Functionalized phosphonium-based ionic liquids as efficient catalysts for the synthesis of cyclic carbonate from expoxides and carbon dioxide

Article abstract of DOI:10.1016/j.apcata.2013.10.060

A series of novel functionalized phosphonium-based ionic liquids (FPBILs) were synthesized by a simple method, and first evaluated as catalysts for the synthesis of cyclic carbonates through the cycloaddition of CO₂ to epoxides in the absence of co-catalyst and solvent. The FPBILs perform well in the cycloaddition reaction, especially the carboxyl-functionalized one. Over [Ph₃PC₂H₄COOH]Br, the yield of propylene carbonate is 97.3% (TOF = 64.9 h^-1) at 130 C and 2.5 MPa in 3 h. The synergistic effects of polarization induced by hydrogen bonding and nucleophilic attack of Br^-anion account for the excellent performance. Furthermore, the FPBILs with moderate methylene chain length show superior catalytic activity. It is because they have both strong acidity and weak electrostatic interaction between phosphonium cation and halide anion. The strong acidity facilitates the ring-opening of epoxyl, and the weak electrostatic interaction enhances the nucleophilic attack capability of Br ^-. It is envisaged that the metal- and solvent-free process has high potential for the catalytic conversion of CO₂ into value-added chemicals.

Full text of DOI:10.1016/j.apcata.2013.10.060

Applied Catalysis A: General 470 (2014) 183–188
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Functionalized phosphonium-based ionic liquids as efficient
catalysts for the synthesis of cyclic carbonate from expoxides
and carbon dioxide
Dai Wei-Lia, Jin Bia, Luo Sheng-Liana,∗, Luo Xu-Biaoa, Tu Xin-Mana, Au Chak-Tonga,b
a Key Laboratory of Jiangxi Province for Persistant Pollutants Control and Resources Recycle, Nanchang Hangkong University , Nanchang, 330063 Jiangxi ,
China
b Department of Chemistry, Hong Kong Baptist University, Kowloon Tong, Hong Kong, China
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel functionalized phosphonium-based ionic liquids (FPBILs) were synthesized by a simple
method, and first evaluated as catalysts for the synthesis of cyclic carbonates through the cycloaddition
of CO2 to epoxides in the absence of co-catalyst and solvent. The FPBILs perform well in the cycloaddition
reaction, especially the carboxyl-functionalized one. Over [Ph3PC2H4COOH]Br, the yield of propylene
carbonate is 97.3% (TOF = 64.9 h−1) at 130 ◦C and 2.5 MPa in 3 h. The synergistic effects of polarization
induced by hydrogen bonding and nucleophilic attack of Br−anion account for the excellent performance.
Furthermore, the FPBILs with moderate methylene chain length show superior catalytic activity. It is
because they have both strong acidity and weak electrostatic interaction between phosphonium cation
and halide anion. The strong acidity facilitates the ring-opening of epoxyl, and the weak electrostatic
interaction enhances the nucleophilic attack capability of Br−. It is envisaged that the metal- and solvent-
free process has high potential for the catalytic conversion of CO2 into value-added chemicals.
© 2013 Elsevier B.V. All rights reserved.
Received 14 September 2013
Received in revised form 25 October 2013
Accepted 28 October 2013
Available online 5 November 2013
Keywords:
Carbon dioxide fixation
Cycloaddition reaction
Phosphonium-based ionic liquid
Cyclic carbonate
1. Introduction
unsatisfactory activity, harsh reaction conditions, being water- or
air-sensitive, and the need of a metal salt as co-catalyst. Unlike
With the fast consumption of fossil fuels, “C1 resource crisis” has
become a concern [1]. Despite blamed for global warming, CO2 is a
C1 resource that is renewable, nontoxic, abundant, and inexpensive
[2–4]. The fixation of CO2 to generate valuable chemicals such as
cyclic carbonates is meaningful. Cyclic carbonates are used as polar
aprotic solvents, as electrolytes in lithium secondary batteries, as
precursors for the formation of polycarbonates, and as intermedi-
ates in the production of pharmaceuticals and fine chemicals [2–4].
A wide range of catalysts have been explored for the synthe-
sis of cyclic carbonates, including metal oxides [5,6], modified
molecular sieves [7,8], alkali metal salts [9], organometallic com-
plexes [10,11], and ionic liquids (ILs) [12–30]. Since the 2000s,
the use of ILs has attracted much interest, such as imidazolium
salts [12–19], quaternary ammonium salts [18,20,21], quaternary
phosphonium salts [22,23], pyridinium salts [25], gunidinium salts
[26–28], and other type ILs [29,30] were tested as catalysts for the
cycloaddition of CO2 into epoxides. The ILs show favorable features
such as good solvating ability, negligible vapor pressure, and vari-
able polarity but most of them suffer from shortcomings such as
imidazolium-based ILs, phosphonium-based ILs has not been well
studied because of such shortcomings. For example, He et al. [22]
used polyfluoroalkyl phosphonium iodides as catalyst for the syn-
thesis of propylene carbonate (PC) from propylene oxide (PO) and
CO2. Although the catalysts can be separated from the PC product,
the harsh reaction conditions (14.0 MPa, 24 h) seriously limit their
application. Sun et al. [23] reported that in the presence of ZnCl2,
traditional quaternary phosphonium halide exhibits good catalytic
activity. Our group also reported a ZnBr2–Ph4PI system that shows
excellent catalytic performance [24], but zinc halide is needed as a
co-catalyst. Although the catalyst systems mentioned above exhibit
good activity using metal halide as co-catalyst, the quaternary
phosphonium ionic liquids (such as PPh3BuBr, PPh3BuCl, PPh3BuI,
PPh3EtBr, PPh3PrBr, and PPh3HeBr) are themselves low in activity
[31]. Despite the presence of water could remarkable improve the
activity of quaternary phosphonium ionic liquids, problems such
as low PC selectivity (<95%) and difficulty in product separation
exist [31]. Hence, it is highly desirable to develop one-component
phosphonium-based ILs as effective catalysts for commercial syn-
thesis of cyclic carbonates.
Recently, it was demonstrated that by introducing certain func-
tional groups into the cation and/or anion of traditional ILs, the
“tailor-made” materials are task-specific in diverse areas [32–36].
∗
Corresponding author. Tel.: +86 791 83953373; fax: +86 791 83953373.
E-mail address: sllou@hnu.edu.cn (L. Sheng-Lian).
0926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.10.060

184
D. Wei-Li et al. / Applied Catalysis A: General 470 (2014) 183–188
805, 802, 742, 690 cm−1. HR-MS (QTOF): m/z = 387.0508, calcd. for
Ph
P
Ph
P
COOH
R
C
20H21OPBr (M + H): 387.0513.
Ph
Ph
n
[Ph3PC3H7]Br (1d): White solid; 1H NMR (400 MHz, DMSO-
Br
Y
Ph
Ph
d6): ı = 8.00–7.70 (m, 15H), 3.69–3.61 (m, 2H), 1.68–1.59 (m, 2H),
1.17–1.13 (m, 3H). IR (neat): n = 3076, 3051, 2925, 2885, 2879, 1585,
1485, 1436, 1380, 746, 690 cm−1. HR-MS (QTOF): m/z = 385.0718,
calcd. for C21H23PBr (M + H): 385.0721.
: n = 1, Y = Br; : n =3, Y = Br;
2a 2b
:R = COOH; : R = OH;
1a
1c
1b
2c:n = 4, Y = Br
:n = 2, Y = Cl; : n = 2,Y = I
:R = NH ; :R = CH ;
1d
2
3
3a
3b
1e:R = H
[Ph3PC2H5]Br (1e): White solid; 1H NMR (400 MHz, DMSO-d6):
ı = 7.97–7.79 (m, 15H), 3.67–3.59 (m, 2H), 1.30–1.21 (m, 3H). IR
(neat): n = 3072, 3045, 2923, 2883, 2876, 1585, 1485, 1436, 1379,
737, 690 cm−1. HR-MS (QTOF): m/z = 371.0552, calcd. for C20H21PBr
(M + H): 371.0564.
Scheme 1. Structures of FPBILs.
For the synthesis of cyclic carbonates by cycloaddition of CO2 to
epoxides, imidazolium-based ILs that contain hydroxyl [18,37] or
carboxyl [19,38,39] group show better catalytic performance than
traditional ILs. To the best of our knowledge, there is no application
of functionalized phosphonium-based ILs (FPBILs) for the synthe-
sis of cyclic carbonates. Herein, we report the synthesis of a series
of phosphonium-based ILs that are functionalized with hydroxyl,
carboxyl, and amino groups (Scheme 1). For the first time, these
novel materials were tested as catalysts for the synthesis of cyclic
carbonates through cycloaddition of CO2 to epoxides. The FPBILs
display high epoxide conversion, and with excellent selectivity
show almost quantitative yield of cyclic carbonates in the absence
of solvent or co-catalyst. In the study, the effect of catalyst struc-
ture and reaction parameters on performance was investigated in
detail.
[Ph3PCH2COOH]Br (2a): Brown solid; 1H NMR (400 MHz,
DMSO-d6): ı = 12.72 (s, 1H), 7.81–7.34 (m, 15H), 3.69 (s, 2H).
IR (neat): n = 3055, 3010, 2937, 2891, 2881, 1716, 1585, 1485,
1436, 742, 692 cm−1. HR-MS (QTOF): m/z = 401.0306, calcd. for
C20H19O2PBr (M + H): 401.0306.
[Ph3PC3H6COOH]Br (2b): White solid; 1H NMR (400 MHz,
DMSO-d6): ı = 12.42 (s, 1H), 8.06–7.77 (m, 15H), 3.79–3.72 (m, 2H),
2.48–2.46 (m, 2H), 1.75–1.69 (m, 2H). IR (neat): n = 3057, 3026,
2947, 2898, 2878, 1727, 1587, 1486, 1433, 738, 686 cm−1. HR-MS
(QTOF): m/z = 429.0623, calcd. for C22H23O2PBr (M + H): 429.0619.
[Ph3PC4H8COOH]Br (2c): White solid; 1H NMR (400 MHz,
DMSO-d6): ı = 12.75 (s, 1H), 7.93–7.75 (m, 15H), 3.85–3.77 (m, 2H),
3.44–3.35 (m, 2H), 2.93–2.86 (m, 2H), 2.61–2.50 (m, 2H). IR (neat):
n = 3062, 3028, 2960, 2904, 2873, 1726, 1583, 1484, 1436, 742,
690 cm−1. HR-MS (QTOF): m/z = 443.0762, calcd. for C23H25O2PBr
(M + H): 443.0775.
2. Experimental
[Ph3PC2H4COOH]Cl (3a): White solid; 1H NMR (400 MHz,
DMSO-d6): ı = 12.77 (s, 1H), 7.93–7.62 (m, 15H), 3.83–3.76 (m, 2H),
2.60–2.56 (m, 2H). IR (neat): n = 3052, 3010, 2989, 2939, 2893, 1739,
1583, 1485, 1436, 740, 690 cm−1. HR-MS (QTOF): m/z = 371.0968,
calcd. for C21H21O2PCl (M + H): 371.0968.
2.1. Chemicals
Propylene oxide was purchased from Sinopharm Chemical
Reagent Co., Ltd. Other epoxides were purchased from Alfa Aesar
China Co., Ltd. Triphenylphosphine (Ph3P), 2-bromoethanol, 2-
bromoethylamine hydrobromide, bromoethane, 3-bromopropane,
3-bromopropionic acid, 3-chloropropionic acid, 3-iodopropanoic
acid, 4-bromobutyric acid, and 5-bromovaleric acid were pur-
chased from Shanghai Jingchun Industry Co., Ltd. All chemicals
were used as received. The CO2 (99.9% purity) purchased from Nan-
chang Guoteng Gas Co. was used without any further treatment.
[Ph3PC2H4COOH]I (3b): White solid; 1H NMR (400 MHz, DMSO-
d6): ı = 12.77 (s, 1H), 7.93–7.77 (m, 15H), 3.84–3.76 (m, 2H),
2.60–2.56 (m, 2H). IR (neat): n = 3059, 3028, 2958, 2896, 2875, 1725,
1583, 1484, 1433, 740, 690 cm−1. HR-MS (QTOF): m/z = 463.0332,
calcd. for C21H21O2PI (M + H): 463.0324.
2.3. Synthesis of cyclic carbonates
2.2. Synthesis of phosphonium-based ionic liquids
The cycloaddition reactions were conducted in a 50 mL high-
pressure stainless-steel autoclave equipped with
a magnetic
A typical procedure for the preparation of [Ph3PC2H4NH2]Br (1c)
is as follows: A solution of triphenylphosphine (5 mmol) and 2-
bromoethylamine hydrobromide (5 mmol) in 20 mL toluene was
heated and subject to reflux for 24 h. After cooling, the resulted
crude solid was filtered out, and stirred in 10 mL triethylamine for
4 h. Afterward, the triethylamine was removed, and the as-obtained
solid was washed three times with ethyl acetate, then dried at
60 ◦C under vacuum for 12 h to give product 1c in the form of a
pale yellow solid. 1H NMR (400 MHz, DMSO-d6): ı = 7.95–7.76 (m,
15H), 3.68 (t, J = 6.0 Hz, 2H), 3.48–3.42 (m, 2H). IR (neat): n = 3396,
3053, 3026, 2976, 2973, 2894, 1583, 1486, 1433, 805, 802, 724,
690 cm−1. HR-MS (QTOF): m/z = 386.0677, calcd. for C20H22NPBr
(M + H): 386.0673.
stirring bar. In a typical run, the reactor was charged with epox-
ide (35.7 mmol), catalyst (0.5 mol%, calculated according to the
amount of ionic liquid), and an appropriate amount of biphenyl
(as internal standard for GC analysis). After the reactor was fed
with CO2 to a desired pressure, the autoclave with its contents
was heated to a selected temperature and stirred for a designated
period of time. Then the reactor was cooled to 0 ◦C in an ice-water
bath, and the remaining CO2 was released. The resulting mixture
was analyzed using a GC-mass spectrometer. The products were
quantitatively analyzed on a gas chromatograph (Agilent 7890A)
that was equipped with a FID and a DB-wax capillary column
(30 m × 0.53 mm × 1.0 ␮m).
Other FPBILs were prepared likewise but without the procedure
of stirring in triethylamine.
3. Results and discussion
[Ph3PC2H4COOH]Br (1a): White solid; 1H NMR (400 MHz,
DMSO-d6): ı = 12.77 (s, 1H), 7.90–7.11 (m, 15H), 3.85–3.77 (m, 2H),
2.61–2.54 (m, 2H). IR (neat): n = 3055, 3020, 2937, 2896, 2883, 1725,
1585, 1485, 1436, 724, 692 cm−1. HR-MS (QTOF): m/z = 415.0473,
calcd. for C21H21O2PBr (M + H): 415.0463.
3.1. Catalytic performance
The synthesis of PC through the cycloaddition of CO2 to PO was
chosen as model reaction to evaluate the catalytic activity of the as-
synthesized FPBILs, and the results are summarized in Table 1. It can
be seen that Ph3P and 3-bromopropionic acid show poor catalytic
activity when used alone (entries 1, 2), whereas the as-synthesized
[Ph3PC2H4OH]Br (1b): White solid; 1H NMR (400 MHz, DMSO-
d6): ı = 7.93–7.75 (m, 15H), 3.59–3.51 (m, 2H), 1.69–1.65 (m, 2H).
IR (neat): n = 3317, 3053, 3026, 2976, 2937, 2894, 1583, 1486, 1433,

D. Wei-Li et al. / Applied Catalysis A: General 470 (2014) 183–188
185
Table 1
100
95
90
85
80
100
90
Catalytic performance of various catalystsa.
Yield
Sel.
Entry
Catalytic results
Catalyst
80
Yield (%)
Sel. (%)
70
60
50
40
1
2
3
4
5
6
7
8
Ph3P
BrC2H4COOH
[Ph3PC2H4COOH]Br (1a)
[Ph3PC2H4OH]Br (1b)
[Ph3PC2H4NH2]Br (1c)
[Ph3PC3H6]Br (1d)
14.4
2.6
82.5
99.0
99.8
99.5
99.6
99.6
99.6
99.5
99.8
99.8
99.6
99.8
99.5
99.4
99.7
99.7
99.6
99.6
99.6
97.3
96.5
95.4
87.3
85.6
90.4
90.5
91.1
86.2
97.9
91.4
17.2
74.1
49.6
98.5
89.0
91.3
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
[Ph3PC2H4]Br (1e)
Catalyst loading(mol%)
[Ph3PC2H4]Br/C2H5COOHb
[Ph3PC2H4]Br/C2H5OHb
[Ph3PC2H4]Br/CH3(CH2)2NH2
[Ph3PCH2COOH]Br (2a)
[Ph3PC3H6COOH]Br (2b)
[Ph3PC4H8COOH]Br (2c)
Ph3P/CH3CH2COOHb
Ph3P/BrC2H5COOHb
[Ph3PC2H4COOH]Cl (3a)
[Ph3PC2H4COOH]I (3b)
[MImC2H4OH]Brc
9
Fig. 1. Influence of catalyst loading on PC yield and selectivity. Reaction conditions:
b
10
11
12
13
14
15
16
17
18
19
PO 35.7 mmol, initial CO2 pressure 2.5 MPa, Temp. 130 ◦C, time 3 h).
activity is 3b (I−) > 1a (Br−) » 3a (Cl−) (entries 3, 16, 17), which coin-
cides with the order of nucleophilicity of the halides. It is noted that
using FPBILs as catalyst, the selectivity to PC is always above 99.5%,
and only a minute amount of 1,2-propanediol generated due to PO
hydrolysis is detected as by-product. Moreover, Sun et al. reported
the imidazolium-based ILs functionalized by hydroxyl [18] and
carboxyl [19] showed good catalytic activity for the cycloaddition
reaction. In order to make a direct comparison, we evaluated the
[MImC2H4OH]Br [18] and [MImC3H6COOH]Br [19] ILs reported by
Sun et al. under the conditions adopted in this study, and the PC
yields observed were 89.0% and 91.3%, respectively (entries 18, 19).
It is noted that the yields are lower than those observed when
FPBILs were used (96.5% 95.4%, and 97.9%, entries 4, 5, and 12,
respectively). From the results, it is apparent that the FPBILs have
great potential for industrial application as a single-component
catalyst for the cycloaddition of CO2 to epoxides.
[MImC3H6COOH]Brc
a
Reaction conditions: PO 35.7 mmol, catalyst 0.5 mol%, initial CO2 pressure
2.5 MPa, Temp. 130 ◦C, time 3 h.
b
Equal catalyst amount (0.179 mmol).
[MImC2H4OH]Br (1-(2-hydroxyl-ethyl)-3-methylimidazolium bromide) and
c
[MImC3H6COOH]Br (1-(3-carboxyl-propyl)-3-methylimidazolium bromide) were
prepared according to ref. [18] and [19], respectively.
[Ph3PC2H4COOH]Br (1a) displays surprisingly high activity (yield
of 97.3%) and selectivity (99.8%) (entry 3). As for [Ph3PC2H4OH]Br
(1b) and [Ph3PC2H4NH2]Br (1c) (entries 4, 5), both show activity
slightly lower than that of 1a, but the catalytic performance can
still be considered as good. As reported, hydrogen bonding has a
positive effect on the ring-opening of epoxide [18,19,37,39]. The
relative low activity of 1b and 1c can be partly due to the polariza-
tion capability of hydroxyl and amino group is not strong enough
for the opening of epoxide ring. Compared with 1a, 1b and 1c, the
traditional ILs [Ph3PC3H6]Br (1d) and [Ph3PC2H4]Br (1e) are much
lower in catalytic activity. It is noted that in the presence of alkyl
compounds that contain NH2, COOH or OH group, there is enhance-
ment of 1e activity (entries 8–10). The results indicate that the
functional groups play an important role in the promotion of the
reaction through hydrogen bonding.
The structure of FPBILs cations has a strong influence on catalytic
activity. With Br−as anion and under equal reaction conditions,
the effect of varying the molecular weight of phosphonium cations
([Ph3PCnH2nCOOH]+, n = 1–4, entries 3, 11–13) was investigated. It
was found that catalytic activity increases with a rise of n from 1 to
3. Above n = 3 (entries 3, 11, 12), there is decrease of PO conversion.
The phenomena may be due to two factors: (i) An increase of methy-
lene chain length weakens the electrostatic interaction between
halide anion and phosphonium cation, enhancing the nucleophi-
lic attack ability of halide anion, and facilitating the cycloaddition
reaction as a result [23]; (ii) The acidity of carboxylic acid with a
shorter methylene chain is stronger than that with a longer methy-
lene chain, thus the cation with shorter methylene chain has higher
ring-opening capability, resulting in rise of PC yield [19]. Therefore,
the competition between the two opposite factors results in the
FPBILs with moderate methylene chain length (n = 2, 3) having the
best catalytic activity.
3.2. Effects of reaction parameters
It is generally accepted that reaction parameters such as cata-
lyst loading, temperature, time and initial CO2 pressure have an
effect on product yield. We adopted [Ph3PC2H4COOH]Br (1a) to
investigate the effects of reaction parameters because of its good
activity. As shown in Fig. 1, with increase of 1a loading from 0.1 to
0.5 mol%, there is a sharp increase of PC yield from 44.0 to 97.3%.
Further rise of 1a loading from 0.5 to 0.8 mol% does not cause sig-
nificant increase of PC yield (to 98.0% only). Obviously, even at a
low 1a loading of 0.5 mol%, excellent PC yield (97.3%) and selectiv-
ity (99.8%) can be obtained. It is observed that upon a change of
catalyst loading from 0.1 to 0.8 mol%, PC selectivity remains higher
than 99.5%.
The dependence of PC yield and selectivity on reaction temper-
ature is shown in Fig. 2. It is apparent that the performance of 1a
is sensitive to reaction temperature. With increase of temperature
from 110 to 130 ◦C, PC yield increases sharply from 69.8 to 97.3%.
However, further rise of temperature to 150 ◦C results in only slight
increase of PC yield. On the other hand, the selectivity to PC is almost
independent of temperature. In the 110–150 ◦C range, PC selectivity
is ≥99.7%.
Fig. 3 shows the dependence of PC yield and selectivity on initial
CO2 pressure. It can be seen that the PC yield is strongly affected by
initial CO2 pressure. In the low-pressure range (1.5–2.5 MPa), there
is rapid rise of PC yield (from 69.2% to 97.3%) with increase of initial
CO2 pressure, but further rise of pressure to 4.0 MPa results in mod-
erate decrease of PC yield. Such an effect of CO2 pressure on catalytic
activity has been observed in other catalytic systems [24,37,39,40].
