NMe2
SiMe2Ph
Li
NMe2
SiMe2Ph
2
0. These reactions illustrate an umpolung of reactivity in the
i
amide 14.
The carbene intermediate 8 could be formed by Brook
+
DNBO
13 40-55%
rearrangement, followed by or concerted with the elimination of
silane oxide (Scheme 1). This pathway is, as far as we are aware,
a new one for reactions taking place within the Brook
rearrangement manifold, and is a new route to carbene or
carbene-like intermediates. The nearest analogy is the formation
of an oxygen-stabilised carbene when the acetals of acylsilanes
1
2a
NMe2
SiPh2Me ii
Li
9
NMe2
SiPh2Me
D
[2H]-11 85% (76% D)
12b
12 Our reaction takes place, presumably,
are heated to 190 °C.
because of the extra electronic push (5, arrows) from the Me
lone pair. It could equally be derived by cheletropic extrusion of
2
PhMe SiO directly from the intermediate 4. It is not clear what
O
2
N
O2N
DNB =
2
the structure of the carbene is in detail—it could be the carbene
itself 8, as we have drawn it here for simplicity, or it could be an
equivalent species such as an a-lithio iminium ion. A carbene
was also invoked by Ogawa and Sonoda in their work using
samarium iodide induced coupling of amides.13 Whatever its
nature, our carbene was not trapped by a silicon hydride—only
the enediamine 2 was formed, and not the silyl amine 9, when
the tetrahedral intermediate 3 was warmed to 220 °C in the
O2N
2 2 6 3 2
Scheme 3 Reagents and conditions: i, 3,5-(O N) C H COCl; ii, D O
and 278 °C, at which temperature it survived long enough to be
2
quenched with D
The PhMe Si intermediate 12a, having been formed at higher
temperature, somewhere between 278 and 220 °C, evidently
found a proton before it could be quenched with D O. The
2
O to give the deuterated a-silyl amine [ H]11.
2
presence of PhMe SiH, nor have we found at any stage products
2
2
that might have been derived by insertion of the carbene into the
source of that proton appears to be, at least in part, THF, which
is known to react with strong bases like BuLi, losing the proton
on C-2 and undergoing a retro-cycloaddition to give the enolate
neighbouring C–H bond, nor into a well-placed CNC bond, as
described in the following paper. The enediamine could be
produced from the carbene or carbenoid by dimerisation, or,
more likely in view of the probable low concentration of such a
species, by attack upon it by the C-nucleophilic intermediate 4
or 5 of the Brook rearrangement, followed by b-elimination of
a second silyloxy anion.
of acetaldehyde.9 We detected a low level of deuterium
2
incorporation when the reaction was carried out in [ H
8
]THF,
and, expecting to trap the enolate of acetaldehyde, we added
,5-dinitrobenzoyl chloride, and obtained instead the 3,5-dini-
3
trobenzoate 13 in yields of 40–55%, with no trace of the
expected product. As far as we are aware, the apparently simple
E2 elimination from THF losing the proton from C-3 has not
been seen before in solution chemistry, although it is thoroughly
established with hydroxide ion and amide ion as bases in the gas
phase,10 and the formation of but-3-enol from the reaction
The following paper describes some other remarkable
reactions that can be ascribed to the presence of intermediate
lithium reagents like 12 and 15. They add further support to this
being the correct mechanism, at least in outline.
We thank the EPSRC and Lilly Industries for a CASE
studentship for S. R. M.
11
between sodium and 3-chlorotetrahydrofuran is also known. It
is presumably unfavourable because the elimination is a retro-
Notes and References
5-endo-trig reaction.
We also found that the presence of a phenyl group in the N,N-
†
E-mail: if10000@cam.ac.uk
dimethylbenzamide 14 stabilised the corresponding inter-
mediate 15, which survived even at 220 °C, and gave a
1 I. Fleming, U. Ghosh, S. R. Mack and B. P. Clark, Chem. Commun.,
1998, 711.
2 I. Fleming and U. Ghosh, J. Chem. Soc., Perkin Trans. 1, 1994, 257.
2
deuterated a-silyl amine 16 on quenching with D O (Scheme
4). The intermediate 15 also reacted with alkyl halides giving
3
D. A. Bravozhitivitovsk, S. D. Pigarev, I. D. Kalikhman, O. A.
the amines 17a and 17b, and with isobutyraldehyde giving,
initially, an alkoxide 18 that undergoes a Peterson elimination
giving the enamine 19, which is easily hydrolysed to the ketone
Vyazankina and N. S. Vyazankin, J. Organomet. Chem., 1983, 114,
5
1.
4
5
A. G. Brook, Acc. Chem. Res., 1974, 7, 77.
A. Wright and R. West, J. Am. Chem. Soc., 1974, 96, 3214, 3222 and
NMe2
SiMe2Ph
NMe2
SiMe2Ph
3
227.
6 R. J. Linderman and A. Ghannam, J. Am. Chem. Soc., 1990, 112, 2392;
I. Fleming, Chemtracts: Org. Chem., 1996, 9, 121.
D
E
7
Similarly, carrying out the original reaction with the cyclohexane-
carboxamide and adding the isobutyramide before warming gave only
the homo-coupled product (5 in the preceding paper).
1
6
68%
17a E = Me 54%
b E = allyl 63%
ii
iii or iv
(
90% D)
8
A. G. Brook, G. E. LeGrow and D. M. MacRae, Can. J. Chem., 1967,
4
5, 239.
NMe2
O
NMe2
SiMe2Ph
Li
9 R. B. Bates, L. M. Kroposki and D. E. Potter, J. Org. Chem., 1972, 37,
560.
10 C. H. DePuy and V. M. Bierbaum, J. Am. Chem. Soc., 1981, 103, 5034;
C. H. DePuy, E. C. Beedle and V. M. Bierbaum, J. Am. Chem. Soc.,
Me2N SiMe2Ph
O–
i
v
1
982, 104, 6483.
11 L. Crombie, J. Gold, S. H. Harper and B. J. Stokes, J. Chem. Soc., 1956,
36; M. Jurjew, Zh. Obshch. Khim., 1948, 18, 1807.
1
1
14
15
NMe2
18
1
O
2 A. G. Brook and P. J. Dillon, Can. J. Chem., 1969, 47, 4347.
3 A. Ogawa, N. Takami, M. Sekiguchi, I. Ryu, N. Kambe and N. Sonoda,
J. Am. Chem. Soc., 1992, 114, 8729; A. Ogawa, T. Nanke, N. Takami,
M. Sekiguchi, N. Kambe and N. Sonoda, Appl. Organomet. Chem.,
vi
1
995, 9, 461.
19
20 73%
Scheme 4 Reagents and conditions: i, PhMe
2
SiLi (2.4 equiv.) THF, 278 ?
Received in Liverpool, UK, 12th November 1997; revised manuscript
received, 21st January 1998; 8/00650D
i
2
20 °C, 1.5 h; ii, D O; iii, MeI; iv, allylBr; v, Pr CHO; vi, HCl, H O
2
2
714
Chem. Commun., 1998