32
J. Tharun et al. / Catalysis Communications 54 (2014) 31–34
employing pyridinium-based ILs as the catalyst (Scheme S1, Supple-
mentary information).
Table 1
Catalyst screening for the cycloaddition of AGE and CO
2
.
Entry
Catalyst
Conversion (%)
Selectivity (%)
TOF (h−1)
2
. Experimental
1
2
3
4
5
6
7
8
9
None
EPyCl
PPyCl
BPyCl
HPyCl
BPyBr
BMPyI
0
79
82
86
78
95
93
10
92
73
0
90
91
96
93
95
87
94
96
89
0
1770
1838
1927
1748
2129
2084
224
2
.1. Materials
1
-Ethylpyridinium chloride (N98%) (EPyCl), 1-propylpyridinium
chloride (N98%) (PPyCl), 1-butylpyridinium chloride (N98%) (BPyCl),
-hexylpyridinium chloride (N98%) (HPyCl), 1-butylpyridinium bro-
mide (N98%) (BPyBr), and 1-butylpyridinium hexafluorophosphate
N98%) (BPyPF ) were purchased from TCI chemicals whereas 1-butyl-
-methylpyridinium iodide (N99%) (BMPyI), 1-butyl-3-methyl-
BPyPF
6
1
BMImBr
TBAB
2062
1636
10
(
4
6
Reaction conditions: AGE = 18.6 mmol, catalyst amount = 1 mmol, PCO2 = 0.96 MPa,
microwave power = 200 W, time = 30 s. TOF: moles of AGE converted per mol of
ionic liquid/h.
imidazolium bromide (N97%) (BMImBr), and tetrabutylammonium
bromide (N98%) (TBAB) were obtained from Aldrich and used without
further purification. Allyl glycidyl ether (AGE) was purchased from
Aldrich and used as received. Carbon dioxide of 99.999% purity was
used without further purification. CH Cl was obtained from SK
2 2
Chemicals, Korea, and used as received.
The anion effect of pyridinium-based ILs was investigated using
6
anions possessing Cl, Br, I, and PF (Table 1). Comparing BPyCl, BPyBr,
and BMPyI (entries 4. 6, 7), catalysts containing I or Br anions
showed high catalytic activities of 93% and 95% with excellent TOF of
084 and 2129 h (respectively) under moderate reaction conditions
0.96 MPa, 200 W, 30 s). This was most likely due to the high nucleophi-
licity and leaving ability of these anions. Compared to I and Br sys-
tems, less nucleophilic anions Cl and PF
lower catalytic activity [24,30]. Various other ILs, such as imidazolium-
based BMImBr and ammonium salt-based TBAB, were also tested
entries 9 and 10) using solvent-free microwave cycloaddition and
they resulted in good activity. The high activity observed for BMImBr
and TBAB confirms the most important criteria for obtaining high cata-
2
lytic activity for cyclic carbonate synthesis from CO and epoxides,
namely, the presence of strong nucleophilic anions with appropriate
alkyl chain length cations. From the above studies and various others
8,24,29,30], bromide or iodide ions with butyl group-containing cat-
ions could serve as the best catalyst for solvent-free cycloaddition of
epoxide and CO under moderate reaction conditions. In most cases,
the TOFs for IL-based cycloaddition of epoxides and CO were lower
2
than 50 h
−
−
−
1
2
(
2
.2. Cycloaddition of epoxides and CO
2
−
−
For the reaction, a multi-mode microwave reactor (KMIC-2KW)
containing a source with a continuously adjustable power setting from
kW to 2 kW (adjusted via a 3-stub tuner) as well as a temperature
−
−
6
(entries 4 and 8) showed
0
controller unit operating at a frequency of 2.450 GHz was used. The
surface temperature of the reactor was measured using an IR tempera-
ture detector. The cut off temperature was kept at 175 °C considering
the subtle stability of Pyrex glass microwave reactor. The synthesis of
AGC from AGE and CO (Scheme S1) using pyridinium based ILs as a cat-
2
alyst was performed with the microwave reactor in a 40 mL Pyrex glass
reactor equipped with a magnetic stirrer. For each typical reaction, ILs
(
[
(
0.1–2.5 mmol) and AGE (18.6 mmol) were charged into the reactor
without a solvent and then purged several times with CO . The reactor
was then pressurized with CO and heated to a desired temperature
2
2
2
by stirring the reaction mixture at 600 rpm. After the completion of re-
action time, cycloaddition was stopped by cooling the reaction mixture
to room temperature and venting the remaining CO . The product was
2
dissolved in dichloromethane and an analysis was carried out using a
gas chromatography/mass spectrometry (GC–MS, Micromass, UK) anal-
ysis. The conversion of epoxides was obtained from gas chromatogra-
phy (GC, HP-5 capillary column, HP 6890, Agilent Technologies, Santa
Clara, CA, USA) data.
