www.chemasianj.org
Jie Xu, Yong-Xin Li et al.
rangement of filled mesopores and hexagonal-like slices. By
combining the SAXS patterns and TEM images of the two
CND materials, it was apparent that the order in the paren-
tal mesoporous architecture had been well-retained in
CND-FDU12-2.5, more so than in CND-SBA15-2.5. This
phenomenon may have been due to the differences between
the mesostructures of SBA-15 and FDU-12. That is, FDU-
materials were much higher than that of CND-SBA15-2.5-
EDA. Moreover, the mesostructures of the original SBA-15
template underwent significant shrinkage or collapse com-
pared to the CND-SBA15-2.5 materials that were fabricated
by using common dispersing agents. One probable reason
for this difference is that DCDA could not be well-dissolved
in common solvents, which inevitably led to poor filling and
diffusion of DCDA in the mesopores of the mesoporous sili-
ceous templates. Indeed, using water, EtOH, and MeCN as
the dispersing media for DCDA and prolonging the stirring
time at elevated temperature only afforded a white suspen-
sion. However, DCDA was easily and well-dissolved in
EDA within 10 min under ambient conditions.
12 had a larger pore size than SBA-15 (see the Supporting
Information, Table S1), which was naturally favorable for
the filling of the precursor into the mesopores.
As mentioned above, although DCDA is widely available
and relatively non-toxic, reports on the use of DCDA for
preparing mesoporous g-CN through a nanocasting ap-
proach are very rare, probably owing to the practical issues
associated with the poor solubility of DCDA in most re-
agents. Herein, besides EDA, we also used other common
solvents, including water, EtOH, and MeCN as dispersing
media for DCDA. The as-synthesized CND-SBA15-2.5-
Besides their crystalline and textual properties, productiv-
ity is also a practical and important issue in the fabrication
of mesoporous carbon nitride materials. The Supporting In-
formation, Table S1, summarizes the product masses for the
mesoporous CND samples; the productivities were about
50%, much higher than the values of mesoporous CN mate-
rials that were obtained by using guanidinium chloride
H O/EtOH/MeCN materials showed inferior adsorption
2
quantities (Figure 5A) to CND-SBA15-2.5-EDA, and their
[8]
(
about 15 wt%) and hexamethylenetetramine as precur-
[14]
sors (about 10 wt%). This result was mainly attributed to
the stable chemical composition of DCDA. Regarding this
point, the utilization of DCDA for the synthesis of mesopo-
rous CND materials has a distinct advantage. Together with
its low toxicity and commercial availability, with the aid of
EDA, DCDA holds great potential for the medium-scale
preparation of mesoporous CN materials.
XPS was employed to analyze the surface chemical prop-
erties of the CND materials. The survey spectrum of CND-
SBA15-2.5 (Figure 6A) showed that the sample was solely
composed of C, N, and O, which may have originated from
adsorbed water molecules (H cannot be detected by XPS).
The absence of Si and F confirmed that the siliceous tem-
plate had been thoroughly etched and that the CND sample
had been well-rinsed. Moreover, the chemical compositions
of CND-SBA15 and CND-FDU12 were calculated from the
XPS data (see the Supporting Information, Table S1). The
molar C/N ratios for the mesoporous CND materials were
about 2.2–2.4:1, whereas the C/N ratios as determined by el-
emental analysis were 1.5–1.6:1. This difference indicated
that the bulk phase of the CNDs had a higher N content
than their surface, owing to the loss of N species on the sur-
face of the as-synthesized CND materials during the calcina-
tion and/or detemplating processes. Notably, the bulk C/N
ratios for these CND materials were much higher than those
of the ideal g-C N compound (C/N=0.75:1). This phenom-
enon could be ascribed to the use of EDA as a solvent,
which may condense with DCDA through their amino
groups, thereby sacrificing a small proportion of the N spe-
cies. In fact, on employing common dispersing agents, such
as water, instead of EDA, with DCDA as a precursor, low
C/N ratios (about 0.82) were obtained for the final CND
samples. However, as stated above, owing to the poor dis-
persion of DCDA and the siliceous templates, the surface
areas and pore volumes in those mesoporous materials were
extremely low (see the Supporting Information, Table S2).
2
Figure 5. N -adsorption/desorption isotherms (A) and the corresponding
pore-size distributions (B) of CND-SBA15-2.5 materials that was pre-
pared by using various dispersing media: water (&), EtOH (*), and
MeCN (~).
2
À1
surface areas were less than 190 m g , much lower than
that of CND-SBA15-2.5-EDA (see the Supporting Informa-
tion, Table S2). Moreover, in sharp contrast to CND-
SBA15-2.5, the three CND-SBA15-2.5 materials presented
very broad hysteresis loops within the range 0.45–0.95 p/p0,
thereby showing that the pores were unevenly distributed.
For the corresponding PSDs, besides the main distribution
at about 3.8 nm, the materials also showed an additional dis-
tribution within the range 7–9 nm (Figure 5B). SAXS was
employed to further examine the order in the mesostructure
of CND-SBA15-2.5. As shown in the Supporting Informa-
tion, Figure S5, all of the CND-SBA15-2.5 samples revealed
two well-defined scattering peaks. However, the intensities
3
4
of CND-SBA15-2.5-H O/EtOH/MeCN were considerably
2
lower than that of CND-SBA15-2.5-EDA. In addition, the
q values for the (100) planes of the three CND-SBA15-2.5
Chem. Asian J. 2014, 9, 3269 – 3277
3272
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim