Full Paper
to exhibit high catalytic performance and the catalytic mecha-
nism was elucidated in detail by means of DFT calculations. In
addition, compounds 2d, 1 f, and 2 f, each with 8 or 12 nucle-
ophiles to promote the nucleophilic attack, were also synthe-
sized. Careful investigation of the catalytic activities by control-
ling temperature, reaction time, and catalyst loading revealed
that zinc(II) complex 2d was highly active and robust, as well
as exhibiting a very high turnover number (TON) and turnover
frequency (TOF) under solvent-free conditions. We investigated
the substrate scope of 2d for not only terminal epoxides, but
also internal epoxides. Although internal epoxides were much
less reactive than terminal epoxides, compound 2d showed
catalytic activities for both of them. Herein, we describe the
synthesis, catalytic activities, and substrate scope of the bifunc-
tional porphyrins (Figure 1).
Figure 2. Schematic representation of the ring opening of an epoxide by
a) meta-, b) para-, and c) ortho-substituted catalysts.
whereas substitution at the ortho positions hinders coordina-
tion of the epoxide, both of which decrease the catalytic activi-
ties (Figure 2). Catalyst 2e, with eight nucleophilic arms at the
ortho positions, showed even lower activity than that of 2c;
this result strongly supports steric hindrance by the ortho sub-
stituents (Table 1, entry 8). To search for a more active catalyst,
we examined zinc(II) porphyrin 2d with eight nucleophiles at
the meta positions. Complex 2d showed even higher catalytic
activity than that of 2a and comparable activity to the most
active catalyst, 1d, under the same conditions (Table 1, en-
tries 3, 6, and 7). For comparison, a two-component catalytic
system composed of ZnTPP and TBAB was evaluated under
the same conditions; this system showed very poor catalytic
activity (Table 1, entry 11). Clearly, the high catalytic activity of
Results and Discussion
We initially attempted to synthesize magnesium(II) porphyrins,
such as para-substituted porphyrin 1b and ortho-substituted
porphyrin 1c. Although 1b was prepared by a magnesium(II)
metalation method similar to that applied to 1a, compound
1
c could not be obtained by the same method or by
[
8]
a method reported in the literature. Instead, we synthesized
zinc(II) porphyrins 2b–e. In addition, we prepared 1 f and 2 f,
with 12 nucleophiles at both meta and para positions (for syn-
thetic details, see the Supporting Information).
The catalytic activities of these bifunctional catalysts were in-
vestigated. Each catalyst (0.002 mol%) was used in the cou-
2
+
2d results from the cooperative action of the Lewis acid (Zn
)
À
[6]
pling reaction of epoxide 3a with CO (1.7 MPa) at 1208C for
and the nucleophile (Br ). The importance of the position of
the quaternary ammonium bromide groups in 2d was sup-
ported by DFT calculations, as described below. We attempted
to further optimize the catalyst, but the activity of 1 f was
almost the same as that of 1d, and the activity of 2 f was
lower than that of 2d, despite the higher number of nucleo-
philes (Table 1, entries 9 and 10). It is likely that the Lewis acidi-
ty of the metal center of 2 f is reduced by the electron-donat-
ing alkoxy groups at the para positions, which leads to the
drop of the catalytic activity of 2 f.
2
3
h in a stainless-steel autoclave. The results are summarized in
Table 1. The catalytic activities increased in the order 2c<
b<2a (ortho<para<meta; Table 1, entries 3–5). Likewise,
compound 1b was less active than 1a (Table 1, entries 1 and
). These results suggest that substitution at the para positions
2
2
directs the nucleophiles away from the coordinating epoxide,
Table 1. Screening of the catalysts for the synthesis of cyclic carbonate
[
a]
4
2
a from epoxide 3a and CO .
To estimate the degree of steric hindrance around the coor-
dination space of the catalysts, we determined the binding
[
b]
Entry
Catalyst
Yield [%]
TON
[
9]
constants (Ka). Zinc(II) porphyrins are suitable for this purpose
1
2
3
4
5
6
7
8
9
1a
1b
2a
2b
2c
1d
2d
2e
1 f
68
54
65
30
19
79
77
11
81
61
5
34000
27000
32500
15000
9500
39500
38500
5500
because they form a 1:1 complex with a ligand, giving a reliable
[
10]
Ka value. Complex formation between zinc(II) porphyrin and
epoxide 3a was monitored by UV/Vis titration in CH Cl2 at
2
2
08C, and the K values were calculated by the nonlinear least-
a
[
10]
squares method (see the Supporting Information). The re-
sults are shown in Table 2. The K values increase in the order
a
40500
30500
2500
2
c<2a<2b (ortho<meta<para; Table 2, entries 1–3), which
1
11
0
2 f
[c]
strongly suggests steric hindrance by the ortho substituents
ZnTPP+TBAB
and even by the meta substituents. The K value for 2b is
a
[
(
a] Reaction conditions: 3a (10 mmol), catalyst (0.002 mol%), CO
1.7 MPa), 1208C, 3 h, in a 30 mL autoclave. [b] Yields were determined by
2
smaller than that for ZnTPP, which suggests that the alkoxy
groups at the para positions in 2b decrease the Lewis acidity
of the metal center to some degree (Table 2, entries 2 and 7).
Notably, catalyst 2a, which showed higher catalytic activity
than that of 2b, has a lower coordination capacity than that of
1
H NMR spectroscopy by using 2-methoxynaphthalene as an internal
standard. [c] (5,10,15,20-Tetraphenylporphyrinato)zinc(II) (ZnTPP;
.002 mol%) and tetra-n-butylammonium bromide (TBAB; 0.016 mol%)
0
were used for comparison.
Chem. Eur. J. 2016, 22, 6556 – 6563
6558
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim