Facile alkali-assisted synthesis of g-C3N4 materials and their high-performance catalytic application in solvent-free cycloaddition of CO2 to epoxides

Article abstract of DOI:10.1039/c6ra10509b

A series of graphitic carbon nitride materials were synthesized using guanidine hydrochloride (GndCl) as a precursor with the aid of alkali treatment. The introduction of alkali successfully enabled GndCl to be transformed into g-C₃N₄ at much lower calcination temperatures (450-475 °C). The g-C₃N₄ samples synthesized under various conditions have been characterized by several techniques including XRD, FT-IR, UV-vis, ¹³C NMR, and XPS spectroscopy. The results confirmed that the alkali could effectively accelerate further condensation of melem-like fragments to g-C₃N₄. Meanwhile, a possible mechanism of alkali-assisted synthesis of g-C₃N₄ from GndCl has been proposed. In solvent-free catalytic cycloaddition of CO₂ to propylene oxide to propylene carbonate (PC), g-C₃N₄-NaOH and g-C₃N₄-KOH materials demonstrated high and stable catalytic performances, affording PC yields of ca. 90% under optimized reaction conditions. Moreover, the activities were superior to those obtained over g-C₃N₄ prepared without alkali treatment. In addition, the catalytic activity along with preparation method for the present g-C₃N₄ has also been compared with other reported g-C₃N₄-based catalysts.

Full text of DOI:10.1039/c6ra10509b

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Facile alkali-assisted synthesis of g-C3N4 materials
and their high-performance catalytic application in
solvent-free cycloaddition of CO2 to epoxides†
Cite this: RSC Adv., 2016, 6, 55382
*
*
Jie Xu, Jie-Kun Shang, Quan Jiang, Yue Wang and Yong-Xin Li
A series of graphitic carbon nitride materials were synthesized using guanidine hydrochloride (GndCl) as
a precursor with the aid of alkali treatment. The introduction of alkali successfully enabled GndCl to be
transformed into g-C3N4 at much lower calcination temperatures (450–475 ꢀC). The g-C3N4 samples
synthesized under various conditions have been characterized by several techniques including XRD, FT-
IR, UV-vis, 13C NMR, and XPS spectroscopy. The results confirmed that the alkali could effectively
accelerate further condensation of melem-like fragments to g-C3N4. Meanwhile, a possible mechanism
of alkali-assisted synthesis of g-C3N4 from GndCl has been proposed. In solvent-free catalytic
cycloaddition of CO2 to propylene oxide to propylene carbonate (PC), g-C3N4-NaOH and g-C3N4-KOH
materials demonstrated high and stable catalytic performances, affording PC yields of ca. 90% under
optimized reaction conditions. Moreover, the activities were superior to those obtained over g-C3N4
prepared without alkali treatment. In addition, the catalytic activity along with preparation method for the
present g-C3N4 has also been compared with other reported g-C3N4-based catalysts.
Received 23rd April 2016
Accepted 30th May 2016
DOI: 10.1039/c6ra10509b
www.rsc.org/advances
signicant chemicals, which are extensively used as excellent
aprotic polar solvents, electrolytes, ne chemical intermediates,
and precursors for polycarbonate materials.8,9
1. Introduction
Carbon dioxide (CO2) is the major byproduct of worldwide
fossil-fuel-based energy production and is believed to be partly
responsible for changes to the global climate.1 At the present
time, carbon capture and storage (CCS) from fossil-fuel power
stations is the only strategy being investigated to reduce the
emissions of CO2, however it demands a high cost of energy.2
Alternatively, CO2 is also an abundant, cheap, and sustainable
C1 resource.3 In this respect, transformation of CO2 as
a renewable building block to valuable chemicals possesses
signicant values in both environmental preservation and
resource utilization, and therefore has recently become the hot
topic both in sustainable chemistry and green catalysis.4 Among
the various schemes proposed for the transformation of CO2,
because of its high atomic economy, the cycloaddition of CO2
with epoxides (e.g. propylene oxide, PO) to the cyclic carbonates
(e.g. propylene carbonate, PC), especially performed under
solvent-free circumstance, has been considered as the most
promising pathway.5–7 More importantly, the products of the
cycloaddition of CO2, i.e. cyclic carbonates, are also very
A wide range of catalysts have been explored to promote
cycloaddition of CO2 to cyclic carbonates. Up to now, the ionic
liquids (ILs) and organometallic complexes (e.g. salen
complexes) have been widely recognized as the most efficient
catalysts for the reactions.6,10 Unfortunately, the crucial issue
associated with the application of such catalysts lies in their
inherently homogeneous nature, being difficult in product
separation and catalyst recovery.11,12 Alternatively, immobiliza-
tion of ILs or metal complexes onto porous solids13,14 could
practically relieve the above problem. Despite this effort, the
high cost of coupling agents and tedious route of immobiliza-
tion still restrain the further practical implementation of the
supported ILs or complexes for catalytic cycloaddition reactions
of CO2.15,16 In this context, in terms of both fundamental
research and industrial application, it is of signicance to
search and develop a new solid catalyst that can allow high
performance for the catalytic cycloaddition reactions of CO2.
Very recently, as an analogy of graphite, graphitic carbon
nitride (g-C3N4) material has being emerged as an appealing
and fascinating metal-free material, and attracted a great deal of
attention in a wide community.17–19 So far, g-C3N4 and its based
materials have been widely applied in various research elds,
including photocatalysis,20–22 heterogeneous catalysis,23–25 fuel
cells,26 gas storage,27,28 etc. The success of g-C3N4 is largely due
to its unique combination of multiple physicochemical prop-
erties.17 Of particular note is that g-C3N4 has abundant N-
Jiangsu Key Laboratory of Advanced Catalytic Materials and Technology, School of
Petrochemical Engineering, Changzhou University, Gehu Road 1, Changzhou,
Jiangsu 213164, PR China. E-mail: shine6832@163.com; liyxluck@163.com; Fax:
+86-519-86330135; Tel: +86-519-86330135
† Electronic
supplementary
information
(ESI)
available:
N2
adsorption–desorption isotherms, surface areas, XPS survey of g-C3N4 materials;
and FT-IR spectra of the fresh and spent g-C3N4 catalysts. See DOI:
10.1039/c6ra10509b
55382 | RSC Adv., 2016, 6, 55382–55392
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
containing species in the forms of amino groups on the edges of
graphitic and tertiary amines bridging heptazine rings
(Scheme 1); such constitution enables g-C3N4 with a typical
solid base.16,29 To date, it has been reported that g-C3N4 mate-
rials demonstrated potential catalysis in several base-mediated
organic reactions, including Knoevenagel condensation30–32 and
transesterication.16,33,34 Moreover, Thomas35 and Park24 et al.
have found that g-C3N4 materials could effectively activate CO2
molecules in some CO2-involving oxidative reactions. Very
recently, Yin et al.36 have synthesized phosphorous-modied g-
C3N4 materials, and revealed that such g-C3N4 samples could
efficiently catalyze the CO2 cycloaddition to epoxides owing to
their acid–base bifunctionality.
Nevertheless, owing to its weak basic intensity, the activity of
pristine g-C3N4 materials (namely prepared via direct thermal
treatment) for the catalytic cycloaddition of CO2 was indeed
limited. To enhance their activity, in our recent work, we have
prepared a series of mesoporous g-C3N4 materials, and then
loaded them with zinc halides37 or functionalized with quater-
nary amines.38 Although the modication has successfully
improved the nal activity of g-C3N4 in the cycloaddition reac-
tions of CO2, it should be noted that the mesoporous g-C3N4
materials were fabricated using silica nanoparticles as hard
2. Experimental
2.1. Preparation of g-C3N4 materials
9 g (0.094 mmol) of guanidine hydrochloride (GndCl) was added
into 50 mL of alkali solution (NaOH/KOH) ethanolic solution
(0.5 mol Lꢁ1), and stirred for ca. 2 h under room temperature.
During the procedure, white precipitation, i.e. alkali metal
chloride was observed. The mixture was ltered to remove the
precipitation, and the ltrate was dried at 65 ꢀC overnight to
evaporate ethanol. Aerwards, the white solid was calcinated
from room temperature to the desired temperature with
a ramping rate of 3 C minꢁ1, and temperate for another 3 h
ꢀ
under N2 atmosphere (20 mL minꢁ1) in a tube furnace. The
resultant pale yellow powder was collected and labeled with g-
C3N4-T-MOH, where MOH and T indicated the alkali metal
hydroxide and thermal treatment temperature, respectively. For
comparison, 9 g of GndCl was directly calcinated without
addition of alkali, and the nal yellow solid was designated as g-
C3N4-T-direct.
2.2. Sample characterization
templates via a nanocasting method. The preparation process is X-ray diffraction patterns were recorded with a Rigaku D/max
time-consuming and complicated, meanwhile demanding 2500 PC X-ray diffractometer equipped with a graphite mono-
volatile and toxic HF solution as an etching agent. Furthermore, chromator (40 kV, 40 mA) using Ni-ltered Cu-Ka radiation (l ¼
˚
the employment of such acidic HF solution would neutralize 1.5418 A).
and thus damage the original basic sites of g-C3N4 materials,
resulting in an inevitable decline in their catalytic activity.16,39
Fourier transform infrared (FT-IR) spectra of the samples
were collected in transmission mode from KBr pellets at room
In continuation of the work, by means of the introduction of temperature ꢁon1 a Bruker Tensor 27 spectrometer with a resolu-
transition metal halides into the precursors of g-C3N4, we tion of 4 cm , using 32 scans per spectrum in the region of
prepared a series of metal doped bulk g-C3N4, which exhibited 400–4000 cmꢁ1. The mass ratio of every sample to KBr was
enhanced catalytic activities in the transesterication reactions. constant at 1 : 100.
However, there was potential leaching of metal species. In order
UV-vis diffuse reectance spectra (DRS) were recorded on
to further develop a robust g-C3N4 solid with high-performance a Shimizu UV-3600 spectrophotometer. BaSO4 was used as
activity, in this work, we have introduced a convenient alkali- a standard reference. Each sample was pressed into a thin tablet
assisted synthetic approach of g-C3N4, which could be oper- and tested under ambient conditions. The absorption spectrum
ated under lower thermal treatment temperatures (450–475 ꢀC) was calculated from the reectance data with Kubelka–Munk
compared with the prevailing methods. More encouragingly, function.
the synthesized materials showed high catalytic activities in the
solvent-free cycloaddition reactions of CO2 to epoxides.
X-ray photoelectron spectroscopy (XPS) measurements were
performed using a Perkin-Elmer PHI 5000C spectrometer
working in the constant analyzer energy mode with Mg Ka
radiation as the excitation source. The carbonaceous C 1s line
(284.6 eV) was used as the reference to calibrate the binding
energies.
The solid-state NMR 13C CP/MAS (cross polarization magic
angle spinning) measurement was conducted using a Varian
400WB 300 spectrometer using a 7.5 mm triple-tuned probe
spinning at 30 kHz.
Elementary analysis (EA) was performed with Elementar
Vario EL III instrument to determine the carbon, nitrogen, and
hydrogen contents of the samples.
Nitrogen adsorption–desorption isotherms were measured
at ꢁ196 ꢀC using a Micromeritics ASAP 2020 analyzer. Before
the analysis, the samples were degassed (10 mmHg) at 150 ꢀC for
at least 4 h. The specic surface area was calculated according
to the Brunauer–Emmett–Teller (BET) method, and pore size
Scheme 1 A possible structure of g-C3N4 constituted by the tri-s-
triazine tectonic units.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 55382–55392 | 55383

View Article Online
RSC Advances
Paper
distribution was determined by the Barrett–Joyner–Halenda
method.
2.3. Catalytic evaluation
The cycloaddition of CO2 to PO was carried out in 80 mL
stainless steel autoclave equipped with a magnetic stirrer. In
a typical reaction process, 10 mL of PO (0.14 mol), 38 mg of ZnI2
(0.12 mmol), and 0.4 g of the catalyst were added into the
reactor. Then, the reactor waꢀs pressurized with CO2 to a desired
pressure and heated to 140 C under stirring for 6 h. Aer the
reaction, the autoclave was cooled down to the room tempera-
ture in ice water and the excess of CO2 was vented. The liquid
product was separated by centrifugation and analyzed using
a GC equipped with a SE-54 capillary column and FID. The
liquid mixture consisted of PO, PC, and 1,2-propylene glycol
(PG) as the byproduct originating from the hydrolysis of PO with
trace H2O. No other substance was detected. The carbon
balance was nearly 100%. Their quantitative calculation (i.e.
conversion of PO, and selectivity to PC) was based on an area-
normalization method.
Fig. 1 XRD patterns of g-C3N4-T materials prepared without (A) and
with (B) treatment of NaOH under different thermal treatment
temperatures (a: 400; b: 425; c: 450; d: 475; and e: 500 ꢀC).
preparation temperatures of g-C3N4, the complex peaks per-
taining to melem disappeared while pronounced sharp peaks
located at 2q ¼ 27.4ꢀ emerged, together with broad peaks with
low intensity at 2q ¼ 13ꢀ. The primary diffraction peaks were
attributed to characteristic interplanar stacking structures of
graphitic materials,23 i.e. (002) planes with a d-spacing of ca.
0.325 nm. While the other minor peaks related to the in-plane
structural packing motif, such as the hole-to-hole distance of
the nitride pores in the crystal17 (Scheme 1), namely (100) planes.
This means that, under higher temperatures above than 475 ꢀC,
the GndCl precursor transformed into g-C3N4. Additionally, the
values of product mass of g-C3N4 powder (Table S1†), prepared
using the same amount of GndCl, have also witnessed the
further decomposition of the precursor under elevated
temperatures.
