Boron-doped melamine-derived carbon nitrides tailored by ionic liquids for catalytic conversion of CO2 into cyclic carbonates

Article abstract of DOI:10.1039/c7gc00279c

A new approach to tailoring edge active groups of graphitic carbon nitrides for catalytic conversion of CO₂ into cyclic carbonates was proposed in this work. To improve the catalytic performance, boron-doped melamine-derived graphitic carbon nitrides (MCNB) with numerous exposed edge defects were prepared by using 1-butyl-3-methylimidazolium tetrafluoroborate (BmimBF₄) as the soft template and boron source. Different mass ratios of BmimBF₄ to melamine were explored for MCNB preparation, MCNB(x) (x is the mass ratio) with different polymerization degree, pore structure and boron doping content were obtained, and the relationship between MCNB properties and the corresponding catalytic activity was then investigated. With low polymerization degree, abundant meso-macroporous structure and small amounts of boron (<1.59 atm%) doped in the skeleton, MCNB(0.01) exhibited better catalytic performance and could be suitable for various epoxide substrates with the yield of cyclic carbonates up to 89.0% at 130°C in 6 h. According to the XPS analysis and DFT calculation results, the active centers were confirmed to be the partially-condensed amino groups in edge defects, which were enhanced by moderate doping of boron in the skeleton.

Full text of DOI:10.1039/c7gc00279c

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: Q. Su, X. Yao, W.
Cheng and S. Zhang, Green Chem., 2017, DOI: 10.1039/C7GC00279C.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
7 7 April 2016 Pages 1821–2242
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Accepted Manuscripts are published online shortly after
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Please do not adjust margins
Page 1 of 10
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00279C
Green Chemistry
PAPER
Boron-doped Melamine-derived Carbon Nitrides Tailored by Ionic
Liquids for Catalytic Conversion of CO2 into Cyclic Carbonates
Qian Su,a,b Xiaoqian Yao, a Weiguo Cheng*a and Suojiang Zhang*a
Received 00th January 20xx,
Accepted 00th January 20xx
A new approach to tailoring edge active groups of graphitic carbon nitrides for catalytic conversion of CO2 into cyclic
carbonates was proposed in this work. To improve the catalytic performance, boron-doped melamine-derived graphitic
carbon nitrides (MCNB) with numerous exposed edge defects were prepared by using 1-butyl-3-methylimidazolium
tetrafluoroborate (BmimBF4) as the soft template and boron source. Different mass ratios of BmimBF4 to melamine were
explored for MCNB preparation, MCNB(x) (x is the mass ratio) with different polymerization degree, pore structures and
boron doping contents were obtained, and the relationship between MCNB properties and corresponding catalytic
activities was then investigated. With low polymerization degree, abundant meso-macroporous structure and small
amounts of boron (<1.59 atm%) doped in the skeleton, MCNB(0.01) exhibited better catalytic performance and could be
DOI: 10.1039/x0xx00000x
www.rsc.org/
o
suitable for various epoxide substrates with the yield of cyclic carbonates up to 89.0% at 130 C in 6 h. According to the
XPS analysis and DFT calculation results, the active centers were confirmed to be the partially-condensed amino groups in
edge defects which were enhanced by moderately doping of boron in the skeleton.
catalysts due to the combination of thermal stability and
adjustable edge active groups in the structure.2, 13-15 In the
respect of CO2 conversion, different methods have been
Introduction
CO2 conversion into high-value products is of great significance
in easing energy and environmental concerns. As an efficient
and atom-economy way for CO2 transformation, the
cycloaddition with epoxides into cyclic carbonates has
attracted extensive attention in both academia and industry,
of which the products could be widely used in production of
electrolytes, glycols, polar aprotic solvents, pharmaceutical
ingredients and fine chemicals in biomedical synthesis.1, 2 In
recent years, the development of the heterogeneous catalysts
has attracted great attention in the literature, such as metal-
based catalysts,3 modified molecular sieve,4-6 supported ionic
explored to enhance the activity of g-C3N4, including increase
of surface area and dispersity,16-19 exposure of edge defects,17,
20, 21 immobilization of functional groups,22 and doping of other
atoms.23, 24 According to these reports, the partially-condensed
amino groups in edge defects of g-C3N4 could function directly
or indirectly for the improvement in catalytic activity.
Therefore, the key point in developing efficient g-C3N4
catalysts is digging more edge defects, keeping integrality of
active sites, and enhancing the active groups. To create more
edge defects and keep their integrality, non-template way has
been reported for urea-derived carbon nitrides adjustion by
changing polymerization temperatures.18, 22 To enhance edge
active groups, the introduction of new active components and
doping of other atoms were respectively investigated.22-24 Up
to date, there is few approach to achieving the above three
purposes simultaneously for CO2 activation.
In recent years, ionic liquids have been used as soft
templates in material preparation owning to the unique
properties of low melting points, negligible vapor pressures
and high decomposition temperatures (some excess 400 oC).25-
27 In carbon nitride synthesis, ionic liquids not only show great
advantages as soft templates for the precursors with high
polymerization temperatures, but also could be used as
dopants for doping other atoms. Zhang et al. found that the
electronic structures of carbon nitrides as well as their surface
properties could be modified by doping phosphorus with 1-
butyl-3-methylimidazolium hexafluorophosphate (BmimPF6) as
phosphorus source.28 Wang et al. successfully synthesized
8
liquids,7, and poly(ionic liquids),9 etc. Though they have
contributed more or less to the catalytic performance, some
existing disadvantages including metal residue and low
thermal stability still perplex the academia and industry. The
stability of catalysts, and the mild and environmental-friendly
reaction conditions have always drawn great attention of
researchers in fields of catalysis and chemical engineering.10-12
The metal-free graphitic carbon nitride (g-C3N4) materials
could effectively make up for the weakness of traditional
a. Key Laboratory of Green Process and Engineering, State Key Laboratory of
Multiphase Complex Systems, Beijing Key Laboratory of Ionic Liquids Clean Process,
Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190,
China. E-mail: wgcheng@home.ipe.ac.cn, sjzhang@home.ipe.ac.cn;
Fax: +86 10 8262 7080; Tel: +86 10 8262 7080
b. University of Chinese Academy of Sciences, Beijing 100049, China
Electronic Supplementary Information (ESI) available: [details of any
supplementary information available should be included here]. See
DOI: 10.1039/x0xx00000x
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 2 of 10
ARTICLE
Journal Name
edge active groups of different MCNB materials were studied.
View Article Online
Then, the influence of these proDpOeIr:t1i0e.s1039o/nC7GcCa0t0a2ly79tiCc
performance for CO2 conversion into cyclic carbonates was
also investigated. Finally, the active centers in edge defects
were figured out and the possible reaction mechanism in
presence of MCNB catalysts was preliminarily proposed.
Results and Discussion
Characterization of MCNB materials
With BmimBF4 as the soft template and boron source, a series
of MCNB materials with different B/M mass ratios (x=0.1, 0.05,
0.03, 0.01) and melamine-derived carbon nitride without
addtion of BmimBF4 (MCN) were prepared. The polymerization
degrees, pore structures and edge active groups of all the as-
prepared MCNB materials were analyzed and detected with
different characterization methods.
First of all, TGA curves of all MCNB materials were obtained.
From the curves shown in Fig. 1, all MCNB materials had high
o
thermal stability above 500 C and they all owned one weight
loss peak which indicated that all the materials might be pure
phase without other unstable impurities. FT-IR spectra and
XRD patterns for all the MCNB samples were shown in Fig. 2
and Fig. 3, respectively. All the characteristic peaks of g-C3N4
phases were detected in FT-IR spectra. The adsorption bands
at 810 cm−1 and 1200–1650 cm−1 could be attributed to the
breathing mode of triazine and the stretching mode of C–N
heterocycles, respectively.31 The broad band at 3000–3700
cm−1 belongs to N–H vibration due to partial condensation and
the adsorbed water molecules.32, 33 The typical vibration of B–
N at 1370 cm-1 is presumably overlapped by that of the C–N
stretch.29 With the increase of B/M mass ratios, both of the
peaks at 810 cm−1 and in the region of 1240–1650 cm−1 became
weaker, illustrating the decrease in polymerization degree of
triazine repeating units.34 Compared with MCN, the peaks
around 3000–3700 cm−1 got more distinct for MCNB(0.01) to
MCNB(0.05), meaning more edge –NH groups were exposed
Scheme 1 Polymerization pathways to graphitic carbon nitrides from
melamine without (left) and with ionic liquids (right).
boron- and fluorine-containing mesoporous carbon nitride
polymers for cyclohexane oxidation by using 1-butyl-3-
methylimidazolium tetrafluoroborate (BmimBF4) as the soft
template.27
Considering the two functions of ionic liquids in modifying
materials, they would be promising in tailoring carbon nitrides
for CO2 conversion both as the soft template and heteroatom
source. From previous research, boron could enter the C–N
condensation scheme replacing carbon as strong Lewis acid
sites,29 which might affect the charge distribution and bond
-
length of adjacent groups. BF4 was thought to be a good
dopant for carbon nitride modification but fluorine was
inevitably mixed in when using dicyandiamide as the precursor,
which was thought to be unfriendly to environment. To obtain
boron-doping and fluorine-free carbon nitrides in this work,
melamine with the structure close to carbon nitride unit was
adopted as the precursor, and BmimBF4 was used as the soft
template and boron source. Compared with more dense
structure in non-template polymerization process of melamine,
which goes more directly to melem with less porosity and edge
defects (left in Scheme 1),30 that structures with ionic liquids as
the soft template could be adjusted for more porosity and
edge defects (right in Scheme 1). By changing mass ratios of
BmimBF4 to melamine (B/M), a series of boron-doped
melamine-derived carbon nitrides, abbreviated as MCNB(x) (x
is the mass ratio), were obtained. First of all, the effect of B/M
ratios on the polymerization degrees, pore structures, and
Fig. 1 TGA curves of different MCNB materials prepared with different B/M
ratios.
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 10
Green Chemistry
Please do not adjust margins
Journal Name
ARTICLE
crystalline structure of g-C3N4 was less damagedVbieyw Aerxticclee sOsnilvinee
BmimBF4 addition though the (100)DOpIe: 1a0k.10o39f/Ci7nGtCra0-0l2a7y9eCr
crystalline structure was barely distinguished. The less effect
on MCNB crystallinity with excess BmimBF4 addition might be
ascribed to self-aggregation of ionic liquids in melamine
solution during pre-treatment which caused more ionic liquids
condensing alone rather than acting as template during
calcination.39-41 Thus, small amounts of BmimBF4 addition were
more suitable to adjust the crystallinity for exposure and
access of edge groups, but too much (B/M ratio above 0.1)
might have little effect.
