Wei Chena,
Tian-Yu Liua,
Ting Huanga,
Xiao-Heng Liu*a,
Guo-Rong Duana,
Xu-Jie Yanga and
Shen-Ming Chen*b
aKey Laboratory of Education Ministry for Soft Chemistry and Functional Materials, Nanjing University of Science and Technology, Nanjing 210094, China. E-mail: xhliu@mail.njust.edu.cn
bElectroanalysis and Bioelectrochemistry Lab, Department of Chemical Engineering and Biotechnology, National Taipei University of Technology, No.1, Section 3, Chung-Hsiao East Road, Taipei 106, Taiwan, Republic of China. E-mail: smchen78@ms15.hinet.net
First published on 11th November 2015
In this report, we first successfully designed and fabricated novel C,N co-doped titanium dioxide nanoparticles/graphite-like carbon nitrogen ultrathin nanosheets (C,N-TiO2 NPs/g-C3N4) heterostructures, wherein the C,N-TiO2 NPs were in situ grown on the porous g-C3N4 ultrathin nanosheets (NSs) by a simple one-pot solvothermal route with the assistance of concentrated nitric acid. The resulting C,N-TiO2 NPs/g-C3N4 nanocomposite photocatalysts were systematically characterized by powder X-ray diffraction (PXRD), Fourier transform infrared spectroscopy (FT-IR), diffuse reflectance spectroscopy (DRS), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), photoluminescence (PL) spectroscopy, transient photocurrent–time (I–t) curves and electrochemical impedance spectroscopy (EIS) Nyquist plots. The photocatalytic ability was evaluated by photocatalytic water splitting for hydrogen evolution. These studies indicate that C,N-TiO2 NPs/g-C3N4 composites exhibit superior ability for hydrogen generation compared to single C,N-TiO2 NPs and pure g-C3N4 NSs under visible light illumination. The optimal composites with 3 wt% C,N-TiO2 NPs/g-C3N4 showed the highest hydrogen evolution rate of 39.18 μmol g−1 h−1, which is about 10.9 and 21.3 times higher than those of C,N-TiO2 NPs and pure g-C3N4 NSs, respectively. The improved photocatalytic H2 evolution can be attributed to improved optical absorption and the lengthening lifetime of charge carrier pairs as a result of the C,N elemental codoping and the construction of intimate heterogeneous interfaces. This simple and feasible method for the fabrication of highly-efficient visible light responsive catalysts provides a great applied potential in energy generation.
So now many practices with different fabrication routes have proven that elemental doping, such as C, N, S, F, Ti doping or co-doping, is a feasible strategy to modify the photocatalytic activity of TiO2 materials, which would not only narrow the band gap of TiO2 to increase the light absorption but also introduce more active sites to lengthen the lifetime of photogenerated holes and electrons pairs.13–15 To further optimize the photocatalytic activity, on the basis of original TiO2 or doped TiO2, researchers have constructed the heterogeneous composites by coupling with other materials.16 Graphitic carbon nitrogen (g-C3N4), with the features of appropriate band gap, abundant available, simple preparation and highly stability at different environments, were widely composited with TiO2 and modified TiO2. Recently, Wang et al.17 reported the fabrication of TiO2/g-C3N4 by calcination route and their highly-efficient photocatalytic activity for hydrogen generation from water splitting. Li et al.18 utilized the similar synthetic strategy to prepare the Ti3+ self-doped TiO2/g-C3N4 heterojunctions, which shows significantly enhanced photocatalytic performance under LED light irradiation. Also, Zhou et al.19 reported obvious photocatalytic enhancement of hybrid TiO2/g-C3N4 via ball milling method. Obviously, although the enhancement of photocatalytic activity has been realized as results of the construction of heterogeneous junctions by coupling between TiO2 (modified TiO2) and g-C3N4, the deficiency of these researches focuses on the composite with bulk g-C3N4, which often exhibits high recombination rate of photo-generated charge carriers due to the formation of a vast grain boundary defects during the thermal polycondensation and the low specific surface area.20 Han et al.21 in situ prepared the N-TiO2/g-C3N4 nanosheets heterostructures by a simple electrospinning process combined with a modified heat-etching method, which present highly efficient degradation activity and photocatalytic H2 production with the exposure of simulated solar light.