In the case of g-C3N4-direct samples prepared using alkali
treatment (Fig. 1B), under the low thermal treatment tempera-
ture of 400 ꢀC, the XRD pattern of g-C3N4-400-NaOH showed
very similar results with that of g-C3N4-400-direct. However,
aer the temperature was elevated slightly, it is of interest that
an intensive peak at 2q ¼ 27.2ꢀ appeared on the pattern of g-
C3N4-425-NaOH, which suggested the presence of graphitic-like
crystalline generated on this material. However, on the other
hand, the pattern of g-C3N4-425-direct still seemed like that of g-
C3N4-400-direct, with no apparent diffraction associated with
graphitic phase. Further raising temperature, the synthesized g-
The PO conversion and selectivity to PC were calculated as
follows:
APC ꢂ fPC þ APG ꢂ fPG
Conv: ¼
Sel: ¼
;
APO þ APC ꢂ fPC þ APG ꢂ fPG
APC ꢂ fPC
;
APC ꢂ fPC þ APG ꢂ fPG
where A, and f were the peak area of GC, and response factor for
each product.
The yield of PC was calculated as
Yield ¼ Conv. ꢂ Sel.
The TOF (mass of synthesized PC per gram catalyst per hour)
for each catalyst was calculated as follows:
mPC
nPO ꢂ Conv:%ðPOÞ ꢂ Sel:%ðPCÞ ꢂ MPC
Wcatal: ꢂ t
TOF ¼
¼
;
Wcatal: ꢂ t
where nPO, MPC, t, and Wcatal. were the molar amount (mmol) of
PO, formula weight (g molꢁ1) of PC, reaction time (h), and the
mass of overall catalyst (mg), respectively.
3. Results and discussions
3.1. Structure characterization
Initially, XRD technique was employed to analyze the crystalline C3N4-NaOH materials revealed major diffraction peak at ca.
structure of g-C3N4 materials prepared using various thermal 27.4ꢀ, resembling the results achieved on g-C3N4-direct speci-
treatment temperatures. As shown in Fig. 1A, when the GndCl mens (Fig. 1A). By this comparison, it can be inferred that the
precursor was calcinated under low temperatures, i.e. 400–425 treatment of NaOH for GndCl has effectively accelerated the
ꢀC, the corresponding XRD patterns of the synthesized materials condensation of melem units, allowing them to smoothly
demonstrated ternary diffraction peaks in the range of 2q ¼ 25 to transform into graphitic layer of g-C3N4 even under relatively
30ꢀ, along with some weak peaks below 2q ¼ 20ꢀ. The patterns thermal treatment temperatures.
with multiple diffraction lines corresponded to the presence of
Fig. 2 depicts FT-IR spectra of g-C3N4 materials. For the g-
melem.40,41 The presence of melem has also been reported when C3N4 materials synthesized under 400–425 C, two insensitive
calcinating dicyandiamide42 and melamine40 as starting mate- peaks at 3342 and 3130 cmꢁ1 have been observed, corre-
rials under moderate temperatures. Upon elevating the sponding to the stretching modes of residual –NH2 attached to
ꢀ
55384 | RSC Adv., 2016, 6, 55382–55392
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
Fig. 3 UV-vis spectra of g-C3N4-450-direct (a), g-C3N4-475-direct
(b), g-C3N4-450-NaOH (c), and g-C3N4-475-NaOH (d) materials.
Fig. 2 FT-IR spectra of g-C3N4-T-direct (A) and g-C3N4-T-NaOH (B)
materials under various thermal treatment temperatures (a: 400; b:
425; c: 450; d: 475; and e: 500 ꢀC).
has reinforced the condensation of g-C3N4 planar packing, in
the sp2 hybridized carbon, and –NH groups at the defect sites of concert with the above XRD and FT-IR characterization results.
the aromatic ring, respectively.43 The dominating multiple
bands in the 1600–1200 cmꢁ1 dominating were attributed to
stretching and rotation vibration of C–N and C]N in hetero-
cycles.42 Also, a sharp peak centered at 807 cmꢁ1 has been found
for the spectrum of each sample, characteristic breathing mode
of the triazine functionalities.42 In addition, there were two the carbon atoms (Ca) connected with –NH2 groups, whereas
weak signals observed at 1016 and 600 cmꢁ1. The two peaks the other resonance at d ¼ 156 ppm was associated with carbon
ought to be due to the uncompleted condensed intermediates of species in heptazine units and linked with bridging N atoms
the GndCl precursor, which have also been previously reported (Cb).45,47 These assignments are in good agreement with those of
The chemical environment of C atoms in g-C3N4 samples was
further investigated by 13C MAS NMR (Fig. 4). For the g-C3N4-
direct samples prepared via direct thermal treatment, there
were two apparent resonances detected for each spectrum. The
peak with much higher intensity at d ¼ 164 ppm was assigned to
by other groups.41,44 Whereas, as the thermal treatment
temperature of GndCl was increased, the acquired FT-IR spectra
showed remarkable variation. Wherein, the bands related to
residual –NH2 groups and uncondensed intermediates became
bulk g-C3N4 reported elsewhere,47 further indicating the g-C3N4-
direct materials were composed by heptazine tectonic units.
Similarly, the spectra of g-C3N4-NaOH materials presented two
signals at the same position. By comparison, the overall inten-
sity of resonance peaks acquired on g-C3N4-NaOH samples was
ꢁ1
ꢀ
weaker above 450 C, while intensity of the band at 807 cm
seemed to be improved and much sharper. Obviously, the much stronger than those obtained in g-C3N4-direct samples,
change resulted from the contribution of high-degree conden- indicating that the former materials possessed a higher degree
sation of melem-like species to wider conjugated graphitic layer
at elevated temperatures. Consequently, the percentage of the
free –NH2 located at the edges of graphitic layers decreased.
The optical properties of the g-C3N4 samples were charac-
terized by means of UV-vis DRS. The representative spectra
obtained over g-C3N4-450 and g-C3N4-475 are given in Fig. 3. It
can be seen that all the materials exhibited strong light
absorption with edges at around 450 nm, which was derived
from bandgap (ca. 2.75 eV) between HOMO and LUMO in the
polymeric melon units of g-C3N4.20 When the temperature was
ꢀ
increased from 450 to 475 C, the absorption edge was moved
towards longer wavelength; the phenomenon has been detected
on samples prepared both with and without alkali treatment.
The red shi originated from the extension of electron delo-
calization in the conjugated layers of g-C3N4 with enhanced
structural condensation.45,46 Compared with those of the pris-
tine g-C3N4-direct samples, the UV-vis spectra of g-C3N4-NaOH
materials showed absorption edges with much longer wave-
lengths. The comparison conrms that, the treatment of NaOH
Fig. 4 Solid-state 13C NMR spectra of g-C3N4-450-direct (a), g-C3N4-
475-direct (b), g-C3N4-450-NaOH (c), and g-C3N4-475-NaOH (d)
materials.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 55382–55392 | 55385

View Article Online
RSC Advances
Paper
of condensation of melon units. Moreover, the intensity ratio of a noticeable improvement subjected to thermal treatment with
peak at 156 ppm to 164 ppm seemed to be obviously enhanced high temperature or alkali treatment.
from g-C3N4-475-direct to g-C3N4-475-NaOH. The nding proves
Since it is widely recognized that there exist several types of N
that, at the same thermal treatment temperatures of GndCl, the species in g-C3N4, in order to probe the detailed information of
aid of NaOH treatment favorably resulted in further condensa- N species of the present g-C3N4 materials, deconvolution of N 1s
tion of melon units in g-C3N4 materials.
spectra of g-C3N4-475-direct and g-C3N4-475-NaOH has been
The bulk chemical composition of g-C3N4 materials was further carried out. As shown in Fig. 5, the N 1s spectra of both
determined by EA measurement. As summarized in Table 1, the the two samples could be separated into three species. The
bulk molar ratios C/N of for the four g-C3N4-direct specimens primary peaks with a binding energy of ca. 398.8 eV were
were in the range of 0.64–0.66, lower than the theoretical value assigned to sp2-hybridized N atoms with a bonding type as
(0.75) of ideal g-C3N4 compound. Not surprisingly, during the C]N–C in heptazine rings.17 These N atoms contributed to
thermal condensation procedures from GndCl to g-C3N4, the a majority of the total N species in g-C3N4-direct and g-C3N4-
melem and/or melon intermediates could not be totally NaOH. The peak located at 400.1 eV were ascribed to sp2
N
condensed, thereby leaving a certain amount of uncondensed atoms bonded to three atoms (i.e. C–N(–C)–C or C–N(–H)–C).49
primary and secondary amino group at the edges of graphitic Such kind of N atoms, especially tertiary N, is a key junction that
layers.48 XPS technique was subsequently applied to study the bridges two or three heptazine rings (Scheme 1). It should be
surface chemical composition of the materials. All the survey also noted that the bridging N atoms have been regarded as the
spectra (Fig. S1†) showed two sharp peaks centered at 288 and active base sites in base-mediated reactions, including CO2-
398 eV, which were assigned to C 1s and N 1s signals, respec- activating reactions50 and transesterication reactions16 since
tively. No apart from the two main elements, a minor peak at they possess higher basic intensity than other N species. The
532 eV could also be observed for each XPS survey, due to the last peak with the highest binding energy corresponded to sp3 N
trace water adsorbed on the surface. It is worth noting that there atoms, i.e. free primary (–NH2) and secondary (–NH–) amino
was no Cl signal (Cl 2s 265–285 eV) detected, indicating that groups at the edges of planar layers of g-C3N4.49 In addition,
there was no residual Cl element derived from GndCl precursor. a broad peak centered at 404.3 eV with a low intensity has also
According to the peak areas of XPS surveys, chemical been detected at each N 1s spectrum (not shown here for
composition for each sample has been calculated (Table 1). The concise), which was generally contributed to the charging effect
C/N molar ratios of on the surface were 0.70–0.75, superior to the of XPS.17
values measured by EA. In our previous work,25 we ascribed the
gap to the considerable but inevitable loss of N species occurring
on the surface of g-C3N4 in the thermal treatment stage. Further
comparing the resultant formulas of the four samples, it can be
found that the detailed C/N molar ratios of both surface and
bulk phase, were very sensitive to the preparation conditions.
That is, using either higher thermal treatment temperature or
alkali treatment could prominently increase the percentage of C
atoms. Unequivocally, as discussed above, the two procedures
enabled further condensation of melem and/or melon frag-
ments by eliminating the terminal amino groups, and eventually
increased the total proportion of C atoms. On the other hand,
owing to the deep condensation, much larger amount of small
molecules (such as NH3) released correspondingly. This is also
responsible for the fact that the surface areas along with pore
volumes (Table S2†) of g-C3N4 materials (Fig. S2† for their cor-
responding N2 adsorption–desorption isotherms) showed
Fig. 5 N 1s spectra of g-C3N4-475-direct (A) and g-C3N4-475-
NaOH (B).
Table 1 Bulk and surface chemical compositions and formulas of g-C3N4 samples
EA
Ca
XPS
Cb
Sample
Na
Ha
Formula
Nb
Ob
Formula
g-C3N4-450-direct
g-C3N4-475-direct
g-C3N4-450-NaOH
g-C3N4-475-NaOH
32.50
33.71
32.45
33.04
58.76
59.20
58.37
58.23
2.607
2.494
2.567
2.402
C0.645NH0.621
C0.664NH0.590
C0.648NH0.616
C0.661NH0.577
40.05
41.49
40.32
41.60
57.75
56.02
56.81
55.12
2.20
2.49
2.87
3.28
C0.694NO0.038
C0.741NO0.044
C0.710NO0.051
C0.755NO0.060
Values expressed by weight percents. b Values expressed by molar percents.
a
55386 | RSC Adv., 2016, 6, 55382–55392
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
According to the areas of deconvoluted peaks, the compo- NaOH remaining in the nal g-C3N4-NaOH materials, it should
nent distributions of various N atoms have been calculated and be noted that, the molar amount of NaOH was 0.025 mol, much
listed in Table 2. The percentages of sp2 N bonded as C]N–C less than the value of GndCl (0.094 mol). According to the
groups achieved on g-C3N4-475-direct and g-C3N4-475-NaOH possible reaction route (GndCl + NaOH / GndOH + NaCl),
were very close, implying that the two samples had similar such NaOH would be totally converted into NaCl. On the other
population of heptazine rings on the surface area. In the cases hand, according to the XRD patterns (Fig. 1) and XPS survey
of percentages of other types of N species, there were remark- (Fig. S1†) of g-C3N4-NaOH materials (e.g. g-C3N4-450-NaOH),
able differences between the two solids. The N atoms bonded to there is no obvious diffraction peaks associated with NaCl (PDF
three atoms accounted for 20.3% on g-C3N4-475-NaOH, #05-0628), or any signal of Na 1s (1086–1072 eV) or Cl 2s (265–
comparatively higher than those of g-C3N4-475-direct, whereas 285 eV), respectively. Therefore, it can be conrmed that NaOH
g-C3N4-475-NaOH owned much less sp3 N atoms than its has been totally reacted with GndCl, and there was no residual
counterpart. By the comparison, it can be veried that, the NaOH existing in the nal g-C3N4-NaOH materials.
melem and melon-like fragments which have abundant
Moreover, if mixing ethanolic solutions of GndCl and KOH,
terminal amino groups, underwent co-condensation of higher white precipitation, namely KCl (PDF #41-1476), was also ob-
degree on g-C3N4-475-NaOH than g-C3N4-475-direct, thus tained (Fig. S3†). Because either NaCl or KCl is hardly dissolved
yielding greater proportion of tertiary bridging N species on g- in ethanol, we surmised that, GndCl and NaOH reacted under
C3N4-475-NaOH.
ethanolic solution, and generated NaCl and guanidine-
Since the above ndings in several characterization tech- hydroxide-like compound (GndOH), as illustrated in Scheme 2.
