Table 2 Catalysed exchange reactions of N,N-diethylallylamine 1 with piperidine 5a
Catalytic
system
Conv.
(%)
b
TOF/h21
Entry
Solvent
Additive (equiv.)
—
t/min
0
1
2
3
4
5
6
Ni /dppb
DMF
DMF
DMF
THF
DMF
DMF
100
100
100
45
100
92
360
150
40
1020
5
35
60
400
4
>> 600
60
0
Ni /dppb
4 4
[NBu ][ClO ] (5)
0
Ni /dppb
AcOH (5)
AcOH (5)
AcOH (50)
AcOH (50)
0
Ni /dppb
0
Ni /dppb
II
Pd /dppb
120
a
[
Metal]/[1]/[5] = 1:50:500; T = 50 °C. b Conversion of 1 into N-allylpiperidine 6 as indicated by quantitative GLC.
beneficial effect of added acetic acid as expected from the
6
5
4
3
2
1
00
00
00
00
00
00
0
b–h
literature.2 Namely, the formation of 6 was best conducted by
2
using Ni(dppb) as the catalyst, DMF as the solvent and by
adding 1 equiv., of AcOH with respect to 1 (Table 2, entry 5).
The most efficient palladium catalyst found [in situ combination
Pd(dppb)
Ni(dppb)
2
Pd(OAc) –dppb, 1:3] took at least 20 times longer for the
2
2
exchange reaction to go to completion than the latter nickel-
based system under the same conditions (entry 6).
In conclusion we have demonstrated that Ni(dppb)
2
easily
generated from Ni(COD) and 2 equiv. of the bisphosphine is
2
Ni0
poisoning
generally a much more active catalyst than comparable
palladium systems for substitution of allylamines with ‘soft’
nucleophiles. Two explanations can be proposed for this trend:
(
i) the higher ability of nickel catalysts to activate allylic
substrates having a poor leaving group, i.e. here the C–N bond,
and/or (ii) the lower propensity of Ni to be coordinated by
amines, thus resulting in larger amounts of active Ni–phosphine
species. The limitation of these nickel-based systems lies in the
sensitivity of nickel intermediates towards acetylacetone.
2a
2b
2c
Fig. 1 Average turnover frequencies of nickel and palladium catalysts for
the coupling of 1 with 2a–c (see Table 1 for reaction conditions)
system for this reaction. A similar trend, with even more marked
differences between nickel and palladium catalysts, was
observed for the alkylation of 1 with methyl acetoacetate 2b.
Footnotes
*
†
E-mail: mortreux@ensc-lille.fr
In a typical experiment (Table 1, entry 13), to Ni(COD) (36 mg, 0.13
2
2
The simple system using Ni(dppb) as the catalyst and DMF as
mmol) in a 50 ml glass reactor equipped with a Teflon cap was added a
degassed solution of dppb (111 mg, 0.26 mmol) in DMF (12.5 ml) under
nitrogen. After 15 min of magnetic stirring, 1 (0.74 g, 6.5 mmol), 2b (1.13
g, 9.75 mmol) and heptane (1.00 g, 10 mmol, GLC internal standard) were
added. The solution was stirred at 50 °C and the reaction was monitored by
quantitative GLC analysis of aliquot samples. Coupling reactions involving
a co-reagent and exchange reactions of 1 with piperidine were carried out in
a similar manner. In situ palladium catalysts were prepared by mixing
the solvent allowed the reaction carried out with 2 mol% to Ni
to be completed within 5 min at 80 °C and with an average
2
1
activity of 100 h at 50 °C (entries 12 and 13).† The in situ
combination Pd(OAc) –dppb proved once again to be the most
2
effective palladium catalyst (entries 14–17), but its performance
at 80 °C was still lower than that of the above nickel system at
5
0 °C. However, nickel catalysts proved inefficient for the
2
Pd(OAc) with a degassed solution (THF or DMF) of the phosphine ligand.
The resulting yellow solution was stirred for 15 min at room temperature
before use.
coupling of 1 with acetylacetone 2c (entry 18), because of a
rapid catalyst poisoning evidenced by the gradual shift in the
colour of the solution from orange–yellow [typical colour of
‡
Such a degradation of nickel intermediates could be rationalised as
II
Ni(dppb)
2
] to pale green [Ni species, possibly Ni(acac)
2
].
follows:
Although it must be pointed out that 2c is the most acidic
compound among active methylene compounds 2a–c, there is
no definitive evidence that this catalyst decay occurs via an
oxidative protolysis of nickel intermediates.‡ Anyway, the
catalyst poisoning was still observed in the presence of an
excess of 1 ([Ni]:[1]:[2c] = 1:75:50). No comparable
phenomenon was observed with palladium catalysts, such as the
Ni (dppb) + 2 acacH ? Ni(acac) + H + dppb
0
2
2
References
1
(a) H. Bricout, J.-F. Carpentier and A. Mortreux, Tetrahedron Lett., 1997,
38, 1053; (b) Y. I. M. Nilsson, P. G. Andersson and J.-E. B a¨ ckvall, J. Am.
Chem. Soc., 1993, 115, 6609.
2 Palladium-catalysed nucleophilic substitution of allylamines and related
derivatives: (a) K. E. Atkins, W. E. Walker and R. M. Manyik,
Tetrahedron Lett., 1970, 43, 3821; (b) B. M. Trost and E. Keinan, J. Org.
Chem., 1980, 45, 2741; (c) A. Hosomi, K. Hoashi, S. Kohra, Y.
Tominaga, K. Otaka and H. Sakurai, J. Chem. Soc., Chem. Commun.,
2
in situ combination Pd(OAc) –dppb, which afforded alkylation
products 3c/4c with performance similar to that observed for the
coupling of 1 with 2b (entry 19). Fig. 1 summarises the overall
performance of best Ni–dppb and Pd–dppb systems.
1
1
987, 570; (d) F. Garro-Helion, A. Merzouk and F. Guib e´ , J. Org. Chem.,
993, 58, 6109; (e) S. Murahashi, Y. Imada and K. Nishimura,
The results of the exchange reaction between 1 and piperidine
to give N-allylpiperidine 6 (eqn. 2) bring further evidence of
5
Tetrahedron, 1994, 50, 453; (f) M. Grellier, M. Pfeffer and G. Van Koten,
Tetrahedron Lett., 1994, 35, 2877; (g) S. Lemaire-Audoire, M. Savignac,
J.-P. Gen eˆ t and J.-M. Bernard, Tetrahedron Lett., 1995, 36, 1267; (h)
S. Lemaire-Audoire, M. Savignac, C. Dupuis and J.-P. Gen eˆ t, Bull. Soc.
Chim. Fr., 1995, 132, 1157.
[
Ni] or [Pd]
NEt2
+
+ HNEt2
(2)
HN
N
3
4
To the best of our knowledge, the cross-coupling of allylamines with
boronic acids (‘hard’ nucleophiles) is the sole example of the use of
nickel catalysts: B. M. Trost and M. D. Spagnol, J. Chem. Soc., Perkin
Trans. 1, 1995, 2083.
H. Bricout, J.-F. Carpentier and A. Mortreux, J. Chem. Soc., Chem.
Commun., 1995, 1863.
1
5
6
the general ability of nickel catalysts to promote nucleophilic
substitution of allylamines (Table 2). We found that the
aforementioned trends for the nickel-catalysed coupling of 1
with active methylene compounds 2a, b proved also to be true
for this exchange reaction. A notable exception arose from the
Received in Cambridge, UK, 30th April 1997; 7/02954C
1394
Chem. Commun., 1997