In situ formation of 2-thiobarbituric acid incorporated g-C3N4 for enhanced visible-light-driven photocatalytic performance.

Embedding heterocycles into the skeleton of g-C3N4 has been proved to be a simple and efficient strategy for improving light response and the separation of photo-excited charges. Herein, 2-thiobarbituric acid incorporated g-C3N4 (TBA/CN) with good photocatalytic efficiency for Rh B degradation and H2 production was successfully achieved via a facile thermal copolymerization approach. The incorporation of aromatics and S atoms into the skeleton of g-C3N4 was identified via systematic characterizations. This unique structure contributed to the narrowed band-gap, extended delocalization of lone pair electrons and changed electron transition pathway, which led to the enhanced visible light utilization, accelerated charge migration and prolonged electron lifetime, subsequently resulting in the significant boost of photocatalytic activity. The optimal TBA/CN-3 sample yielded the largest Rh B degradation rate constant k value of 0.0273 min-1 and simultaneously highest rate of H2 evolution of 0.438 mmol g-1 h-1, which were almost 3.5 and 3.8 folds as fast as that of the pristine CN, respectively. Finally, the photocatalytic mechanism was proposed for the detailed elucidation of the process of Rh B degradation coupled with H2 production. This journal is © The Royal Society of Chemistry.

Introduction

In recent decades, emerging global energy dilemma and environmental pollution have become the two urgent threats to human existence and social development.[1,2] Semiconductor photocatalysis has widely proven to be one of the most encouraging and prospective alternative techniques for effectively easing up the energy crisis and environmental issues due to its incorporated merits of high efficiency, energy-saving ability and eco-friendliness.[3-5] Numerous photocatalysts have been constructed and successfully employed for H2 evolution, pollutant degradation and the conversion of CO2 to CO and CH4. However, the practical application of traditional photocatalysts, such as TiO2 and WO3, was mostly suppressed by the limitations of toxicity, high cost, low efficiency and insufficient visible light utilization.[6-8]

Recently, graphitic carbon nitride (g-C3N4, CN), a metal-free semiconductor polymer, has received unprecedented attention due to its unique traits of visible-light-driven, tunable electronic structure, physical–chemical stability, good-durability, inexpensive nature and facile synthesis, which are perfect fit for the H2 production, pollutant degradation, and CO2 reduction.[9-11] Nevertheless, the practical applicability of CN is still restricted by its inherent bottlenecks of limited visible light harvesting and ultrafast recombination rate of photogenerated charge carriers.[12,13] To overcome the above-mentioned defects, considerable efforts have been devoted to boost the photocatalytic property of CN, including porous morphology formation,[14] heteroatom doping,[15] heterostructure construction,[16] aromatic ring integration[17] and dye sensitization.[18]

Among these, embedding heteroaromatic molecules into the skeleton of CN has been confirmed to be an effective strategy, which resulted in an extended π-conjugated system and new localized internal field formation, could expand the visible light response range and facilitate the separation and transportation of charges, leading to an increased photocatalytic efficiency.[17,19,20] To date, numerous small molecules containing aromatic rings, such as tris(p-fluorophenyl)phosphine,[21] thiophene,[22] benzene,[23] pyridine[24] and salicylic acid,[25] have been copolymerized successfully into the framework of CN, which not only resulted in the facilitation of visible-light utilization and increasing charge separation efficiency, but also modulated the redox band potentials, making the effective use of photo-excited charge carriers.

Motivated by the above-mentioned discussion, in this study, 2-thiobarbituric acid incorporated CN was synthesized via a thermal copolymerization approach. Fascinatingly, the as-synthesized TBA/CN composites endowed improved visible-light utilization and accelerative separation efficiency of photo-excited charge carriers, resulting in a substantially boosted photocatalytic performance towards the degradation of Rh B and H2 evolution. Furthermore, the possible mechanism for the outstanding improved photocatalytic property was deeply interpreted.

Experimental section

Photocatalyst preparation and characterization

Analytical pure grade urea and 2-thiobarbituric acid purchased from Aladdin Chemical Reagent Co., Ltd (Shanghai, China) were used as precursors. The modified CN composites were synthesized via a direct annealing route. In detail, the mixture of 20 g urea and desired amounts (0.05, 0.1, 0.2 and 0.3 g) of 2-thiobarbituric acid was completely dissolved in 5 mL ultrapure water via magnetic stirring. After drying at 60 °C overnight, the obtained residues were placed into a covered crucible and heated at 520 °C for 120 min at an increasing rate of 5 °C min−1. When restoring to room temperature, the resulting composites were collected and labeled as TBA/CN-1, TBA/CN-2, TBA/CN-3 and TBA/CN-4 correspondingly.

For comparison, pristine g-C3N4 was also obtained via the direct calcination of urea under the same procedure, and the final product was named as CN.

The crystal and molecular structures, the surface morphologies and elemental composition and the light response behaviors of the as-synthesized composites were systematically characterized via X-ray diffraction (XRD), Fourier transformed infrared (FT-IR) spectroscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM), UV-Vis diffuse reflectance spectroscopy (UV-Vis DRS), X-ray photoelectron spectroscopy (XPS), photoluminescence (PL) spectroscopy and time-resolved photoluminescence (TRPL) spectroscopy.

The photoelectrochemical behaviors of these samples, including photocurrent (PC) response and electrochemical impedance spectra (EIS), were recorded on a CHI 760E electrochemical workstation.

Photocatalytic performance measurements

The visible-light driven photocatalytic properties of the as-synthesized samples were assessed by Rh B degradation coupled with H2 production. A 300 W Xe lamp (λ > 420 nm) was chosen as the radiation source and ultrapure water was employed throughout the experiments.

In a typical degradation experiment, 50 mg of the photocatalyst was fully mixed with 100 mL of Rh B (10 mg L−1) by ultrasonic treatment for 5 min. Before exposure to light, the reaction system achieved adsorption–desorption equilibrium by magnetic stirring for 30 min in darkness. At designated intervals, 3.0 mL suspension was collected and centrifuged at 12 000 rpm for 10 min. Finally, the concentration of Rh B in the supernatant was estimated using a Unico UV-2800A spectrophotometer at the characteristic wavelength of 552 nm.

In the photocatalytic H2 generation test, 30 mg of sample deposited with 1.0 wt% Pt was fully suspended into 50 mL ultrapure water containing 10% (vol/vol) triethanolamine (TEOA), where TEOA was used as the hole scavenger and Pt was used as the co-catalyst. H2 evolution during the photocatalytic reaction was calculated using an online gas chromatograph (Agilent, GC-8890) equipped with a thermal conductivity detector and high-purity nitrogen was used as the carrier gas.

Results and discussion

Material characterization

The possible reaction route for the TBA/CN framework is shown in Fig. 1. Here, the crystal and molecular structures of the as-synthesized composites were first identified by the XRD and FT-IR analyses. As depicted in Fig. 2a, all the samples possessed two typical diffraction peaks of graphitic materials at around 12.9° and 27.5°, corresponding to the (100) and (002) crystal facets (JCPDS 87-1526),[26] which suggested that the original crystal structure of CN was largely intact after the introduction of small amounts of TBA. Specifically, the front peak was related to the in-plane repeating motifs of tri-s-triazine, while the subsequent peak was attributed to the repeated interfacial stacking of the conjugated aromatic system in the CN nanosheets, respectively.[27,28]

Fig. 1

Presentation of the possible formation processes for 2-thiobarbituric acid modified g-C3N4.

Fig. 2

XRD patterns (a) and FT-IR spectra (b) for pristine CN, TBA/CN-1, TBA/CN-2, TBA/CN-3 and TBA/CN-4.

