Yubin Chen*,
Zhixiao Qin,
Xixi Wang,
Xu Guo and
Liejin Guo*
International Research Center for Renewable Energy, State Key Laboratory of Multiphase Flow in Power Engineering, Xi'an Jiaotong University, Shaanxi 710049, P. R. China. E-mail: ybchen@mail.xjtu.edu.cn; lj-guo@mail.xjtu.edu.cn
First published on 4th February 2015
Developing efficient noble-metal-free photocatalysts is of great importance for the large-scale application of photocatalytic hydrogen production. Herein, low-cost and environment-friendly p-type Cu2S was successfully loaded on n-type CdS photocatalyst by an in situ method to achieve efficient Cu2S/CdS hybrid photocatalysts. Cu2S nanoparticles of ca. 50 nm were intimately assembled on the surface of polyhedral CdS nanocrystals, giving rise to the formation of numerous nanoscale p–n junctions between p-type Cu2S and n-type CdS. Compared to single CdS, Cu2S/CdS exhibited increased photocatalytic hydrogen production under visible light irradiation. The generated nanoscale p–n junctions in Cu2S/CdS, leading to the enhanced charge separation efficiency and better utilization of visible light, were crucial to the improved photocatalytic activity. During the photocatalytic reaction, Cu2S nanoparticles captured the photogenerated holes in CdS and served as the active sites for the surface oxidation reaction. The photocatalytic property of Cu2S/CdS photocatalysts was influenced by the Cu/Cd molar ratio, with the optimal one of 0.05. P-type Cu2S could also be utilized for improving the photocatalytic activities of n-type ZnIn2S4 and n-type TiO2 by forming efficient p–n junctions, indicating the general applicability of p-type Cu2S. This work demonstrates that forming p–n junction structure was a useful strategy for developing efficient noble-metal-free hybrid photocatalysts.
The most common and effective cocatalyst is Pt, which is widely applied for various photocatalysts. Other noble metals, such as Ru, Rh, Au and Pd, are also effective cocatalysts owing to the similar physicochemical properties like Pt.5,6 Nonetheless, the high price and low reserve of noble metals limit the large-scale application. Therefore, seeking efficient noble-metal-free photocatalysts is of critical importance. Recently, it has been demonstrated that a series of transition metal sulfides could promote the photocatalytic hydrogen production instead of Pt. Li's group reported that loading MoS2 or WS2 on CdS photocatalysts could give rise to efficient charge separation by forming intimate heterojunctions to achieve high activities.7,8 Xu's group revealed that NiS could act as efficient cocatalyst to significantly improve the hydrogen production of CdS photocatalysts. The apparent quantum efficiency of NiS/CdS at 420 nm reached 51.3%.9 Xie's group discovered that CuS could accelerate the photogenerated charge separation of CdS photocatalysts and subsequently the photocatalytic hydrogen production.10 Besides of CdS, the photocatalytic properties of various semiconductor photocatalysts such as CdxZn1−xS, CuGa3S5, TiO2, ZnO and C3N4 could also be enhanced by the deposition of such metal sulfides, indicating their general applicability.11–15 However, up to now, highly efficient transition-metal sulfide promoters with low price and earth-abundant elements are still limited. Meanwhile, the reported action mechanisms of such metal sulfides are quite different. As a consequence, it is necessary to study the general coupling principle between photocatalysts and such transition-metal sulfides and as a guidance to develop efficient noble-metal-free hybrid photocatalysts.
The basic requirement of coupling different semiconductors to enhance the photocatalytic activity is to form efficient heterojunction structure between different components, which is crucial to the improvement of charge separation efficiency. For example, as noble metal is deposited on semiconductor photocatalysts, the Schottky junction is generated between the noble metal and the semiconductor photocatalyst, which can lead to the efficient charge separation.4 Coupling semiconductors with different band structures to generate heterojunction structure can also result in the efficient charge separation.16 When p-type semiconductor is combined with n-type semiconductor, p–n junction can be formed between the two semiconductors. Across the p–n junction region, a built-in electric field is achieved, which can efficiently promote the separation of photogenerated electrons/holes.17 As we know, a number of reported photocatalysts are n-type semiconductors, such as CdS, ZnIn2S4 and TiO2.18–20 Therefore, Coupling suitable p-type semiconductors to promote the photocatalytic properties of n-type semiconductor photocatalysts can be expected. P-type NiO and p-type NiS have been demonstrated to be efficient cocatalysts to promote the photocatalytic activities of n-type semiconductor photocatalysts by forming p–n junctions.21,22 However, more efficient p-type semiconductor promoters, especially the ones with wide applicability should be explored.
