Communication
bamates (entries 18–19), which
were found to be difficult for the
Os-mediated method,[2d] also
participated readily in the reac-
tion, albeit with diminished dia-
stereoselectivity for the trisubsti-
tuted olefin (entry 19). Further
investigation revealed that N-aryl
amides were also excellent sub-
strates (entries 20–21).
Table 1. Optimization of electrolysis conditions for aminooxygenation.[a]
Entry
Electrolysis conditions
Yield [%][c]
1
2
3
4
Bu4NBF4 (0.1m), H2O/CH3CN (1:19), 608C, 2.5 FmolÀ1
70
81
80
90
89
10
83
61
Bu4NBF4 (0.1m), LiOH (1 equiv), H2O/CH3CN (1:19), 608C, 2.8 FmolÀ1
Bu4NBF4 (0.1m), Cs2CO3 (1 equiv), H2O/CH3CN (1:19), 608C, 3.2 FmolÀ1
Bu4NBF4 (0.1m), Na2CO3 (1 equiv), H2O/CH3CN (1:19), 608C, 1.5 FmolÀ1
Et4NOTs (0.1m), Na2CO3 (1 equiv), H2O/CH3CN (1:19), 608C, 3.9 FmolÀ1
Bu4NBF4 (0.1m), Na2CO3 (1 equiv), CH3CN, 608C, 5.2 FmolÀ1
The stereochemistry of the
aminooxygenation was deter-
mined by a single-crystal X-ray
diffraction study of 2h and 2q.
In each case, a clear trans ar-
rangement of the nitrogen atom
relative to the alkoxyamine
group (Figure 1) was found in
the product. These findings were
consistent with the preferential
5
6[d]
7
Bu4NBF4 (0.1m), Na2CO3 (1 equiv), H2O/CH3CN (1:9), 608C, 3.0 FmolÀ1
Bu4NBF4 (0.1m), Na2CO3 (1 equiv), H2O/CH3CN (1:19), RT, 3.3 FmolÀ1
8
[a] Reaction conditions: 1 (0.35 mmol), TEMPO (0.70 mmol), H2O (0.5 mL), CH3CN (9.5 mL), 608C. [b] Determined
1
by H NMR spectroscopic analysis of the crude reaction mixture. [c] Isolated yield. [d] 1a was recovered in 70%
yield.
entry 1). The addition of a base such as LiOH, Cs2CO3, or
Na2CO3 improved the yield to 81, 80, or 90%, respectively (en-
tries 2–4). Moreover, Na2CO3 also reduced the amount of elec-
tricity needed for the complete consumption of 1a
(1.5 FmolÀ1). However, substituting Bu4NBF4 with Et4NOTs
(entry 5), altering the water composition of the solvent system
(entries 6–7), and conducting the reaction at a lower tempera-
ture (entry 8) all led to a less efficient reaction. Based on these
results, it was concluded that the optimal reaction conditions
would involve adding one equivalent of Na2CO3 to the
Bu4NBF4 electrolyte solution in H2O/CH3CN (1:19) and conduct-
ing the electrolysis at 608C (entry 4).
The N-phenyl ring of 1a was modified with various function-
al groups to investigate their effects on alkene aminooxygena-
tion (Table 2). A broad spectrum of substituents with diverse
electronic properties were found to be well-tolerated, includ-
ing electron-donating groups, such as Me (Table 2; entry 1)
and OMe (entries 2–3), halogens, such as F (entries 4–6) and Cl
(entry 7), and electron-withdrawing groups such as CF3
(entry 8), Ac (entry 9), and CN (entry 10). Carbamates with
a 2,6-disubstituted phenyl group, previously reported to be in-
compatible with hypervalent iodine-mediated cyclization,[3a]
also served as efficient substrates (entries 11–12). The introduc-
tion of a nitro group, however, led to less satisfactory re-
sults,[3a,12] which was probably caused by the decomposition of
the product 2n during the electrolysis.[13]
Further examination of the substrate scope revealed a broad
tolerance of the electrochemical method to cyclic alkenes. Tri-
substituted olefins could be efficiently difunctionalized to gen-
erate tetrasubstituted stereogenic centers that are difficult to
construct with existing methods[4] (Table 2; entries 14–16). No-
tably, the electrolysis of 1o was carried out on a gram-scale
without difficulty and the reaction exhibited exclusive prefer-
ence for the proximal double bond over the distal one
(entry 14). Introduction of a nitrogen functionality into the cy-
clohexene ring was tolerated and led to the formation of
a highly functionalized piperidine (entry 17). Cyclopentenyl car-
approach of TEMPO to the cyclization-derived cyclic carbon
radical from the less hindered direction in the transition state
of the radical-radical combination reaction. Hence, using cyclic
alkenes as substrates, this electrochemical approach offers an
efficient access to trans-configured products with a high level
of regio- and stereochemical control, complementing the cis-
selective Os-based strategy.[2c–e]
Following the success with cyclic alkenes, effort was extend-
ed to testing acyclic systems for their suitability as substrates
(Table 3). As expected, the reaction proceeded in high yield
with a multitude of tri-, and disubstituted olefins. Aminooxyge-
nation of 1,2-disubstituted alkenes was, however, nonselective
and afforded a mixture of diastereomers (entries 3–4). Further-
more, the monosubstituted alkene 3 f (Table 3; entry 6)
seemed to necessitate significantly higher current usage com-
pared to the other more substituted substrates.
As mentioned earlier, the TEMPO-derived alkoxyamine
moiety can provide facile access to a variety of synthetically
useful functionalities. For instance, when using 2a and 2p as
the starting materials, the alkoxyamine group in these com-
pounds can be readily transformed into a hydroxyl group by
reductive cleavage of the NÀO bond, a carbonyl group by oxi-
dation,[3d] or
(Scheme 2).
a
double bond by thermal elimination[10]
To confirm the hypothesis that the aminooxygenation reac-
tion underwent a radical mechanism, compound 9 bearing
a cyclopropane group as the radical probe[12,14] was synthe-
sized and electrolyzed in the presence of TEMPO (Scheme 3).
Indeed, the ring-opening product 10 was obtained in 52%
yield, providing solid evidence for the involvement of a nitro-
gen-centered radical.
To give rise to the nitrogen radicals, the substrates could
either be directly oxidized at the anode, or they could be oxi-
dized by an oxoammonium ion that in turn originates from
anodic oxidation of TEMPO[15] (Scheme 4). To test these possi-
bilities, the oxidation potentials of TEMPO and 1a were mea-
sured by cyclic voltammetry,[16] which indicated that the former
Chem. Eur. J. 2014, 20, 12740 – 12744
12741
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim