Communication
Table 1. Selected optimization of the carbonylative cross-coupling be-
tween 1a and 2a.
Table 2. Selected optimization of the reaction conditions using in situ-
generated N-chloropiperidine.
[
a]
[a]
[
b]
Entry
x [equiv]
y [bar]
T [8C]
Yield [%]
[
b]
Entry
Pd (mol%)
Base (equiv)
Solvent
Yield [%]
[
c]
1
2
3
4
5
6
7
8
2
2
2
3
3
3
3
3
50
40
50
50
50
80
50
50
40
35
45
45
60
45
45
45
66 (46 )
33
[
[
[
[
c]
c]
c]
c]
1
2
3
4
5
6
7
8
9
1
[Pd(PPh
[Pd (dba)
Pd(OAc)
PdCl (3)
Pd/C (2)
Pd/C (2)
Pd/C (2)
Pd/C (2)
Pd/C (4)
Pd/C (4)
Pd/C (4)
Pd/C (2)
3
)
4
3
] (3)
] (3)
(3)
K
K
K
K
2
2
2
2
CO
CO
CO
CO
3
3
3
3
(2)
(2)
(2)
(2)
THF
THF
THF
THF
dioxane
toluene
DME
MTBE
MTBE
MTBE
MTBE
MTBE
24
21
13
29
33
27
21
48
66
32
41
[d]
2
67 (0 )
[e]
2
73 (70 )
71
2
NaHCO
NaHCO
NaHCO
NaHCO
NaHCO
Na
NaOAc (3)
NaHCO (3)
3
3
3
3
3
(2)
(2)
(3)
(3)
(3)
81
41
0
[
[
f]
g]
[
(
[
a] Reaction conditions: 1a (0.25 mmol), piperidine (0.5 mmol), NCS
0.525 mmol), 10 wt% Pd/C (5.32 mg, 5 mmol), tBuOMe (2 mL), 16 h.
b] GC yield determined by using hexadecane as the internal standard.
0
2 3
CO (3)
1
1
[
d]
[e]
[c] 1 mol% Pd/C was used. [d] Either Pd/C or NaHCO
3
was omitted.
12
3
69 (26 )
[
e] Yield of isolated product. [f] 0.8 equivalents of TCCA (trichloroisocya-
[
(
a] Conditions (unless otherwise stated): 1a (0.25 mmol), 2a [0.5 mmol
entries 1–8) or 0.75 mmol (entries 9–12)], 408C, CO (50 bar), 16 h. [b] GC
nuric acid) was used in place of NCS. [g] NBS (N-bromosuccinimide) was
used in place of NCS.
yields were determined by using hexadecane as the internal standard.
c] T=508C, 40 bar CO. [d] No ligand was added. [e] 1 mol% Pd/C was
[
used. DME=1,2-dimethoxyethane; MTBE=methyl tert-butyl ether.
base amount, reaction temperature and CO pressure led to
a much higher yield of 3a, if CO pressure is 80 bar (81%,
Table 2, entry 6). Much like the reactions with with preformed
N-chloropiperidine, other homogeneous palladium precursors,
such as [Pd(PPh ) ], Pd(dba) and Pd(OAc) , were again much
peridine 3a in 24% yield (Table 1, entry 1). Other palladium
precursors, such as [Pd (dba) ], Pd(OAc) , and PdCl were also
2
3
2
2
3 4
2
2
tested, yet the yields remained very low (Table 1, entries 2–4).
After screening different palladium precursors, bases, and sol-
vents (for details, see the Supporting Information, Table S1),
we found that by using 2 mol% Pd/C (10 wt%) as the catalyst
less effective for this reaction (see the Supporting Information,
Table S2). Other N-halogenation reagents, TCCA (trichloroiso-
cyanuric acid) and NBS (N-bromosuccinimide), were much less
reactive than NCS (Table 2, entry 7 and 8). Importantly, no for-
mation of 3a was detected when NCS was replaced with oxi-
dants such as BQ (benzoquinone), DDQ (2,3-dichloro-5,6-dicya-
no-1,4-benzoquinone), K S O , tBuOOtBu, or tBuOOH (see the
and NaHCO as the base in methyl tert-butyl ether, the yield of
3
3
a could be improved to 48% (Table 1, entries 5–8). Further in-
creases in the catalyst loading (4 mol%) and amounts of 2a
3 equiv) and base (3 equiv) led to a 66% yield of 3a (Table 1,
entry 9). Other bases, such as Na CO , NaOAc, Cs CO , K PO ,
2
2
4
(
Supporting Information, Table S3). These results can rule out
[10]
the occurrence of the usual oxidative carbonylation process.
2
3
2
3
3
4
and DABCO, were all much inferior for this reaction (Table S1).
During exploration of the ligand effects for this reaction, we
discovered that the yields of 3a was almost unaffected by dif-
ferent types of ligands (see the Supporting Information,
Table S1, entries 23–25). Accordingly, we carried out a control
reaction with 2 mol% Pd/C under ligand-free conditions; to
our surprise, 69% yield of 3a was obtained (Table 1, entry 12).
However, decreasing the catalyst loading to 1 mol% caused
a significant decrease in the yield (26%).
To gain further mechanistic insight into the reaction, several
control experiments were carried out. In all of the above opti-
mization reactions, and even in the absence of CO gas, we de-
tected no N-phenylpiperidine (4a, Scheme 3), which would
likely be produced from the reductive elimination of intermedi-
[4a]
ate B in Scheme 2.
During the optimization process, we also observed that car-
bamoyl chloride (5a) and urea were produced as byproducts
in some cases. Specifically, piperidine-1-carbonyl chloride 5a
was obtained in 87% GC yield when no base was added
(Scheme 4a). Nevertheless, carbamoyl chloride did not under-
go cross-coupling with phenylboronic acid (1a) under our con-
To aid the practicality of this reaction, we attempted to use
N-chloropiperidine that was generated in situ by chlorinating
reagents (Table 2). Under similar conditions to those for
Table 1, entry 12, we were glad to find that 66% yield of 3a
was obtained by using two equivalents of piperidine and 2.1
equivalents of N-chlorosuccinimide (NCS) at 408C (Table 2,
entry 1). Lesser amounts of piperidine, reduced catalyst load-
ings, or lower temperatures all diminished the yield of 3a to
various extents (Table 2, entries 1 and 2). Control experiments
confirmed that both the base and Pd/C were indispensable for
the carbonylation reaction, as no product was detected when
either of them was omitted (Table 2, entry 3). Variations in the
Scheme 3. Cross-coupling between 1a and 2a under CO-free conditions.
&
&
Chem. Eur. J. 2015, 21, 1 – 6
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!