niques have unanimously conrmed the induction of NaOH On the other hand, as we reported previously, GndCl could be
could nally improve the degree of condensation of g-C3N4, it is transformed into g-C3N4 subjected to four-step poly-
of interest to elucidate the interaction between the GndCl condensation.48 Likewise, GndOH containing guanidine group
precursor and NaOH. Either GndCl or NaOH is easily to be should be smoothly converted into g-C3N4 aer the same
dissolved in ethanol. However, aer the contact of the two thermal treatment procedure as GndCl. Additionally, we think
ethanolic solutions, as expressed in the Experimental section, that the presence of hydroxide anions accelerated the conden-
white precipitation has been clearly observed. By the analysis of sation of melem and melon fragments to the ultimate planar
XRD characterization (Fig. S3†), it was proved that such a white graphitic polymer of g-C3N4. Thereby, it can be well understood
solid was pure NaCl (PDF #05-0628). Regarding the possible why g-C3N4-NaOH sample possessed higher amount of bridging
N atoms yet lower amount of terminal amines than g-C3N4 under
the same thermal treatment temperatures. Indeed, such similar
Table 2 Distributions of various N species for g-C3N4 materialsa
effect of alkali compounds has also been reported in the work
involving the synthesis or posttreatment of g-C3N4 materials.19,51
Percentage of N components (mol%)
3.2. Catalyst activity
401.2 eV
400.1 eV
398.8 eV
The g-C3N4 materials prepared without and with alkali assistant
were applied as catalysts to evaluate their activity in the cyclo-
addition reactions of CO2 to PO (Table 3). Initially, a blank test
without any catalyst was performed and the corresponding
conversion of PO was less than 3% (entry 1). Also, the catalytic
activity acquired under either ZnI2 (entry 2) or bare g-C3N4
Sample
–NH2
C–N(–C)–C
C]N–C
g-C3N4-475-direct
g-C3N4-475-NaOH
9.9
6.0
14.8
20.3
75.3
73.7
a
Determined by N 1s XPS proles.
Scheme 2 A possible mechanism of alkali-assisted synthesis of g-C3N4-MOH.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 55382–55392 | 55387

View Article Online
RSC Advances
Paper
Table 3 Catalytic performances various g-C3N4 samples in cycload-
In order to optimize reaction conditions and thus obtain
high PC productivity, the effects of reaction conditions on the
catalytic performances have been subsequently investigated
using g-C3N4-475-NaOH as a reference catalyst. As depicted in
Fig. 6A, using more amounts of catalyst could effectively bring
out a progressive increase of catalytic activity; however the PO
conversion leveled off when the weight of catalyst was above 0.4
g. Also, the catalytic performance was found to be dependent on
reaction temperature (Fig. 6B). Under the low temperatures
dition reactions of CO2 to POa
Entry
Catalyst
Conv. (%)
Sel. (%)
Yield (%)
1
2
3
4
5
6
7
8
—
ZnI2
2.5
12.8
3.0
58.3
60.6
65.2
79.6
82.1
3.8
95.2
94.9
97.2
99.7
99.8
99.5
99.6
99.6
97.4
99.1
99.4
99.4
2.4
12.1
3.7
58.1
60.5
64.9
79.3
81.8
3.7
b
g-C3N4-450-directc
g-C3N4-450-direct
g-C3N4-475-direct
g-C3N4-500-direct
g-C3N4-525-direct
g-C3N4-550-direct
g-C3N4-450-NaOHc
g-C3N4-450-NaOH
g-C3N4-475-NaOH
g-C3N4-500-NaOH
ꢀ
(100–110 C), the activity was very poor. Meanwhile, the corre-
sponding selectivity was lower than 80%, mainly owing to the
hydrolysis of PO to propylene glycol (PG) in the presence of
residual water in the reactor. Further raising the temperature
resulted in drastic improvements in both PO conversion and PC
selectivity. Although the PC yield could be achieved as ca. 93%
9
10
11
12
79.9
90.0
92.6
79.2
89.5
92.0
a
Reaction conditions: VPO ¼ 10 mL, pCO ¼ 2.0 MPa, T ¼ 140 ꢀC, and t ¼
ꢀ
at 150 ꢀC; however considering the reaction economy, 140 C
2
6 h. Unless specied, each entry was added with 0.4 g of g-C3N4 and 38
mg of ZnI2. b Without g-C3N4. c Without ZnI2.
was selected as the optimal temperature. In addition to catalyst
amount and reaction temperature, the catalytic performances
under various CO2 pressures and reaction time have been also
evaluated. The use of higher CO2 pressures could clearly
improve the catalytic activity (Fig. 6C). Instead, excessive CO2
pressures resulted in a regressive PO conversion. This is prob-
ably because under that harsh circumstance, acidic CO2 might
dissolve in PO and then generate strong CO2–PO complex.52
Additionally, it was found that, aer 6 h of reaction time, the
catalytic activity presented no apparent increase (Fig. S4†).
Besides, recyclability and reproductivity is also an important
issue to examine a heterogeneous catalyst. Taking into account
this point, a series of consecutive catalytic tests have been
carried out over g-C3N4-475-NaOH, on the basis of the above
optimized reaction conditions. As presented in Fig. 7, during
the repetitious ve runs, the selectivities to PC were all higher
than 99%. Simultaneously, it is observed that the catalytic
conversion of PO underwent no apparent decline within the
successive experiments. Furthermore, XRD patterns and FT-IR
characterization results (Fig. S5†) of the spent g-C3N4-475-
NaOH catalysts exhibited the same proles as the fresh one. On
the other hand, for those mesoporous g-C3N4 materials either
loaded with transition metal halides or functionalized with
quaternary amines reported by us,37,38 they all showed apparent
decrease in catalytic activity in the initial several runs, mainly
because they possessed a certain amount of uncondensed
fragments that were liable to leach in liquid phase. Hence, given
the fact, it is proved that the present g-C3N4-NaOH catalysts are
relatively stable heterogeneous catalysts.
As NaOH can be utilized to induce the polymerization of g-
C3N4 under relatively low thermal treatment temperatures. In
this work, we have also tentatively introduced KOH into the
preparation routes of g-C3N4 material, and produced g-C3N4-
475-KOH. The corresponding XRD pattern and FT-IR spectrum
(Fig. S6†) of the sample are in the same as that of g-C3N4-475-
NaOH, with no apparent diffraction lines indexed as KCl
compound. The catalytic performance for the cycloaddition of
CO2 was tested thereaer. As summarized in Table 4, the
synthesized g-C3N4-475-KOH sample exhibited high PO
conversions along with high selectivities to the desired PC.
Besides PO, the catalytic tests using ethylene oxide and
(entries 3 and 9) was very poor. By contrast, aer addition of g-
C3N4 and ZnI2 into the reaction (entry 4), the catalytic produc-
tivity was remarkably improved, affording a PO conversion of
58.3% along with a high selectivity to PC. This means that there
was a potential synergy between g-C3N4 and ZnI2 for the cata-
lytic cycloaddition of CO2 with PO (discussed below).
With elevating the thermal treatment temperatures of GndCl
(entries 4–8), the nal g-C3N4-direct materials exhibited
a monotonous increase in terms of the PO conversions.
Wherein, g-C3N4-550-direct sample offered the highest PC yield
as much as 82% (entry 8). As mentioned above, there exist at
least three types of N species in g-C3N4. Among them, it has
been widely recognized the bridging tertiary amine that
connects the heptazine rings has superior basic intensity to
other free amino groups (i.e. uncondensed primary and
secondary amines located at the graphitic layer of g-C3N4).39,50
During the transformation of GndCl to g-C3N4, high thermal
treatment temperatures would favor the further condensation
of oligomeric melon species by releasing small molecules such
as NH3. Consequently, the amount of tertiary amines has been
improved, thus facilitating the whole activity in activating CO2
molecules. Interestingly, under the same reaction conditions,
the g-C3N4-NaOH materials synthesized under lower tempera-
tures but with the alkali assistant also demonstrated high
catalytic activity (entries 10–12); the PO conversion received
were in the range of 79.9–92.6%. Indeed, if compared based on
the same thermal treatment temperatures, the activity acquired
over g-C3N4-NaOH, for instance g-C3N4-475-NaOH (90.0%), was
much higher than that of its simple counterpart (g-C3N4-475-
direct, 60.6%). As proved in FT-IR and XPS, the aid of NaOH
effectively induced further polymerization of the original
uncondensed melon-like fragments. In this sense, the
threshold temperature for higher polymerized g-C3N4
compound has been reduced; therefore even under low thermal
treatment temperatures, g-C3N4 materials with high amounts of
important basic trigonal
synthesized.
N atoms could be favorably
55388 | RSC Adv., 2016, 6, 55382–55392
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
Fig. 6 Effects of reaction conditions on the catalytic performances of g-C3N4-475-NaOH in cycloaddition reactions of CO2. All the data were
collected at 6 h. Reaction conditions: VPO ¼ 10 mL and 38 mg of ZnI2 added. (A) pCO ¼ 2.0 MPa and T ¼ 140 ꢀC. (B) Wcatal. ¼ 0.4 g and pCO ¼ 2.0
2
2
MPa. (C) Wcatal. ¼ 0.4 g and T ¼ 140 ꢀC.
I). On the other hand, as suggested by Dai et al.54 the amino
groups of g-C3N4 were able to adsorb oxygen-containing small
molecules, such as cycloalkene oxide, via hydrogen bonds.
Therefore, we surmise that PO could also be adsorbed and
activated by the graphitic edges of g-C3N4. Additionally, Zn2+
cation coordinated with oxygen atom of PO and thus triggered
the polarization of C–O bonds (Step II). Next, a carbon atom of
PO was attacked by Iꢁ anion, leading to a ring opening and
generating an iodoalkoxy anion (Step III).37 Aer that, the
Table
4 Catalytic performance of a series of g-C3N4 materials
prepared using various alkali compoundsa
Conv.
(%)
Sel.
(%)
Yield
(%)
Catalyst
Epoxide
Fig. 7 Recycling tests for g-C3N4-475-NaOH in the cycloaddition
reactions of CO2. Reaction conditions: VPO ¼ 10 mL, pCO ¼ 2.0 MPa, T
2
¼ 140 ꢀC, t ¼ 6 h, Wcatal. ¼ 0.4 g and 38 mg of ZnI2 added.
g-C3N4-475-NaOH
90.0
91.7
99.4
99.1
89.5
90.9
chloropropylene oxide as substrates also demonstrated high-
performance catalytic results under identical reaction condi-
tions. Therein, it can be inferred that the present alkali-assisted
synthetic approach is a versatile and practical strategy for the
fabrication of high-performance g-C3N4 materials for the cata-
lytic synthesis of cyclic carbonates from cycloaddition of CO2 to
epoxides.
g-C3N4-475-KOH
g-C3N4-475-NaOH
g-C3N4-475-NaOH
92.3
86.4
99.5
99.2
91.8
85.7
According to the characterization and catalytic results in this
study as well as other related work,52,53 a possible mechanism
for cycloaddition of CO2 catalyzed by g-C3N4 was proposed, as
described in Scheme 3. Compared with the uncondensed
terminal amino groups at the edges of graphitic layers of g-
C3N4, bridging tertiary N atoms possessed higher intensity of
basicity. Therefore, for an acidic CO2 molecule, it is preferen-
tially adsorbed and then activated by the tertiary N atoms (Step
g-C3N4-475-NaOH
83.1
99.0
82.3
a
Reaction conditions: Vepoxide ¼ 10 mL, pCO ¼ 2.0 MPa, T ¼ 140 ꢀC, t ¼
6 h, Wcatal. ¼ 0.4 g, and 38 mg of ZnI2 adde2d.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 55382–55392 | 55389

View Article Online
RSC Advances
Paper
g-C3N4 materials. Given this point, we summarize the catalytic
results of various g-C3N4 materials in the cycloaddition of CO2
with PO in Table 5. To the best of our knowledge, all the re-
ported g-C3N4-based catalysts have been covered. Because the
reaction conditions (including catalyst amount, reaction time,
and PO feeding volume) deviate from each other whereas the
reaction temperatures adopted are ca. 140 ꢀC, for a fair
comparison, TOF value (based on g-C3N4 component) has been
calculated for each entry. As listed in Table 5, the most efficient
one is g-C3N4 catalyst supported on mesoporous SBA-15 (entry
2), followed by mesoporous g-C3N4 materials functionalized by
quaternary ammonium (entry 3) and loaded with ZnBr2 (entry
4). It is worth noting that such materials (also including entry 1)
were prepared using mesoporous silicas as catalytic supports or
sacricial templates, which demanded time-consuming and
complicated procedures. Also, the so templates (i.e. pluronic
surfactant) applied for the mesoporous silicas are relatively
expensive. In sharp contrast, the present g-C3N4 catalysts were
prepared via a much more convenient way, however also
showing pretty good catalytic activities. Furthermore, the
feeding amount of PO is 10 mL (entries 6–7), comparatively
higher than other values. In addition to g-C3N4-based materials,
the catalytic performances of some heterogeneous catalysts
have also been provided in Table 5. Entries 8–11 are typical ionic
liquids catalyst immobilized on various solids, including gra-
phene oxide sheets, silica gel, and cross-linked polymer. The
active sites of these catalysts were solely halide anions, and the
TOF values acquired them were in the range of 2 to 8.8 hꢁ1
(note: the catalyst amount for each entry was based on the total
catalyst). As noted above, the preparation of such graed IL
catalyst was very complicated and also demanded expensive
silicane-type coupling agents. Also the activity of MFI-type
zeolite containing Iꢁ is also lower than that of g-C3N4-475-
Scheme 3 A possible reaction pathway for cycloaddition of CO2
catalyzed by g-C3N4.
iodoalkoxy anion reacted with the activated CO2 molecule on
the surface of g-C3N4, yielding a linear iodocarbonate anion
(Step IV). Finally, the anion intermediate transformed into PC
through a ring closing process (Step V). In a short summary, the
role of ZnI2 was solely to activate epoxide by inducing its ring
opening, while g-C3N4 took in charge of adsorbing and then
activating CO2 by N-containing basic sites. With this hypothetic
mechanism, it can be well explained why with ZnI2 or g-C3N4
alone, the corresponding catalytic activity was limited. Once
again, because g-C3N4-NaOH and g-C3N4-KOH owned more
proportion of crucial bridging N species than g-C3N4-direct, the
former naturally offered higher catalytic activity than the latter.
