Table 1: Palladium-catalyzed synthesis of 2a.[a]
conditions, palladium black is still generated, thus limiting the
overall product yield. The latter can be inhibited by the use of
an excess of phenyl iodide (entry 5). In addition to PtBu3,
several bulky phosphine ligands provide the correct balance
of the ability to mediate aryl iodide oxidative addition and
lability to favor the carbonylation via 5 (entries 6–8), whereas
smaller ligands inhibit catalysis (entries 9–11). In addition to
the use of the palladacycle 7, commercial [{Pd(allyl)Cl}2] is
also a viable catalyst precursor for this reaction, and avoids
the need to pre-synthesize the catalyst (entry 12).
Interestingly, the reactions that employ bulky phosphine
ligands are found to proceed at comparable rates (Table 1,
entries 5–8). Insight into this effect can be obtained by
monitoring the catalytic reaction of entry 12, in Table1 by
31P NMR spectroscopy. This data shows the immediate
Entry
Pd cat.
PR3
Yield
[%][b]
1[c]
2[c]
Pd(OAc)2
PPh3
–
–
7
6
12
3[c]
–
47
4[c]
5
7
7
PtBu3
PtBu3
46
83
63
76
63
–
–
–
75
6
7
P(o-tolyl)3
formation of the palladium benzoyl complex
6 (d =
7
7
tBu2P(2-biphenyl)
8
7
tBu2P(2,4,6-iPr3C6H2)
72.3 ppm), thus confirming the generation of this intermedi-
ate in the catalysis (Scheme 2; cycle A). However, this
complex rapidly disappears, and the phosphine is quantita-
tively converted into its protonated form after one catalytic
cycle (d = 54 ppm).[15,16] Thus, phosphine plays a critical role
in the initial stage of catalysis to presumably inhibit the
immediate generation of inactive palladium black, but once
the catalysis is underway it is a nonphosphine-bound palla-
dium that mediates this reaction. Consistent with this
observation, is the monitoring of the stoichiometric reaction
of the palladium benzoyl complex 6 and imine [Eq. (1)] that
shows 6 cleanly converting into this same protonated
phosphine and imidazoline 2a. No palladium intermediates
are observed in this stoichiometric reaction. This observation
is consistent with the coupling of the palladium benzoyl
complex with the imine as a slow step in imidazoline
formation, and once 3 (Scheme 2) is formed it rapidly
converts into product.
Overall, this reaction provides a straightforward approach
to assemble 2-aryl-substituted imidazolines from five separate
components.[17] As shown in Table 2, a number of N-alkyl-, N-
benzyl-, or N-PMP-substituted (PMP = para-methoxyphenyl)
imines are viable in this reaction, as are variously function-
alized C-aryl and even C-heteroaryl imines (e.g., 2l;
entry 12). More sterically encumbered N-substituted imines
react slowly in this reaction (entries 4 and 11), which is
consistent with the imine coupling with 6 as a slow step in
catalysis. Both electron-rich and electron-poor aryl iodides
are compatible with the reaction, although the latter leads to
lower yields (entry 5 versus entry 8).
9
7
7
7
PPh3
dppe
PCy3
10
11
12[d]
[{Pd(allyl)Cl}2]
PtBu3
[a] Reaction conditions: iodobenzene (102 mg, 0.50 mmol), imine
(21 mg, 0.10 mmol), Pd cat. (2.5 mmol), PR3 (7.5 mmol), and 5 atm CO in
0.8 mL CD3CN. [b] Yields determined by 1H NMR spectroscopy using
benzyl benzoate as an internal standard. [c] 10.2 mg, 0.050 mmol
iodobenzene. [d] 12.5 mmol PR3. Bn=benzyl, dppe=1,2-bis(diphenyl-
phosphino)ethane.
entry 1). In considering the mechanism shown in Scheme 2,
the presence of the strongly coordinating PPh3 and acetate
ligands likely inhibit the formation of the four-coordinate,
CO-bound intermediate 5 for the second carbonylation step.
Similarly, catalysis under the ligandless conditions that favor
the formation of intermediate 5 and carbonylation of N-acyl
iminium salts also yielded only trace amounts of product
(entry 2).[11a] In this case, the lack of stabilizing ligands results
in palladium black precipitation before the phenyl iodide can
undergo oxidative addition to the ligandless palladium(0).
Stoichiometric experiments provide some insight into how
these carbonylation reactions might be coupled. Bulky
phosphines such as PtBu3 have recently been shown to
stabilize both three-coordinate aryl and aroyl palladium
complexes.[14] Similarly, the reaction of [Pd(PtBu3)2] and PhI
with CO allows the generation of the complex 6 [Eq. (1)]. In
addition, subjecting 6 to reaction with imine and CO leads to
the formation of the heterocycle 2a in 72% yield (19 h, 558C).
These stoichiometric results suggest that catalysis should
be viable provided that the palladium acyl intermediate such
as 6 can be generated with labile ligands. This can be achieved
by employing the palladium complex 6 itself as the catalyst
(Table 1, entry 3). Alternatively, the palladacycle 7 is also a
viable catalyst with PtBu3 (entry 4). Under these reaction
As an illustration of the potential utility of this reaction,
triaryl-substituted imidazoline derivatives of the general form
8 [Eq. (2)] have recently been identified as potent cancer cell
sensitizers as well as potential therapeutics for rheumatoid
arthritis.[18] As an alternative to their typical synthesis by
cycloaddition to give preformed mꢁnchnones, these core
structures can be generated through this palladium-catalyzed
multicomponent coupling and subsequent hydrogenolysis to
Angew. Chem. Int. Ed. 2011, 50, 8948 –8951
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Bontemps, Sebastien
