.
Angewandte
Communications
for the hydrogenation of 4-methyl-1,3-dioxolan-2-one in the
presence of a catalytic amount of RuII complex 1a,[14h]
including the solvent, substrate concentration, tBuOK con-
centration, hydrogen pressure, and the reaction temperature,
clearly established the feasibility of this catalysis (for details,
see the Supporting Information, Tables S1–S5). Subsequently,
we examined the catalytic activity of a variety of RuII and IrIII
complexes (1 and 2) in the hydrogenation of ethylene
carbonate. The reactions were conducted in THF at 1408C
under 50 atm of H2 in the presence of the catalytic RuII or IrIII
complex and the base tBuOK (substrate/catalyst/tBuOK =
1000–100000:1:1), and the results are shown in Table 1.
the reaction time to 1 h, which afforded both methanol and
EG in > 99% yield (entry 7). At a catalyst (1a) loading of
0.02 mol%, similar results were obtained in 10 h (entry 8).
When the catalyst loading was decreased to 0.01 mol%, the
reaction could be completed in 48 h to give methanol and EG
in > 99% yield (entry 9). When the catalyst loading was
further decreased to 0.001 mol%, the reaction still proceeded
smoothly at 60 atm of H2 with a 89% conversion of ethylene
carbonate in 72 h, to afford methanol and EG in 84% and
87% yields, respectively (with a TON of 87000 and a TOF of
1200 hÀ1; entry 10). This result demonstrates the high effi-
ciency and good stability of the catalyst.
The application of the optimized catalyst 1a was then
extended to the hydrogenation of a variety of cyclic carbo-
nates (Table 2). Heating a solution of propylene carbonate
(2.92 g, 28.6 mmol) and H2 (50 atm) with a catalytic amount of
complex 1a (3.4 mg, 0.0057 mmol, 0.02 mol%) at 1408C in
THF (20 mL) in a 125 mL autoclave for 10 h selectively
produced methanol and propylene glycol in quantitative
yields (entry 1). Under similar conditions, other cyclic carbo-
nates, including 5-membered 1,3-dioxolan-2-ones with one,
two, or even four substitutents, as well as a six-membered
carbonate, can also be efficiently and selectively hydro-
genated into methanol and their corresponding diols. The
steric hindrance of the substituents at the backbone of cyclic
carbonates has a significant impact on the activity of the
reaction, as longer reaction time and/or higher catalyst
loading are required to reach a complete conversion of
sterically more demanding substrates (entries 2–10 vs. 1). The
cyclic carbonate with a six-membered ring, 1,3-dioxan-2-one,
was also selectively hydrogenated in the presence of
0.05 mol% of 1a to afford methanol and 1,3-propanediol in
quantitative yield (entry 11).
In another unprecedented reaction, complex 1a was
found to catalyze the hydrogenation of polycarbonate into
methanol and the corresponding diol, thus realizing a hydro-
genative degradation of polycarbonate. As shown in
Scheme 2, a sample of poly(propylene carbonate) (PPC;
2.69 g) with a weight-average molecular weight (Mw) of
100698 (Mw/Mn = 1.77, > 99% carbonate linkages), which
had been prepared by an alternative copolymerization of
propylene oxide and CO2,[15] was readily depolymerized by
hydrogenation in the presence of 1a (15.8 mg, 0.1 mol%) to
afford methanol and 1,2-propyleneglycol in high yields (both
99%). Therefore, the present catalytic system may provide
a potential approach to the use of recovered waste poly-
carbonate as a resource.
Table 1: Hydrogenation of ethylene carbonate in the presence of RuII and
IrIII catalysts.[a]
Entry Cat.
(mol%)
t [h] Conv.
[%][b]
Yield of EG
[%][b]
Yield of MeOH
[%][b]
1
2
3
4
5
6
7
8
9
1a (0.1)
1b (0.1)
1c (0.1)
1d (0.1)
1e (0.1)
2 (0.1)
0.5 >99
>99
74
15
76
24
>99
45 (18)[c]
13
0.5
0.5
0.5
0.5
1
74
16
76
24
11
48 (20)[c]
22
10
1
>99
98
>99
84
1a (0.05)
1a (0.02)
1a (0.01)
1
>99
>99
>99
>99
87
10 >99
48 >99
10[d] 1a (0.001)
72
89
[a] Standard reaction conditions: substrate (28.6 mmol), 1 equiv of
KOtBu to the RuII or IrIII complex, THF (20 mL), in a 125 mL autoclave.
[b] Determined by GC analysis using p-xylene (50 mL) as the internal
standard. [c] The data in parentheses are the GC yields of methyl formate
(HCOOMe). [d] 60 atm of H2 was applied.
It was found that at a catalyst loading of 0.1 mol%, RuII
complexes 1a–e are active for the hydrogenation, simulta-
neously affording methanol and EG in various yields
(entries 1–5), whereas Ir complex 2 is much less effective
(entry 6). In the reaction with 1b and 1d some amount of
methyl formate is also formed along with the expected
alcohol products (entries 2 and 4). This is probably due to
their lower activity for the hydrogenation of the 2-hydrox-
yethylformate intermediate, which may undergo a fast ester
exchange with methanol to give methyl formate. Among the
Ru catalyst series (1a–e), complex 1a, which has a readily
available and air-stable HN(CH2CH2PPh2)2 ligand, turned out
to be optimal in terms of catalytic performance (entries 2–5
vs. 1).
Following the optimized reaction conditions mentioned
above, a deuterium-labeling study was carried out by using D2
instead of H2 to get insight into the mechanistic aspects of the
catalysis. As shown in Scheme 3a, in the 1a-catalyzed
deuteration of ethylene carbonate, methanol is produced
with only 87% deuterium content at the methyl group, as
1
determined by H NMR analysis of its benzoate ester, which
indicates a partial transfer of hydrogen atoms from the
ethylene backbone of the carbonate to the carbon of
methanol. This is also confirmed by a significant deuterium
substitution at the carbon atoms of EG (49% deuteration),
which implicates the catalytic activity of complex 1a in both
With the optimized catalyst in hand, we subsequently
turned our attention to improving the catalyst turnover
numbers (entries 7–10). Complete conversion of the substrate
was realized in the presence of 0.05 mol% of 1a by extending
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 13041 –13045