As stated above, owing to their abundant N-containing
species, g-C3N4 materials have been applied as catalysts in
many CO2-activating reactions. Therefore, it would be of
interest to compare the catalytic performances offered by the
present g-C3N4 samples with those of previously reported
Table 5 Comparison of the catalytic performances of various g-C3N4-based materials and other heterogeneous catalysts in cycloaddition of
CO2 with PO
Entry
Catalyst
T (ꢀC)
t (h)
VPO (mL)
Wcatal. (mg)
Yield (%)
TOFk (hꢁ1
)
1a24
2b39
3c38
4d37
5e50
6
MS-MCN
140
150
140
140
130
140
140
140
100
140
130
140
10
1.5
6
6
4
6
6
4
5
1.5
3
10
7
1.5
10
10
15
0.7
2
20
100
200
200
50
400
400
600
100
200
200
100
30.6
96.1
87.7
98.9
23.7
89.5
90.9
96
96.7
93.2
96.0
97.1
3.3
28.1
10.7
8.4
2.6
5.4
5.5
8.8
2.0
2.7
4.1
2.5
g-C3N4/SBA-15
n-ButBr/mp-C3N4
ZnBr2/mp-C3N4
u-g-C3N4-480
g-C3N4-475-NaOH
g-C3N4-475-KOH
GO-[SmIm]I
PDDA-Br-ZnBr2-2/SiO2
P-DVB-HEImBr
PS-DHPIMBr
7
8f15
9g55
10h5
11i6
12j56
5
3
4
1.75
0.7
LMFI-I
a
b
Prepared using disk-shaped 2D hexagonal mesoporous silica as a hard template and melamine as a precursor. Prepared using SBA-15 as
a catalytic support and dicyandiamide as a precursor through a chemical-vapor-deposition method. Zn2+ was further doped into g-C3N4/SBA-15
as an additive. Prepared using silica nanoparticles as hard templates and cyanamide as a precursor. The synthesized mp-C3N4 was graed
c
d
e
with n-bromobutane. ZnBr2 supported on mp-C3N4 material. Prepared using urea as a starting material without addition of any template.
f
g
Graphene oxide sheets immobilized with functionalized imidazolium ionic liquid with iodide ion. Polymeric quaternary ammonium salt
h
i
supported on silica gel. Cross-linked polymer graed with functionalized ionic liquid with bromide ion. Polystyrene supported with diol
functionalized ionic liquids. Organic–inorganic hybrid zeolites with the MFI-type lamellar structure. The detailed calculation method for TOF
j
k
was expressed in the Experimental section.
55390 | RSC Adv., 2016, 6, 55382–55392
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
MOH catalyst under the same reaction conditions. Hence, in 11 L. Han, H. Li, S.-J. Choi, M.-S. Park, S.-M. Lee, Y.-J. Kim and
viewpoint of both catalyst preparation and catalytic activity, the D.-W. Park, Appl. Catal., A, 2012, 429, 67–72.
present g-C3N4 materials could serve potential heterogeneous 12 X. Zhang, D. Wang, N. Zhao, A. S. Al-Ari, T. Aouak, Z. A. Al-
catalysts for efficient cycloaddition of CO2.
Othman, W. Wei and Y. Sun, Catal. Commun., 2009, 11, 43–
46.
13 W. Cheng, X. Chen, J. Sun, J. Wang and S. Zhang, Catal.
Today, 2013, 200, 117–124.
4. Conclusion
14 S. Udayakumar, M.-K. Lee, H.-L. Shim, S.-W. Park and
D.-W. Park, Catal. Commun., 2009, 10, 659–664.
15 J. Xu, M. Xu, J. Wu, H. Wu, W.-H. Zhang and Y.-X. Li, RSC
Adv., 2015, 5, 72361–72368.
16 J. Xu, K.-Z. Long, Y. Wang, B. Xue and Y.-X. Li, Appl. Catal., A,
2015, 496, 1–8.
In summary, we have developed a facile and fast strategy to
fabricate g-C3N4 materials. The using of NaOH or KOH could
effectively reduce the threshold temperatures of the formation
of g-C3N4, enabling GndCl to be smoothly converted into g-C3N4
at 450–475 ꢀC. Due to the condensation with a higher level, the
g-C3N4-NaOH materials had much more bridging tertiary N
atoms than g-C3N4-direct samples. In the cycloaddition of CO2
to PO, both g-C3N4-NaOH and g-C3N4-KOH samples showed
good catalytic performances, higher than the results received
over g-C3N4-direct sample under the same reaction conditions.
Meanwhile, the g-C3N4-NaOH can be reused for at least ve
times without any loss of catalytic activity. Compared with other
mesoporous and supported g-C3N4 materials, the present g-
C3N4 samples were synthesized with an easier method, while
manifesting a promising catalytic application for the cycload-
dition of CO2 to PC.
17 A. Thomas, A. Fischer, F. Goettmann, M. Antonietti,
¨
¨
J.-O. Muller, R. Schlogl and J. M. Carlsson, J. Mater. Chem.,
2008, 18, 4893–4908.
18 Y. Gong, M. Li, H. Li and Y. Wang, Green Chem., 2015, 17,
715–736.
19 J. Zhu, P. Xiao, H. Li and S. A. C. Carabineiro, ACS Appl.
Mater. Interfaces, 2014, 6, 16449–16465.
20 X. Wang, K. Maeda, A. Thomas and K. Takanabe, Nat. Mater.,
2009, 8, 76–80.
21 F. Su, S. C. Mathew, G. Lipner, X. Fu, M. Antonietti,
S. Blechert and X. Wang, J. Am. Chem. Soc., 2010, 132,
16299–16301.
Acknowledgements
22 S. Cao, J. Low, J. Yu and M. Jaroniec, Adv. Mater., 2015, 27,
2150–2176.
23 F. Goettmann, A. Fischer, M. Antonietti and A. Thomas,
Angew. Chem., Int. Ed., 2006, 45, 4467–4471.
24 M. B. Ansari, B.-H. Min, Y.-H. Mo and S.-E. Park, Green
Chem., 2011, 13, 1416–1421.
This work was supported by National Natural Science Founda-
tion of China (21203014 and 21376032), Jiangsu Key Laboratory
of Advanced Catalytic Materials and Technology (BM2012110),
and the Project Funded by the Priority Academic Program
Development of Jiangsu Higher Education Institutions.
25 J. Xu, T. Chen, Q. Jiang and Y.-X. Li, Chem.–Asian J., 2014, 9,
3269–3277.
26 Y. Zheng, J. Liu, J. Liang, M. Jaroniec and S. Z. Qiao, Energy
Environ. Sci., 2012, 5, 6717–6731.