To our knowledge, there is no report involving the synthesis of C,N-TiO2 NPs/g-C3N4 ultrathin nanosheets heterostructures until now. Herein, we demonstrate the simple and novel strategy to prepare the C,N-TiO2 NPs/g-C3N4 NSs nanocomposites with highly efficient catalytic performance for hydrogen generation. The fabrication strategy of C,N-TiO2 NPs/g-C3N4 is simple and exclusive of calcination treatment at high temperature condition. Different from the previous reports of C-TiO2 and N-TiO2,5,7 here the anion group (–C4H9O) in tetrabutyl titanate (TBT) serve as carbon doping agent and the addition of concentrated HNO3 is used as inhibitor to control the rate of hydrolysis of TBT and employed as the source of N doping reagent.
![]() | ||
Fig. 1 (a) Typical XRD pattern, (b) FT-IR spectra, (c) UV-vis diffuse reflectance spectra and (d) corresponding to Kubelka–Munk transformed function of different photocatalysts. |
The FT-IR spectra of pure g-C3N4 NSs and a series of C,N-TiO2 NPs/g-C3N4 composites are displayed in Fig. 1b. For pristine g-C3N4 NSs, a wide absorption band between 3000 cm−1 and 3600 cm−1 corresponds to stretching vibration of O–H of the absorbed water molecule and N–H stretching vibration modes of residual NH2 attached to the sp2 hybridized carbon or NH groups at the defect sites of the aromatic ring.25 While several strong characteristic peaks located between 1200 cm−1 and 1700 cm−1 are explained as typical stretching vibration of CN heterocycles.26 Additionally, a sharp peak at 808 cm−1 is assigned to breathing mode of triazine units.27 The FT-IR spectra of C,N-TiO2 NPs/g-C3N4 composites exhibit the overlap of the spectra of both C,N-TiO2 NPs (Fig. S2†) and g-C3N4 NSs, which suggests the coexistence of C,N-TiO2 NPs and g-C3N4 NSs in composite samples. These results are in well accordance with XRD analysis.
The optical properties of as-prepared C,N-TiO2 NPs, pure g-C3N4 NSs as well as C,N-TiO2 NPs/g-C3N4 composites were detected by UV-vis diffuse reflectance spectra (DRS) and the corresponding band gap energies were estimated by Kubelka–Munk equation. Fig. 1c shows the DRS spectra of the hybrid C,N-TiO2 NPs/g-C3N4 photocatalysts, from which, we can clearly found that g-C3N4 exhibits typical light absorption in the region ranging from 200 nm to 450 nm, which is ascribed to the charge transfer response of g-C3N4 NSs from the valence band formed with N 2p orbits to the conduction band populated by C 2p orbits.28 Interestingly, the C,N-TiO2 NPs is able to absorb visible light with the maximum wavelength up to 460 nm as a result of the doping level generated by C and N atom co-doping. Compared with that of single component, all the C,N-TiO2 NPs/g-C3N4 samples exhibit significantly improved light-harvesting capability. Such red shifts of light absorption might be attributed to the beneficial interaction between C,N-TiO2 NPs and g-C3N4 NSs resulted from the construction of close heterojunction interface. The band gap energies (Fig. 1d) of C,N-TiO2 NPs and pure g-C3N4 NSs calculated by Kubelka–Munk formula are 2.95 eV and 2.70 eV, respectively.
