Enhanced Photocatalytic Performance under Visible and Near-Infrared Irradiation of Cu1.8Se/Cu3Se2 Composite via a Phase Junction

A novel Cu1.8Se/Cu3Se2 composite photocatalyst was prepared by the simple precipitation method. This composite possesses a wide photoabsorption until the range of near-infrared light, and exhibits significantly enhanced photocatalytic activity for methyl orange degradation under visible and near-infrared light irradiation compared with bare Cu1.8Se and Cu3Se2. The mechanism of this outstanding photocatalytic behavior can be explained by the calculated energy band positions. The efficient charge separation via a phase junction of Cu1.8Se/Cu3Se2 composite would make a great contribution to its much-enhanced photocatalytic efficiency.


Introduction
Semiconductor photocatalysis for the degradation of organic pollutants using solar light energy is a promising approach to solving environmental issue. Since TiO 2 was first reported on photocatalytic water splitting under ultraviolet (UV) light [1], photocatalysts have been widely studied in recent years [2][3][4][5]. Recently, photocatalysts with outstanding photocatalytic properties have been found, such as TiO 2 and ZnO in the UV spectral range [6,7], and CdS and Bi 2 WO 6 in the visible range [8,9], and only few semiconductors present near-infrared (NIR) photoactivity, such as Cu 2 (OH)PO 4 and WS 2 [10,11]. Nowadays, the development of effective ways to exploit near-infrared light (which accounts for more than 50% of solar energy) and develop UV-visible-near-infrared (UV-Vis-NIR) broad light spectrum photocatalysts is an important emerging topic [12]. Forming heterogeneous structures with matched band potentials is one approach [13][14][15]. For example, Au-tipped PbSe/CdSe/CdS core/shell/shell heterostructure nanocrystals are active NIR photocatalysts for the degradation of methylene blue (MB) [16], and Bi 2 WO 6 /TiO 2 heterojunction can harness UV, visible, and near-infrared light, and exhibits enhanced broad-spectrum photocatalytic properties to decompose methyl orange (MO) [17]. Thus, the design of a broad light spectrum photocatalyst with heterogeneous structure is of great significance.

Results and Discussion
The composition of products was controlled by varying the addition of NaOH. As shown in Figure 1, when 0.4 g NaOH was added, the X-ray diffraction (XRD) pattern of Figure 1a matched the JCPDS reference file for tetragonal Cu 3 Se 2 [PDF#65-1656]. When 4 g NaOH was included, the diffraction profile of Figure 1b shows that the composite was composed of tetragonal Cu 3 Se 2 [PDF#65-1656] and cubic Cu 1.8 Se [JCPDF 06-0680]. By increasing the addition of NaOH to 8 g, the XRD pattern of Figure 1c indicates that all the peaks can be well indexed to cubic Cu 1.8 Se [JCPDF 06-0680]. No characteristic peaks of any impurities were detected in the patterns, and the high diffraction intensity reveals that all the samples had good crystallinity. In addition, the particle sizes calculated from Scherrer's equation and the BET (Brunauer-Emmett-Teller) specific surface area of different samples are shown in Table 1. There are notbig difference in their particle sizes, and Cu 1.8 Se displays a larger specific surface area than Cu 1.8 Se/Cu 3 Se 2 composite and Cu 3 Se 2 , which could lead to the exposure of more active sites for the photocatalytic experiment. Nanomaterials 2017, 7, 19 2 of 10 microwave-assisted heating, sonochemical method, hydrothermal method, and vacuum evaporation [31][32][33][34].
Herein, we report on the aqueous precipitation synthesis of Cu1.8Se/Cu3Se2 heterogeneous structure, bare Cu1.8Se, and Cu3Se2 based on the redox reaction between alkaline selenium solution and cupric aqueous solution under atmospheric pressure. Compared with other reported methods, this method established in aqueous solution is rather simple and safe. The photocatalytic degradation of MO under visible and near-infrared (NIR) light demonstrated that Cu1.8Se/Cu3Se2 composite exhibits much enhanced photocatalytic activity compared with bare Cu1.8Se or Cu3Se2. The mechanism of enhanced photocatalytic activity for Cu1.8Se/Cu3Se2 composite based on the calculated energy band positions was also proposed.

Results and Discussion
The composition of products was controlled by varying the addition of NaOH. As shown in Figure 1, when 0.4 g NaOH was added, the X-ray diffraction (XRD) pattern of Figure 1a matched the JCPDS reference file for tetragonal Cu3Se2 [PDF#65-1656]. When 4 g NaOH was included, the diffraction profile of Figure 1b shows that the composite was composed of tetragonal Cu3Se2 [PDF#65-1656] and cubic Cu1.8Se [JCPDF 06-0680]. By increasing the addition of NaOH to 8 g, the XRD pattern of Figure 1c indicates that all the peaks can be well indexed to cubic Cu1.8Se [JCPDF 06-0680]. No characteristic peaks of any impurities were detected in the patterns, and the high diffraction intensity reveals that all the samples had good crystallinity. In addition, the particle sizes calculated from Scherrer's equation and the BET (Brunauer-Emmett-Teller) specific surface area of different samples are shown in Table 1. There are notbig difference in their particle sizes, and Cu1.8Se displays a larger specific surface area than Cu1.8Se/Cu3Se2 composite and Cu3Se2, which could lead to the exposure of more active sites for the photocatalytic experiment.  The morphology and microstructure of Cu1.8Se/Cu3Se2 composite are shown in the TEM (transmission electron microscopy) image of Figure 2c, with homogeneous size distribution of about 50-80 nm (in agreement with the calculated particle size in Table 1   The morphology and microstructure of Cu 1.8 Se/Cu 3 Se 2 composite are shown in the TEM (transmission electron microscopy) image of Figure 2c, with homogeneous size distribution of about 50-80 nm (in agreement with the calculated particle size in Table 1 Figure 2d), which is consistent with the spacing of (111) and (101) planes of cubic Cu 1.8 Se and tetragonal Cu 3 Se 2 , respectively. The high-quality image of the interface between Cu 1.8 Se and Cu 3 Se 2 can be observed clearly, which could be considered as a phase junction structure [35].  Figure 2d), which is consistent with the spacing of (111) and (101) planes of cubic Cu1.8Se and tetragonal Cu3Se2, respectively. The high-quality image of the interface between Cu1.8Se and Cu3Se2 can be observed clearly, which could be considered as a phase junction structure [35]. The UV-Vis diffuse reflection spectrum (DRS) of the as-prepared samples is shown in Figure 3a. It can be seen that these samples present effective optical absorption from the UV light region to the whole range of NIR light, which can be attributed to their small bandgaps. As shown in Figure 3b (converted from the DRS spectrum according to the Kubelka-Munk function [36][37][38]), the bandgap energies (Eg) of Cu1.8Se/Cu3Se2 composite, Cu3Se2, and Cu1.8Se were estimated to be around 1.42, 1.45, and 1.5 eV, respectively, which are similar to the reported bandgap [31,32]. It is worth noting that these bandgap values are very close, and Cu1.8Se/Cu3Se2 composite has a relatively narrower band gap and stronger light absorption than that of Cu3Se2 and Cu1.8Se. The UV-Vis diffuse reflection spectrum (DRS) of the as-prepared samples is shown in Figure 3a. It can be seen that these samples present effective optical absorption from the UV light region to the whole range of NIR light, which can be attributed to their small bandgaps. As shown in Figure 3b (converted from the DRS spectrum according to the Kubelka-Munk function [36][37][38]), the bandgap energies (E g ) of Cu 1.8 Se/Cu 3 Se 2 composite, Cu 3 Se 2 , and Cu 1.8 Se were estimated to be around 1.42, 1.45, and 1.5 eV, respectively, which are similar to the reported bandgap [31,32]. It is worth noting that these bandgap values are very close, and Cu 1.8 Se/Cu 3 Se 2 composite has a relatively narrower band gap and stronger light absorption than that of Cu 3 Se 2 and Cu 1.8 Se.
The photocatalytic activities of these samples were evaluated by the photodegradation of MO under visible and NIR light irradiation. As shown in Figure 4a,b, under identical experimental conditions, the Cu 1.8 Se/Cu 3 Se 2 composite exhibited strongly enhanced photocatalytic activity compared to both bare Cu 1.8 Se and Cu 3 Se 2 . The photodegradation rate of MO reached 80% and 75% in the presence of Cu 1.8 Se/Cu 3 Se 2 composite after 3 h of visible and NIR light irradiation, respectively. Thus, the Cu 1.8 Se/Cu 3 Se 2 composite shows relatively good visible-NIR broad light spectrum photocatalytic property, which is comparative with that of WS 2 [11] or Bi 2 WO 6 /TiO 2 [17].
By contrast, 50% and 46% of MO was degraded by bare Cu 1.8 Se, whereas only 20% and 17% of MO was degraded by bare Cu 3 Se 2 within the same time. Moreover, photodegradation of MO under full solar light without optical filter was performed. As shown in Figure 4c, after 2 h of full solar light irradiation, 82%, 52%, and 20% of the MO was degraded with Cu 1.8 Se/Cu 3 Se 2 composite, pure Cu 1.8 Se, and Cu 3 Se 2 as catalysts, respectively. In addition, the stabilities of these photocatalysts are shown in Figure 5. After four successive cycles, the photodegradation rate of MO under full solar light slightly decreased to 70%, 42%, and 18% within 120 min for Cu 1.8 Se/Cu 3 Se 2 composite, Cu 1.8 Se, and Cu 3 Se 2 , respectively, indicating that these materials could be re-used without appreciable loss of photocatalytic ability. The UV-Vis diffuse reflection spectrum (DRS) of the as-prepared samples is shown in Figure 3a. It can be seen that these samples present effective optical absorption from the UV light region to the whole range of NIR light, which can be attributed to their small bandgaps. As shown in Figure 3b (converted from the DRS spectrum according to the Kubelka-Munk function [36][37][38]), the bandgap energies (Eg) of Cu1.8Se/Cu3Se2 composite, Cu3Se2, and Cu1.8Se were estimated to be around 1.42, 1.45, and 1.5 eV, respectively, which are similar to the reported bandgap [31,32]. It is worth noting that these bandgap values are very close, and Cu1.8Se/Cu3Se2 composite has a relatively narrower band gap and stronger light absorption than that of Cu3Se2 and Cu1.8Se.  conditions, the Cu1.8Se/Cu3Se2 composite exhibited strongly enhanced photocatalytic activity compared to both bare Cu1.8Se and Cu3Se2. The photodegradation rate of MO reached 80% and 75% in the presence of Cu1.8Se/Cu3Se2 composite after 3 h of visible and NIR light irradiation, respectively. Thus, the Cu1.8Se/Cu3Se2 composite shows relatively good visible-NIR broad light spectrum photocatalytic property, which is comparative with that of WS2 [11] or Bi2WO6/TiO2 [17]. By contrast, 50% and 46% of MO was degraded by bare Cu1.8Se, whereas only 20% and 17% of MO was degraded by bare Cu3Se2 within the same time. Moreover, photodegradation of MO under full solar light without optical filter was performed. As shown in Figure 4c, after 2 h of full solar light irradiation, 82%, 52%, and 20% of the MO was degraded with Cu1.8Se/Cu3Se2 composite, pure Cu1.8Se, and Cu3Se2 as catalysts, respectively. In addition, the stabilities of these photocatalysts are shown in Figure 5. After four successive cycles, the photodegradation rate of MO under full solar light slightly decreased to 70%, 42%, and 18% within 120 min for Cu1.8Se/Cu3Se2 composite, Cu1.8Se, and Cu3Se2, respectively, indicating that these materials could be re-used without appreciable loss of photocatalytic ability.
To further understand the reaction kinetics of MO degradation under full solar light, we applied the Langmuir−Hinshelwood model, which is well-designed for photocatalytic experiments when the concentration of the organic pollutant is in the millimolar range [39], as expressed by where C0 and Ct are the concentrations of pollutant in solution at time t0 and t, respectively, and k is the kinetic constant, which is calculated to be 0.0145, 0.0062, and 0.0018 min −1 for Cu1.8Se/Cu3Se2 composite, bare Cu1.8Se, and Cu3Se2, respectively ( Figure 4d). Then, the apparent reaction rate constant (K) was calculated to get deeper insight into the photodegradation rate per unit surface area according to the following formula: where k is the kinetic constant from Formula (1), m is the mass of photocatalyst (0.1 g), and S is the specific surface area from Table 1. The apparent reaction rate constant (K) was calculated to be 0.0248, 0.0072, and 0.0027 min −1 for Cu1.8Se/Cu3Se2 composite, bare Cu1.8Se, and Cu3Se2, respectively. In other words, the photocatalytic activity per unit surface area of Cu1.8Se/Cu3Se2 is around 3.4 and 9.1 times higher than that of bare Cu1.8Se and Cu3Se2.   On the other hand, X-ray photoelectron spectroscopy (XPS) analysis was carried out to investigate the chemical binding states of Cu1.8Se/Cu3Se2 composite, as shown in Figure 6a,b. The Cu 2p peaks of Cu3Se2 and Cu1.8Se/Cu3Se2 composite broaden and undergo splitting, while pronounced satellite peaks (marked as Sat.) form due to the existence of Cu 2+ vacancy [40,41]. Two strong peaks are consistent with the literature data of Cu + 2p3/2 and 2p1/2 [42][43][44]. On the sides of Cu + 2p3/2 and 2p1/2 peaks, two low-intensity components appeared, which can be assigned to the Cu 2+ oxidation state [45,46]. As illustrated in Figure 6c, the Cu 2p3/2 and Cu 2p1/2 peaks are symmetric, narrow, and devoid of satellite peaks, which is indicative of monovalent copper of Cu + for bare Cu1.8Se. Therefore, the existence of Cu3Se2 leads to the presence of Cu 2+ in the Cu1.8Se/Cu3Se2 composite. To further understand the reaction kinetics of MO degradation under full solar light, we applied the Langmuir−Hinshelwood model, which is well-designed for photocatalytic experiments when the concentration of the organic pollutant is in the millimolar range [39], as expressed by where C 0 and C t are the concentrations of pollutant in solution at time t 0 and t, respectively, and k is the kinetic constant, which is calculated to be 0.0145, 0.0062, and 0.0018 min −1 for Cu 1.8 Se/Cu 3 Se 2 composite, bare Cu 1.8 Se, and Cu 3 Se 2 , respectively ( Figure 4d). Then, the apparent reaction rate constant (K) was calculated to get deeper insight into the photodegradation rate per unit surface area according to the following formula: where k is the kinetic constant from Formula (1), m is the mass of photocatalyst (0.1 g), and S is the specific surface area from On the other hand, X-ray photoelectron spectroscopy (XPS) analysis was carried out to investigate the chemical binding states of Cu 1.8 Se/Cu 3 Se 2 composite, as shown in Figure 6a,b. The Cu 2p peaks of Cu 3 Se 2 and Cu 1.8 Se/Cu 3 Se 2 composite broaden and undergo splitting, while pronounced satellite peaks (marked as Sat.) form due to the existence of Cu 2+ vacancy [40,41]. Two strong peaks are consistent with the literature data of Cu + 2p 3/2 and 2p 1/2 [42][43][44]. On the sides of Cu + 2p 3/2 and 2p 1/2 peaks, two low-intensity components appeared, which can be assigned to the Cu 2+ oxidation state [45,46]. As illustrated in Figure 6c, the Cu 2p 3/2 and Cu 2p 1/2 peaks are symmetric, narrow, and devoid of satellite peaks, which is indicative of monovalent copper of Cu + for bare Cu 1.8 Se. Therefore, the existence of Cu 3 Se 2 leads to the presence of Cu 2+ in the Cu 1.8 Se/Cu 3 Se 2 composite. satellite peaks (marked as Sat.) form due to the existence of Cu 2+ vacancy [40,41]. Two strong peaks are consistent with the literature data of Cu + 2p3/2 and 2p1/2 [42][43][44]. On the sides of Cu + 2p3/2 and 2p1/2 peaks, two low-intensity components appeared, which can be assigned to the Cu 2+ oxidation state [45,46]. As illustrated in Figure 6c, the Cu 2p3/2 and Cu 2p1/2 peaks are symmetric, narrow, and devoid of satellite peaks, which is indicative of monovalent copper of Cu + for bare Cu1.8Se. Therefore, the existence of Cu3Se2 leads to the presence of Cu 2+ in the Cu1.8Se/Cu3Se2 composite. In view of the mentioned XRD result (Figure 1b), Cu 1.8 Se is the main phase of Cu 1.8 Se/Cu 3 Se 2 composite, resulting in the smaller relative concentration ratio of Cu 2+ to Cu + (1.04) than that (2.32) of bare Cu 3 Se 2 . Combined with the photodegradation results (Figure 3), it could be inferred that decreasing the concentration ratio of Cu 2+ to Cu + is conducive to enhancing photoactivity. It is worth noting that Cu 1.8 Se/Cu 3 Se 2 composite has a larger concentration ratio of Cu 2+ to Cu + (1.04) than that (0) of bare Cu 1.8 Se, but the best photocatalytic activity-mainly due to the existence of phase junction between Cu 1.8 Se and Cu 3 Se 2 mentioned in the HRTEM image (Figure 2d).
The relative band positions of the two semiconductors were determined to account for the enhanced photocatalytic activity of the Cu 1.8 Se/Cu 3 Se 2 composite. The energy of conduction band (CB) bottoms (E CB ) were calculated empirically according to formula [47] where E g is the band gap of the semiconductor, E 0 is −4.5 eV for a normal hydrogen electrode, and X is the electronegativity of the semiconductor, expressed as the geometric mean of the absolute electronegativity of the constituent atoms [48]. The X values for Cu 1. As shown in Figure 7, Cu 1.8 Se and Cu 3 Se 2 are easily excited by visible or NIR light, and photoinduced electrons and holes are generated. The CB edge potentials of the two phases enable photogenerated electrons to easily transfer from Cu 1.8 Se to Cu 3 Se 2 . Simultaneously, holes on the valence band of Cu 3 Se 2 can be transferred to that of Cu 1.8 Se under the band energy potential difference. In such a way, long-lived reactive photogenerated carriers can be yielded, and thus enhanced charge separation efficiency through the phase junction can be achieved.
Cu3Se2 were calculated to be −0.31 and −0.22 eV, respectively. Correspondingly, the valence band tops (EVB) of Cu1.8Se and Cu3Se2 are 1.19 and 1.23 eV, respectively. Thus, both the conduction band bottom (ECB) and the valence band top (EVB) of Cu1.8Se are higher than that of Cu3Se2. The calculated band positions of Cu1.8Se/Cu3Se2 composite was conducive to the separation and transportation of photogenerated carriers.
As shown in Figure 7, Cu1.8Se and Cu3Se2 are easily excited by visible or NIR light, and photoinduced electrons and holes are generated. The CB edge potentials of the two phases enable photogenerated electrons to easily transfer from Cu1.8Se to Cu3Se2. Simultaneously, holes on the valence band of Cu3Se2 can be transferred to that of Cu1.8Se under the band energy potential difference. In such a way, long-lived reactive photogenerated carriers can be yielded, and thus enhanced charge separation efficiency through the phase junction can be achieved. To further confirm the effect of phase junction, the photoelectrochemical (PEC) behavior of Cu1.8Se/Cu3Se2 composite has been explored. As shown in Figure 8, the photocurrent responses were recorded under visible light irradiation. The electrodes (1 × 1 cm 2 ) demonstrated photocurrent responses around 1 μA/cm 2 and 2.5 μA/cm 2 for Cu1.8Se and Cu1.8Se/Cu3Se2 composite, respectively, while the photocurrent responses of Cu3Se2 were not apparent. This result provides strong evidence that the formation of a phase junction between Cu1.8Se and Cu3Se2 would efficiently accelerate the separation efficiency of charge carriers. Therefore, the Cu1.8Se/Cu3Se2 composite presents highly enhanced performance as compared to bare Cu1.8Se and Cu3Se2. To further confirm the effect of phase junction, the photoelectrochemical (PEC) behavior of Cu 1.8 Se/Cu 3 Se 2 composite has been explored. As shown in Figure 8, the photocurrent responses were recorded under visible light irradiation. The electrodes (1 × 1 cm 2 ) demonstrated photocurrent responses around 1 µA/cm 2 and 2.5 µA/cm 2 for Cu 1.8 Se and Cu 1.8 Se/Cu 3 Se 2 composite, respectively, while the photocurrent responses of Cu 3 Se 2 were not apparent. This result provides strong evidence that the formation of a phase junction between Cu 1.8 Se and Cu 3 Se 2 would efficiently accelerate the separation efficiency of charge carriers. Therefore, the Cu 1.8 Se/Cu 3 Se 2 composite presents highly enhanced performance as compared to bare Cu 1.8 Se and Cu 3 Se 2 .

Preparation
All reagents were of 99.9% purity and were used without further purification. A typical synthesis process of Cu1.8Se/Cu3Se2 composite was described as follows: 4 g NaOH, 0.76 g NaBH4, and 0.3 g elemental Se were added to 100 mL distilled water under constant stirring. The mixture reached about 80 °C in a few minutes to form alkaline selenium aqueous solution. Meanwhile, 10 mL Cu(NO3)2 aqueous solution (0.5 M) was prepared, and the mixture was combined with the alkaline selenium solution through rapid stirring. Finally, 0.1 g SDS was included as surfactant to control the morphology. After stirring for 8 h, the resulting products were separated by filtration, washed several times with distilled water and absolute alcohol, and then dried at 60 °C for 6 h. For comparison, bare Cu3Se2 and Cu1.8Se were also prepared by the hydrothermal method under the same conditions mentioned above by adding 0.4 g and 8 g NaOH, respectively.

Preparation
All reagents were of 99.9% purity and were used without further purification. A typical synthesis process of Cu 1.8 Se/Cu 3 Se 2 composite was described as follows: 4 g NaOH, 0.76 g NaBH 4 , and 0.3 g elemental Se were added to 100 mL distilled water under constant stirring. The mixture reached about 80 • C in a few minutes to form alkaline selenium aqueous solution. Meanwhile, 10 mL Cu(NO 3 ) 2 aqueous solution (0.5 M) was prepared, and the mixture was combined with the alkaline selenium solution through rapid stirring. Finally, 0.1 g SDS was included as surfactant to control the morphology. After stirring for 8 h, the resulting products were separated by filtration, washed several times with distilled water and absolute alcohol, and then dried at 60 • C for 6 h. For comparison, bare Cu 3 Se 2 and Cu 1.8 Se were also prepared by the hydrothermal method under the same conditions mentioned above by adding 0.4 g and 8 g NaOH, respectively.

Photocatalytic Test
The photocatalytic activities of as-prepared samples were investigated by photodegradation of MO under visible and NIR light. A 300 W xenon lamp (CEL-HXF300, Beijing, China) with cutoff filters (420 nm, 800 nm) was used as the light source. The specific process was as follows: 0.1 g photocatalyst was added into 100 mL MO (50 mg/L). Prior to irradiation, the slurry was continuously stirred in the dark for 1 h to ensure an adsorption-desorption equilibrium between photocatalyst and MO. Ice-water bath and magnetic stirring always existed to prevent thermal effect in the photocatalytic process. At a given time interval, 3 mL suspensions were centrifuged to remove the photocatalyst. The concentration of MO was analyzed by the UV-Vis spectrophotometer by recording the variations of the absorption band maximum (465 nm).

Photoelectrochemical Testing
Photocurrent under visible light was measured with a standard three electrode system on an electrochemical workstation (CHI 660, ChenHua, Shanghai, China). Ag/AgCl and Pt plate were used as reference electrode and counter electrode, respectively, in Na 2 SO 4 solution (0.5 mol/L) as electrolyte. The working electrode was made by depositing photoctalyst on the FTO (fluorine-doped tin oxide) substrate (Beijing, China).

Conclusions
The Cu 1.8 Se/Cu 3 Se 2 heterogeneous structure prepared by the simple precipitation method exhibits excellent photocatalytic activity in the degradation of MO under visible and NIR light irradiation, which is 3.4 and 9.1 times higher than that of bare Cu 1.8 Se and Cu 3 Se 2 . We also present that decreasing the concentration ratio of Cu 2+ to Cu + is advantageous for the enhancement of photoactivity. The mechanism of enhanced photocatalytic activity for the Cu 1.8 Se/Cu 3 Se 2 composite was investigated based on the calculated energy band positions. The formation of a phase junction between the two semiconductors leads to the effective separation of photogenerated electron-hole pairs, which mainly account for the strongly enhanced photoactivity.