N. Aoyagi et al. / Tetrahedron Letters 54 (2013) 7031–7034
7033
Nucleophilicity in
Table 3
Effect of structure of catalyst for the synthesis of 2a under ambient conditionsa
ROH
protic conditions
O
O
I
> Br > Cl
O
PhO
PhO
1a- f•MeI
(5 mol%)
O
O
O
B
+ X HOR
O
O
+
CO2
(1 atm)
2-Propanol
25 °C, 24 h
Catalyst
Solvent
ROH
HOR
B+
O
B+•X
PhO
HOR
2a
3a
(B = MePh3P ) (excess)
X
O
Entry
Catalyst
Yieldb (%)
O
1
2
3
4
5
6
1aÁMeI
1bÁMeI
1cÁMeI
1dÁMeI
1eÁMeI
1fÁMeI
97
87
84
89
87
O
O
O
O
CO2
PhO
PhO
B+
X
Product yield (ROH = IPA)
I (97%) > Br (37%) > Cl (7%)
Leaving ability
Br > Cl
I
>
Not detected
a
Reaction conditions: 1 mmol of 2a, 0.05 mmol of 1a–fÁMeI, 0.2 mL of 2-propa-
Scheme 2. A plausible mechanism for catalyzed synthesis of cyclic carbonates from
epoxides and CO2.
nol, 25 °C under 1 atm of CO2 for 24 h.
b
Determined by 1H NMR.
of the ring-opening reaction. In fact, the product yields decreased
with increasing acceptor number (AN) of solvents, that is, IPA
(33.8) < MeOH (41.3) < HFIP (63.0), which are known to be propor-
tional to the solvation ability.13
respectively (entries 5 and 6). The use of phosphine ligands with-
out counter anion gave no cyclic carbonate (entries 7 and 8). The
catalytic activity is thus highly affected by the counter anion of
the catalysts.
A plausible mechanism for the catalytic synthesis of carbonate
is shown in Scheme 2. Initially, the protic solvent activates the
epoxide, and at the same time solvates the halide anion of the cat-
alyst through hydrogen bonds. Then, the ring-opening intermedi-
ate is formed by the nucleophilic attack of the halide anion on
the activated epoxy group, and subsequent nucleophilic attack on
CO2 leads to the alkylcarbonate anion. Finally, the ring-closure
through the elimination of the halide anion gives the cyclic carbon-
ate. In the reaction, the protic solvents play two opposite roles, that
is, activation of the epoxy ring via hydrogen bonds and diminishing
the reactivity of the halide anion through solvation. The latter is
the dominant factor for the catalytic activity, which can be roughly
indicated by the acceptor numbers. The counter anion of the cata-
lysts is another important factor of the catalytic activity, which was
in the order of iodide > bromide > chloride. This could be ac-
counted for by the nucleophilicity of halide anions in protic sol-
vents at the ring-opening step as well as the leaving ability at
the ring-closing step, both in the order of IÀ > BrÀ > ClÀ.
Furthermore, we examined various phosphonium iodides
(1a–fÁMeI) in the carbonate formation from 2a and CO2 in 2-propa-
nol at 1 atm and ambient temperature (Table 3). Catalytic activities
of the phosphonium salts bearing substituents more electron-rich
than 1aÁMeI (1b–dÁMeI) were slightly less than that of 1aÁMeI (en-
tries 1–4). The phosphonium iodide with dimethylamino groups
resulted in more or less the same yield as 1b–dÁMeI (entry 5). In
contrast, the phosphonium iodide with phenoxy substituents
(1fÁMeI) gave no cyclic carbonate (entry 6). Thus, the methyltri-
phenylphosphonium iodide (1aÁMeI) was the best catalyst among
the phosphonium salts investigated here (1a–fÁRX).
Next, the carbonate formation of 2a with CO2 was carried out in
several secondary alcohols and other solvents using 1aÁMeI
(Table 4). As in the case of 1aÁHI, THF and methanol gave 2a in
low yields (entries 1 and 2), and the yield of 2a was not obtained
in 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) (entry 3). In contrast,
both 1-methoxy-2-propanol and 2-propanol resulted in almost
quantitative yields (entries 4 and 5). The reaction efficiency was
thus significantly affected by the structures of the secondary alco-
hols (entries 3–5). These results could be explained in terms of the
solvation of the halide anions. The protic solvents activate the
epoxide via hydrogen bonds as described above, while they
strongly solvate iodide anion, thereby diminishing the reactivity
Finally, we examined the carbonate-forming reactions of other
epoxides with 1aÁMeI and 1-methoxy-2-propanol at 25 or 45 °C
(Table 5).14 The carbonate (3a) was isolated quantitatively by short
Table 5
Synthesis of various cyclic carbonates by using 1aÁMeI under ambient conditionsa
O
1a•MeI
O
(5 mol%)
O
Table 4
+
CO2
(1 atm)
Effect of solvent on the synthesis of 2a by using 1aÁMeI under ambient conditionsa
O
1-Methoxy-2-propanol
25 or 45 °C, 24-48 h
R
R
O
2a-d
3a-d
O
O
1a•MeI
a: R = PhOCH2, b: R = ClCH2, c: R = CH2=C(CH3)CO2CH2, d: R = n-Bu
O
(5 mol%)
O
O
+
CO2
(1 atm)
Entry
Epoxide
Temp (°C)
Time (h)
Yieldb (%)
Org. Solv.
25 °C, 24 h
1
2
3
4
5
6
7
8
2a
2b
2b
2c
2c
2d
2d
2d
25
25
25
25
25
25
25
45
24
24
36
24
36
24
48
24
99 (99)c
90
2a
3a
98 (97)c
96
Entry
Solvent
Yieldb (%)
1
2
3
4
5
THF
MeOH
(CF3)2CHOH
35
30
>99 (>99)c
65
Not detected
99
97
86
(MeOCH2)(CH3)CHOH
(CH3)2CHOH
92 (91)c
a
Reaction conditions: 1 mmol of 2a–d, 0.05 mmol of 1aÁMeI, 0.2 mL of 1-meth-
a
Reaction conditions: 1 mmol of 2a, 0.05 mmol of 1aÁMeI, 0.2 mL of solvent,
oxy-2-propanol, 25 or 45 °C under 1 atm of CO2 for 24–48 h.
b
25 °C under 1 atm of CO2 for 24 h.
Determined by 1H NMR.
Determined by 1H NMR.
Isolated yields are in parentheses.
b
c