Surface crystal feature-dependent photoactivity of ZnO–ZnS composite rods via hydrothermal sulfidation

ZnO–ZnS core–shell composite rods were synthesized using a two-step facile hydrothermal methodology wherein different sulfidation durations were employed. The effects of sulfidation duration on the morphology and crystalline quality of ZnS shell layers on the surfaces of ZnO rods were investigated. A ZnS shell layer with visible granular features was obtained in the adequately controlled 3 h sulfidation process. A structural analysis demonstrated that the ZnS shell layers of ZnO–ZnS composite rods synthesized after 3 h sulfidation were in a well-defined crystalline cubic zinc blend phase. Moreover, optical properties revealed that these composite rods had a higher light-harvesting ability than those obtained after 1 and 2 h sulfidation. The density of surface crystal defects and the photoexcited charge separation efficiency of the composite rods were associated with changes in the microstructure of the synthesized ZnS shell layers. The optimal sulfidation duration of 3 h for the ZnO–ZnS composite rods resulted in the highest photocatalytic activity for the given photodegradation test conditions. The improved light harvesting and charge transport at the ZnO–ZnS heterointerface accounted for the enhanced photocatalytic activity of the ZnO–ZnS composite rods synthesized after 3 h sulfidation.


Introduction
The development of oxide semiconductor photocatalysts for environmental remediation has been widely studied. [1][2][3] Zinc oxide (ZnO) has received global research interest for application as a highly efficient photocatalyst. However, low quantum efficiency, short-range disorder, and photocorrosion continue to affect the performance of ZnO in practical applications. 4,5 Several studies have been proposed to enhance the properties of ZnO by covering the surface with another semiconductor. [6][7][8][9][10] Moreover, formation of a heterostructure is benecial to improve the photocorrosion effect of ZnO. 11,12 Among various heterostructured systems, the core-shell system, where the shell plays a crucial role as a physical barrier between the optically active core and the surrounding medium, was proposed for ZnO-based composites. 6,13 In most of the studies, heterogeneous core-shell structures were formed by oxideoxide and oxide-sulde composites. Ruosong et al., reported the hydrothermal synthesis of ZnO-TiO 2 core-shell nanorods for photocatalysts and investigated the photocatalytic decomposition of rhodamine B. An enhanced photocatalytic activity was observed for ZnO-TiO 2 compared with pure ZnO. 14 Moreover, Guifeng et al., showed that ZnO-CdS core-shell nanowires exhibit an enhanced light-harvesting ability in both UV and visible light ranges. The increased shell surface roughness led to the improved photocatalytic decomposition of methylene blue (MB) compared with bare ZnO. 15 ZnO-Ag 2 S core-shell nanoparticles synthesized using a low-temperature chemical method demonstrated a superior photocatalytic activity than ZnO nanoparticles at both acidic and basic pH values. 16 These results demonstrate that integrating oxide-sulde to form a ZnO-based core-shell composite is promising to improve the light-harvesting properties of ZnO and enhance the photocatalytic reaction efficiency.
Zinc sulde (ZnS) is an important semiconductor, and nanostructured ZnS offers unique optical and catalytic properties. It is nontoxic and water insoluble. 17 In addition, ZnS is an effective photocatalyst because of the rapid generation of electron-hole pairs by photoexcitation and the highly negative reduction potentials of the excited electrons. However, the band gap of ZnS is large and some studies have reported that its band gap can reach up to 3.66 eV, 18 which is extremely large and thus decreases the light-harvesting ability of ZnS. In this regard, a core-shell heterogeneous structure was an innovative strategy for designing highly efficient photocatalysts incorporated with ZnS. However, another important factor that affects the photocatalytic performance is the band alignment between core and shell materials. A type-II band alignment is suitable for an efficient charge separation between core and shell semiconductors. Based on the aforementioned discussions, the integration of ZnS into ZnO to form a ZnO-based core-shell heterostructure is a promising approach for improving the photocatalytic performance of both ZnO and ZnS. However, only a few studies have reported the synthesis and photocatalytic applications of such ZnS-ZnO core-shell composites. In the present study, a ZnO-ZnS core-shell structure was synthesized using a facile two-step hydrothermal methodology. The crystalline feature of the constructed shell layer generally has a profound effect on the photocatalytic performance of core-shell composites. However, this topic has rarely been investigated for ZnO-ZnS core-shell composites synthesized in a hydrothermal suldation reaction. The photocatalytic decomposition of MB dyes by using ZnO-ZnS core-shell composites hydrothermally derived at different suldation durations was studied. The correlation between the structure and the optical and photocatalytic performance of these heterogeneous ZnO-ZnS core-shell composites was discussed.

