1
409
However, we envisaged that the catalysis of Sn(OMe)4 would be
much better than that of R2Sn(OMe)2, comparing the plausible
reaction pathways using Sn(OMe)4 and R2Sn(OMe)2. In the
100
80
1
5
Sn(OMe) system, the intermediate species corresponding to 1 in
4
the R2Sn(OMe)2 system would be 2, in which each methoxy-
carbonate (OCOOMe) group on one tin atom can coordinate
to the other tin to stabilize the six-coordinated tin complex with
a local minimum energy (Scheme 2).15 Therefore, the CO2
insertion into an SnOMe bond of a dimer of Sn(OMe)4 would
be promoted by the formation of the stabilized 2. It is also
expected that the rate-limiting step with a lower activation
energy than that of R2Sn(OMe)2 will follow the formation of 2.
It should be noted that 2 has more exchangeable spectator
ligands, which only exert electronic and steric influences on the
tin atoms, than 1, which has only one exchangeable spectator
ligand other than inert alkyl groups. Therefore, intermediate 2
can be tuned with some additives that lead to a better catalysis
for the DMC synthesis than intermediate 1.
60
4
2
0
0
0
0
5
10
15
time/h
20
25
30
Figure 1. The time course of DMC formation from CO2 and
MeOH using Sn(Ot-Bu) ( ) and n-Bu Sn(OMe) ( ).
4
2
2
Table 1. DMC synthesis from CO and MeOH using Sn(Ot-
2
a
Sn(OMe)4 was first synthesized and used for the DMC
synthesis without additives or dehydrating agents. However, no
reproducible results were obtained probably because Sn(OMe)4
Bu) or Bu Sn(OMe) with various additives
4
2
2
Run Catalyst
Additive
DMC yield to Sn/%
1
6
1
2
3
4
5
6
7
8
9
Sn(Ot-Bu)4
None
Pyridine
PhOH
PhOH + Pyridine
Ph NH OTf
C6F5OH
C6F5OH + Pyridine
24
32
42
45
3.4
39
60
1.1
20
was apt to be hydrolyzed (see Supporting Information). Next,
we turned our attention to Sn(Ot-Bu) , because Sn(Ot-Bu) is
4
4
monomeric unlike other tin primary alkoxides and easily
purified by distillation. In addition, Sn(Ot-Bu)4 is in situ
convertible into the corresponding tin methoxides in contact
with MeOH.
2
2
Sn(Ot-Bu)4 was prepared from SnCl4 and t-BuOH in the
1
7
presence of Et2NH, purified by distillation (bp: 85 °C/40 Pa),
and obtained as a white waxy solid (mp: ca. 40 °C). The catalyst
Bu2Sn(OMe)2 None
Ph2NH2OTf
(
(
0.25 mmol), additives (50 mol % of catalyst), and dry MeOH
2 mL, 49 mmol) were added to a 10-mL autoclave. After
aReaction time is 3 h.
introducing 4 MPa of CO , the autoclave was heated at 150 °C,
2
and the interior pressure was adjusted to a given value. After the
reaction mixture in the autoclave was stirred for a given time, the
The effects of acid or base additives were then investigated
by comparing the yields at 3 h (Table 1). While Ph2NH2OTf was
1
9
autoclave was cooled to ¹78 °C, and the CO gas was gently
reported to be the most effective for Bu Sn=O, it gave a poor
2
2
released. The reaction products were transferred in a vacuum in
order to separate off the high-boiling components and collected.
The DMC yield was determined by GC analysis.
result (Run 5 vs. 9), implying that the reaction mechanism of
Sn(OR)4 seems to be different from that of R2Sn(OMe)2. The
addition of PhOH or pyridine increased the reaction rates (Runs
2 and 3). The addition of both PhOH and pyridine gave a slight
enhancement (Run 4). Using C6F5OH instead of PhOH had
almost the same effect (Run 6), but the combined use of
C F OH and pyridine gave the highest yield of 60% (Run 7).
The reaction using Sn(Ot-Bu) proceeded much faster than
4
that using n-Bu2Sn(OMe)2 as shown in Figure 1 and Table 1,
Runs 1 and 8. Based on a GC-MS analysis of the reaction
mixture, it was confirmed that almost all the t-BuO groups
6
5
(
>3.5 equiv/Sn) were displaced and detected as t-BuOH. No
Other additives, such as N,N-dimethylaminopyridine, bipyri-
dine, stearic acid, and catechol, were ineffective (not shown).
t-BuO(CO)OMe or t-BuO(CO)Ot-Bu were observed, while
RO(CO)OMe and RO(CO)OR were reported to be produced
when M(OR)4 (M = Ti or Sn, R = Et, n-Bu) were applied.
The dependence of the catalytic activity of Sn(Ot-Bu) on
4
6
,7a
the CO2 pressure is shown in Figure 2. Although it was reported
that the catalytic activity of R2Sn(OMe)2 monotonously im-
This implies that the t-BuO group was not able to occupy the
bridging position between the two tin atoms and that the
insertion of CO2 into an SnOt-Bu bond would be very slow or
would not take place.
The time course of the DMC formation in the absence of
dehydrating agents revealed that the reaction had an induction
period of ca. 1 h (Figure 1).18 It is reasonable that tin butoxide
takes some time to be converted into the corresponding
methoxides. The reaction rate then abruptly increased but
slowed down when the yield reached 90% at 24 h. Therefore,
the catalytic activities under the following different conditions
were estimated by comparing the yields at 3 h as the “pseudo
reaction rate”.
12
proved as the CO pressure increased up to 20 MPa, that of
2
Sn(Ot-Bu)4 showed its maximum at 7.5 MPa. This is possibly
because the insertion of CO2 into the SnOMe bond more
readily takes place in the Sn(OMe)4 system than that in
R Sn(OMe) due to the formation of the stabilized intermediate
2
2
2 by the double coordination of the carbonate groups around the
tin atoms.
The addition of a dehydrating agent of 2,2-dimethoxypro-
pane (Me2C(OMe)2; DMP) was also found effective in the case
of Sn(Ot-Bu)4. The yield at 24 h in the presence of C6F5OH and
pyridine as well as DMP at 20 MPa was 236%, which was twice
as high as that without the dehydrating agent. This indicates the
Chem. Lett. 2011, 40, 14081410
© 2011 The Chemical Society of Japan