Moreover, the obtained TBA/CN samples exhibited essentially similar FT-IR feature to that of pristine CN. In detail, the wide band located in the region between 3000 and 3400 cm−1 was derived from the terminal N–H and surface-adsorbed O–H stretching vibrations.[29] Series of characteristic signals in the fingerprint region from 1150 to 1700 cm−1 were related to the different types of aromatic C–N or C–C stretching vibrations in the tri-s-triazine ring and heptazine heterocyclic ring units.[30,31] Finally, the sharp peak at 809 cm−1 was attributed to the unique breathing vibration of the heptazine units.[32] In addition, no evident absorption peaks associated with the bonds of the S atom were detected in the FT-IR spectra mainly due to its low modification level and relatively weak absorption.[33]

The morphology of TBA/CN-3 was observed by SEM and TEM. Fig. 3a indicated that TBA/CN-3 displayed an aggregated multilayer porous two-dimensional structure, which was in good accordance with the typical urea-derived CN structure.[34] The corresponding elemental mapping (Fig. 3b and c) clearly reveal that C, N, O, and S elements were evenly dispersed throughout the TBA/CN-3 samples, demonstrating that TBA was successfully integrated into the CN skeleton. This result could be further confirmed by EDS, exhibiting the presence of C, N, O and S in the TBA/CN-3 sample. The representative TEM image of TBA/CN-3, as given in Fig. 4, also elucidated the ultrathin nanosheet-like structure equipped with abundant nanosized pores, which was largely due to the gas templating effects of urea and TBA.[35] The ultrathin nanosheet morphology and enriched nanoholes on the surface of TBA/CN-3 would provide more active sites and might lead to significant acceleration of the photo-excited charge separation and migration, subsequently resulting in much improved photoactivity.[36,37]

Fig. 3

SEM image (a), the corresponding elemental maps (b and c) and EDS pattern (d) of TBA/CN-3.

Fig. 4

TEM images of TBA/CN-3.

The porous feature of TBA/CN-3 was further identified by N2 adsorption–desorption isotherms. As depicted in Fig. 5a, the isotherms of TBA/CN-3 as well as CN exhibited typical type IV curves featured with H3 hysteresis loops, suggesting their unaltered mesoporous structures.[38] The BJH pore size distribution plots in Fig. 5b demonstrated that both TBA/CN-3 and pure CN had porous structures with average pore sizes of about 4.0 nm, which was quite consistent with the results of TEM. The almost similar parameters of BET specific surface area, pore volume and average pore size further confirmed that the structural integrity of CN remained intact even after the introduction of TBA. Thus, the enhancement of photocatalytic efficiency was mainly due to the incorporation of TBA rather than the surface morphology change of CN.

Fig. 5

N2 adsorption–desorption isotherms of CN, TBA/CN-3 (a) and the corresponding pore-size distribution curves (b).

The XPS analysis was carried out to gain insight into the surface composition and valence states of different elements in CN and TBA/CN-3 samples, and the results are depicted in Fig. 6. From the XPS survey spectra (Fig. 6a), the peaks of C 1s, N 1s and O 1s could be clearly detected in both the samples, while the S 2p peak was only presented in the TBA/CN-3 sample. The weak absorption peaks of S 2p in the survey spectrum of TBA/CN-3 are mainly due to the small introduced content. The C 1s high-resolution XPS spectrum of TBA/CN-3 (Fig. 6b) could be well fitted into three peaks centered at 284.88, 286.60 and 288.13 eV, which correspond to the different chemical environments of carbon including C–C, C–NH2 and N–CN in the heterocycle, respectively.[39,40] In addition, with respect to the C 1s high-resolution XPS spectrum of CN provided in Fig. S1b,† no difference between TBA/CN-3 and CN was detected, indicating the occurrence of little change in the basic skeleton structure of CN. Similarly, the N 1s high-resolution XPS spectrum of TBA/CN-3 (Fig. 6c) was also consistent with that of the pristine CN (Fig. S1c†). Specifically, the three peaks with bond energies of 398.44, 399.80 and 400.94 eV were assigned to C–NC, (C)3–N and terminal C–N–H amino functional groups, respectively.[41] In terms of the S 2p XPS spectrum (Fig. 6d), TBA/CN-3 presented three peaks at 164.05, 165.89 and 169.29 eV, corresponding to the C–S bond, SC bond and oxidized S (S–O) generated during annealing.[42,43] The S atom with extra electrons introduced into the CN skeleton could lead to a significant electronic polarization effect, which would be highly conducive to the migration and conductivity of the charge carriers.[44]

Fig. 6

Survey XPS spectra (a) and high-resolution XPS spectra of C 1s (b), N 1s (c) and S 2p (d) for TBA/CN-3.

The optical behaviors of the as-synthesized samples were evaluated by UV-Vis DRS and the corresponding band gaps were obtained according to the method of Kubelka–Munk.[45] As presented in Fig. 7, pristine CN had a fundamental absorption edge at about 440 nm, with the calculated value of band gap being 2.70 eV. In comparison, the TBA/CN samples displayed a noticeable gradual red-shift and stronger response in the visible light region with the increase in TBA.

Fig. 7

UV-Vis DRS (a), the plot of (αhv)1/2versus the energy of light (hv) spectra (b) and the band edge potentials (c) of the as-prepared samples.

The electronic band structure of VB and CB for the as-synthesized samples could be calculated according to the following formula:[46]where X is the semiconductor absolute electronegativity (CN = ∼4.73 V vs. NHE), Ee is the free electron energy (∼4.5 V vs. NHE).

Therefore, the EVB values of CN and TBA/CN samples were calculated to be 1.58, 1.51, 1.42, 1.32 and 1.27 eV, while the ECB values were −1.12, −1.05, −0.96, −0.86 and −0.82 eV, respectively (Fig. 7c). Hence, it can be clearly deduced that the TBA modification could narrow the band gap of CN, which resulted in more visible light harvest, consequently elevating the generation of photo-excited charges and thus improving the photocatalytic properties.

The separation efficiency and migration behaviors of the photo-excited charges over CN and TBA/CN samples were compared via the PL analysis. Fig. 8a presents the particular strong emission peak of pristine CN at 448 nm, suggesting the immense recombination of photo-excited charges.[47] Compared with CN, the peak intensities of TBA/CN samples significantly reduced and gradually redshifted to 456, 477, 490 and 492 nm for TBA/CN-1, TBA/CN-2, TBA/CN-3 and TBA/CN-4 correspondingly, which were consistent with the results of the UV-Vis DRS analysis. TBA/CN-3 possessed the weakest PL peak intensity, which indicated that the recombination of the photoinduced carriers was most effectively restrained.[48] As shown in Fig. 8b and c, based on Gaussian fitting, the PL emission of CN could be well deconvoluted into three peaks centered at 436.1, 459.6 and 500.6 nm, corresponding to the electron transition pathways of σ*–LP (lone pair state), π*–p, and π*–LP.[49,50] However, these three peaks of TBA/CN-3 clearly redshifted to 471.7, 508.5 and 574.2 nm in comparison to those of CN, suggesting that the electronic transition pathway has changed. Moreover, the improved ratio of peaks B and C of TBA/CN-3 revealed that the corresponding electronic transition types of π*–p, and π*–LP dominated, which was mainly due to the increase in the sp2 CC bonds and S atoms with more valence electrons by the incorporation of TBA.[51]

Fig. 8

PL spectra of the as-prepared samples (a). Schematic of the electron transition form of CN (b). The Gaussian fitting of the PL spectra of CN (c) and TBA/CN-3 (d).

Therefore, these results implied that the loading of TBA with abundant aromatic CC bonds and S heteroatoms could change the electron transition path, thus preventing the charge recombination, which would result in an enhancement of the photocatalytic performance.