Cu2S is a p-type semiconductor with a bulk bandgap of 1.2 eV.23 Due to its low cost and earth-abundant elements, Cu2S has been utilized for solar cells, optoelectronic devices and photocatalytic degradation of organic pollutants.24–28 However, the utilization of Cu2S in the area of photocatalytic hydrogen production has been seldom reported.29 Herein, we firstly prepared n-type CdS nanocrystals with polyhedral morphology. Then Cu2S nanoparticles were deposited by an in situ method to obtain Cu2S/CdS photocatalysts. A number of well-dispersed nanoscale p–n junctions were developed in the hybrid photocatalysts, which could efficiently improve the charge separation, leading to the much enhanced photocatalytic activity. P-type Cu2S was also coupled with n-type TiO2 and n-type ZnIn2S4. Interestingly, the improved activities were also achieved by forming nanoscale p–n junction structure, indicating the good applicability of p-type Cu2S.
![]() | (1) |
To further analyse the valance states of copper in CuxS and CuxS/CdS, XPS measurements were carried out. As indicated in Fig. 2a, the Cu 2p XPS spectrum of CuxS revealed that the binding energies of Cu 2p3/2 and Cu 2p1/2 peaks were respectively located at 931.8 and 951.8 eV, which corresponded to the typical values for Cu+ in Cu2S.33,34 Meanwhile, there were no “shake-up” peaks in the higher binding energy direction, indicating the absence of CuS. In addition, the generation of Cu2S could be demonstrated by the associated Auger line (Cu LMM) at 569.6 eV in Fig. 2b, which was the typical value of Cu2S.33 It was noted in Fig. 2 that the observed Cu 2p XPS and Auger Cu LMM peaks of CuxS/CdS were at the same positions with those for single CuxS, which revealed that copper in CuxS/CdS was also in the form of Cu+. On the basis of XRD and XPS results, we can summarize that Cu2S and Cu2S/CdS photocatalyst were finally synthesized in the present study.
As presented in the experimental section, Cu2S was obtained by adding 0.005 mol of Cu(NO3)2·3H2O into 190 mL of 0.25 M Na2SO3/0.35 M Na2S solution. In order to clarify the reaction process, 0.005 mol of Cu(NO3)2·3H2O was respectively added into 190 mL of 0.25 M Na2SO3 or 0.35 M Na2S solution for comparison. When Cu2+ was added into Na2SO3 solution and stirred for 0.5 h, the solution was still transparent and no solid products could be obtained. In the case of Cu2+ into Na2S solution, black powders could be achieved, which could be denoted as CuxS (Na2S). The XRD pattern of CuxS (Na2S) showed the typical diffraction peaks assigned to both Cu2S and CuS (Fig. S2†). It is inferred that as Cu2+ was added into the solution with Na2S, the eqn (2) would occur to form CuS due to its low solubility product (Ksp) of 6.3 × 10−36.35 Since the amount of S2− (0.0665 mol) in the solution was much higher than the amount of added Cu2+ (0.005 mol), parts of CuS could be reduced to Cu2S by excess S2− according to the eqn (3). Therefore, both Cu2S and CuS were generated by adding Cu2+ into Na2S solution. However, when Na2SO3 was present in the reaction solution, SO32− would react with S22− by the eqn (4), accelerating the formation of Cu2S by the eqn (3). Meanwhile, CuS could be reduced to Cu2S by SO32− according to the eqn (5). As a consequence, only Cu2S was achieved by adding Cu2+ into the solution with both Na2SO3 and Na2S.
Cu2+ + S2− → CuS | (2) |
2CuS + S2− → Cu2S + S22− | (3) |
S22− + SO32− → S2O32− + S2− | (4) |
2CuS + SO32− + 2OH− → Cu2S + SO42− + S2− + H2O | (5) |
In order to further confirm the successful deposition of Cu2S nanoparticles, EDX-SEM elemental mapping analysis of Cu2S/CdS-0.05 was carried out. As displayed in Fig. 4, such mapping results demonstrated the presence of cadmium, copper and sulfur. In addition, it was observed in Fig. 4c that copper was uniformly dispersed, which indicated that well-dispersed heterojunctions between CdS and Cu2S could be generated. The deposition process of Cu2S nanoparticles could be illustrated in Fig. S3.† CdS polyhedrons, serving as the substrates to deposit Cu2S nanoparticles, were firstly dispersed into Na2SO3/Na2S aqueous solution. Appropriate amount of aqueous solution containing Cu2+ ions was then dripped into the above suspension. Cu2S would subsequently nucleate on the surface of polyhedral CdS nanocrystals and grew into crystallized nanoparticles. The Cu/Cd molar ratios of different Cu2S/CdS photocatalysts were also determined by EDX system coupled with FESEM. As shown in Table S1,† with the increase of added Cu precursor, the Cu/Cd molar ratio by EDX gradually increased. Meanwhile, the measured Cu/Cd molar ratios are a little higher than the added ratios in the synthesis of Cu2S/CdS samples, possibly due to Cu2S nanoparticles deposited on the surface of polyhedral CdS nanocrystals (the analysis diameter and depth for the used EDX system is ca. 1.0 μm).
![]() | ||
Fig. 4 (a) FESEM image of Cu2S/CdS-0.05 and (b–d) the corresponding EDX-SEM mapping images of Cd, Cu and S in Cu2S/CdS-0.05. The scale bar is 2 μm in each image. |
TEM and HRTEM measurements were performed to identify the detailed structure of Cu2S/CdS-0.05 hybrid photocatalysts. Fig. 5a shows a typical CdS polyhedron deposited by the highly dispersed Cu2S nanoparticles with a diameter of ca. 50 nm (marked by red circles), which led to the formation of nanoscale heterojunctions between the CdS nanocrystal and Cu2S nanoparticles. A close examination in Fig. 5b revealed that the contact between the two semiconductors was intimate, where the nanoscale heterojunction was at a lattice level. It was also discovered that the Cu2S nanoparticle was completely crystalline. The clear lattice fringes with the interlayer spacing of 0.32 nm could be indexed to the (111) plane of cubic Cu2S.31,32
![]() | ||
Fig. 6 UV-vis absorption spectra of Cu2S/CdS hybrid photocatalysts with different Cu/Cd molar ratios. |
Fig. 7 shows the room-temperature PL spectra of CdS and Cu2S/CdS-0.05 under excitation at 420 nm. It was observed that the as-prepared CdS photocatalysts displayed a broad spectrum in the range of 500–850 nm with the main peak located at ca. 570 nm, which was close to the absorption onset of CdS (ca. 540 nm). The broad PL band generally revealed that a number of trapped states were formed in the forbidden band of CdS.37 Wang et al. reported that the luminescence centered at 750 nm observed from the synthesized CdS nanowires should be trapped emission, which was mainly due to the surface defects (such as sulfur vacancies).38 In our PL study, the emission between 650 and 850 nm was also considered to result from the surface defects of CdS. Such surface defects could act as the recombination centers of photoexcited charges to lower the photocatalytic efficiency for hydrogen production. As depicted in Fig. 7, the PL spectrum of Cu2S/CdS-0.05 was similar to that of single CdS. However, the intensity was lower, verifying the higher charge separation efficiency in Cu2S/CdS-0.05.39 In particular, the emission intensity between 650 and 850 nm corresponding to the surface defect states of CdS was apparently decreased. It was reported by Kamat's group that CuxS could interact with defect states on the CdSe surface and quench emission resulting from selenide vacancies by forming p–n interfacial junction.40 In the hybrid Cu2S/CdS photocatalysts, similar mechanism can be expected. It was considered that Cu2S could interact with the surface defects of CdS (such as sulfur vacancies), which could quench the emission between 650 and 850 nm from the surface trap states, leading to the much more efficient surface charge separation in Cu2S/CdS hybrid photocatalysts.
As analysed from the PL results of CdS and Cu2S/CdS photocatalysts (Fig. 7), Cu2S deposition could lead to the more efficient charge separation in CdS, which was considered to be the key factor for the improved activity of Cu2S/CdS. The schematic diagram of the charge separation process in Cu2S/CdS could be illustrated in Fig. 9. As mentioned above, CdS and Cu2S are respectively n-type and p-type semiconductors. Meanwhile, as shown in Fig. S5,† a negative slope of the linear region in the Mott–Schottky plot of Cu2S film indicated the p-type semiconductor character of Cu2S.27 When n-type CdS polyhedrons were coupled with p-type Cu2S nanoparticles, large amounts of nanoscale p–n junctions could be generated between the two semiconductors (Fig. 5). Across the p–n junction a built-in electric field is formed, where the p-type Cu2S region is negatively charged, and the n-type CdS region is positively charged (Fig. 9a). Considering the band structures of Cu2S and CdS, it has been reported that both the conduction band (CB) and the valance band (VB) of Cu2S are higher than those of CdS.25,42 When Cu2S and CdS formed p–n heterojunction, the electron transfer occurred from Cu2S to CdS while the hole transfer occurred from CdS to Cu2S due to the different Fermi levels until the system attained equilibration. The proposed band profile for Cu2S/CdS heterojunction was shown in Fig. 9b.24,43
When Cu2S/CdS hybrid photocatalysts were irradiated with visible light, the photoexcited electrons and holes would be both obtained in CdS and Cu2S. As a result of the built-in electric field, the photogenerated electrons in the CB of Cu2S would diffuse into the CB of CdS through the p–n junction, giving rise to the accumulation of photogenerated electrons in the CdS nanocrystal. Likewise, the photogenerated holes in the VB of CdS would diffuse into the VB of Cu2S, leading to the accumulation of photogenerated holes in the Cu2S nanoparticle. As a consequence, the efficient charge separation was successfully achieved by the nanoscale p–n junctions in Cu2S/CdS. Meanwhile, The close interaction between CdS and Cu2S (Fig. 5b) could facilitate the charge transfer.44 Subsequently, the accumulated electrons would transfer to the surface of CdS to reduce H+ for hydrogen production, and the accumulated holes would transfer to the surface of Cu2S to oxidize the sacrificial reagents (S2−, SO32−). The improved photocatalytic activity was acquired through the efficient charge separation. During the photocatalytic reaction, Cu2S nanoparticles captured the photogenerated holes in CdS and functioned as the active sites for the surface oxidation reaction.
Although the charge separation could be achieved for Cu2S/CdS with different Cu/Cd molar ratios through the nanoscale p–n junctions, the efficiency of charge separation was not the same. As the Cu/Cd molar ratio is below 0.05, Cu2S nanoparticles were uniformly dispersed on the surface of CdS polyhedron. More nanoscale p–n junctions could be generated with the increased Cu2S nanoparticles deposited, resulting in higher charge separation efficiency and better photocatalytic performance. However, when the Cu/Cd molar ratio was above 0.05, the aggregation of Cu2S nanoparticles occurred, which restricted the formation of more p–n junctions. Meanwhile, the grain boundaries of Cu2S nanoparticles would act as the recombination centers, leading to lower charge separation efficiency.45 Further increasing the amount of Cu2S would give rise to the decreased photocatalytic activity.
Single Cu2S could not produce hydrogen possibly owing to its low charge separation efficiency. However, in Cu2S/CdS hybrid photocatalysts, the efficient charge separation from nanoscale p–n junctions was achieved, which could promote the utilization of photoexcited charges in Cu2S for photocatalytic reaction. This result implied that more visible light could be harvested by Cu2S/CdS for photocatalytic hydrogen production compared to single CdS, which should also contribute to the enhanced activity.
In order to demonstrate whether p-type Cu2S can enhance the activities of other n-type semiconductor photocatalyst. Cu2S was respectively coupled with n-type ZnIn2S4 and n-type TiO2 to generate Cu2S/ZnIn2S4 and Cu2S/TiO2 hybrid photocatalysts with p–n junction heterostructures. The photocatalytic hydrogen production of different photocatalysts with and without Cu2S were tested. The detailed synthesis and photocatalytic process of Cu2S/ZnIn2S4 and Cu2S/TiO2 photocatalysts can be found in the ESI.† As shown in Table 1, the hydrogen production rates of single ZnIn2S4 and TiO2 were rather low. After loading Cu2S, both photocatalytic activities were apparently improved. This result indicated that Cu2S was suitable for various photocatalysts. No matter the photocatalyst is either metal sulfide or metal oxide, Cu2S could efficiently promote the photocatalytic hydrogen production. The nanoscale p–n junctions between p-type Cu2S and n-type semiconductor, leading to more efficient charge separation, are considered to be the key factor for the improved activities. The present study thus provides a general strategy to develop efficient Cu2S-modified photocatalysts by forming p–n junction structure.
Sample | Hydrogen production rate without Cu2S | Hydrogen production rate with Cu2S |
---|---|---|
ZnIn2S4 | 5.0 μmol h−1 | 30.9 μmol h−1 |
TiO2 | 0.5 μmol h−1 | 16.0 μmol h−1 |
Footnote |
† Electronic supplementary information (ESI) available: The detailed synthesis and photocatalytic process of Cu2S/ZnIn2S4 and Cu2S/TiO2 photocatalysts; the calculation method of the number of incident photons; the Cu/Cd molar ratios determined by EDX of different Cu2S/CdS samples; the schematic diagram of the photocatalytic reactor for hydrogen production; XRD patterns of CuxS (Na2S) sample and Cu2S (Na2SO3/Na2S) sample (Cu2S achieved by adding Cu2+ into Na2SO3/Na2S solution); illustration of the deposition process of Cu2S nanoparticles on CdS polyhedrons; Cu 2p XPS spectra of Cu2S/CdS-0.05 sample before and after long-time photocatalytic reaction; Mott–Schottky plot of Cu2S film. See DOI: 10.1039/c5ra00091b |
This journal is © The Royal Society of Chemistry 2015 |