[c]
Entry
Catalyst
Loading [mol%]
d.r.
Yield [%]
[
[
d]
d]
1
1
2
3
4
5
6
7
8
[Fe(TPP)Cl]
[Fe(TPP)Cl]
3
1
3
3
5
5
1.5
5
>95:5
>95:5
>95:5
86:14
>95:5
–
89
70
50
56
51
n.r.
74
n.r.
analysis of the H NMR spectra (Table 2). Even a disubsti-
tuted styrene derivative (Table 2, entry 6) afforded the
product in good yield and diastereoselectivity. Extension of
the method to alkyl-substituted olefins gave the product in
[
d]
[Co(TPP)]
[Ru(TPP)CO]
[
e]
5
6
1
0% yield (Table 2, entry 8). This poor yield might result
from the lower reactivity of this class of olefin, thus making
the decomposition of the metal–carbene faster than reaction
[Rh (esp) ]
1:1
–
2
2
7
[
11]
with the olefin. Although the process is currently limited to
styrenes, it is important to note that most of the products in
Table 2 are either unknown (Table 2, entries 2–4 and 6), have
only been previously synthesized in low yields (24–50%;
[
[
[
a] General procedure: alkene (0.22 mmol, 1 equiv), H O (1.3 mL).
b] Determined by H NMR spectroscopy. [c] Yield of isolated product.
d] DMAP (3 equiv relative to catalyst). [e] N-methylimidazole
10 mol%). DMAP=4-dimethylaminopyridine, n.r.=no reaction.
2
1
(
Angew. Chem. Int. Ed. 2010, 49, 938 –941
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
939
Morandi, Bill