Experiments
In this study, ZnO-ZnS core-shell composite nanorods with different ZnS shell layer crystal qualities were fabricated. Hydrothermally synthesized high density ZnO nanorods on the 200 nm-thick SiO 2 /Si (100) substrates were used as templates for further suldation treatment to grow ZnO-ZnS core-shell nanorods. The synthesis of vertically aligned ZnO rods consisted of two steps corresponding to the formation of ZnO seed layer and the growth of rods. The detailed experiment on the synthesis of hydrothermally synthesized ZnO rods has been described elsewhere. 19 Theas-synthesized ZnO rods were immersed in a Teon autoclave containing 0.05 M thioacetamide (TAA) aqueous solution. The reaction system was heated to 130 C and kept for 1, 2, and 3 h to grow ZnO-ZnS core-shell rods. The ZnO-ZnS-1, ZnO-ZnS-2, and ZnO-ZnS-3 are used to present the as-synthesized ZnO-ZnS core-shell rods aer 1, 2, and 3 h suldation, respectively in this study. Finally, the reaction system was cooled to room temperature naturally, and then the nal material was washed in deionized water and dried in an oven.
Crystal structures of the as-synthesized ZnO-ZnS core-shell rods were investigated by X-ray diffraction (XRD) using Cu Ka radiation. The surface morphologies of the various samples were characterized by scanning electron microscopy (SEM), and high-resolution transmission electron microscopy (HRTEM) was used to investigate the detailed microstructures of the coreshell rods. Room-temperature-dependent photoluminescence (PL) spectra were obtained using the 325 nm line of a He-Cd laser. The analysis of absorbance spectra of the core-shell rods were conducted by using UV-Vis spectrophotometer. To measure photocurrent properties of the samples, silver glues were laid on the surfaces of the rods to form two contact electrodes and the applied voltage was xed at 5 V during electric measurements under irradiation. Photocatalytic activity of various ZnO-ZnS rod samples were performed by comparing the degradation of aqueous solution of methylene blue (MB; 5 Â 10 À6 M) containing various ZnO-ZnS rods as catalysts under solar light irradiation excited from the 100 W Xe arc lamp. The solution volume of MB is 20 ml and the sample size for tests has a xed coverage area of 1.0 cm Â 1.0 cm for photocatalytic test use.  Fig. 1(a) and (d), the smooth surface of the hexagonal ZnO rods turned into a rough surface aer 1 h suldation. Layered aggregates homogeneously covered the ZnO rods and the hexagonal faces of the ZnO rods became rounder aer suldation. These observations suggested the partial conversion of ZnO into ZnS, thus leading to a composite structure. When the suldation duration was further increased to 2 h, the surface of the ZnO-ZnS rods became rougher than that of the rods subjected to only 1 h suldation ( Fig. 1(b) and (e)). The surface of the layered aggregates of the composite rods became rougher aer 2 h suldation. Fig. 1(c) and (f) show that the surface of the composite rods obtained aer 3 h suldation was granular. This might be associated with an increased crystallization of the surface ZnS crystallites with the increasing suldation duration in the hydrothermal synthesis process. Notably, increasing the suldation duration to 4 h caused consumption of the ZnO rods and the formation of ZnS tube-like features (Fig. S1 †). Fig. 2(a) shows XRD pattern of ZnO-ZnS-1 rods. An intense peak located at approximately 34 , which is corresponded to the ZnO (002) plane (JCPDS no. 36-1451). In addition, no peaks except the (00l) Bragg reections are observed in the whole measurement range, indicating the highly c-axis oriented hexagonal ZnO rods were used for suldation processes. Moreover, several tiny and broaden Bragg reections centered at approximately 29.4 and 48.1 corresponding to (111) and (220) Bragg reections of cubic zinc blend ZnS, respectively (JCPDS no. 05-0566) were visible in the XRD pattern. This reveals that ZnS crystallites were formed on the surfaces of ZnO rods aer 1 h suldation. Fig. 2(b) shows the XRD pattern of ZnO-ZnS-2 rods. The intensity of ZnS (111) and (220) Bragg reections increased and ZnS (311) Bragg reection became visible. Further increased suldation duration to 3 h ( Fig. 2(c)), all the Bragg reections from ZnS became more intense and the full-width at half maximum of the ZnS Bragg reections became narrower compared with those ZnO rods treated with shorter suldation duration. Moreover, the Bragg reection originated from ZnO rods is still intense aer 3 h suldation. Based on the XRD observations, the ZnO-ZnS composites were formed when the ZnO rods were subjected to suldation processes with various durations (1 to 3 h) and the crystallinity of the as-formed ZnS layer was improved