The sample morphology and microstructure information were investigated by TEM and HRTEM, as shown in Fig. 2. As expected, the C,N-TiO2 NPs with the diameter of several nanometers were homogeneously immobilized on the surface of ultrathin g-C3N4 NSs by one-pot solvothermal route, which is favorable to the formation of intimate heterogeneous interface. The size of C,N-TiO2 NPs in C,N-TiO2 NPs/g-C3N4 heterojunction photocatalysts is similar to that of pristine C,N-TiO2 NPs and no agglomeration structures of C,N-TiO2 NPs located on the surface of g-C3N4 NSs have be observed. This good dispersion is benefit to the separation of photo-induced holes and electrons pairs and thus is responsible for the enhancement of photocatalytic performance. From the high-resolution image of C,N-TiO2 NPs/g-C3N4 composites, the sharp-edged lattice images ascribed to C,N-TiO2 NPs are presented with the lattice spacing of 0.352 nm, corresponding to (101) crystallographic planes of anatase TiO2.
To verify the elemental categories and chemical status in C,N-TiO2 NPs/g-C3N4 composites, the X-ray photoelectron spectra (XPS) of C,N-TiO2 NPs, g-C3N4 NSs as well as C,N-TiO2 NPs/g-C3N4 composites were measured. Fig. 3a displays the XPS survey of these three samples. It is obvious that C,N-TiO2 NPs/g-C3N4 composites is composed of C, N, Ti and O elements, which well confirms the coexistence of C,N-TiO2 NPs and g-C3N4 NSs in C,N-TiO2 NPs/g-C3N4 hybrid samples. The typical high resolution XPS spectra of C 1s, N 1s and Ti 2p in different samples are also shown in Fig. 3b–d. The C 1s high-resolution spectrum displays two peaks centering at 284.8, and 288.0 eV for pure g-C3N4 NSs, 284.8 and 287.9 eV for composite sample, which can be attributed to the contaminated carbon and the sp2-bonded carbon (–NC–N) in graphitic carbon nitride, respectively.29 The C 1s peak, originating from the sp2 carbon, shifts to a lower binding energy direction after coupling with C,N-TiO2 NPs, indicating the variation of chemical environment of sp2 carbon via chemical bonds. While the N 1s spectrum of g-C3N4 NSs can be deconvoluted into four different Gaussian–Lorentzian peaks centered at the binding energies of 398.3, 398.9, 400.5 and 404.2 eV. The main peak at 398.3 eV could be ascribed to nitrogen atoms sp2-hybridized to carbon atom (C
N–C).30 The characteristic peak at 398.9 eV is related to either N–(C)3 groups linking structural motif (C6N7) or amino groups ((C)2–N–H) connecting with structural defects and incomplete condensation.25,31 The weak peak with the binding energy of 400.5 eV is resulted from N bonded to three carbon atoms to develop the N–(C)3 group in the aromatic cycles.32 However, due to the coupling with C,N-TiO2 NPs, the peaks belonging to N 1s at 398.9 and 400.5 eV separately shift to 399.0 and 400.7 eV, which means the presence of interaction between C,N-TiO2 NPs and g-C3N4 NSs. We attribute this interaction to the generation of C–N–Ti bonds as evidence provided by high-resolution XPS spectra of Ti 2p region. The binding energies of Ti 2p3/2 and Ti 2p1/2 are located at 458.5 and 464.2 eV, which are very close with values of Ti4+ in N-TiO2.33 The obvious shifts of Ti 2p peaks in hybrid sample with respect to that of Ti4+ in C,N-TiO2 NPs further confirm the existence of C–N–Ti bonds at the interface of C,N-TiO2 NPs and g-C3N4 NSs.