Thereafter, the morphology and pore structures were
respectively characterized by SEM and TEM. The obtained
MCN and MCNB images were respectively shown in Fig. 4 and
Fig. 5. All the materials exhibited porous platelet-like structure,
consistent with previous reports about g-C3N4.42 Different from
dense layer structure of MCN in Fig. 4, MCNB materials were
more bouffant with expanded space between layers (Fig. 4),
which agreed well with XRD analysis. The space expansion
might result from boron doping in g-C3N4 skeletons. Apart
from that, all the MCNB samples presented meso-
macroporous structures, which could be further verified
through N2 adsorption-and-desorption analysis. Both the
isotherms and pore size distribution curves of all MCNB
samples were displayed in Fig. S1. The isotherms were type IV
with H3 hysteresis loops, indicating the presence of mesopores
(mesopores have widths between 2 and 50 nm).43 Besides, the
pore-size distribution curves (inset in Fig. S1) of all samples
were very wide around pore diameters of 50 nm, indicating
the co-existence of meso- and macropores.43 With B/M ratios
between 0.01 and 0.05, the proportion of meso-/macropores
(pore diameter around 50 nm) in MCNB was higher than that
with B/M ratio of 0.1, suggesting that bigger pores tended to
occur with less BmimBF4 addition. This trend could further be
proved in Table 1 that both the BET surface area and BJH
Desorption average pore diameters decreased with more
BmimBF4 addition and MCNB(0.01) owned the highest BET
surface area and average pore diameter. Otherwise, with B/M
mass ratio in excess of 0.1, a sudden increase in BET surface
area was observed, which might be attributed to extra self-
condensed products out of g-C3N4 pore structures from ionic
Fig. 2 FT-IR spectra of different MCNB materials prepared with
different B/M ratios.
Fig. 3 XRD patterns of different MCNB materials prepared with different
B/M ratios.
after BmimBF4 addition.
In addition, XRD patterns of all the as-prepared MCNB
materials were recorded in Fig. 3. All materials owned two
typical diffraction peaks at around 27.5° and 13.1°,
demonstrating the formation of g-C3N4 phases. The peak at
around 27.5° could be indexed to the (002) crystal plane of g-
44-47
liquids aggregates pre-formed in melamine solution.40,
C3N4 due to the interlayer stacking of melon networks of g-
36
C3N4,35,
while a minor peak at around 13.1°could be
attributed to the (100) crystal plane of g-C3N4 due to the
repetitive in-plane tri-s-triazine units.37 Compared with MCN,
there was a slight broadening trend for all the characteristic
peaks and a slight shift to lower angle for (002) peak after
BmimBF4 addition, indicating the decreased polymerization
degree and the expanded interlayer space after boron
doping.27 Among MCNB samples, the shift for (002) peak of
MCNB(0.01) was more obvious than others, meaning the
larger space distance between layers. The lower
polymerization degree and larger space distance indicated that
more edge defects were easier to reach.38 Whereas, when the
B/M mass ratio was raised to 0.10, the (002) peak intensity at
Fig. 4 TEM (a) and SEM (b) images of melamine-derived carbon nitride
(MCN) without BmimBF4 addition.
27.5° got
a sudden increase, indicating the inter-layer
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 4 of 10
ARTICLE
Journal Name
N1s and B1s were presented in Fig. 6. As seenViferwoAmrticFleiOg.nli6ne,
DOI: 10.1039/C7GC00279C
there existed three kinds of nitrogen components in MCNB
structures, including the uncondensed primary nitrogen
groups N–H2 (400.8–401.1 eV), condensed bridge groups C–
N(–C)–C or partially-condensed C–N(–H)–C (399.1–399.5 eV)
and trazine skeleton nitrogen C
＝N–C (398.3–398.6 eV), which
was consistent with previous reports about g-C3N4 edge
groups.48 Beyond that, B1s spectra confirmed that boron was
doped with the binding energy (BE) peak at 191.5 eV, which
was reported to be HN
＝B(–N)2 structure resulting from boron
replacement for the original carbon dot.29 Further evidence of
the boron doping came from the 11B solid-state MAS NMR
measurements. As 11B solid-state MAS NMR displayed in Fig.
S3, one sharp peak at -11.69 ppm was observed in MCNB(0.01)
and MCNB(0.05), which was more likely corresponded to the
bay boron position in the framework structure.29 In addition,
the relative percentages of nitrogen components in different
MCNB networks were calculated from corresponding XPS
spectra. Both of the percentages of amino groups and boron
contents were listed in Table 2. Table 2 showed that the sum
of uncondensed N–H2 and partially-condensed C–N(–H)–C (or
C–N(–C)–C) percentages increased after BmimBF4 addition,
further confirming the decreased polymerization degree. With
B/M ratios raised from 0.01 to 0.05, boron content got higher,
while the percentage of uncondensed nitrogen groups became
Fig. 5 SEM and TEM images of MCNB materials prepared with different B/M
mass ratios.
Table 1 BET surface area, BJH Desorption average pore diameters and C/H
mol ratios of MCNB materials
BET surface area
(m2/g)a
Average pore
C/H mol
ratioc
Materials
diameter (nm)b
MCN
10.5
11.1
9.9
7.8
10.3
17.8
28.3
23.4
21.6
17.4
1.398
1.193
1.173
1.164
1.116
MCNB(0.01)
MCNB(0.03)
MCNB(0.05)
MCNB(0.1)
a BET surface area was obtained from Brunauer–Emmett–Teller (BET)
method; b Average pore diameters were obtained from Barrett–Joyner–
Halenda (BJH) approach; C C/H mole ratios were obtained from elemental
analysis.
These self-condensed structures of MCNB(0.1) might account
for the higher BET surface area than others. That also
explained the XRD results that the crystallinity of MCNB was
less affected at B/M ratio of 0.1 compared with others.
