Organometallics
Article
a
Table 1. Pd(OAc)2-Catalyzed Carbonylative Transformation of Allyl Alcohol to CA in Different Solvents
selectivity (%)
b
entry
solvent
conversion (%)
γ-butyrolactone
allyl ester
3-butenoic acid
crotonic acid
TON for CA
1
2
3
4
5
benzene
dioxane
DME
THF
CH2Cl2
>99
>99
>99
>99
70
<1
ND
ND
2
ND
ND
<1
8
ND
ND
23
30
30
20
99
77
70
60
37
24
17
16
12
5
43
a
b
Reaction conditions: 0.1 M allyl alcohol (1 mmol, 10 mL of solvent), 4 mol % Pd(OAc)2, 4 mol % dppb, 50 bar CO, 110 °C, 18 h. Determined
1
using H NMR using mesitylene as an internal standard; TON for CA = mol of CA formed per mol of Pd.
RESULTS AND DISCUSSION
Table 2. Pd(OAc)2-Catalyzed Carbonylative
Transformation of Allyl Alcohol to CA
■
a
It has been reported that a catalytically active Pd0 complex can
be spontaneously formed from a PdII precursor in the presence
of phosphine ligands.18,19 Therefore, the allyl alcohol carbon-
ylation was initially performed with the readily available, air-
stable Pd(OAc)2 and dppb in different solvents under 50 bar
CO at 110 °C for 18 h in a 100 mL stainless steel tube reactor
(Table 1, entry 1−5). The crude carbonylation products from
b
c
entry CO (bar) T (°C) time (h) S/C CA yield (%)
TON
1
2
3
4
5
5
5
5
5
5
110
80
140
110
110
18
18
18
48
48
25
25
25
100
250
93
48
90
92
63
23.3
12.0
22.5
92.0
157.5
1
each reaction were analyzed by H NMR spectroscopy with
a
Reaction conditions: 1 mmol of allyl alcohol, 10 mL of benzene,
1
b
1
mesitylene as an internal standard. The H NMR spectral
ligand = dppb, Pd/dppb = 1:1. CA yield, calculated using H NMR
spectroscopy. TON = mol of CA formed per mol of Pd.
c
results showed that a mixture of products including CA, 3-
butenoic acid, allyl esters of CA and 3-butenoic acid, and γ-
butyrolactone (GBL) were formed in different ratios in
different solvents studied (Scheme S1). Interestingly, a 99%
conversion of allyl alcohol and 99% selectivity to CA was
observed upon using the nonpolar solvent benzene (Table 1,
entry 1). When ethereal solvents such as 1,4-dioxane,
dimethoxyethane (DME), and tetrahydrofuran (THF) were
used, the selectivity toward CA was decreased to 77, 70, and
60%, respectively (Table 1, entries 2−4). 3-Butenoic acid was
formed as the byproduct in both 1,4-dioxane and DME;
however, in the case of THF, along with 3-butenoic acid
(30%), the ally ester of CA (8%) and a negligible amount of
GBL (2%) were also formed. The use of CH2Cl2, a
noncoordinating polar solvent, resulted in a reduced
conversion (70%) and very low selectivity for CA (37%)
that was observed along with the formation of GBL (43%) and
3-butenoic acid (20%) (Table 1, entry 5). These byproducts
are believed to be the intermediates during the carbonylative
transformation of allyl alcohol to CA (vide infra). The large
excess of the ethereal solvents might interfere with the allyl
alcohol coordination to the Pd-site and might reduce the
reaction rate, resulting in the incomplete conversion of the
observed intermediates to CA. Whereas, the coordination of
allyl alcohol is expected to be facilitated in benzene and a
complete conversion of allyl alcohol to CA is observed.
However, the reason for the formation of GBL as a major
product in the CH2Cl2 solvent is currently not clear. Thus, for
the complete carbonylative transformation of allyl alcohol to
CA, benzene was used as the solvent in subsequent
experiments.
transformation of allyl alcohol to CA. At a slightly lower
temperature (80 °C), the yield of CA was reduced to 48% with
a TON of 12.0 (Table 2, entry 2). It is noticeable that
homogeneous Pd complexes tend to form Pd-black at elevated
temperatures. Similarly, the formation of Pd-black has been
observed frequently in Pd(OAc)2/dppb-catalyzed carbon-
In line with this, the TON for CA was reduced to 22.5 at a
higher temperature of 140 °C, due to undesired side reactions
and Pd-black formation (Table 2, entry 3). In this study, the
optimum reaction temperature was 110 °C.
To improve the productivity of the Pd(OAc)2/dppb
catalytic system, the substrate to catalyst (S/C) ratio was
increased from 25 to 100 and 250. Interestingly, a yield of 92%
CA with a TON of 92.0 was obtained with an S/C ratio of 100
in 48 h (Table 2, entry 4). Additionally, with an S/C ratio of
250, only a 63% yield of CA was obtained after 48 h, possibly
because of the gradual decomposition of active Pd-species into
Pd-black (Table 2, entry 5). Here, the use of heterogeneous
phosphine-based ligands as supports for the Pd precursors may
help to avoid both the aggregation of leached Pd metal and the
formation of Pd-black in the carbonylation reaction.
To further understand the mechanism of the conversion of
1
allyl alcohol to CA, detailed H NMR analysis of the reaction
mixture obtained at different time intervals was performed
using the Pd(OAc)2/dppb catalytic system at 5 bar CO in
benzene. Structures of the intermediates and product are
shown in Scheme 2, which were determined by comparing
their 1H NMR spectra to those of synthesized standard
samples or literature data. Peak assignments were made using
the structures shown in the inset of Figure 1 and in Figure S2.
Product amounts were estimated by integrating the area of the
peaks assigned to the corresponding products. In particular,
the content of allyl 3-butenoate, 3-butenoic acid, allyl
crotonate, CA, and allyl alcohol/ether were calculated
according to the integrated areas of peaks b, c, d, e, and a/f,
CA yield was unaffected by varying the CO pressure in the
range 5−50 bar; reactions at a low CO pressure (5 bar)
resulted in similar conversion rates (93% yield, TON = 23.3)
to those at a higher CO pressure (50 bar), under otherwise
identical reaction conditions (Table 2, entry 1). It may
originate from the fast CO insertion in this transformation, and
low CO pressure is enough to effectuate the transformation.
The reaction temperature significantly influenced the
reactivity and outcome of the Pd-catalyzed carbonylative
B
Organometallics XXXX, XXX, XXX−XXX