![]() | ||
Fig. 3 XPS spectra of original C,N-TiO2 NPs, pure g-C3N4 and 3 wt% C,N-TiO2 NPs/g-C3N4: (a) survey spectra, (b) C 1s, (c) N 1s and (d) Ti 2p region. |
The photocatalytic activities of those C,N-TiO2 NPs/g-C3N4 composites were investigated by hydrogen production from water splitting with the addition of 10 vol% TEOA as sacrificial reagent under the visible light irradiation. Controllable experiments suggested that no appreciable H2 produced in the absence of either photocatalyst or light irradiation. Fig. 4 displays the H2 generation curves (Fig. 4a) and the average rate of H2 evolution (Fig. 4b) over C,N-TiO2 NPs/g-C3N4 hybrid samples with different concentration of C,N-TiO2 NPs loadings. As shown in Fig. 4a, the total amount of H2 evolution during the 8 h photoredox reaction increases upon enhancing the C,N-TiO2 NPs contents from 0 to 3 wt% and then decreases with further enhancement to 5 wt%. Obviously, the pure C,N-TiO2 NPs and g-C3N4 NSs alone show negligible photocatalytic performance for H2 production with average value of 3.59 μmol g−1 h−1 and 1.84 μmol g−1 h−1, respectively. The extremely low photocatalytic activity of g-C3N4 NSs possibly originates from some vast grain boundary defects formed during the high temperature preparation process, which hugely promote the faster recombination of photo-induced holes and electrons pairs.19 However, after assembly of C,N-TiO2 NPs onto the g-C3N4 NSs, the obtained C,N-TiO2 NPs/g-C3N4 nanocomposites can response broader visible light region and form more photo-induced holes and electrons, thus benefiting the photocatalytic activities for H2 evolution. In this hybrid system, the optimal content of C,N-TiO2 NPs immobilized on the g-C3N4 NSs to achieve the highest photocatalytic water splitting is determined to 3 wt%. The average H2 evolved rate over 3 wt% C,N-TiO2 NPs/g-C3N4 is 39.18 μmol g−1 h−1, which is ca. 10.9 times higher than that of sole C,N-TiO2 NPs and 21.3 times higher than that of pure g-C3N4 NSs.
![]() | ||
Fig. 4 (a) Hydrogen evolution and (b) corresponding to hydrogen production rate from water splitting by different samples under visible light irradiation. |
The important factors for 3 wt% C,N-TiO2 NPs/g-C3N4 nanocomposites with the highest photocatalytic activity lie in the synergistic effect of element doping of C and N into TiO2 and the construction of intimate heterojunction structure between C,N-TiO2 NPs and g-C3N4 NSs, namely, the enhancement of light absorption and the lengthen of charge carrier lifetime. The C,N-TiO2 NPs in hybrid system could remarkably strengthen the light absorption to extend into visible light region, from which, the doped TiO2 NPs here was similar with the role cocatalyst played. Moreover, due to the construction of intimate heterojunction interface and well-matched band edge, the C,N-TiO2 NPs also could act as electron acceptors and transfer channels, thus improving the charge separation efficiency. As can be seen in Fig. 5, upon photo-irradiation, C,N-TiO2 NPs and g-C3N4 could easily absorb visible light to generate hole–electron pairs. According to previous reports, the conduction band (CB) and valence band (VB) potential could be determined to be −1.12 and +1.58 eV for g-C3N4, −0.29 and 2.91 eV for pristine TiO2.18,22 Therefore, the photo-generated electrons in conduction band (CB) of the g-C3N4 would directly inject into that of C,N-TiO2 NPs, meanwhile the holes could migrate from C,N-TiO2 NPs to VB of g-C3N4. However, the electrons accumulated in the CB of C,N-TiO2 NPs would be extracted by oxygen vacancy originated from elemental doping.34 Such electron transfer paths effectively suppress their recombination thus benefiting the photocatalytic activity.
![]() | ||
Fig. 5 Schematic illustration of charge transfer and separation in C,N-TiO2 NPs/g-C3N4 photocatalysts for hydrogen production under visible light irradiation. |
Besides photocatalytic activity, the photo-stability is also an important parameter to estimate the catalysts for practical application. Therefore, recycling experiments of water splitting for H2 generation were carried out at same photocatalytic condition. Fig. 6 shows the H2 evolution curve of 3 wt% C,N-TiO2 NPs/g-C3N4 photocatalyst in four consecutive days cycling photocatalytic runs. The results indicate that the photocatalytic performance of the composite presents no apparent deactivation with the exposure of irradiation for 32 h, indicating the excellent stability of C,N-TiO2 NPs/g-C3N4 photocatalysts.
