Highly efficient pollutant removal of graphitic carbon nitride by the synergistic effect of adsorption and photocatalytic degradation

Environmental remediation based on semiconducting materials offers a green solution for pollution control in water. Herein, we report a novel graphitic carbon nitride (g-C3N4) by one-step polycondensation of urea. The novel g-C3N4 material with a surface area of 114 m2 g−1 allowed the repetitive adsorption of the rhodamine B (RhB) dye and facilitated its complete photocatalytic degradation upon light irradiation in 20 min. This study provides new insights into the fabrication of g-C3N4-based materials and facilitates their potential application in the synergistic removal of harmful organic pollutants in the field of water purification.


Introduction
The problems of environmental pollution have invigorated growing awareness all over the world. 1,2To date, a number of treatments, such as biodegradation, 3 adsorption, 4 and photocatalytic degradation, 5 have been studied to remove organic pollutants in water.Photocatalysis based on the conversion of solar into chemical energy has been regarded to be one of the most promising technologies to remove environmental pollutants. 5,62][13][14][15][16] As a metal-free polymeric photocatalyst, g-C 3 N 4 exhibits a number of excellent characteristics, such as facile synthesis, high chemical and thermal stability, reasonable cost, abundant and inexpensive building elements, appropriate electronic band structure for visible-light response, and exible supermolecular networks for ne-tuning material properties, for photocatalysis. 13,17However, the practical applications of g-C 3 N 4 are still hindered by the several obstacles and shortcomings, especially its low specic surface area, limited active sites, poor adsorption ability, and the serious aggregation observed during a photocatalytic process, of common bulk g-C 3 N 4 prepared via the direct polycondensation of nitrogenrich precursors. 17,18To overcome these drawbacks, many attempts, such as doping with heteroatoms, 19,20 constructing heterostructures, 21,22 fabricating copolymers, 23,24 and thermal etching, 25,26 have been dedicated towards improving the photocatalytic capability of g-C 3 N 4 .However, preparation of a highly active g-C 3 N 4 material using a facile and eco-friendly strategy is still desirable.8][29] However, adsorption cannot solve these problems drastically because organic pollutants just can be concentrated rather than degraded to non-polluting molecules.In addition, the materials need to undergo tedious desorption processes before being recycled. 30To combine photocatalysis with adsorption, a number of efforts have been devoted towards developing g-C 3 N 4 -based materials with strong adsorption.Chen prepared a g-C 3 N 4 /activated carbon composite photocatalyst with good efficiency in the photodegradation of phenol. 31An agar-C 3 N 4 hybrid hydrogel photocatalyst with a 3D network structure was also prepared, and the hybrid hydrogel showed highly efficient pollutant removal ability via the synergistic effect of adsorption and photocatalytic degradation. 32In addition, polyaniline/ carbon nitride nanosheets were shown to be highly excellent in removing organic pollutants on account of the cooperation of adsorptive preconcentration and a following photocatalytic oxidation reaction. 33Very recently, Panneri et al. reported bifunctional granules of carbon-doped g-C 3 N 4 for the efficient removal of the antibiotic tetracycline. 34Inspired by these fruitful ndings, we envisaged the preparation of a novel g-C 3 N 4based material with high photocatalytic activity and strong adsorption capability using a facile method.
As illustrated, the characteristic properties and chemical structures of g-C 3 N 4 are strongly affected by the reaction atmosphere through inducing disordered structures, defects, and carbon and nitrogen vacancies. 13In our previous study, we have successfully prepared modied g-C 3 N 4 with high photocatalytic activity under a self-producing atmosphere.However, we found no enhancement in the adsorption capacity. 35In this study, we have developed novel g-C 3 N 4 nanosheets from a urea precursor under a self-producing atmosphere by controlling the addition of N 2 , as shown in Fig. 1.When compared with conventional g-C 3 N 4 , the novel g-C 3 N 4 nanosheets demonstrate an improved crystal structure, larger specic surface area, and more regular and homogeneous morphology.Moreover, the novel g-C 3 N 4 material exhibited strong adsorption and efficient photocatalytic activity for the removal of rhodamine B (RhB) and could act as an excellent candidate for pollutant removal via the synergistic effect of adsorption and photocatalytic degradation.Our study provides new insights into the fabrication of g-C 3 N 4 -based materials and facilitates their potential application for the synergistic removal of various organic pollutants in the eld of water purication.