Furthermore, the C/H mole ratios were also obtained by
elemental analysis (Table 1). With addition of BmimBF4, the
C/H mole ratio was reduced from 1.398 to 1.116, indicating
that more hydrogen retained because of the lower
polymerization degree than MCN. Among MCNB samples, the
declining trend of C/H mole ratio while raising B/M ratios
might result from the decreased carbon amounts in g-C3N4
skeleton because of the boron replacement.29 To further
explore the active groups in MCNB structure. XPS analysis of
both MCN and MCNB materials were conducted for
comparison. First of all, no peak was detected from F1s
spectrum in Fig. S2, indicating that fluorine atom was scarcely
doped in MCNB structure in this work. Hence, the following
discussion is mainly about N- and B-containing groups. Three
representative carbon nitrides, MCN, MCNB(0.01) and
MCNB(0.05), were chosen for comparison. The XPS spectra of
Fig. 6 XPS spectra of different MCNB materials prepared with different B/M
ratios.
Table
2 Percentages of different nitrogen species and boron contents
calculated from XPS analysis
Relative percentage of N components
B content
(atm%)b
(atm%)a
Samples
MCN
N–H2
C–N(–C)–C or C–N(–H)–C
C＝N–C HN＝B(–N)2
12.7
35.6
62.9
51.7
24.0
0.32
1.06
MCNB(0.01) 13.1
MCNB(0.05) 7.2
50.0
42.8
1.59
a The percentage was calculated according to the relative magnitude of peak
area in N1s spectra from Fig. 6. b B content was determined by XPS analysis.
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 10
Green Chemistry
Please do not adjust margins
Journal Name
ARTICLE
lower and MCNB(0.01) owned the largest percentage.
Therefore, only a small amount of boron content (boron<1.59
atm%) could be efficient for acquiring more partially-
condensed amino groups in edge defects. Too much boron
species doped in the g-C3N4 skeleton might cause unwanted
negative effect to active groups.
Table 3 Catalysts screening for CO2 cycloaddition with PO intoVPieCw Article Online
DOI: 10.1039/C7GC00279C
PC yield
[%]
PC selectivity
[%]
TOFf
[h-1]
Entry
Catalysts
1
2
Blank
0.6
0.4
98.9
7.9
–
Melaminea
MCNa
0.02
0.24
1.51
1.38
0.63
0.24
1.14
0.97
0.54
0.07
0.38
3
4.8
42.1
99.9
98.6
65.2
99.9
99.9
92.5
93.2
48.8
99.2
99.9
Catalytic evaluation of MCNB materials
4
MCNB(0.01)a
MCNB(0.03)a
MCNB(0.05)a
MCNB(0.1)a
MCNB(0.01)b
MCNB(0.03)b
MCNB(0.05)b
MCNB(0.1)b
MCNB(0.01)c
MCNB(0.01)d
31.0
28.4
13.0
4.9
All the MCNB materials prepared under different B/M ratios
were tested for comparison in catalytic activity for CO2
conversion into cyclic carbonates. With propylene oxide (PO)
as model reactant, the cycloaddition reactions were conducted
under identical conditions for screening out the best MCNB
catalyst. In addition, the suitability of the best catalyst for
other epoxides’ conversion was also investigated. All yields
and selectivities were determined by GC using biphenyl as an
internal standard. Yields were given based on epoxides.
5
6
7
8
46.9
39.8
22.1
3.1
9
10
11
12
13
First, all the MCNB catalysts were screened by using the
model reaction of CO2 cycloaddition with propylene oxide (PO)
into propylene carbonate (PC) under identical reaction
conditions. The results were listed in Table 3. Without catalyst
participation, PO was hardly converted. With MCN as catalyst,
higher catalytic activity was obtained than melamine precursor.
After BmimBF4 addition, the catalytic activity could be
improved further, but with the increase of B/M mass ratios
from 0.01 to 0.1, the catalytic activity of MCNB exhibited a
decreasing trend. Among all the MCNB catalysts, MCNB(0.01)
94.2
95.7
1.16
Reaction Condition: PO: 14.3 mmol, temperature: 130 oC, CO2 pressure: 2.8
MPa; a catalyst (50 mg), 6 h; b catalyst (100 mg), 6 h; c catalyst (100 mg), 36
h; d EO, catalyst (50 mg), 24 h. TOF: mass of synthesized PC per gram
f
catalyst per hour. All yields and selectivities were determined by GC. Yields
were given based on PO reactant.
Table 4 Suitability of MCNB(0.01) for catalytic conversion of other epoxides
Yield
[%]
Selectivity
[%]
TOFc
[h-1]
o
was the best one with PC yield up to 46.9% at 130 C for 6 h,
Entry
1a
Epoxide
Product
and MCNB(0.1) was the worst. To further investigate the
suitability of MCNB(0.01) as catalyt for other substrates, the
cycloaddition of CO2 with different epoxides, ethylene oxide,
epichlorohydrin and isobutylene oxide, were tested
successively under given conditions. As the catalytic evaluation
results in Table 4 displayed, all the tested substrates exhibited
good product yields with MCNB(0.01) as catalyst, especially for
epichlorohydrin whose product yield was up to 89.0% at 130
oC for 6 h. Considering the high catalytic activity and suitability
for various epoxides, MCNB(0.01) was taken as a promising
catalyst for the synthesis of cyclic carbonates in future industry.