Photocatalytic property

The photodecomposition of Rh B was first carried out to evaluate the photocatalytic activity of the as-synthesized samples. As presented in Fig. 9a, a negligible change of Rh B after 90 min visible light exposure demonstrated that it was very stable under the experimental conditions. After adding the photocatalyst, the degradation efficiency significantly enhanced, and the TBA/CN samples showed better degradation effect than the pristine CN. As expected, TBA/CN-3 showed the optimal photocatalytic efficiency and about 92.0% of Rh B could be removed after 90 min of visible light illumination. However, the TBA/CN-4 sample with the narrowest band gap of 2.09 eV exhibited reduced degradation activity, which can be possibly due to the formation of recombination centers and the reduction of charge separation efficiency caused by excessive TBA incorporation.[52] The corresponding degradation kinetics was calculated by the following pseudo-first-order equation: ,[53] and the results are shown in Fig. 9b. The largest rate constant k value of 0.0273 min−1 for TBA/CN-3 was obtained, which was almost 3.5-fold as that of the pristine CN. Fig. 9c depicts the UV-Vis absorption change of Rh B in a given reaction time over TBA/CN-3. With prolonged reaction time, the UV-Vis absorption intensity of Rh B markedly declined, and the characteristic absorption wavelength at 552 nm shifted to 518 nm, indicating the complete disruption of the conjugated aromatic system and chromogenic groups.

Fig. 9

Photocatalytic degradation of Rh B (a), the first-order-kinetic plots of as-prepared samples (b), the UV–vis absorption spectral changes of Rh B (c) and the stability (d) over the TBA/CN-3 sample.

Thereafter, the recycling experiments were performed to evaluate the practical applications of TBA/CN-3. As shown in Fig. 9d, it was noticeable that there was only a slight decline in the photocatalytic degradation efficiency over TBA/CN-3 after three consecutive recycles, representing excellent photostability and reusability.

To reveal the dominant oxidative species involved in the photodecomposition of Rh B, active radical capturing experiments were carried out, in which p-benzoquinone (BQ), isopropanol (IPA) and ammonium oxalate (AO) were introduced as the quenchers of superoxide radicals (˙O2−), hydroxyl radicals (˙OH) and photo-excited holes (h+), respectively.[54,55] As shown in Fig. 10a, the degradation efficiency of Rh B over TBA/CN-3 significantly reduced from 92.0% to 40.5% with the addition of BQ, reduced to 47.9% with the addition of AO and slightly reduced to 75.4% with the addition of IPA. These results indicate that ˙O2− and h+ played the decisive roles with ˙OH being partially involved in the photodegradation of Rh B. To further confirm the above-mentioned results, ESR experiments were conducted to detect the signals of ˙O2− and ˙OH during the photocatalytic process. It could be found from Fig. 10b and c that strong signals typical of ˙O2− and ˙OH were detected in TBA/CN-3, but the signals were relatively weak in CN. It signified that the TBA modification could observably benefit the yield of ˙O2− and ˙OH radicals in the photocatalytic degradation system.

Fig. 10

Effect of different scavengers on the photodegradation of Rh B (a), ESR spectra of DMPO-˙O2− (b) and DMPO-˙OH (c) over TBA/CN-3.

The photocatalytic H2 evolution efficiency of the five as-synthesized samples was determined and the results are illustrated in Fig. 11. It can be seen in Fig. 11a that all the samples achieved stable near-linear generation of H2 with marked difference in the evolution rate. The pristine CN exhibited smallest photocatalytic H2 evolution rate of about 0.115 mmol g−1 h−1 (Fig. 11b), while the TBA/CN samples exhibited gradually enhanced rates with the increase in the concentration of the introduced TBA. The TBA/CN-3 sample possessed optimal H2 evolution rate of 0.438 mmol g−1 h−1, which was almost 3.8 times than that of CN. The enhanced H2 evolution of TBA/CN-3 was largely due to the stronger visible light harvesting and faster photo-exited charge transfer with lower recombination.

Fig. 11

Photocatalytic H2 evolution (a), corresponding H2 evolution rates (b) of as-prepared samples, H2 evolution cyclicity performance (c) and XRD patterns (d) of TBA/CN-3 before and after photocatalysis.

The photocatalytic H2 generation experiments were also repeated three times to ensure the durability of TBA/CN-3. As shown in Fig. 11c, the TBA/CN-3 sample displayed a steady H2 evolution without any apparent decline over three cycles (for 12 h in total), indicating a good photocatalytic stability and reproducibility. Furthermore, it can be seen in Fig. 11d that almost identical XRD patterns of TBA/CN-3 before and after the photocatalytic experiments further confirmed its durability.

