Spatial Separation of Electrons and Holes among ZnO Polar {0001} and {101̅0} Facets for Enhanced Photocatalytic Performance.

Spatial separation of electrons and holes is critical for improving their photocatalytic performance, which is ascribed to the suppressed photoinduced carriers' recombination among facets. In this work, the ZnO-Au-MnO x heterogeneous nanostructure photocatalyst was prepared by photodepositing Au and MnO x on the ZnO polar {0001} and {101̅0} crystal facets, respectively. The photocatalytic performance of ZnO-Au-MnO x was higher than ZnO and ZnO-Au for the degradation of rhodamine B dye under UV light irradiation. Due to the potential difference between different crystal planes of zinc oxide, electrons and holes will migrate to different crystal planes of zinc oxide. This will lead to the deposition of Au and MnO x on different crystal facets of zinc oxide. The efficient photoinduced carrier separation of ZnO-Au-MnO x resulted in the high photocatalytic activity, which is well supported by photoelectrochemical and photoluminescence analyses. The intermediated species formed during the reaction were investigated by high performance liquid chromatography. The reaction mechanism was investigated by radical trapping experiments and electron spin resonance analysis. The special structure of selective deposition of redox cocatalysts on the different facets should be promising and intriguing for designing highly efficient photocatalysts.

Introduction

Photocatalysis is a green technology, which has attracted attention for applications in energy shortage and environmental pollution. However, the practical application of photocatalysis is greatly restricted by the low quantum yield.[1]

Researchers had made numerous efforts to solve the issue of fast photogenerated carriers’ recombination.[2,3] In recent years, the facet-induced charge separation is found to be an effective strategy to increase photocatalytic performances.[4,5] Atom composition and surface band structures of different facets will influence the transfer pattern of photoinduced carriers.[6,7] Recently, the internal polar field mechanism has attracted much attention since it provides a driving force for charge separation.[8,9]

In this regard, the facet-induced charge separation of some semiconductors, such as BiVO4,[10,11] TiO2,[12] BiOCl,[13] SrTiO3,[14] Bi2WO6,[15] and BiOI[16] have been reported. However, in all the reported cases, there are few different experimental results that the spontaneous separation of charge carriers occurs between the different facets.[17,18] Thus, the facet selective charge separation mechanism is still obscure. It remains a challenge to investigate the internal relation between the selectively spatial charge separation and crystal structure of the semiconductors.

Polar modification of photocatalysts was an efficient strategy to facilitate the separation of the photogenerated electron and hole.[19−21] ZnO is a kind of typical polar crystal. There is an internal electric field between positively Zn2+-terminated {0001} and negatively O2–-terminated {0001̅} polar planes due to the spontaneous polarization.[22]

Herein, inspired by the abovementioned consideration, we reported on the controllable growth of ZnO nanocrystals with polar facet exposure. Then photochemical labeling is used to obtain the information about charge separation. An internal electric field is existed along direction from Zn–ZnO surface to O–ZnO surface as the intrinsic driving force for charge separation.[23] We present evidence for the spatial separation of electrons and holes among the {0001} and {101̅0} crystal facets of ZnO crystals under proper photodeposition conditions. The results indicated that the Au particles were selectively deposited on ZnO {0001} facets by the photoreduction method. When Au and MnO are deposited at the same time, electrons and holes will transfer spontaneously toward {0001} and {101̅0} facets, respectively, which is beneficial to increase the light harvesting and the more effective charge transfer and separation. So, the photocatalytic activities are enhanced greatly.

Our results reveal the great potential of using an internal polar field to construct the photocatalysts with highly photocatalytic activities and unearth a new understanding of the effect of internal polar field to offer guidance to design more effective photocatalysts.

Experimental Section

Materials

Zinc acetate [Zn(Ac)2·2H2O], hexamethylenetetramine (HMTA), sodium borohydride (NaBH4), chloroauric acid tetrahydrate (AuCl3·HCl·4H2O), manganous nitrate hexahydrate [Mn(NO3)2·6H2O], potassium iodate (KIO3), rhodamine B (RhB), 4-p-chlorophenol (4-p-CP), and ethanol were of analytical grade and were used without further purification (Sinopharm Chemical Reagent Company). Throughout this study, deionized water was used.

Synthesis

The ZnO sample terminated with {0001}and {101̅0} facets was synthesized by a hydrothermal procedure according to our previous work, which is also displayed in Supporting Information.[22] Single reduction, single oxidation, as well as simultaneous reduction and oxidation were carried out for the facet-selective photo-depositions, respectively. Normally, 0.15 g of ZnO powder and a designated amount of metal or metal oxide precursors were mixed in 45 mL of deionized water. HAuCl4 (0.0486 mol/L) and Mn(NO3)2 (0.0797 mol/L) were used as precursors for photo-depositions. The suspension was kept for 30 min with stirring in the dark to establish adsorption and desorption equilibrium. Then the suspension was irradiated by four UV lamps of 365 nm (4 W, Philips TL/05) under continuous stirring. After 4 h photo-deposition, the suspension was filtered, washed with deionized water for several times, and finally dried at 60 °C overnight. The as-prepared samples are denoted as ZnO–Au, ZnO–MnO, and ZnO–Au–MnO, indicating the Au or MnO loading amount in the photocatalyst. For comparison, 1 wt % Au-loaded ZnO photocatalyst was prepared by impregnation followed by a chemical reduction process according to a reported ref (24). The as-prepared products are denoted as ZnO–Au–C.D. and ZnO–Au–P.D. (C.D.: chemical reduction; P.D.: photochemical reduction), respectively.

Characterizations

Crystal structures of the products were characterized by the X-ray diffractometer (Bruker D8 Advance).The morphologies were recorded by the field emission scanning electron microscope (FEI Nova NanoSEM 230). Microstructures of the samples were performed by transmission electron microscopy (TEM, Tecnai G2F20 S-TWIN, FEI Company). The chemical composition was operated using X-ray photoelectron spectroscopy (XPS, ESCALAB 250, Thermo Fisher Scientific). The optical properties of the samples were collected via a diffuse reflection spectrophotometer (DRS, Varian Cary 500). The BASi Epsilon workstation was utilized to measure photocurrents and electrochemical impedance spectroscopy (EIS). The photoluminescence (PL) spectra were performed on the F-7000 FL spectrophotometer. HPLC–mass spectrometry (HPLC–MS, Agilent 1290-6545) was carried out to monitor the intermediate products. The radicals during the photocatalytic reaction were detected by the electron spin resonance (ESR) spectrometer (Bruker model A300). The Mn and Au leaching amount into the reaction solution was detected by inductively coupled plasma analysis (ICP, Agilent 7800).

Photocatalysis Measurement

The photocatalytic performance was estimated by photodegradation of 10 ppm RhB under the ultraviolet light illumination. The lamps were cooled down by four mini fans fixed around the lamp to prevent the overheating the lamp. Furthermore, the irradiating lamp and the photoreactor were positioned adequate distance to keep the temperature at a constant level. For each reaction, 80 mg of the catalyst was dispersed in 80 mL of RhB solution in a quartz tube, which was surrounded by four UV lamps of 365 nm (4 W, Philips TL/05). Before illumination, the suspension was sonicated for 5 min. After adsorption and desorption equilibrium, the lights were turned on. About 4 mL of suspension was collected and centrifuged at certain irradiation time intervals. The concentration of the supernatant liquid was analyzed on the UV–vis spectrometer (Varian Cary 50).

Results and Discussion

XRD patterns of the as-prepared ZnO, ZnO–Au and ZnO–MnO samples are shown in Figure . All observed strong and sharp diffraction peaks can be clearly indexed to the wurtzite ZnO crystal structure (JCPDS card no. 36-1451),[25] which indicates that photo-deposition Au or/and MnO did not affect the crystal structure of ZnO.[16] For comparison, chemical reduction was applied to load Au nanoparticles onto ZnO NCs. XRD patterns of ZnO prepared by chemical reduction are shown in Figure S1. Au(111) characteristic peaks located at 38.23° were detected on both samples (JCPDS card no. 04-0784).[26] The results indicated that Au NPs were successfully deposited onto ZnO NCs.

Figure 1

XRD pattern of pure ZnO and the sample photodeposition with Au and MnO, respectively.

However, the characteristic diffraction peaks of the MnO are not inspected due to the low content on the surface of ZnO.[27]

To visually probe the microstructure and morphology of as-prepared samples, Figure shows the typical SEM images of ZnO NCs with Au–MnO by photodeposition processes. The morphology of ZnO is a hexagonal prism. Furthermore, the main exposed facets of ZnO correspond to {0001} and {101̅0} facets.[22] It can be seen that the Au nanoparticles are mainly deposited on the {0001} facets rather than {101̅0} facets (Figure a,b), which indicated that the Au3+ are photo-reduced on the {0001} facets.[28] The photogenerated electrons tend to transfer spontaneously toward {0001} facets. The photo-reduction equation can be described as follows[10]

Figure 2

Photoinduced selective deposition (a,b) Au, (c,d) MnO, and (e,f) Au and MnO on the respective facets of ZnO prism.