The catalytic evaluation results showed the same trend with
the above analyzed properties. As a result, BmimBF4 addition
enhanced the catalytic activity of melamine-derived carbon
nitrides from two aspects. On one hand, the polymerization
degree and crystallinity of MCNB were adjusted with BmimBF4
as the role of soft template, which created more exposed edge
defects in carbon nitride structure and the partially-condensed
groups therein, such as primary amine (NH2) groups and
secondary amino (C–NH–C) groups. These partially-condensed
31.0
59.7
89.0
99.9
99.2
98.0
1.51
2.50
5.79
2a
3a
4b
34.5
82.9
0.48
Reaction Condition: catalyst (50 mg), epoxides (14.3 mmol), temperature:
a
b
c
130 oC, CO2 pressure: 2.8 MPa. Reaction time: 6 h; 24 h. TOF: mass of
synthesized PC per gram catalyst per hour. All yields and selectivities were
determined by GC. Yields were given based on epoxides.
amino groups were active centers for CO2 cycloaddition with
epoxides. Furthermore, with increase of B/M ratios, the
proportion of meso-/macropores around 50 nm in MCNB
structure declined and MCNB(0.01) owned the largest
proportion of meso-/macropores (see Fig. S1). The larger
proportion of meso-/macropores is another factor for
enhancement of catalytic activity because molecule
transportation between different amino groups could be easily
done with less diffusion resistance therein. On the other hand,
the active centers were also influenced by doped boron atoms
with BmimBF4 as a role of boron source. Both the XPS analysis
and catalytic testing results showed that only a small amount
of boron doped content (B<1.59 atm%) could be efficient for
Scheme 2 The reaction of CO2 cycloaddition with epoxides into cyclic
carbonates.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 6 of 10
ARTICLE
Journal Name
Recyclability studies of MCNB catalyst
View Article Online
DOI: 10.1039/C7GC00279C
To examine the recyclability of MCNB catalyst, recycling
experiments were conducted by using PO cycloaddition with
CO2 as the model reaction with MCNB(0.01) as catalyst at
130 °C, 2.8 MPa for 24 h. After each cycle experiment, the
catalyst was easily recovered by simple filtration, and then
used for the next cycle. As displayed in Fig. 7, almost no
decrease was detected in PC yield (about 72%) and selectivity
(over 99%) after five recycle times. indicating that MCNB(0.01)
maintained its catalytic activity after use.
In addition, the chemical structure and thermal stability of
the catalyst after recycling experiments were also studied. FT-
IR spectra and TGA curves of MCNB(0.01) before and after use
for five runs were respectively shown in Fig. 8 and Fig. 9. By
comparison in FT-IR spectra, all the characteristic peaks of
MCNB(0.01) structure remained and there was no obvious
change between fresh and used catalyst. Furthermore, TGA
curves showed that the decomposition temperature of used
Fig. 7 PC yield and selectivity after each cycle run under reation condition:
PO (14.3 mmol), MCNB(0.01) (100 mg), CO2 pressure (2.8 MPa),
temperature (130 °C), time (24 h).
o
catalyst stayed high over 500 C, which was consistent with
fresh one.
The recyclability of catalytic activity, chemical structure and
decomposition temperature demonstrated the high stability of
MCNB catalyst, which was beneficial for the industrial
application.
DFT calculation for edge active groups of MCNB materials
According to the previous report, two kinds of boron sites
(corner and bay boron) might exist in MCNB structures.29 To
figure out the specific function of boron sites in g-C3N4
skeleton, DFT calculations were carried out for these two kinds
of boron sites of MCNB scheme to compare with that of MCN.
The B3LYP/6-311++G(d,p) method has been used for structure
optimizations, and subsequent frequency calculations at the
same level verify the optimized structures to be ground states
without imaginary frequencies (NImag=0). In order to make
the calculation more representative, three existing structures
of edge groups, with no hydrogen (MCN(C2)), one hydrogen
(MCN(CH)), and two hydrogen connected (MCN(H2)), were
proposed for comparison. MCNB skeletons with corner and
bay boron for different hydrogen-containing structures were
respectively marked as MCNB(C2)-I, MCNB(C2)-II, MCNB(CH)-I,
MCNB(CH)-II and MCNB(H2)-I, MCNB(H2)-II in Fig. 10. The NPA
(natural population analysis) charge distributions by NBO
calculations of nitrogen adjacent to boron on the MCNB
structures were depicted in Table 5. For the same hydrogen-
containing edge groups, the NPA charge values and bond
lengths exhibited the same increasing trends from no boron
doping to bay boron doping. Obiviously, the charge values
changed from -0.450 to -0.710 for MCNB(C2)-II, from -0.575 to
-0.848 for MCNB(CH)-II and from -0.744 to -1.032 for
MCNB(H2)-II, indicating that electrons are more inclined to
transfer from the bay boron site to the nitrogen. According to
previous reports,49-51 CO2 could interact with an amine via N–
Fig. 8 FT-IR spectra of MCNB(0.01) catalyst before use (fresh) and after use
for five runs (used).
Fig. 9 TGA curves of MCNB(0.01) catalyst before use (fresh) and after use
for five runs (used).
improving the catalytic activity. Too much addition might be H···O hydrogen bonding or via an interaction between the N
detrimental for the catalytic performance, similar to previous lone pair and the C atom of CO2 and tertiary amines with no
research.29
hydrogen atom attached to the nitrogen atom have a low
reactivity with respect to CO2. Thus, the increased electron
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 10
Green Chemistry
Please do not adjust margins
Journal Name
ARTICLE
density around the nitrogen atom is especially favorable for
View Article Online
DOI: 10.1039/C7GC00279C
interaction of primary (NH2) and secondary (C–NH–C) amino
52
groups with CO2.49,
The interaction between boron and
nitrogen will also bring changes in the bond lengths of B–N.