with increased suldation duration. Fig. 3(a)-(c) present low-magnication TEM images of the ZnO-ZnS composite rods synthesized with different suldation durations. With 1 h suldation, the ZnS shell layer appeared to be composed of loose sponge-like aggregates, as shown in Fig. 3(a). With an increase in the suldation duration, particlelike aggregates of the ZnS shell layer became more predominant ( Fig. 3(b)). When the suldation duration reached 3 h (Fig. 3(c)), the ZnS shell layer structure exhibited a highly granular surface with a distinguishable particle size and boundaries with adjacent grains. The ZnO-ZnS composite rods synthesized with 1 h suldation exhibited a smoother surface than the composite rods synthesized with higher suldation durations. Fig. 3(d) presents a high-resolution TEM image of the outer region of the ZnS shell for the ZnO-ZnS-1 composite rod. Ordered lattice fringes were distinguishable and distributed within a short range of the selected region in the ZnS shell layer synthesized aer 1 h suldation. Furthermore, grain boundaries and ordered lattice fringes with an interval of approximately 0.312 nm were visible and assigned to the (111) plane of cubic ZnS, as shown in Fig. 3(e)-(f). This observation veries the previously reported structure of hydrothermally derived cubic ZnS crystallites. 20 The TEM analyses revealed that the ZnS shell layer synthesized aer extending the suldation duration from 1 h to 3 h exhibited an improved crystallization and an increased particle size. These observations are consistent with the XRD results.

Results and discussion
The difference in the chemical binding status of the ZnS shell layers synthesized with different durations of suldation of ZnO rods was examined through XPS measurements. Fig. 4(a)-(c) show the XPS surface narrow scan Zn 2p spectra of the ZnO-ZnS composite rods synthesized with 1, 2, and 3 h suldation, respectively. In the Zn spectrum, the peaks at binding energies of 1044.4 and 1021.3 eV are assigned to the Zn 2p 1/2 and Zn 2p 3/2 peaks of Zn 2+ , respectively. 21 Moreover, the Zn 2p doublet separations in Fig. 4(a)-(c) appeared at 23.0-23.1 eV, which is in good agreement with previously reported values for zinc ions binding to sulfur ions in ZnS lattices. 22 The asymmetric S 2p peak in Fig. 4(d)-(f) was deconvoluted into two subpeaks corresponding to S 2p 3/2 and S 2p 1/2 and were located at 161.1 and 162.2 eV, respectively. The binding energy of 161.1 eV originated from S 2À in the ZnS structure. The subpeak at approximately 162.2 eV might be due to surface defects of the S-S species in the ZnS shell layer and was previously reported in chemically derived ZnS nanorods. 23 Notably, the intensity ratio of the spin-orbit splitting peaks for S 2p 3/2 and S 2p 1/2 was approximately 2 : 1 for the ZnS shell layers aer 1 and 2 h sul-dations. This is in agreement with previous reports. 24 However, the intensity ratio of the spin-orbit splitting peaks for S 2p 3/2 and S 2p 1/2 was determined to be approximately 1 : 1 for the ZnS shell layer aer 3 h suldation. This can be attributed to the increased bonding states of sulfur-related surface defects in ZnS. 25    increased light scattering. 6 In microwave-assisted surface sul-dation synthesis of ZnO-ZnS heterostructured microowers, the introduction of more S 2À ions during synthesis creates a higher number of surface defect centers and, consequently, improved light harvesting and photocatalytic activity. 26 Furthermore, an increase in the ZnS particle load on ZnO in the chemical bath synthesis of ZnO-ZnS nanowires showed an increased electronic interaction between ZnO and ZnS, and therefore a pronounced red shi of the edge of the absorbance band. 27 Fig. 5(b) shows the PL spectra of the ZnO-ZnS composite rods obtained with different suldation durations. In Fig. 5(b), the ZnO-ZnS-1 and ZnO-ZnS-2 composite rods have similar PL spectra. A sharp UV emission band centered at 378 nm and a distinct broad green emission band centered at 567 nm were clearly visible. Furthermore, the PL peak in the UV region was still present for the ZnO-ZnS-3 rods, but an additional peak in the blue region and a visible light emission band centered at 520 nm appeared. The peak in the PL spectrum of the ZnO-ZnS-3 composite rods differed from that of the composite rods obtained aer 1 and 2 h suldation. The UV emission band was attributed to the nearest band edge emission originating from the recombination of free excitons of ZnO through an exciton-exciton collision process. 2,6 The green emission band centered at 567 nm was previously attributed to the contribution of the surface defects in ZnO. 2 Notably, the origin of the green emission centered at 520 nm for the ZnO-ZnS-3 composite rods be assigned to the deep states from Zn vacancy in the ZnS crystallites. By contrast, the emission band in the blue region was due to the sulfur vacancy and interstitial lattice defects in the ZnS nanostructures. 28 The difference in the PL spectrum for the ZnO-ZnS rods synthesized with 3 h suldation might be associated to the suldation durationdependent crystal quality differences in the ZnS shell layers. The ZnS shell layer formed aer 3 h suldation showed clear granule-like features and the crystalline quality increased with the suldation duration. The prolonged crystal growth due to an increased duration of the hydrothermal suldation reaction might create more point defects in the solid-state ZnS crystallites. The aforementioned reasons might account for the PL spectrum of the ZnO-ZnS composite rods obtained aer 3 h suldation being different from those of the composite rods obtained with a shorter suldation time. According to the PL spectra, the intensity of the PL peak in the UV region clearly decreases when the suldation duration increases. The decrease in the intensity of the UV emission band could be attributed to the different heterointerface states of band alignment between ZnO and ZnS. 29 Moreover, our results show that the ZnO-ZnS composite rods had a higher charge separation efficiency for a prolonged suldation duration. The photoresponses of the ZnO-ZnS composite rods obtained with various suldation durations were also investigated to determine the charge separation efficiency of photoexcited carriers at the ZnO-ZnS heterointerface. The corresponding photocurrent responses to several light cycles for the ZnO-ZnS composite rods obtained with different suldation durations are shown in Fig. 6(a). The fast and uniform photocurrent responses suggested a quick and stable charge transport in the samples. In the dark, the measured current for the ZnO-ZnS composite rods obtained with various suldation durations were low and were approximately 5.7 Â 10 À6 , 6.1 Â 10 À6 , and 6.4 Â 10 À6 A, respectively. Under light irradiation, the photocurrent increased sharply to reach a value as high as 3.2 Â 10 À5 , 5.2 Â 10 À5 , and 1.0 Â 10 À4 A for each prepared sample. Moreover, it decreased quickly as soon as the light was turned off. The photocurrent responses showed that the ZnS-ZnO composite rods obtained aer 3 h suldation achieved a 15-fold increase in current when exposed to light irradiation. This increase was higher than that for the ZnO-ZnS composite rods synthesized with 1 h (5.6-fold increase) and 2 h (8.5-fold increase) suldation. Fig. 6(b) presents the band alignment of the ZnO-ZnS heterostructure. 30 The mechanism underlying the photoresponse of the ZnO-ZnS composite rods could be explained through the photogenerated electron-hole pairs in the ZnS shell layer. The pairs separate and electrons are injected into the conduction band of the ZnO core. Moreover, the photogenerated holes in the ZnO core are transferred to the valence band of the ZnS shell because of the type-II band alignment in the ZnO-ZnS system. 30 A similar band alignment of the oxide-sulde heterostructures for improved photoactivity was observed in WO 3 -CdS. 31 The relatively high photoresponse level of the ZnO-ZnS composite rods obtained with 3 h suldation in this study implies a fast separation and transport of the photogenerated holes and electrons at the ZnO-ZnS heterointerface.  The photoactivity of the ZnO-ZnS composite rods obtained with different suldation durations was evaluated through the catalytic degradation process of an aqueous MB solution under solar irradiation. Fig. 7(a)-(d) show the absorbance spectra of the MB solution containing ZnO rods and various ZnO-ZnS composite rods under solar irradiation for various durations. The intensity of the absorbance peaks gradually decreased with increasing exposure time. This result indicates that MB is gradually degraded under the light irradiation in the presence of ZnO rods and ZnO-ZnS composite rods. The decrease in the absorbance peaks in both the UV and visible regions was more intense for the MB solution containing the ZnO-ZnS composite rods than for the MB solution containing the ZnO rods. Moreover, a relatively higher decrease in the intensity of the absorbance peak of the MB solution containing various ZnO-ZnS composite rods was observed in both the UV and visible regions for the composite rods obtained with 3 h suldation for a given irradiation time. The variation in the intensity of the absorbance peaks centered at 663 nm was further used to evaluate the photoactivity of various ZnO-ZnS composite rods. The photodegradation level of the MB solution containing various ZnO-ZnS composite rods is dened as (C/C o ), where C o is the concentration of aqueous MB without irradiation aer the dark adsorption equilibrium and C is the concentration of aqueous MB corresponding to a given solar light irradiation duration. 