.
Angewandte
Communications
a very small amount of cyclic nitrone 2a (0–6%; Table 1,
entry 1). By contrast, the treatment of oxime 1a with K3PO4
(10 mol%) in toluene at 1108C brought the reaction to
completion in 12 h, thereby yielding cyclic nitrone 2a in 98%
Table 1: Hydroamination of g,d-unsaturated ketoxime 1a.[a]
Entry Base
[equiv]
Solvent
T
[8C]
t
[h]
Yield of 2a [%][b]
1
none
various
solvents[c]
toluene
o-xylene
PhCl
DMSO
DMF
PhCl
110–150 24
0–6 (58–92)[d]
2
3
4
5
6
7
8
9
K3PO4 (0.1)
K3PO4 (0.1)
K3PO4 (0.1)
K3PO4 (0.1)
K3PO4 (0.1)
K2CO3 (0.1)
KOtBu (0.1)
KOtBu (0.1)
110
120
120
120
120
120
120
60
12
10
4
24
24
24
0.6 96
20
7
98
98
98
59 (29)[d]
12 (72)[d]
21 (77)[d]
PhCl
PhCl
97
10
11
NaOMe (0.1) PhCl
NaH (0.1) PhCl
120
120
67 (30)[d]
88
7
[a] The reactions were carried out using 0.2 mmol of 1a in the solvents
(0.1m). [b] Yield of isolated product. [c] The reactions were examined in
toluene (1108C, no product, 92% recovery of 1a), o-xylene (1508C, 4%
yield of 2a with 70% recovery of 1a), PhCl (1308C, 5% yield of 2a with
78% recovery of 1a), and DMSO (1208C, 6% yield of 2a with 58%
recovery of 1a). [d] Yield of recovered 1a shown in parentheses.
Figure 1. Energy diagrams (in kcalmolÀ1) for the reaction of neutral
oxime 1a (a) and anionic oxime 1a’ (b), as obtained at the B3LYP-
(SCRF)/6-311+G(d,p) level. Zero-point-energy (ZPE) effects are
included. Key bond distances are given in ꢀ in the three-dimensional
models of the transition states.
yield (entry 2). Solvent screening (entries 3–6) revealed that
chlorobenzene was optimal for this transformation (at 1208C
for 4 h, 98% yield; entry 4) and the reactions in polar solvents
such as DMSO and DMF became very sluggish (entries 5 and
6). With K2CO3 as the catalyst, the reaction was less efficient
than with K3PO4 (entry 7). The use of a stronger base, KOtBu,
as the catalyst allowed the reaction temperature to be lowered
to 608C (entries 8 and 9). Other stronger inorganic bases such
as NaH and NaOMe also worked reasonably well (entries 10
and 11).
To obtain detailed information pertaining to the role of
the inorganic base in the formation of cyclic nitrone 2a from
oxime 1a, we performed density functional theory (DFT)
calculations to analyze the energy profiles for the reactions of
oxime 1a in its neutral and anionic forms at the B3LYP-
(SCRF)/6-311 + G(d,p) level (Figure 1), taking into account
the chlorobenzene solvent effect.[15–18] In the cyclization of
neutral oxime 1a (in the absence of the base), the oxime
nitrogen and the hydroxy hydrogen attack the two carbon
atoms of the alkene in a concerted fashion (i.e. through retro-
Cope hydroamination) to yield the five-membered cyclic
nitrone 2a. However, the transition state for this reaction (TS-
1) is highly unstable, and thus the energy barrier is very high
(31.3 kcalmolÀ1; Figure 1a).
ure 1b). Interestingly, in the transition state TS-2, K+ interacts
with the oxime oxygen atom and stays close to the ipso-
carbon atom of the phenyl group. While the ipso-carbon has
a positive charge according to the Mulliken population
analysis (Figure S1 in the Supporting Information), the
adjacent three carbon atoms are negatively charged. The K+
ion thus effectively stabilizes these negative charges that
accumulate as the cyclization reaction progresses. The
resultant carbanion intermediate 2a’ will subsequently
accept a proton from oxime 1a to yield 2a and regenerate
anionic oxime 1a’. Our computational results therefore
suggest an essential role for metal cations such as K+ and
Na+ in this unprecedented stepwise hydroamination reaction.
We next explored the substrate scope of this K3PO4-
mediated hydroamination with regard to the g,d-unsaturated
ketoximes 1 for the synthesis of 5-membered cyclic nitrones 2
(Scheme 3). For the substituent R1, the reactions tolerated the
installation of not only electron-rich and electron-deficient
benzene rings (Scheme 3, 2b and 2c) but also sterically bulky
aryl groups (2d and 2e), as well as a 2-thienyl moiety (2 f).
Alkyl groups (2g and 2h) and a morpholino moiety (2i) could
also be introduced at R1, although the reaction of aldoxime 1j
(R1 = H) afforded nitrone 2j in only 24% yield along with the
corresponding nitrile in 47% yield through the dehydration of
aldoxime 1j. Investigation into the effect of the substituents
attached to the alkene (R2 and R3) revealed that the
By contrast, the energy barrier for the reaction of anionic
oxime 1a’ (with K+) is much lower (15.8 kcalmolÀ1; Fig-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1959 –1962