Angewandte
Chemie
Table 1: Tandem N,N,C-trialkylation reaction from aniline 2a.[a]
[17]
À
phosphinesulfonate ligand with an Ir O bond of 2.140 ꢀ.
The catalyst B was successfully involved in the tandem
process under the reaction conditions used for A, thus
affording the trialkylated amine 7a as the major product
(entries 4–7). This result demonstrated that B was active for
the base-free N,N-dialkylation of 2a with 3a to give 6, and
that catalytic activity was maintained for the oxidant-free
direct dehydrogenation of 6, thus leading to a reactive
nucleophilic enamine. It is noteworthy that [{IrCl2Cp*}2]
was totally inactive toward C alkylation but led to 5 as a major
compound (entry 3). With B, increasing the concentration
resulted in lower conversion and higher relative amount of 6,
and the best result was obtained at 0.4m (entries 4 and 5). A
slight excess of 4a ensured better selectivity for 7a (entry 6).
Finally, the presence of 1 mol% of camphorsulfonic acid
(CSA) as a Brønsted acid was required to reach complete
conversion and optimized reaction conditions, which afforded
a 70% yield of compound 7a as determined by GC analysis
(entry 8). The tedious separation of the piperidines 7a and 6
by column chromatography using silica gel led to the isolation
of pure 7a in 55% yield.
The overall transformation corresponds to three dehy-
drogenation and three hydrogenation steps, thus six formal
catalytic hydrogen-transfer processes, and also two conden-
sations of the intermediate aldehyde with aniline and one
condensation reaction between an enamine and an aldehyde.
If we assume that the organic reactions proceed quantita-
tively, each individual catalytic step presents an excellent
yield (> 90%). These results show that only one organome-
tallic precatalyst is sufficient for the overall transformation.
Moreover, the reaction allows the first oxidant-free direct
endo dehydrogenation of the cyclic saturated 6, and does not
require more-reactive substrates (i.e. containing activated
benzylic positions as in N-benzyl amines or tetrahydroisoqui-
nolines), for dehydrogenation[14] and thus highlights the
improved activity of this new iridium precatalyst B.
Entry Catalyst
Conc. 2a/3a/4a[c] 5/6/7a Yield
[m][b]
7a [%][d]
1
[{Ru(p-cymene)Cl2}2]
0.6
0.6
0.6
0.9
0.4
0.6
0.4
0.4
1.1:1:1.2
1.1:1:1.2
1.1:1:1.2
1.1:1:1
1.1:1:1
1:1:1.2
1:0:0
7:83:10
82:13:5
3:45:52 42
4:33:63 50
1:22:77 50
3:27:70 58
2:10:88 70(55)
n.d.
5
1
2
A
3
[{Ir(Cp*)Cl2}2]
4
5
6
B
B
B
B
B
7
1.2:1.1:1
1:1:1.1
8[e]
[a] All reactions were carried in toluene. Step 1 was run for 16 h, step 2
for 19h, and step 3 for 2 h under inert an atmosphere using
a thermostated oil bath (1508C). [b] Molar concentration of the limiting
substrate; see column 4. [c] Ratio of the products. [d] Yield of 7a
determined by GC analysis using tetradecane as an internal standard.
Value in parentheses is the yield of the isolated product. [e] 1 mol% of
CSA was added in step 2. n.d.=not determined.
for 16 hours, and after cooling benzaldehyde (4a) was added
and the reaction was heated for an additional 19 hours at
1508C (Table 1). Based on our previous results obtained with
ruthenium catalysts for C3 functionalization of cyclic amines
with aldehydes, formic acid was added as a reducing agent at
the last stage (step 3; 1508C for 2 h) of the procedure to
ensure complete reduction of the unsaturated intermediates.
The expected formation of 3-benzyl-N-phenylpiperidine (7a)
was observed, but N-benzylaniline (5) and N-phenylpiper-
idine (6) were also detected. The {ruthenium(II)(p-cymene)}
dimer did not afford the C-alkylated piperidine 7a but the
presence of 5, arising from N-alkylation of 2a by 4a, was
observed as the main product (entry 1). Surprisingly, our
previously reported ruthenium(II) catalyst A, which exhib-
ited a high efficiency for N or C alkylation of N-benzylpiper-
idine, showed very low activity for the tandem protocol, thus
affording less than 50% conversion of the diol 3a and leading
mainly to the piperidine derivative 6 and only 5% of 7a
(entry 2). We then investigated the reactivity of the new
iridium(III) complex featuring the chelating phosphane
sulfonate. The complex B was obtained in good yield upon
treatment of the deprotonated diphenylphosphinobenzene
sulfonic acid 1 with the cyclopentadienyl dichloride iridium-
(III) dimer (Scheme 2). The solid-state structure of B
revealed a piano-stool geometry around the iridium center
and a six-membered ring arising from the chelation of the
To clarify the difference of reactivity between the
ruthenium catalyst A and the iridium catalyst B, their
efficiencies in the first step were evaluated. Thus, 2a was
reacted with 3a in the presence of 3 mol% of catalyst A at
1508C for 16 hours. A 50% conversion of 2a into the
monoalkylated amine 8 as the major product was observed
(Scheme 3). Addition of 4a, which also acts as hydrogen
acceptor, to this mixture enhanced only the cyclization rate
Scheme 3. Monoalkylation of 2a using the catalyst A versus dialkyla-
tion with the catalyst B.
Scheme 2. Preparation of the catalyst B.
Angew. Chem. Int. Ed. 2012, 51, 8876 –8880
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8877