Sample preparation
The conventional g-C 3 N 4 sample was prepared using a previously reported thermal polymerization method. 36In a typical procedure, the material was synthesized by heating 5 g of urea in a semi-closed alumina crucible under a ow of N 2 gas to 550 C for 4 h at a heating rate of 5 C min À1 and then cooled down to room temperature.The product was obtained and ground into powder and denoted as U-N.The novel g-C 3 N 4 material was prepared using a similar method as used for U-N with the exception that the addition of N 2 at 300 C was stopped.The obtained sample was denoted as U-300.

Catalyst characterization
The X-ray diffraction (XRD) patterns were obtained via a Rigaku Dmax/Ultima IV diffractometer using Cu Ka radiation (l ¼ 1.5418 A) to evaluate the crystal structure and phase purity.
Fourier transform infrared (FTIR) spectra were obtained using a Perkin-Elmer model 2000 FTIR spectrophotometer.Scanning electron microscopy (SEM) images were acquired using an FEI QUANTA F250 scanning electron microscope.X-ray photoelectron spectroscopy (XPS) measurements were carried out using an ESCALAB 250 Xi with a high-performance Al monochromatic source (hn ¼ 1486.6 eV, 150 W).All binding energies were calibrated by setting the C 1s peak to 284.8 eV for surface adventitious carbon, and the elemental compositions were determined from the peak area ratios aer correction for the sensitivity factor for each element.Brunauer-Emmett-Teller (BET) surface area measurements were conducted using the N 2 adsorption-desorption isotherms obtained at 77 K using Quantachrome Instruments version 3.0.UV-vis diffuse reectance spectra (DRS) were obtained via a Shimadzu UV-3600 spectrophotometer using BaSO 4 as the reference sample.The photoluminescence (PL) spectra of the photocatalysts were obtained using a Varian Cary Eclipse spectrometer with the excitation wavelength of 350 nm.The RhB adsorption and degradation were monitored by a Shimadzu UV-2550 UV-vis spectrophotometer at denite time intervals.

Adsorption and photocatalytic activity
The adsorption capacities and photocatalytic activities of the synthesized samples were evaluated using the adsorption and photodegradation of RhB at ambient temperature in air under magnetic stirring.To nd out the differences in the adsorption performance, 25 mg of the photocatalysts was dispersed in a RhB solution (10 mg L À1 ).The resulting suspension was rst sonicated for 10 min and then stirred continuously in the dark for 50 minutes.Aliquots were taken every 10 minutes to measure the adsorbed concentration using a UV-vis spectrophotometer.The photocatalytic activities of the synthesized samples were then evaluated through degradation of RhB under irradiation.8][39] In a typical photocatalytic experiment, 25 mg of photocatalyst was added to 50 mL of RhB aqueous solution (10 mg L À1 ) at room temperature.Prior to irradiation, the mixed suspension was rst sonicated for 10 min and then magnetically stirred for 20 min in the dark to obtain an adsorptiondesorption equilibrium.At given intervals during irradiation, about 3 mL of sample was taken out from the reaction system and centrifuged to remove the photocatalyst powders.The absorbance of RhB in the supernatant was measured using a UV-vis spectrophotometer at 554 nm.The efficiency of degradation was calculated by C/C 0 , where C is the concentration of remaining dye solution at time t and C 0 is the initial concentration.
The stability of U-300 was investigated using recyclability studies.Aer each cycle, the catalyst was obtained by centrifugation, washed with distilled water, and dried at 60 C overnight.Then, the recovered catalyst was directly used for the next cycle of photocatalytic degradation of RhB as abovementioned.

Structural characteristics
The XRD patterns of the common and novel g-C 3 N 4 are displayed in Fig. 2a.The two materials show distinctively different XRD patterns.U-300 shows two typical diffraction peaks at around 12.8 and 27.6 , as reported previously, [11][12][13] which are due to the in-plane structural packing motif and periodic stacking of layers along the c-axis, respectively.However, the two sharp peaks become very weak in the U-N pattern, which demonstrate the absence of long-range order in the atomic arrangements. 40,41We may infer the multiple effects of too many gas bubbles, which are produced during thermal condensation of urea, and the additional N 2 signicantly disrupts the longrange atomic order in both the perpendicular and parallel directions to the g-C 3 N 4 layers.Aer stopping the addition of N 2 , the self-producing atmosphere originates from the interior of the reaction system.The condition of polycondensation was more homogeneous; thus, we obtained an improved crystal structure.Cui et al. have reported that the well-condensed crystallinity of g-C 3 N 4 indicates a larger surface area and improved transport of photogenerated carriers in its network. 41t is reasonable to expect the U-300 sample may show enhanced photocatalytic activity.Moreover, the main peak was slightly shied to higher angles: U-N at 27.1 and U-300 at 27.6 , corresponding to a reduction in the stacking distance of the graphitic-layered structure.Merschjann et al. have shown that electronic transport is predominantly perpendicular to the sheets in g-C 3 N 4 . 42Thus, the decreased interlayer distance may be in favor of charge transport and thus improve the photocatalytic activity.
The microstructures of the conventional and novel g-C 3 N 4 samples were further revealed using Fourier transform infrared (FT-IR) spectroscopy, which was sensitive to the local (or shortrange) structure of the materials (Fig. 2b).The peak at approximately 810 cm À1 originates from the characteristic breathing mode of the tri-s-triazine units. 43The bands in the range from ca. 1645 to 1235 cm À1 can be attributed to the typical stretching modes of C-N heterocycles.The broad absorption bands located in the range from 3700 to 3000 cm À1 are assigned to the stretching vibration of N-H and O-H bonds, associated with the uncondensed amino groups and surface-bonded H 2 O molecules, respectively. 41,43We can see that the FTIR spectra of the two g-C 3 N 4 materials show similar characteristic vibrational peaks; however, the peak intensity and stretching modes of the skeletal U-300 network are enhanced and better resolved probably due to the better organization of the conjugated system, which are also in agreement with the previously reported results. 41,44,45he compositions and surface chemical states of the two different samples were studied using X-ray photoelectron spectroscopy (XPS).In the XPS survey spectra (Fig. S2  exhibited similar C 1s and N 1s spectra without any signicant peak shis; this indicated similar chemical states.The deconvoluted C 1s spectra (Fig. 3a and c) showed three peaks at 284.8, 288.4,and 293.6 eV, which corresponded to surface carbon contamination, sp 2 hybridized carbon of the tri-s-triazine rings, and p / p* satellite band. 48The N 1s XPS spectra (Fig. 3b and  d) can be tted into four peaks.The main peaks at 398.9 eV, 400.1 eV, and 401.1 eV correspond to the sp 2 -hybridized nitrogen in the heterocycle (C-N]C), tertiary N in the form of N(-C) 3 , and uncondensed amino functional groups (NH 2 or NH), 45,49 respectively.A weak peak at 404.5 eV of N 1s was assigned to the p / p* satellite band, which was very much like the satellite component for its C 1s signal. 48The absorption associated with the different nitrogen moieties identies the defect types to some degree, as well as the uncondensed amino groups. 26,49The N-associated species were quantied using the deconvoluted N 1s spectra (Table 1).A larger value of NH x / N(-C) 3 demonstrates the smaller degree of polymerization and more uncondensed amino groups.It can be found that U-300 shows a larger value of NH x /N(-C) 3 .The results illustrated that aer stopping the entrance of N 2 , the pyrolysis-generated self-producing atmosphere inuenced the process of thermal polymerization, resulted in a decreased degree of polymerization and a large number of amino groups, which concurred with our FTIR results.