Following the selective photo-deposition of single metals, selective photo-deposition of metal oxide was further investigated. It can be seen that the {0001} face of the ZnO/MnO is smooth, indicating that sponge-like MnO is selectively deposited on the {101̅0} facet (Figure c,d). In addition, we further investigated the photodeposition of dual precursors simultaneously. It is clear to see from Figure e,f that the Au particles are loaded on the {0001} facets, while the MnO particles are photodeposited selectively on the {101̅0} facets of ZnO. The reason for the facet selective photodeposition can be ascribed to the simultaneous reduction and oxidation reactions regardless of their combination order.[10]

In contrast, in the chemical reduction process (Figure ), Au nanoparticles were deposited on all exposed facets of ZnO NCs, including the {0001} and {101̅0} facets, thus indicating that the chemical reduction process of Au nanoparticles is nonselective deposition.

Figure 3

SEM pattern of the sample with Au by chemical reduction: (a) front sections of the SEM image, (b) cross sections of the SEM image.

To obtain the microscopic structure and component, TEM and selected area EDS spectra analysis were carried out, as shown in Figure . It can be seen that MnO nanosheets (Figure a,b) appeared on the surface of ZnO–Au–MnO samples, which was in accordance with the EDS analysis (Figure c). As shown in Figure d, the Au nanoparticles appeared on the surface of ZnO–Au–MnO samples. The lattice spacing of 0.24 nm was indexed to the {111} planes of the Au crystal (Figure e).[24] The result was in agreement with the EDS data, as shown in Figure f. It indicated that the dual cocatalysts loading ZnO–Au–MnO system was successfully established.

Figure 4

TEM image and selected area EDS spectra of the ZnO–Au–MnO sample: (a–c) MnO layer on the surface of ZnO and (d–f) Au nanoparticles on the surface. Insets in (c,f) are tables of the element contents.

In order to confirm the surface elemental component and the chemical states of the Au and MnO on the surface of ZnO–Au–MnO samples, XPS was used, as shown in Figure . Figure a shows the full XPS spectra of the ZnO–Au–MnO samples. Figure b shows the high-resolution XPS spectra of Zn 2p, which revealed that Au–MnO photodeposited on the surface of ZnO did not affect the valance state of ZnO. The high-resolution XPS spectra of Au 4f and Mn 2p are displayed in Figure c,d, respectively. The peaks at about 84.1 and 87.8 eV corresponded to the Au 4f 7/2 and Au 4f 5/2 peaks, respectively, which indicated that the Au signal was photo-reduced on the {0001} facets of ZnO with respect to the metal ion precursors.[29] The XPS signals located at 642.2 and 654.9 eV can be assigned to Mn 2p3/2 and Mn 2p1/2 peaks, respectively. The broad peaks of Mn 2p3/2 and 2p1/2 suggested that the Mn ion existed in different oxidation states (Mn2O3 and MnO2).[30] Thus, the Mn species was responsible for the amorphous structure of MnO. The MnO from photo-oxidation of Mn2+ [Mn(NO3)2] was deposited on the {101̅0} facets of ZnO. The photo-oxidation equation can be described as follows

Figure 5

XPS spectra of the sample photodeposited by Au and MnO: (a) survey spectra, (b) Zn 2p, (c) Au 4f, and (d) Mn 2p.

The photocatalytic performance first depends on their light harvest ability. In order to compare the optical properties of ZnO, ZnO–Au, ZnO–MnO, and ZnO–Au–MnO, the UV–vis absorption spectra are illustrated in Figure . Compared with ZnO, after Au loading on the surface of ZnO, the absorption edges provided a slight red shift and the absorption intensity increased due to the surface plasmon resonance of Au excited at the wavelength of about 575 nm.[31] After MnO or Au–MnO deposited on the surface of ZnO, the noble metal–ZnO samples showed great light absorption intensity in the UV–vis region, which suggests that they could have higher photocatalytic activity.

Figure 6

UV–vis diffuse reflectance spectra of ZnO, ZnO–MnO, ZnO–Au, and ZnO–Au–MnO.

So far, some research results have shown that selective deposition of cocatalysts on different facets may enhance the separation of photogenerated electrons and holes.[23,32] In order to prove our hypothesis about the spatial transport of photo-generated electrons and holes between ZnO facets, the photocatalytic decomposition of RhB is opted as a probe reaction under the UV light (365 nm) irradiation. After reaching the adsorption equilibrium in the dark, the photocatalytic degradation was detected. As shown in Figure , compared to pure ZnO, the samples with deposited Au particles showed enhanced photocatalytic performance. Owing to the lower Fermi energy, deposition of Au onto the surface of ZnO NCs facilitates the transfer of electrons from ZnO to noble metals and reduces recombination between photogenerated electrons and holes.[33] While selective deposition of Au onto ZnO {0001} facets further enhanced the photocatalytic activities, the photogenerated electrons tend to unidirectionally transfer to {0001} facets, which may further enhance spontaneous separation between photogenerated electrons and holes.[34] In the case of Au chemical deposition, electrons and holes will transfer nonselectively toward {0001} and {101̅0} facets, and Au particles on the {0001} and {101̅0} facets may cause the recombination of some e– and h+ pairs. Hence, the photocatalytic activity of ZnO NCs with Au selectively deposition was improved greatly.

Figure 7

RhB degradation vs irradiation time in the presence of pure ZnO, ZnO–1% Au–P.D. and ZnO–1% Au–C.D.

Herein, inspired by the abovementioned results, spatial separation of electrons and holes is critical to improve their photocatalytic performance. Considering the improved light absorption intensity and the spatial separation of electrons and holes, it can be concluded that the ZnO–Au–MnO composite may exhibit higher photocatalytic performance. The photocatalytic performance of the samples with different Au and MnO content is displayed in Figure S2. The optimized photocatalytic activities were the sample with the mass ratio of 1% Au and 1% MnO, respectively. The rate constant of ZnO–Au–MnO is determined to be 0.04333 min–1, which is 6.7 times higher than that of the pure ZnO (Figure a,b). The photocatalytic decomposition rates of RhB were about 42% for pure ZnO; 71.5% for ZnO–MnO; and 90.2% for ZnO–Au–MnO. Furthermore, 4-p-chlorophenol was used as the colorless pollutant for the photocatalytic degradation experiments under UV light illumination (Figure S3). The result shows that the ZnO–Au–MnO also shows the best photoactivity for 4-p-CP degradation among these samples.

Figure 8

(a) Concentration of RhB as a function of time and (b) re-plot of the concentration of RhB in the −ln(c/c0) vs t.

These results suggest that the ZnO–Au–MnO exhibited the best photocatalytic activity, which may be ascribed to a synergistic effect between the cocatalysts. When Au and MnO are deposited at the same time, electrons and holes will transfer spontaneously toward {0001} and {101̅0} facets, respectively, which is beneficial for increasing the light harvesting and the more effective charge transfer and separation. So, the photocatalytic activities are enhanced greatly.

The stability of the photocatalyst was evaluated through five cycles for the photocatalytic reaction, and the results are shown in Figure S4. The photocatalytic efficiency of ZnO–Au–MnO is stable after five cycles, confirming the stability and reusability of the catalyst.

After the end of five cycles of reuse, the reaction solution was detected by ICP analysis to detect the soluble Mn and Au species. The results indicated that there was no Mn–Au leaching after the photocatalytic reactions.[35] Therefore, it can be concluded that the ZnO–Au–MnO composite is relatively stable during the photocatalytic reaction.

To better understand the origin of the enhanced photocatalytic activity, transient photocurrent measurements, EIS, and PL analysis were explored to characterize transportation and separation of photogenerated charge carriers.

As observed in Figure , the transient photocurrent response of ZnO–Au–MnO is significantly higher than that of the other samples, suggesting more effective charge transfer and separation.[36] The result was further confirmed by the following EIS analysis. EIS Nyquist plots of ZnO, ZnO–MnO, ZnO–Au, and ZnO–Au–MnO are displayed in Figure S5. The arc radius of the EIS Nyquist plot of ZnO–Au–MnO was smaller than those of the other catalysts, which was attributed to the decreased charge transfer resistance.[37]

Figure 9

Photocurrent responses of ZnO, ZnO–MnO, ZnO–Au, and ZnO–Au–MnO in 0.2 M Na2SO4 aqueous solutions.

Figure shows the PL spectra. It is clear to see that the intensity of the peaks follows the order: ZnO > ZnO–MnO > ZnO–Au > ZnO–Au–MnO. The reduction of the PL intensity can be assigned to the lower recombination rate of photogenerated electrons and holes.[38]

Figure 10

PL spectra of all the samples.

