yield of 73% after chromatographic purification. Tertiary
allylamines bearing a variety of substituents were smoothly
catalytically deallylated by Grubbs' carbene to give the
corresponding N-deprotected amines 2 (Table 1).10 Aromatic
as well as aliphatic amines were amenable to this novel
deallylation reaction. The only exception was 1-allyl-2-
phenyl indole, which gave recovered starting material
together with a complex mixture of products in which the
deallylated product was a minor component; this could be
due to steric hindrance. Attempts to effect the reaction at
temperatures lower than 50 °C (refluxing dichloromethane)
notably increased the reaction time. The deallylation reactions
of compounds (+)-1b (entry 2) and (+)-1c (entry 3) with
the ruthenium catalyst deserve special mention, because we
have previously reported the ring-closing metathesis of
related â-lactam diene substrates11 and 1a-b are susceptible
to both reaction pathways. Of interest is the ability of Grubbs’
carbene for the selective deprotection of allylamines in the
presence of allyl ethers (entry 3),12 competing favorably with
the π-allyl palladium deallylation methodology.
Scheme 3
intermediate 5 or 7 in the presence of a catalytic amount of
Grubbs’ carbene. This mechanistically informative result was
provided by monitoring the reactions between 1d and 1h
with Cl2(Cy3P)2RudCHPh by 1H NMR spectroscopy. Indeed,
we observed disappearance of the terminal vinyl group and
a comparable rate of appearance of a methyl group, pointing
to Scheme 3 as the correct mechanism. Intermediate metalla-
cyclobutanes 6 evolve through a retro metathesis-like reaction
involving the cleavage of bonds other than those accounting
for the cycloreversion in the metathesis reaction to give
enamines 7 and the metal-carbene complex 3. Fragment 3
generated in this way could react with a new molecule of
allylamine 1, being involved again in the catalytic cycle. Not
unexpectedly, enamines 7d and 7h were characterized by
It may be reasonable to postulate that a nitrogen-assisted
ruthenium-catalyzed isomerization to a more stable olefin
took place,13 followed by hydrolysis under chromatographic
work up of the enamine intermediate to the NH-amine. From
a mechanistic point of view, our results could be explained
as illustrated in Scheme 2 or Scheme 3. Intermediates 3-7
can account for this catalytic cleavage reaction.
1
the H NMR spectra of the crude reaction mixtures as a
mixture of two isomers.
Scheme 2
Grubbs’ catalyst is known to be moderately termally
unstable, and the thermolytic half-live of Grubbs' carbene
was reported to be 8 days at 55 °C.14 It should be mentioned
that we used an experimental trick that proved to be very
efficient in our previous report on RCM,11 namely, Grubbs’
carbene was added in small portions every 20 min (5 mol
% is the overall amount of all the portions). In this way, the
catalytic species is continuously being renewed by fresh
Grubbs’ carbene.
In conclusion, Grubbs’ carbene efficiently catalyzes the
deprotection of tertiary allylic amines. In addition to the
novelty of the method, it is general, selective, and syntheti-
cally simple, offering the first ruthenium-catalyzed deallyl-
ation of allylamines. We believe that this C-N bond cleavage
implicates a ruthenium-catalyzed isomerization to a more
stable olefin, followed by hydrolysis of the resulting enamine.
It may be involved an unprecedented mode of ring opening
of the metathesis intermediate metalla-cyclobutane.
To probe the correct mechanism (Scheme 3) we must show
that an allylamine does give the corresponding enamine
(4) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996,
118, 100.
(5) For the unexpected reactivity of Grubbs’ catalyst for Kharasch
addition, see: Tallarico, J. A.; Malnick, L. A.; Snapper, M. L. J. Org. Chem.
1999, 64, 344.
(6) For an example of a C-N bond cleavage in a Mo(IV) bispyridine
complex, see: Cameron, T. M.; Abboud, K. A.; Boncella, J. M. Chem.
Commun. 2001, 1224
(7) (a) Alcaide, B.; Almendros, P. Chem. Soc. ReV. 2001, 30, 226. (b)
Alcaide, B.; Almendros, P. Org. Prep. Proced. Int. 2001, 33, 315.
(8) Alcaide, B.; Almendros, P.; Aragoncillo, C.; Rodr´ıguez-Ranera, C.
Unpublished observations.
(11) Alcaide, B.; Almendros, P.; Alonso, J. M.; Aly, M. F.; Redondo,
M. C. Synlett 2001, 773.
(12) Conjugation of the new double bond with the lone pair of the
nitrogen atom is believed to promoted the enamine intermediate formation
in allylamines. This ability is excluded in allyl ethers.
(9) We believed that the higher stability of enamides compared with
enamines favors the double bond isomerization, preventing from N-allyl
cleavage.
(10) For the utility of related piperidinyl â-lactams to (+)-2a-c in the
preparation of enantiopure indolizidines, see: (a) Alcaide, B.; Almendros,
P.; Alonso, J. M.; Aly, M. F., Torres, M. R. Synlett 2001, 1531. For the
preparation of 4-oxopipecolic acid from related systems to (+)-1f and (-)-
1g, see: (b) Badorrey, R.; Cativiela, C.; D´ıaz de Villegas, M. D.; Ga´lvez,
J. A. Tetrahedron 1999, 55, 7601.
(13) Unexpected ruthenium-catalyzed isomerizations to the more stable
internal olefin have been recently noted. See: (a) Kinderman, S. S.; van
Maarseveen, J.-H.; Schoemaker, H. E.; Hiemstra, H.; Rutjes, F. P. J. T.
Org. Lett. 2001, 3, 2045. (b) Hoye, T. R.; Zhao, H. Org. Lett. 1999, 1, 169.
(c) Fu¨rstner, A.; Thiel, O. R.; Ackermann, L.; Schanz, H.-J.; Nolan, S. P.
J. Org. Chem. 2000, 65, 2204. (d) Miller, S. J.; Blackwell, H. E.; Grubbs,
R. H. J. Am. Chem. Soc. 1996, 118, 9606. (e) Joe, D.; Overman, L. E.
Tetrahedron Lett. 1997, 38, 8635.
(14) Ulman, M.; Grubbs, R. H. J. Org. Chem. 1999, 64, 7202.
Org. Lett., Vol. 3, No. 23, 2001
3783