Textural properties
The pore structures and BET surface areas of two different samples were obtained using the N 2 adsorption-desorption measurements conducted at 77.4 K.As shown in Fig. 4a, U-300 exhibits type IV isotherms with an extremely high adsorption capacity in the high relative pressure region (P/P 0 : from 0.8 to 1); this indicates the presence of abundant mesopores and Table 1 The relative content of the various nitrogen species obtained from the N 1s XPS data (%)  macropores. 45,50The BET surface area of U-300 was calculated to be 114.96m 2 g À1 , which was about 5 times that of U-N (23.12 m 2 g À1 ).
To further analyze the pore structures, the pore size distribution using the BJH method (Fig. 4b) provided the pore sizes.Accordingly, a sharp peak at about 3.8 nm and a broad distribution in the range from 10 to 170 nm were identied in the pore size distribution curve for U-300.The higher BET surface area and larger pore size distribution may be caused by the effective prevention of the aggregation of the g-C 3 N 4 nanosheets by the self-producing atmosphere.The large surface area and pore volume can provide more reactive sites and more edge structures and may effectively adsorb more reactants and be conducive to mass transfer and charge carrier transfer during the photocatalytic process. 45,50

Morphology information
The typical SEM images of the as-prepared samples are illustrated in Fig. 5.As can be seen, the U-N sample displays disorderly stacked irregular clusters and is mainly composed of interconnected thin layers with some pores that may result from the gas bubbles formed during the pyrolysis of urea.Contrary to U-N, U-300 exhibits an obviously layered platelet-like surface morphology.Furthermore, the images of U-300 display smaller particle sizes, better dispersion, and a more regular and homogeneous morphology.Based our previous study, N 2 can offer an inert, sole positive atmosphere before polycondensation, whereas aer stopping the addition of N 2 , the atmosphere of polycondensation originates from the interior of Fig. 5 The SEM patterns of the two different samples.the reaction system.The conditions for polycondensation are more homogeneous; thus, we obtain a regular and homogeneous morphology, which is in agreement with our XRD and FTIR results.Obviously, this is a facile way to tune the microstructures of g-C 3 N 4 .

Optical properties
The electronic band structures and photoelectric properties of the two materials were analyzed by UV-vis diffuse reectance spectroscopy (DRS) (Fig. 6a) and photoluminescence spectroscopy (PL) (Fig. 6b).The UV-vis DRS spectra indicate that the absorption edge of the novel g-C 3 N 4 sample displays a remarkable blue shi.This result is consistent with the pale yellow color (Fig. S3 †) and indicates the low degree of polycondensation.Accordingly, the electronic band gaps derived from the Tauc plots (Fig. 6a) are 2.94 eV for U-N and 3.04 eV for U-300.The band gap of U-300 was widened by 0.10 eV as compared to that of U-N.The increase in the band gap by 0.10 eV improves the redox ability of the charge carriers generated in the CN nanosheets. 40This result was further conrmed by the blue-shi in its uorescence emission peak in Fig. 6b; this could be attributed to the quantum connement effect. 41,45he photoluminescence emission (PL) spectra originating from the recombination of free charge carriers are important to reveal the separation, migration, and recombination of the photogenerated charge carriers.As shown in Fig. 6b, all the excitation spectra were monitored at room temperature with an excitation wavelength of 350 nm and match well with UV-vis diffuse reectance spectra.Note that U-N has a low PL emission peak.According to the literature, surface defects may act as e À /h + recombination sites that result in increasing the nonradiative recombination rates and reduce the free carrier concentration; this leads to a decreased radiative PL intensity. 47or U-300, the enhanced PL intensity can be attributed to the improved crystal structure. 41,47,51,52Furthermore, the blue-shied wavelength of the uorescence emission peak observed for U-300 illustrates the smaller degree of polycondensation as compared to the case of U-N and the presence of sub-gap defects in the material. 47he XPS valence band (VB XPS) spectra were analyzed to investigate the band edges of the two different samples.In  Fig. 7a, it can be seen that the samples have the same VB of $2.06 eV.When combined with the UV-vis DRS results, we found that the CB of U-300 was up-shied; this indicated its stronger redox abilities. 41

