Organic Letters
Letter
Table 1, to minimize the eluted DMAP’s hazardous effect in
the downstream carbene reaction, 5 equiv of DMAP mixed
Batch Conditions
Table 1. Optimization of Aryldiazoacetate Formation in
isolated yield of 2 (%)
a
a
a
entry
eluent
DCM
DMAP/DCM
DMAP/DCM
DMAP (equiv)
first
second
third
entry
condition variation
additive
yield (%)
ee (%)
b
1
2
3
5
5
10
30
44
96
22
35
88
24
1
2
3
4
−
−
silica
4 Å MS
HFIP
HFIP
18
64
67
72
77
97
98
97
b
c
c
b
39
a
82
b
a
b
silica
HFIP
The column was recycled three times. The column was flushed with
air for 10 min before the third batch of hydrazone was loaded. The
eluent concentration is 0.06 mol/L DMAP in DCM.
c
a
The first reaction vial was charged with 40 mg of silica powder. The
procedure.
with 10 equiv of Cu(OAc) ·H O and 4 g silica were initially
2
2
packed in the column (Table 1, entries 1 and 2). However, the
hydrazone was not fully converted, although flushing the
column with air before the third run and including DMAP in
the eluent (0.06 M) did increase the conversion. We
hypothesized that without enough base to entirely activate
the Cu(OAc) ·H O catalyst to accelerate the oxidation
but this could be eliminated by adding silica to the copper
acetate in the first oxidation step (Table 2, entry 3). Further
enhancement of the yield of 3 (72%) was achieved by adding
the crude aryldiazoacetate 2 dropwise to the rhodium-
catalyzed reaction. These results indicate that a copper
acetate-impregnated silica column should be a useful solid
phase for generating aryldiazoacetates and would likely trap
water from passing through the rhodium-catalyzed reaction.
ReactIR studies were conducted to understand further the
role of DMAP and HFIP in the rhodium-catalyzed reaction
2
2
3
5
reaction, the column efficiency was insufficient given the
limited residence time (1 min). We therefore packed 10 equiv
of DMAP with Cu(OAc) ·H O in the column and used
2
2
DMAP/DCM as the eluent to keep the Cu(OAc) ·H O
2
2
saturated with base coordination. The new conditions (Table
, entry 3) gave full hydrazone conversion for three sequential
1
batches. Hydrazone 1 (0.2 mmol) was added on the top of the
column, and aryldiazoacetate 2 was obtained at the bottom of
1
the column in 1 min at room temperature. From the crude H
NMR, the reaction was shown to be very clean with only a
trace amount of azine dimer formed. This high column
efficiency was maintained for three batches with no require-
ment of air flushing or catalyst recharging.
The next step was to combine the upstream hydrazone
oxidation with the downstream cyclopropanation reaction to
determine the compatibility of the two reactions. This was
initially explored as a batch-to-batch procedure, and the key
results are summarized in Table 2. Hydrazone oxidation was
conducted in a vial under catalytic conditions of Cu(OAc)2·
H O (10 mol %) and DMAP (60 mol %) in dichloromethane
2
(
DCM). The oxidation reaction mixture was stirred open to
air, and once the reaction had reached completion, the formed
crude aryldiazoacetate 2 was directly injected into a second vial
with styrene, Rh (R-p-Ph-TPCP) , and 4 Å molecular sieves
2
4
(
MS) in dichloromethane under N for the tandem cyclo-
2
propanation reaction. Under these conditions, the cyclo-
propane 3 was formed in only 18% yield and the level of
enantioselectivity was moderate (77% ee). Most of the
aryldiazoacetate 2 remained unreacted (Table 2, entry 1),
presumably because the DMAP required for the hydrazone
oxidation suppresses the reactivity of the Rh(II) catalyst in the
second step. We have recently shown that HFIP as an additive
in rhodium-catalyzed reactions can limit interference by
Figure 1. Kinetic investigation of the effect of DMAP and HFIP on
Rh(II)-catalyzed cyclopropanation.
in 0.1 mL of dichloromethane) was injected into a solution of
the aryldiazoacetate 2, styrene, and DMAP in dichloro-
methane, the distinctive IR signal for the diazo compound
did not change. However, when HFIP (10 equiv) was injected,
the reaction was initiated and proceeded to completion,
generating the cyclopropane 3 in 93% yield and 98% ee (note
the apparent rapid decrease in the level of aryldiazoacetate 2
upon addition of HFIP is due to a change in the concentration
37
nucleophilic heterocycles. Indeed, this was also the case
here, as repeating the reaction with 20 equiv of HFIP
generated the cyclopropane 3 in 64% yield with 97% ee (Table
2
, entry 2). Insertion of O−H into water was a side product,
5
364
Org. Lett. 2021, 23, 5363−5367
Wei, Bo