The phenomenon can be explained based on the concepts of reac-
tion equilibrium and mass transfer of reactant. PC is in its liquid
form under the adopted reaction conditions, and an increase of
CO2 pressure is a positive factor for PO conversion. On the other
hand, acidic CO2 dissolves in basic epoxide and liquefies due to the
We also investigated the effect of counter anions on catalytic
performance. One can see that in the presence of propionic acid,
there is a slight increase of Ph3P activity (entry 14). However when
propionic acid is replaced by 3-bromopropionic acid, there is large
enhancement of Ph3P activity (entry 15). It is deduced that nucleo-
philic attack on the epoxide by Br−is crucial for the synthesis of PC.
We tried Cl−and I−as counter anions, and found that the order of

186
D. Wei-Li et al. / Applied Catalysis A: General 470 (2014) 183–188
100
90
80
70
60
100
95
90
85
80
Yield
Sel.
110
120
130
140
150
o
Temperture ( C)
Fig. 2. Effect of temperature on PC yield and selectivity. Reaction conditions: PO 35.7 mmol, 1a 0.5 mol%, initial CO2 pressure 2.5 MPa, time 3 h.
formation of CO2-epoxide complexes [40,41]. Hence, rather than
promoting the interaction between PO and catalyst, a high CO2
pressure enhances CO2–PO complex formation, consequently lead-
ing to a decline of catalytic activity. It is noted that PC selectivity is
always above 99.5% and can be considered as independent of CO2
initial pressure.
The plots of PC yield and selectivity versus reaction time are
shown in Fig. 4. It is observed that the cycloaddition reaction pro-
ceeds rapidly within the first 3 h, reaching a PC yield of 97.3%.
Further the reaction results in only a slight rise of yield. Hence,
3 h was chosen as the reaction time in this study. Furthermore, the
selectivity to PC stays above 99.6% throughout.
quantitative yield of PC (97.2%) is achieved (entry 3). Furthermore,
the reactivity of styrene oxide (entry 5) is nearly equal to that of PO
(entry 2), but higher than that of phenyl glycidyl ether (entry 6) and
1,2-hexene oxide (entry 4). It is observed that much longer time is
required to complete the conversion of cyclohexene (entry 7), plau-
sibly due to the fact that the two rings of cyclohexene oxide hinder
the nucleophilic attack of Br−. From the results, one can see that
the reactivity of the selected substrates is in the order of ethylene
oxide > epichlorohydrin > stytrene oxide ≈ PO > phenyl glycidyl
ether > 1, 2-hexene oxide > > cyclohexene oxide.
3.4. Possible reaction mechanism
As shown in Table 1, the FPBILs with carboxyl, hydroxyl, and
amino groups exhibit better catalytic activities than those with-
out functional groups. One can enhance the activity of traditional
ILs by adding alkyl compounds that contain carboxyl, hydroxyl, and
amino groups. The results indicate that those functional groups play
an important role in the cycloaddition reaction. Moreover, nucleo-
philic attack of bromide anion on epoxide is essential. As reported
[18,19,37,39], there is hydrogen bonding between carboxyl group
and epoxide that facilitates the ring-opening of epoxide. In addi-
tion, the effect of CO2 pressure on the activity of FPBILs suggests
3.3. Cycloaddition of CO2 to various epoxides
In order to demonstrate the scope of FPBILs application, we
examined the cycloaddition reaction of CO2 to selected epoxides
over 1a, and the results are depicted in Table 2. As shown, the cat-
alyst 1a exhibits good activity and selectivity towards a variety of
terminal epoxides (entries 1–6). Ethylene oxide is the most reac-
tive substrate, and 100% yield of carbonate is achieved in 1 h (entry
1). The reactivity of epichlorohydrin is lower than that of ethylene
oxide but higher than that of the other epoxides. Within 2 h, almost
100
95
100
90
100
95
90
85
80
100
90
80
70
60
Yield
Sel.
80
90
Yield
Sel.
70
85
60
80
1
2
3
4
1
2
3
4
5
6
Initial CO2 pressure (MPa)
Time (h)
Fig. 3. Effect of initial CO2 pressure on PC yield and selectivity. Reaction conditions:
PO 35.7 mmol, 1a 0.5 mol%, Temp. 130 ◦C, time 3 h.
Fig. 4. Effect of time on PC yield and selectivity. Reaction conditions: PO 35.7 mmol,
1a 0.5 mol%, initial CO2 pressure 2.5 MPa, Temp. 130 ◦C.

D. Wei-Li et al. / Applied Catalysis A: General 470 (2014) 183–188
187
Table 2
Cycloaddition of CO2 to various epoxides catalyzed by 1aa.
Entry
Catalytic results
Yield (%)
Epoxide
Product
Time (h)
Sel. (%)
O
O
O
1
1
100
100
O
O
O
O
O
2
3
97.3
99.8
CH3
CH3
O
O
3
4
2
5
97.2
99.4
99.5
99.7
O
O
Cl
Cl
O
O
O
O
O
CH3
( )3
CH3
( )3
O
O
O
5
6
3
99.5
99.8
O
O
O
O
O
4
99.3
82.3
99.8
99.8
Ph
O
Ph
O
O
O
7
20
O
a
Reaction conditions: PO 35.7 mmol, catalyst 0.5 mol%, initial CO2 pressure 2.5 MPa, Temp. 130 ◦C.
O
that the activation of PO by FPBILs is a key step, and the competitive
formation of CO2–PO complex also suggests the coupling of PO and
FPBILs is essential. Based on the results reported by others as well
as those obtained in the present study, we suggest a mechanism
for the synthesis of cyclic carbonates over FPBILs (Scheme 2). First,
there is hydrogen bonding between the H atom of functional group
and the O atom of epoxide, resulting in polarization of epoxide C-O
bond and formation of intermediate I. Then, there is nucleophi-
lic attack of Br−on the less sterically hindered ␤-carbon atom of
epoxide, facilitating opening of epoxy ring and formation of inter-
mediate II. Following the insertion of CO2 into intermediate II,
there is the generation of halocarbonate III that transforms to cyclic
carbonate through intermolecular displacement of the bromide
ion.
O
R
O
O
O
COOH
Br
R
Ph3P
O
C
O
C
O
H
Ph3P
O
H
Ph3P
(III)
(I)
O
O
O
Br
R
R
Br
O
C
O
H
Ph3P
(II)
Br
CO2
O
R
4. Conclusion
We developed a facile, reliable, and highly selective protocol for
the synthesis of cyclic carbonate using FPBILs as catalysts under
Scheme 2. Proposed mechanism for the cycloaddition reaction.

188
D. Wei-Li et al. / Applied Catalysis A: General 470 (2014) 183–188
solvent- and metal-free conditions. The FPBILs are easily synthe-
sized, air-stable, cheap, and effective for the cycloaddition reaction
under metal- and solvent-free conditions. Compared to traditional
ILs, the FPBILs show superior performance, especially the carboxyl-
functionalized one. The excellent performance is ascribed to the
synergetic effect of hydrogen bonding formed between functional
group and epoxide, and the nucleophilic attack of halide anion
on the epoxide. The FPBILs that are moderate in methylene chain
length (n = 2, 3) exhibit performance superior to those that are
shorter or longer in chain length. The FPBILs catalysts are efficient
for the fixation of CO2 by means of cycloaddition to epoxides. They
are environment-benign and can be used as substitutes for the
Lewis-acid/Lewis-base binary catalytic systems.
[10] X.B. Lu, B. Liang, Y.J. Zhang, Y.Z. Tian, Y.M. Wang, C.X. Bai, H. Wang, R. Zhang, J.
Am. Chem. Soc. 126 (2004) 3732–3733.
[11] L.L. Jin, H.W. Jing, T. Chang, X.L. Bu, L. Wang, Z.L. Liu, J. Mol. Catal. A: Chem. 261
(2007) 262–266.
[12] J.J. Peng, Y.Q. Deng, New. J. Chem. 25 (2001) 639–641.
[13] H. Kawanami, A. Sasaki, K. Matsui, Chem. Commun. 39 (2003) 896–897.
[14] J.M. Sun, S.I. Fujita, F.Y. Zhao, M. Arai, Green Chem. 6 (2004) 613–616.
[15] H.S. Kim, J.J. Kim, H. Kim, H.G. Jang, J. Catal. 220 (2003) 44–46.
[16] J.K. Lee, Y.J. Kim, Y.S. Choi, H. Lee, J.S. Lee, J. Hong, E.K. Jeong, H.S. Kim, M. Cheong,
Appl. Catal., B 111–112 (2012) 621–627.
[17] J.Y. Kim, R.S. Varma, J. Org. Chem. 20 (2005) 7882–7891.
[18] J. Sun, S.J. Zhang, W.G. Cheng, J.Y. Ren, Tetrahedron Lett. 49 (2008) 3588–3591.
[19] J. Sun, L.J. Han, W.G. Cheng, J.Q. Wang, X.P. Zhang, S.J. Zhang, ChemSusChem. 4
(2011) 502–507.
[20] V. Calo, A. Nacci, A. Monopoli, A. Fanizzi, Org. Lett. 4 (2002) 2561–2563.
[21] J.M. Sun, S.I. Fujita, F.Y. Zhao, M. Arai, Appl. Catal., A 287 (2005) 221–226.
[22] L.N. He, H. Yasuda, T. Sakakura, Green Chem. 5 (2003) 92–94.
[23] J. Sun, L. Wang, S.J. Zhang, Z.X. Li, X.P. Zhang, W.B. Dai, R. Mori, J. Mol. Catal. A:
Chem. 256 (2006) 295–300.