The B–N bond length becomes longer than that of MCN(CH)
(from 1.350 to 1.411) and MCN(H2) (from 1.348 to 1.411),
which causes easier access to CO2. Both the enhancement on
electron density and bond length of –N groups after boron
replacement was another reason for the increased catalytic
activity after BmimBF4 addition. The reaction mechanism with
MCNB as catalysts was proposed in Scheme 3. The boron-
enhanced nitrogen groups (primary and secondary amino
groups) in edge defects of MCNB fragments could activate CO2
via nucleophilic attack and the primary amino (NH2) groups
might induce C–O bond to easy opening by formation of
hydrogen bond. With the contribution of meso-/macropores
therein to molecule transportation, the active groups in pore
Scheme 3 The possible mechanism for CO2 conversion into propylene
carbonates over MCNB fragments.
edge defects could be easily accessible and this constituted
another reason for the effective catalytic performance of
MCNB catalysts.
Conclusions
A series of boron-doped melamine-derived carbon nitrides
(MCNB) were successfully prepared by adding different
amount of BmimBF4 (B/M mass ratio=0.1, 0.05, 0.03, 0.01) into
the preparation process of carbon nitrides. Compared with
non-doping melamine-derived carbon nitride (MCN), the
addition of BmimBF4 led to the decrease of polymerization
degree with more generated edge defects and more meso-
macroporous structures, as well as boron doping in MCNB
networks. In addition, MCNB owned better catalytic activity for
CO2 cycloaddition with epoxides than MCN, of which
MCNB(0.01) with boron doped content less than 1.59 atm%
was the best one and could be suitable for catalytic conversion
of various epoxide substrates. The improvement on catalytic
activity could be attributed to the increase and enhancement
of uncondensed amino groups in edge defects. First, the
decreased polymerization degree and the existence of more
meso-macroporous structures made more amino groups in
edge defects exposed and accessible. Secondly, boron doping
enhanced the electron density of hydrogen-containing amino
groups which was the active sites for CO2 and epoxide
activation. Thus, it is the combination of edge defects creation
and boron doping that contributed to the better catalytic
performance. The boron-enhanced amino groups, including
primary (–NH2) and secondary amino groups (C–NH–C) were
finally considered to be active centers for activation of
epoxides and CO2 and the boron doping form was confirmed
MCN(C2)
MCNB(C2)-I
MCN(C2) -II
MCN(CH)
MCNB(CH)-I
MCNB(CH)-II
MCN(H2)
MCNB(H2)-I
MCNB(H2)-II
Fig. 10 Optimized structures by B3LYP/6-311++G(d,p) for Structure A:
MCN(C2), MCNB(C2)-I, MCNB(C2)-II; Structure B: MCN(H2), MCNB(H2)-I,
MCNB(H2)-II and Structure C: MCN(H2), MCNB(H2)-I, MCNB(H2)-II.
Table 5 NPA charge distributions and the changes of bond length (Å) for
different structures
Structure A
NPA Charge
MCN(C2)
-0.450
MCNB(C2)-I
-0.470
MCNB(C2)-II
-0.710
bond length (Å)
Structure B
1.356
1.360
1.416
MCN(CH)
-0.575
MCNB(CH)-I
-0.599
MCNB(CH)-II
-0.848
to be HN＝B(–N)2 structure with the bay boron sites. This work
NPA Charge
proposed an effective way to tailor carbon nitride formation
for high catalytic activity. Considering the good recyclability in
catalytic activity, chemical structure and thermal stability,
MCNB is promising to be an efficient heterogeneous catalyst
for CO2 conversion into cyclic carbonates in future industry.
bond length (Å)
Structure C
1.350
1.357
1.411
MCN(H2)
-0.744
MCNB(H2)-I
-0.770
MCNB(H2)-II
-1.032
NPA Charge
bond length (Å)
1.348
1.361
1.411
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 8 of 10
ARTICLE
Journal Name
More relevant exploitation on further improvement of MPa pressure before being heated. When theVrieewaAcrttiocler Ownlianse
DOI: 10.1039/C7GC00279C
catalytic performance is in progress.
heated to the target temperature, the CO2 pressure was then
increased to 2.8 MPa and holded at this constant pressure for
a certain time. After the reaction, the reactor was cooled down
to room temperature and the excess CO2 was released slowly.
The products were separated from the remaining mixture and
analyzed by GC and GC-MS. The cycloaddition of CO2 with
other epoxides, ethylene oxide, epichlorohydrin and
isobutylene oxide, were conducted with the same procedure.
Experimental Section
Preparation of MCNB materials
According to a typical synthetic procedure as described in the
literature,27 certain amount (0.1g, 0.05g, 0.03g, 0.01g) of
BmimBF4 and 15 mL deioned water were first added to a 50
mL flask and the mixture was stirred for 10 min. Then 1 g
melamine was added in. The final mixture was unsealed and
Acknowledgements
o
stirred at 80 C for 5 min. Thereafter, the flask was heated
o
o
through 90 oC to 100 C and hold at 100 C until the water in
flask was dried up. The obtained dried white solid was
transferred into a crucible and slowly heated over 2 h to a
temperature of 360 oC. The material was hold at this
temperature for 10 min, heated over 1 h to 500 oC, and
tempered at this higher temperature for another 3 h. After
cooling in the oven to room temperature, the sample was in
sequence washed by diluted hydrogen chloride, deioned water
and methanol for the removal of some remaining impurities.