Notes and references
1 J. Roeser, K. Kailasam and A. Thomas, ChemSusChem, 2012, 27 S. S. Park, S.-W. Chu, C. Xue, D. Zhao and C.-S. Ha, J. Mater.
5, 1793–1799.
Chem., 2011, 21, 10801–10807.
2 M. North, R. Pasquale and C. Young, Green Chem., 2010, 12, 28 Q. Li, J. Yang, D. Feng, Z. Wu, Q. Wu, S. S. Park, C.-S. Ha and
1514–1539. D. Zhao, Nano Res., 2010, 3, 632–642.
3 J. Ma, N. Sun, X. Zhang, N. Zhao, F. Xiao, W. Wei and Y. Sun, 29 F. Su, M. Antonietti and X. Wang, Catal.: Sci. Technol., 2012,
Catal. Today, 2009, 148, 221–231. 2, 1005–1009.
4 W.-L. Dai, S.-L. Luo, S.-F. Yin and C.-T. Au, Appl. Catal., A, 30 S. N. Talapaneni, S. Anandan, G. P. Mane, C. Anand,
2009, 366, 2–12.
5 W.-L. Dai, B. Jin, S.-L. Luo, S.-F. Yin, X.-B. Luo and C.-T. Au, J.
CO2 Util., 2013, 3–4, 7–13.
6 R. A. Watile, K. M. Deshmukh, K. P. Dhake and
B. M. Bhanage, Catal.: Sci. Technol., 2012, 2, 1051–1055.
7 A. Ion, V. Parvulescu, P. Jacobs and D. De Vos, Appl. Catal., A,
2009, 363, 40–44.
D. S. Dhawale, S. Varghese, A. Mano, T. Mori and A. Vinu,
J. Mater. Chem., 2012, 22, 9831–9840.
31 J. Xu, K. Shen, B. Xue and Y.-X. Li, J. Mol. Catal. A: Chem.,
2013, 372, 105–113.
32 M. B. Ansari, H. Jin, M. N. Parvin and S.-E. Park, Catal.
Today, 2012, 185, 211–216.
33 J. Xu, T. Chen, X. Wang, B. Xue and Y.-X. Li, Catal.: Sci.
Technol., 2014, 4, 2126–2133.
8 J.-Q. Wang, J. Sun, W.-G. Cheng, C.-Y. Shi, K. Dong,
X.-P. Zhang and S.-J. Zhang, Catal.: Sci. Technol., 2012, 2, 34 X. Jin, V. V. Balasubramanian, S. T. Selvan, D. P. Sawant,
600–605.
M. A. Chari, G. Q. Lu and A. Vinu, Angew. Chem., Int. Ed.,
2009, 48, 7884–7887.
35 F. Goettmann, A. Thomas and M. Antonietti, Angew. Chem.,
Int. Ed., 2007, 46, 2717–2720.
9 X.-Y. Liu, L.-B. Sun, X.-D. Liu, A.-G. Li, F. Lu and X.-Q. Liu,
ACS Appl. Mater. Interfaces, 2013, 5, 9823–9829.
10 M. H. Anthofer, M. E. Wilhelm, M. Cokoja, I. I. Markovits,
¨
¨
A. Pothig, J. Mink, W. A. Herrmann and F. E. Kuhn, Catal.: 36 D.-H. Lan, H.-T. Wang, L. Chen, C.-T. Au and S.-F. Yin,
Sci. Technol., 2014, 4, 1749–1758.
Carbon, 2016, 100, 81–89.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 55382–55392 | 55391

View Article Online
RSC Advances
Paper
37 J. Xu, F. Wu, Q. Jiang, J.-K. Shang and Y.-X. Li, J. Mol. Catal. A: 47 Y. Cui, Z. Ding, X. Fu and X. Wang, Angew. Chem., Int. Ed.,
Chem., 2015, 403, 77–83. 2012, 51, 11814–11818.
38 J. Xu, F. Wu, Q. Jiang and Y.-X. Li, Catal.: Sci. Technol., 2015, 48 J. Xu, H.-T. Wu, X. Wang, B. Xue, Y.-X. Li and Y. Cao, Phys.
5, 447–454. Chem. Chem. Phys., 2013, 15, 4510–4517.
39 Z. Huang, F. Li, B. Chen, T. Lu, Y. Yuan and G. Yuan, Appl. 49 F. Dong, L. Wu, Y. Sun, M. Fu, Z. Wu and S. C. Lee, J. Mater.
Catal., B, 2013, 136, 269–277. Chem., 2011, 21, 15171–15174.
40 S. Chu, C. Wang, J. Feng, Y. Wang and Z. Zou, Int. J. Hydrogen 50 Q. Su, J. Sun, J. Wang, Z. Yang, W. Cheng and S. Zhang,
Energy, 2014, 39, 13519–13526. Catal.: Sci. Technol., 2014, 4, 1556–1562.
41 L. S. Rangel, J. R. de la Rosa, C. J. L. Ortiz and M. J. Castaldi, 51 X. Bai, S. Yan, J. Wang, L. Wang, W. Jiang, S. Wu, C. Sun and
J. Anal. Appl. Pyrolysis, 2015, 113, 564–574. Y. Zhu, J. Mater. Chem. A, 2014, 2, 17521–17529.
42 M. J. Bojdys, J.-O. Muller, M. Antonietti and A. Thomas, 52 L.-F. Xiao, F.-W. Li, J.-J. Peng and C.-G. Xia, J. Mol. Catal. A:
Chem.–Eur. J., 2008, 14, 8177–8182. Chem., 2006, 253, 265–269.
43 E. Z. Lee, Y.-S. Jun, W. H. Hong, A. Thomas and M. M. Jin, 53 J. Sun, L. Wang, S. Zhang, Z. Li, X. Zhang, W. Dai and
Angew. Chem., Int. Ed., 2010, 49, 9706–9710. R. Mori, J. Mol. Catal. A: Chem., 2006, 256, 295–300.
44 L. Shi, L. Liang, F. Wang, J. Ma and J. Sun, Catal.: Sci. 54 J. Ding, Q. Liu, Z. Zhang, X. Liu, J. Zhao, S. Cheng, B. Zong
Technol., 2014, 4, 3235–3243. and W.-L. Dai, Appl. Catal., B, 2015, 165, 511–518.
45 Q. Gu, Y. Liao, L. Yin, J. Long, X. Wang and C. Xue, Appl. 55 B. Song, L. Guo, R. Zhang, X. Zhao, H. Gan, C. Chen, J. Chen,
Catal., B, 2015, 165, 503–510. W. Zhu and Z. Hou, J. CO2 Util., 2014, 6, 62–68.
46 J. Xu, Y. Li, S. Peng, G. Lu and S. Li, Phys. Chem. Chem. Phys., 56 C.-G. Li, L. Xu, P. Wu, H. Wu and M. He, Chem. Commun.,
¨
2013, 15, 7657–7665.
2014, 50, 15764–15767.
55392 | RSC Adv., 2016, 6, 55382–55392
This journal is © The Royal Society of Chemistry 2016