Acknowledgments
[24] S.S. Wu, X.W. Zhang, W.L. Dai, S.F. Yin, W.S. Li, Y.Q. Ren, C.T. Au, Appl. Catal., A
341 (2008) 106–111.
[25] H.Y. Ju, J.Y. Ahn, M.D. Manju, K.H. Kim, D.W. Park., Korean J. Chem. Eng. 25
(2008) 471–473.
This work was financially supported by the National Natural Sci-
ence Foundation of China (Grant Nos. 51104089, 51238002, and
51272099), Jiangxi Province Science and Technology Project (Grant
Nos. 2010AIB00400); Jiangxi Province Science and Technology Sup-
port Program (Grant Nos. 20111BBG70003-4); and Natural Science
Foundation of Jiangxi Province (Grant Nos. 20122BAB213013). CTA
thanks NCHKU for an adjunct professorship.
[26] H.B. Xie, H.F. Duan, S.H. Li, S.B. Zhang, New J. Chem. 29 (2005) 1199–1203.
[27] H.B. Xie, S.H. Li, S.B. Zhang, J. Mol. Catal. A: Chem. 250 (2006) 30–34.
[28] X.Y. Dou, J.Q. Wang, Y. Du, E. Wang, L.N. He, Synlett 19 (2007) 3058–3062.
[29] B. Ochiai, T. Iwamoto, T. Endo, Green Chem. 8 (2006) 138–140.
[30] Z.Z. Yang, L.N. He, C.X. Miao, S. Chanfreau, Adv. Synth. Catal. 352 (2010)
2233–2240.
[31] J. Sun, J.Y. Ren, S.J. Zhang, W.G. Cheng, Tetrahedron Lett. 50 (2009) 423–426.
[32] V.I. Pârvulescu, C. Hardacre, Chem. Rev. 107 (2007) 2615–2665.
[33] M.P. Atkins, M.J. Earle, K.R. Seddon, M.S. Kwany, L. Vanoye, Chem. Commun. 46
(2010) 1745–1747.
References
[34] S.W. Tang, A.M. Scurto, B. Subramaniam, J. Catal. 268 (2009) 243–250.
[35] N. Yan, Y. Yuan, R. Dykeman, Y. Kou, P.J. Dyson, Angew. Chem. 122 (2010)
5681–5685.
[36] Z.F. Fei, D.B. Zhao, T.J. Geldbach, R. Scopelliti, P.J. Dyson, Chem. Eur. J. 10 (2004)
4886–4893.
[37] W.L. Dai, L. Chen, S.F. Yin, W.H. Li, Y.Y. Zhang, S.L. Luo, C.T. Au, Catal. Lett. 137
(2010) 74–80.
[38] L.N. Han, S.J. Choi, M.S. park, S.M. Lee, Y.J. Kim, M.I. Kim, B.Y. Liu, D.W. Park,
React. Kinet. Mech. Catal. 106 (2012) 25–35.
[39] Y.Y. Zhang, S.F. Yin, S.L. Luo, C.T. Au, Ind. Eng. Chem. Res. 51 (2012) 3951–3957.
[40] L.F. Xiao, F.W. Li, J.J. Peng, C.G. Xia, J. Mol. Catal. A: Chem. 253 (2006) 265–269.
[41] R. Nomura, M. Kimura, S. Teshima, A. Ninagawa, H. Matsuda, Bull. Chem. Soc.
Jpn. 55 (1982) 3200–3203.
[1] Y.H. Zheng, Z.F. Li, S.F. Feng, M. Lucas, G.L. Wu, Y. Li, C.H. Li, G.M. Jiang, Renew.
Sust. Energ. Rev. 14 (2010) 3132–3139.
[2] T. Sakakura, J.C. Choi, H. Yasuda, Chem. Rev. 107 (2007) 2365–2387.
[3] M. North, R. Pasquale, C. Young, Green Chem. 12 (2010) 1514–1539.
[4] X.B. Lu, D.J. Darensbourg, Chem. Soc. Rev. 41 (2012) 1462–1484.
[5] B.M. Bhanage, S.I. Fujita, Y. Ikushima, M. Arai, Appl. Catal., A 219 (2001)
259–266.
[6] W.L. Dai, S.F. Yin, R. Guo, S.L. Luo, X. Du, C.T. Au, Catal. Lett. 136 (2010) 35–44.
[7] E.J. Doskocil, J. Phys. Chem. B 109 (2005) 2315–2320.
[8] R. Srivastava, D. Srinivas, P. Ratnasamy, Catal. Lett. 91 (2003) 133–139.
[9] T. Sako, T. Fukai, R. Sahashi, Ind. Eng. Chem. Res. 41 (2002)
5353–5358.