Plausible mechanisms of the enhanced photocatalytic properties

In order to reveal the enhanced photocatalytic properties of TBA/CN-3, TRPL, PC value and EIS measurements were carried out. TRPL decay spectra in Fig. 12a illustrate that TBA/CN-3 presented an evident longer average lifetime of 7.44 ns compared to pure g-C3N4 (5.29 ns), suggesting the highly efficient migration of the charge carriers.[56,57] The prolonged lifetime of photo-excited charge was possibly attributable to the established unique electron transfer pathway of N–C–S–C–N (or N–C–C–C–N) via the introduction of TBA (Fig. 1). Thereafter, the PC value and EIS behavior of TBA/CN-3 further consolidated the observed results. As clearly depicted in Fig. 12b, the PC density of TBA/CN-3 was more robust than that of CN, suggesting dramatical enhancement of the photo-excited charge carrier migration.[58] Similarly, TBA/CN-3 exhibited a much reduced EIS Nyquist plot radius than that of pure CN, further confirming the smaller electron transfer resistance and faster charge transfer.[59]

Fig. 12

Comparison of the TRPL decay spectra (a), photocurrent response (b) and EIS Nyquist curves between CN and TBA/CN-3 (c).

In light of the above results, the possible mechanism of the enhanced photocatalytic Rh B degradation and H2 evolution over TBA/CN composites is schematically predicted in Fig. 13. When TBA is incorporated into the skeleton of CN, extended delocalization of the lone pair electrons and unique electron transfer pathways would lead to enhanced visible light absorption, prolonged electron lifetime and facilitated charge migration. Irradiated by visible light, TBA/CN-3 was immediately activated, resulting in the production of photo-excited electrons and holes. The electrons on VB could effortlessly transfer along the bandgap to the CB of TBA/CN-3. Because of the prolonged lifetime, the stored electrons in the CB could have more opportunities to generate ˙O2− by reducing dissolved O2, which was benefited from the more negative potential (−0.86 V vs. NHE) of TBA/CN-3 than that of O2/˙O2− (−0.33 V vs. NHE).[60] Furthermore, these ˙O2− could be further partially converted to ˙OH through complex chain reactions. Moreover, these accumulated electrons could also easily transfer to the Pt co-catalyst and have sufficient reduction capability to product H2 by water splitting on the surface of TBA/CN-3. Because of the lower VB potential (+1.32 V vs. NHE) compared to the standard redox potential of H2O/˙OH (+1.99 V vs. NHE) and OH−/˙OH (+2.80 V vs. NHE),[61] the retained holes were not able to oxidize H2O or OH− to yield ˙OH, but could directly oxidize Rh B or react with TEOA to form TEOA+. As a result, the electrons in the CB and the holes on the VB could be separated efficiently, and these holes and electrons with high strong redox capacity strongly contributed to the enhancement of the photocatalytic activity of TBA/CN-3.

Fig. 13

Possible mechanisms on the photocatalytic degradation of Rh B with simultaneous H2 production over TBA/CN-3 under visible light irradiation.

Conclusion

Overall, the TBA modified g-C3N4 was synthesized by copolymerizing the mixture of urea and TBA. The systematic characterization results proved that the structural units of TBA were well incorporated into the skeleton of g-C3N4 nanosheets. The PL, TRPL and photoelectrochemical experimental results suggested that the TBA/CN samples possessed prolonged lifetime of the photo-excited carriers, enhanced carrier transfer efficiency and reduced photocurrent resistance, which were in favor of the Rh B degradation and H2 evolution. The optimized TBA/CN exhibited 3.5 and 3.8-folds improved visible-light-driven Rh B degradation and H2 evolution activities compared to the pristine CN, respectively. This research may provide a new insight into the rational design of g-C3N4-based photocatalysts for efficiently visible light capture and conversion.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.