Then, the sample was dried in the vacuum oven overnight.
Through the above steps, a series of MCNB materials with
different B/M mass ratios (x=0.1, 0.05, 0.03, 0.01) and MCN
(without addition of BmimBF4) were successively prepared.
We wish to thank the financial support of the National Key
Projects for Fundamental Research and Development of China
(No. 2016YFB0600903), National Natural Science Foundation
of China (No. 21676274), Key Program of National Natural
Science Foundation of China (No. 91434203), Key Research
Program of Frontier Sciences, CAS (QYZDY-SSW-JSC011)
and the Fund of State Key Laboratory of Multiphase Complex
Systems, IPE, CAS (No. MPCS-2015-A-05).
References
1
J. Bayardon, J. Holz, B. Schäffner, V. Andrushko, S.
Verevkin, A. Preetz and A. Börner, Angew. Chem. Int. Ed.,
2007, 46, 5971-5974.
Characterization of MCNB materials
2
3
4
S. Yin, J. Han, T. Zhou and R. Xu, Catal. Sci. Technol., 2015,
The chemical structures of the materials were analyzed by
Fourier transform infrared (FT-IR) spectroscopy (Thermo
Nicolet 380) with anhydrous KBr as the standard. X-ray
diffraction (XRD) were conducted by a SmartLab (9) X-ray
diffractometer (Rigaku, Japan) with Cu Kα radiation (λ = 0.1541
nm). Thermogravimetric analysis (TGA) were performed on a
DTG-60H thermal analyzer (Shimadzu, Japan) at a heating rate
of 10 °C min−1 under a N2 atmosphere using an empty crucible
as the reference. Scanning electron microscopic (SEM) and
transmission electron microscopic (TEM) images were
respectively taken by a SEM (JSM6700F, JEOL) and a TEM
instrument (JEM2100F, JEOL). Isotherms of N2 adsorption and
BET surface areas were obtained by N2 physisorption test using
a Micromeritics Porosimeter (ASAP-2020 HD88). Elemental
analysis (C/H/N) was operated on a Vario EL Cube elemental
analyzer. The distribution of surface active groups was
achieved through an X-ray photoelectron spectroscope (XPS)
(ESCALAB250Xi). 11B solid-state MAS NMR was recorded on a
Bruker Avance III HD 500M NMR. All the yields and selectivities
were examined via a GC (Agilent 6820) with a TCD detector.
The composition of the products was identified by using GC–
MS (Agilent 6890/5973B). The density functional theory (DFT)
calculations were carried out using the Gaussian 09 program.
5
, 5048-5061.
J. W. Comerford, I. D. V. Ingram, M. North and X. Wu,
Green Chem., 2015, 17, 1966-1987.
B. Mousavi, S. Chaemchuen, B. Moosavi, Z. Luo, N.
Gholampour and F. Verpoort, New J. Chem., 2016, 40
,
5170-5176.
5
6
S. Verma, R. B. N. Baig, M. N. Nadagouda and R. S. Varma,
Green Chem., 2016, 18, 4855-4858.
R. R. Kuruppathparambil, R. Babu, H. M. Jeong, G.-Y.
Hwang, G. S. Jeong, M.-I. Kim, D.-W. Kim and D.-W. Park,
Green Chem., 2016, 18, 6349-6356.
7
Q. He, J. W. O'Brien, K. A. Kitselman, L. E. Tompkins, G. C.
T. Curtis and F. M. Kerton, Catal. Sci. Technol., 2014, 4,
1513-1528.
8
9
B.-H. Xu, J.-Q. Wang, J. Sun, Y. Huang, J.-P. Zhang, X.-P.
Zhang and S.-J. Zhang, Green Chem., 2015, 17, 108-122.
X. Wang, Y. Zhou, Z. Guo, G. Chen, J. Li, Y. Shi, Y. Liu and J.
Wang, Chem. Sci., 2015, 6, 6916-6924.
10 R. Babu, A. C. Kathalikkattil, R. Roshan, J. Tharun, D.-W.
Kim and D.-W. Park, Green Chem., 2016, 18, 232-242.
11 R. Ma, L.-N. He and Y.-B. Zhou, Green Chem., 2016, 18
,
226-231.
12 N. Pasupulety, K. Gunda, Y. Liu, G. L. Rempel and F. T. Ng,
Appl. Catal. A- Gen. , 2013, 452, 189-202.
13 Y. Wang, X. Wang and M. Antonietti, Angew. Chem. Int.
Ed., 2012, 51, 68-89.
General procedure for catalytic conversion of epoxides by MCNB
catalysts
First, 14.3 mmol of propylene oxide and certain amount of
MCNB were put in a 25 ml stainless steel reactor equipped
with an automatic temperature control system at room
temperature. Then the reactor was charged with CO2 to 1.0
14 Y. Gong, M. Li, H. Li and Y. Wang, Green Chem., 2015, 17
,
715-736.
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 10
Green Chemistry
Please do not adjust margins
Journal Name
ARTICLE
43 Q. Xiang, J. Yu and M. Jaroniec, J. Phys. Chem.VCie,w2A0rt1ic1le,O1n1lin5e
15 J. Liu, H. Wang and M. Antonietti, Chem. Soc. Rev., 2016,
45, 2308-2326.
16 F. Goettmann, A. Thomas and M. Antonietti, Angew.
Chem. Int. Ed., 2007, 46, 2717-2720.
,
DOI: 10.1039/C7GC00279C
7355-7363.