−1
[30–34]. To date, the highest TOF obtained for an IL-
−1
based system (HETBA) is in the range of 3000–14,000 h for the syn-
thesis of propylene carbonate from propylene oxide and CO in a
micro-reactor (3.5 MPa, 180 °C, 14 s) [35]. Pyridinium-based ILs have
been employed earlier for the cycloaddition of butyl glycidyl ether and
CO
of 15 h
2
2
to yield butyl glycidyl carbonate, but it resulted in a very low TOF
−1
under conventional reaction conditions of 0.82 MPa at
1
40 °C for 1 h [19]. Utilizing microwave-assisted cycloaddition of
AGE and CO to yield AGC in the presence of pyridinium-based ILs,
2
−
1
3
. Results and discussion
high TOFs ranging from 200 to 7000 h were achieved under moderate
reaction conditions of 0.96 MPa, 200 W, and 30 s. In this study, we
achieved a 450-fold increase in the TOF by using microwaves, which is
much larger than the TOF achieved using conventional synthesis tech-
niques [19].
3
.1. Catalytic activity
The catalytic activity of pyridinium-based ILs was tested using a
microwave-assisted solvent-free cycloaddition reaction using AGE as
the epoxide substrate. The observed results are tabulated in Table 1.
Entry 1 shows the need for an appropriate catalyst because no product
was formed in the absence of a catalyst. When the size of the cations
in the pyridinium IL was increased from ethyl to butyl (entries 2–4),
catalytic activity also increased reaching a maximum of 86% conversion
3.2. Effect of catalyst amount
Since BPyBr showed good catalytic activity, it was used to examine
the effects of reaction parameters (catalyst amount, time, microwave
power, and CO
2
pressure) on the cycloaddition reaction of AGE with
−
1
for BPyCl with a high turn-over frequency (TOF) of 1927 h under a
moderate reaction condition of 0.96 MPa of CO pressure, 200 W of mi-
crowave power, and reaction time of 30 s. The rate-determining step of
the epoxide-CO cycloaddition reaction involved a nucleophilic attack of
CO . Table 2 shows the relationship between the AGE conversion and
2
2
the amount of catalyst. The conversion increased from 31% to 95%
when the amount of catalyst increased from 0.1 mmol to 1.0 mmol,
−
1
−1
2
whereas the TOF decreased from 6947 h to 2129 h (entries 1–3).
This can be explained by the effect of the catalytically active sites in
AGE. The conversion efficiency of the substrate is mainly determined
by the collision frequency between the catalytically active sites and
the substrate. As the number of catalytically active sites increases with
the increase of catalyst concentration under a constant substrate vol-
ume, the collision frequency based on the catalytically active sites
anions on the epoxide ring. Therefore, bulky ILs, having longer distances
between cations and anions, have a higher anion activation capacity
[13]. However, a further increase in cationic size (entry 5) resulted in
a slight reduction of catalytic activity. This could be explained by the ste-
ric hindrance of the bulky ILs, which compensated for the increase of the
anion activation ability by increasing the alkyl chain length [29].