![]() | ||
Fig. 6 Recycling tests of hydrogen evolution over 3 wt% C,N-TiO2 NPs/g-C3N4 nanocomposites under visible light irradiation. |
It was well known that many factors, including the crystallinity, morphology, phase structure, light absorption, separation rate of photogenerated hole–electron pairs, play a key role in affecting the photocatalytic activity.35,36 In the current work, the similarity of crystallinity, phase structure and microstructure should not be responsible for the remarkable enhancement of photocatalytic performance for H2 production. The most important point deciding the photocatalytic activity may be the construction of close heterojunction interface, which dramatically faster the separation rate of charge carrier pairs. Conceivably, the importance of the heterogeneous interface in prompt charge migration was also illustrated in several previous reports.37–41 To further prove this hypothesis, the faster separation rate of charge carrier pairs of C,N-TiO2 NPs/g-C3N4 composites were characterized by photocurrent responses, EIS measurements and PL spectra. Fig. 7a exhibits the photocurrent–time (I–t) curves of C,N-TiO2 NPs, pure g-C3N4 as well as C,N-TiO2 NPs/g-C3N4 hybrid material electrodes with several on–off cycles of intermittent irradiation. Obviously, the photocurrent obtained using 3 wt% C,N-TiO2 NPs/g-C3N4 is significantly enhanced as comparison with the unitary component. As we known, the photocurrent is formed mainly by the diffusion of the photogenerated electrons to the back contact and meanwhile the photoinduced holes are taken by the hole acceptor in the electrolyte.42 As a result, the enhanced photocurrent responses over the 3 wt% C,N-TiO2 NPs/g-C3N4 hybrid composites reveals better charge separation and efficient electron transfer within the hybrid structure compared with those of bare C,N-TiO2 NPs and pure g-C3N4. The more efficient transfer of charge carriers could be presented by the result of electrochemical impedance spectroscopy (EIS). The smaller frequency semicircle of the arc in an EIS Nyquist plot indicates the smaller resistance at the interface and the smaller charge-transfer resistance on the electrode surface.43 The results of description in Fig. 7b present the sequence of the frequency semicircles is: g-C3N4 > C,N-TiO2 NPs > 3 wt% C,N-TiO2 NPs/g-C3N4, which suggests the highest separation rate of charge carrier pairs over 3 wt% C,N-TiO2 NPs/g-C3N4 photoanode, lowest over g-C3N4 photoanode. This also confirms that the construction of heterogeneous interface induces a more efficient separation of the photo-excited electron–hole pairs and lengthens the lifespan of the photo-generated charge carriers.
![]() | ||
Fig. 7 (a) Transient photocurrent responses and (b) electrochemical impedance spectroscopy (EIS) Nyquist plots of C,N-TiO2 NPs, g-C3N4 and C,N-TiO2 NPs/g-C3N4 nanocomposites in 0.1 M Na2SO4 solution. |
Photoluminescence (PL) spectroscopy originated from the recombination of photo-induced holes and electrons, the intensity of which could reflect the recombination rate of those free charge carrier pairs.44,45 Fig. 8 shows the fluorescence emission spectroscopy of pure g-C3N4 sample and 3 wt% C,N-TiO2 NPs/g-C3N4 hybrid sample. It is clear that the fluorescence intensities of 3 wt% C,N-TiO2 NPs/g-C3N4 is smaller than that of g-C3N4, signifying that the hybrid sample has lower recombination rate of the photoinduced electron–hole pairs than pure g-C3N4 sample. These results are in well agreement with the photocurrent test and EIS result, which further confirm it is the heterojunction interface that promotes the enhancement of separation rate of photo-generated hole–electron pairs, thereby significantly improving photocatalytic efficiency.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra18302b |
This journal is © The Royal Society of Chemistry 2015 |