Adsorption and photocatalytic performance
Fig. 8a presents the change in the concentration of RhB with time for the two samples.RhB does not undergo self-photolysis as its concentration remains unchanged with time.Aer reaching an adsorption equilibrium under the same conditions, the U-300 material showed a higher percentage of adsorption (40%) as compared to U-N (8%).The adsorption of RhB was monitored for 60 min, and it was observed that the adsorption equilibrium was reached within 20 min (Fig. S4 and S5 †).The porous morphology and large surface area of U-300 induced high rates of adsorption, and the adsorbed RhB was degraded by more than 90% within 15 min under UV-visible light illumination (Fig. 8a).The U-300 sample was tested through ve consecutive trials to test its reusability and stability.There was hardly any loss of activity in the adsorption and photocatalytic performance of U-300 upon prolonging the reaction time (Fig. 8b).To further test the photocatalytic performance, half mass of U-300 also showed excellent photocatalytic degradation (Fig. S8 †).To conrm the universality of the catalyst, we have studied the photocatalytic degradation of tetracycline hydrochloride (TC-HCl), which is a common antibiotic and colorless pollutant.The results also veried the excellent ability of U-300 to remove organic pollutants in water (Fig. S9 †).Based on the abovementioned experimental results, U-300 can be regarded as a stable high-performance photocatalyst for the photodegradation of organic pollutants, possessing great prospects in environmental protection.
As shown in Fig. 9, the mechanism of the adsorptionenrichment and photocatalytic degradation synergistic effect is proposed.At rst, the organic contaminant was adsorbed and enriched onto the surface of the novel g-C 3 N 4 nanosheets.The large surface area, small aggregation, and porous structure of the novel g-C 3 N 4 nanosheets provide more reactive sites and more edge structures and might effectively adsorb more reactants. 50,53Furthermore, the improved graphitic-like structure and shorter interlayer distance may accelerate charge transport and thus improve the photocatalytic activity. 41,42Finally, acting as an absorbent as well as photocatalyst, aer being absorbed onto the surface of the novel g-C 3 N 4 nanosheets, the organic contaminant was then degraded in situ under light irradiation to realize the synergistic effect of adsorption and photocatalysis.Overall, the adsorbed organic contaminants as well as the Fig. 9 A schematic of the synergistic adsorption and photocatalytic degradation processes.organic contaminants retained in solution were subsequently degraded under light irradiation.

Conclusions
In conclusion, a novel g-C 3 N 4 with strong adsorption capability and efficient photocatalytic activity was prepared by heating urea via a facile method.The self-producing atmosphere during the calcination process induced the condensation process, offering extra structural control for the synthesis of the g-C 3 N 4 networks.Compared to that of the conventional g-C 3 N 4 , the texture of the novel g-C 3 N 4 nanosheets was optimized to have higher surface area, an improved crystal structure, more homogeneous morphology, and smaller particles.The large surface area, porous structure, well-condensed crystallinity, and enlarged band gap are proposed to be primarily responsible for the enhanced adsorption and photocatalytic activity of U-300.Overall, this study demonstrates a convenient route to synthesize a novel g-C 3 N 4 catalyst with attractive photocatalytic activity without the need of any complex modication steps.

Fig. 1 A
Fig.1A schematic of the preparation of novel g-C 3 N 4 and its excellent degradation efficiency via synergistic adsorption and degradation.
, ESI †), only C, N, and O were detected.The very weak O 1s peak may be due to the surface absorbed H 2 O or O 2 . 46,47Both samples

Fig. 2
Fig. 2 The (a) XRD patterns and (b) FT-IR spectra of the two samples.

Fig. 3
Fig.3The high-resolution N 1s XPS spectra and C 1s XPS spectra of the two samples.

Fig. 4
Fig. 4 The (a) N 2 adsorption-desorption isotherms and (b) BJH pore-size distribution curves obtained for two different samples.

Fig. 7
Fig. 7 The (a) VB XPS spectra and (b) electronic band structures of the two different samples.

Fig. 6
Fig. 6 The (a) UV-visible diffuse reflectance spectra and (b) PL emission spectra at an excitation wavelength of 350 nm of the two different samples.

Fig. 8
Fig. 8 (a) The adsorption and photocatalytic degradation of RhB.(b) The cycling tests for the adsorption and photocatalytic degradation of RhB using U-300.