44 M. Blesic, M. H. Marques, N. V. Plechkova, K. R. Seddon, L.
P. N. Rebelo and A. Lopes, Green Chem., 2007, , 481-490.
9
17 M. B. Ansari, B.-H. Min, Y.-H. Mo and S.-E. Park, Green
Chem., 2011, 13, 1416-1421.
45 Y. Wang and G. A. Voth, J. Am. Chem. Soc., 2005, 127,
12192-12193.
18 M. B. Ansari, H. Jin and S.-E. Park, Catal. Sci. Technol.,
46 S. Chen, S. Zhang, X. Liu, J. Wang, J. Wang, K. Dong, J. Sun
and B. Xu, Phys. Chem. Chem. Phys., 2014, 16, 5893-5906.
47 J. Bowers, C. P. Butts, P. J. Martin, M. C. Vergara-Gutierrez
and R. K. Heenan, Langmuir, 2004, 20, 2191-2198.
48 D. R. Miller, J. Wang and E. G. Gillan, J. Mater. Chem.,
2002, 12, 2463-2469.
2013, 3, 1261-1266.
19 Z. Huang, F. Li, B. Chen, T. Lu, Y. Yuan and G. Yuan, Appl.
Catal. B-environ, 2013, 136–137, 269-277.
20 J. Mao, T. Peng, X. Zhang, K. Li, L. Ye and L. Zan, Catal. Sci.
Technol., 2013,
21 Q. Su, J. Sun, J. Wang, Z. Yang, W. Cheng and S. Zhang,
Catal. Sci. Technol., 2014, , 1556-1562.
22 J. Xu, F. Wu, Q. Jiang and Y.-X. Li, Catal. Sci. Technol., 2015,
, 447-454.
23 Z. Huang, F. Li, B. Chen and G. Yuan, Catal. Sci. Technol.,
2016, , 2942-2948.
3, 1253-1260.
49 R. Vaidhyanathan, S. S. Iremonger, G. K. H. Shimizu, P. G.
Boyd, S. Alavi and T. K. Woo, Science, 2010, 330, 650-653.
4
50 P. D. Vaidya and E. Y. Kenig, Ind. Eng. Chem. Res., 2008, 47
,
5
34-38.
51 K. R. Roshan, B. M. Kim, A. C. Kathalikkattil, J. Tharun, Y. S.
Won and D. W. Park, Chem. Comm., 2014, 50, 13664-
13667.
6
24 D.-H. Lan, H.-T. Wang, L. Chen, C.-T. Au and S.-F. Yin,
Carbon, 2016, 100, 81-89.
52 M. D. Soutullo, C. I. Odom, B. F. Wicker, C. N. Henderson,
25 J. Wang, J. Luo, S. Feng, H. Li, Y. Wan and X. Zhang, Green
A. C. Stenson and J. H. Davis, Chem. Mater., 2007, 19,
Energ. Environ., 2016,
26 Y. Wang, X. Wang, M. Antonietti and Y. Zhang,
ChemSusChem, 2010, , 435-439.
1
, 43-61.
3581-3583.
3
27 Y. Wang, J. Zhang, X. Wang, M. Antonietti and H. Li,
Angew. Chem. Int. Ed., 2010, 49, 3356-3359.
28 Y. Zhang, T. Mori, J. Ye and M. Antonietti, J. Am. Chem.
Soc., 2010, 132, 6294-6295.
29 Y. Wang, H. Li, J. Yao, X. Wang and M. Antonietti, Chem.
Sci., 2011,
30 Y. Zheng, J. Liu, J. Liang, M. Jaroniec and S. Z. Qiao, Energ.
Environ. Sci., 2012, , 6717-6731.
2, 446-450.
5
31 B. V. Lotsch and W. Schnick, Chem. Mater., 2006, 18
,
1891-1900.
32 J. Hong, X. Xia, Y. Wang and R. Xu, J. Mater. Chem., 2012,
22, 15006-15012.
33 Y. Wang, J. Hong, W. Zhang and R. Xu, Catal. Sci. Technol.,
2013, 3, 1703-1711.
34 J. Xu, Y. Li, S. Peng, G. Lu and S. Li, Phys. Chem. Chem.
Phys., 2013, 15, 7657-7665.
35 C. Fettkenhauer, G. Clavel, K. Kailasam, M. Antonietti and
D. Dontsova, Green Chem., 2015, 17, 3350-3361.
36 A. Kumar, P. Kumar, C. Joshi, S. Ponnada, A. K. Pathak, A.
Ali, B. Sreedhar and S. L. Jain, Green Chem., 2016, 18
,
2514-2521.
37 B. Lin, G. Yang, B. Yang and Y. Zhao, Appl. Catal. B-Environ,
2016, 198, 276-285.
38 X. Song, H. Tao, L. Chen and Y. Sun, Mater. Lett., 2014,
116, 265-267.
39 Z. Miskolczy, K. Sebők-Nagy, L. Biczók and S. Göktürk,
Chem. Phys. Lett., 2004, 400, 296-300.
40 T. Singh and A. Kumar, J. Phys. Chem. B, 2007, 111, 7843-
7851.
41 J. Wang, H. Wang, S. Zhang, H. Zhang and Y. Zhao, J. Phys.
Chem. B, 2007, 111, 6181-6188.
42 G. Xin and Y. Meng, New J. Chem., 2013, 2013, 1-5.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Green Chemistry
Page 10 of 10
View Article Online
DOI: 10.1039/C7GC00279C
Boron-doped melamine-derived carbon nitrides were tailored with BmimBF4 as soft template and boron source for CO2
conversion into cyclic carbonates.