In addition, the adsorption capacity of sample is also investigated in Figure S6. The adsorption efficiencies of ZnO, ZnO–Au, ZnO–MnO, and ZnO–Au–MnO are approximately 3.9, 10.0, 10.9, and 14.7%, respectively, after 30 min. The decoration of Au and MnO2 could improve adsorption capacity of the samples. This is due to the fact that the Au and MnO2 possess much smaller particles size than that of ZnO, which would be beneficial for providing more active sites for RhB adsorption. Considering that the adsorption performance is consistent with the activity sequence of the sample, the improvement of adsorption performance is also one of the factors to improve the activity.

In order to understand the reaction mechanism for the enhanced photocatalytic activity, the trapping experiment were carried out to reveal more information about the active species. The trapping experiments of active species are shown in Figure . Benzoic acid (BA), benzoquinone (BQ), and ammonium oxalate (AO) were employed as scavengers to trap hydroxyl radicals (•OH), superoxide anion radicals (•O2–), and photogenerated holes (h+), respectively.[39,40] In the presence of AO, the activity was not significantly inhibited, indicating that photogenerated holes (h+) were not the major active species. However, after adding BA or BQ, the photocatalytic efficiency decreased significantly. Therefore, it is revealed that hydroxyl radicals (•OH), and superoxide anion radicals (•O2–) were the main active species for photodegradation RhB in our experimental conditions.

Figure 11

Photodegradation RhB with and without trapping agents.

ESR spectra analysis was used to investigate the generation of the active radical species, as shown in Figure . The typical characteristic peaks of the DMPO–•O2– adducts and DMPO–OH• adducts were both observed.[41,42] Besides, the ESR signal intensity in ZnO–Au–MnO is much stronger than that in the pure ZnO, suggesting a higher photocatalytic activity toward the degradation of RhB.

Figure 12

Results of ESR measurements: (a) DMPO–•O2– adducts and (b) DMPO–OH• adducts in the suspensions of ZnO–Au–MnO.

The degradation intermediate products were determined by HPLC–MS of 0, 60, and 120 min samples (Figure S7). HPLC analysis results indicated that RhB was identified at the retention time of 9.3 min corresponding to the m/z value 443.2.[43] In positive ion mode, the peaks at retention time 8.3 min were N-de-ethylated intermediates.[44]

Based on the abovementioned results and discussions, the possible photocatalytic enhancement mechanism of dual-cocatalysts selectively photodeposited on ZnO is illustrated in Figure . ZnO is the typical polar crystal, which consist of a positive Zn-terminated {0001} surface and a negative O-terminated {0001̅} surface.[45] An internal electric field thus is generated between Zn2+{0001} and O2–{0001̅} planes due to the spontaneous polarization. So, under UV light irradiation, the photogenerated electrons and holes can be separately accumulated on the {0001} and {101̅0} facets to achieve charge separation. Then the photo-induced electron and hole can migrate to the Au nanoparticles and MnO layer, respectively. The metal ions are photo-reduced on the {0001} facets. At the same time, the MnO as oxidation cocatalysts from photo-oxidation of Mn2+, were deposited on {101̅0} facets. The formation of Au and MnO is an indication that the photo-reduced and photo-oxidation selectively takes place on the {0001} and {101̅0} facets, respectively. The results are agreement with the abovementioned SEM and TEM results. The good charge separation reduces the recombination of the photo-induced electron and hole pair. Thus the photocatalytic performance can be most greatly enhanced.

Figure 13

Scheme of selectively photo-deposition on ZnO different facets.

Conclusions

In summary, a ZnO–Au–MnO heterogeneous nanostructure photocatalyst was prepared by photodepositing the Au and MnO on the ZnO polar {0001} and {101̅0} crystal facets, respectively. The samples were characterized by XRD, SEM, TEM, EDS, XPS, and UV–vis DRS. It was demonstrated that electrons and holes would transfer spontaneously toward {0001} and {101̅0} facet, respectively. The internal electric field along the ⟨001⟩ direct was the intrinsic driving force for the spatial charge separation. In comparison with the bare ZnO and ZnO-single cocatalysts, the ZnO–Au–MnO exhibited the highest photocatalytic activity for photocatalytic degradation of RhB dye under UV light irradiation. The enhanced photocatalytic performance was attributed to the increased light harvesting and the more effective charge transfer and separation. The research strategy of facet-selective charge separation on polar crystalline may potentially be extended to other semiconductor photocatalysts with enhanced photocatalytic performance.