32 Fig. 7(e) displays the value of C/C o as a function of the irradiation duration for MB solutions containing ZnO rods and ZnO-ZnS composite rods obtained with different suldation durations. From Fig. 7(e), all of the ZnO-ZnS composite rod photocatalysts achieved the discoloration of the MB solution and showed relatively higher photocatalytic activities than the ZnO rods. The photodegradation level aer a 75 min irradiation of the MB solution containing the ZnO-ZnS composite rods synthesized with 1, 2, and 3 h suldation durations reached approximately 45.2%, 34.9%, and 21.6%, respectively. Notably, the decrease in the absorbance peak intensity in the UV region is associated with the mineralization of the MB dye during photodegradation. 15 The substantial decrease in the absorbance peak intensity with the irradiation duration for the MB solution containing ZnO-ZnS composite rods synthesized with 3 h sul-dation in the UV region supported the observation of the superior discoloration ability of these ZnO-ZnS composite rods among the various composite rod photocatalysts tested. During photodegradation of a MB solution containing semiconductors, hydroxyl radicals produced in the solution are strong oxidizing agents and effectively decompose MB dyes. 33 The photoexcited electrons or holes in the ZnO and ZnS semiconductors were transferred to the active surface where they participate in redox reactions with water molecules to produce a large number of hydroxyl radicals to decompose the MB dyes. The ZnS shell layer grown for the ZnO-ZnS composite rods obtained with 3 h sul-dation was well crystallized, and such a crystal structure achieved an efficient photoexcited charge separation through the heterointerface between the ZnO core and the ZnS shell of the composite rods, as shown in the PL results. Similarly, hydrothermally derived Bi 25 FeO 40 -Bi 2 WO 6 heterostructures with close interfacial connections and good crystalline contacts at the heterojunction interface also exhibit an efficient photoexcited charge separation because of the matching band positions. 34 The distinct granule-like surface features and the relatively high density of surface crystal defects obtained for the ZnS shell layer aer 3 h suldation were benecial to promote surface dye adsorption on the composite rods. Surface crystal defects in hydrothermally derived CdWO 4 -Bi 2 O 2 CO 3 core-shell heterostructure photocatalysts help enhance solar light activity. 35 Moreover, studies have shown that the photocatalytic efficiency of a CuBi 2 O 4 photocatalyst is strongly affected by the surface morphology and a distinct extrusion of the surface structures shows satisfactory photodegradation activity for MB dyes. 36 These factors accounted for the superior photoactivity observed for the ZnO-ZnS composite rods obtained aer 3 h suldation. To further understand the role of active species in the photocatalytic processes, trapping experiments were carried out to detect the active components. [37][38][39][40] In Fig. S2, † the photodegradation level of the MB solution decreases slightly aer the addition of tertbutanol (TBA; a hydroxyl radical scavenger). Notably, the photodegradation of the MB solution was signicantly inhibited aer the addition of the benzoquinone (BQ; a superoxide anion radical scavenger) and ammonium oxalate (AO; a hole scavenger). This means that superoxide anion radicals and holes act as the main active species in the photocatalytic process herein. The stability of these ZnO-ZnS composite rods was further evaluated. Successive experimental cycles were conducted for the photocatalytic degradation of the MB solution, and the results are shown in Fig. 7(f). When the ZnS-ZnO-3 composite rods were used for the rst time, approximately 78.4% of the MB solution was degraded aer 75 min. The ZnO-ZnS rod catalysts were quite stable even aer ve cycles of photocatalytic degradation of the MB solution. Aer ve cycles test, approximately 75.1% of the MB solution was degraded aer 75 min, as shown in Fig. 7(f).

Conclusions
Hydrothermally derived ZnO rods decorated with various crystalline features forming a ZnS shell layer were synthesized using a two-step hydrothermal methodology with 1, 2, and 3 h suldation durations. Increasing the suldation duration increased the crystallinity and granularity of the ZnS shell layer surface of the ZnO-ZnS core-shell composite rods. The longest suldation duration (3 h) resulted in a distinct granular ZnS shell layer. Moreover, our results demonstrate that the effective charge separation at the ZnO-ZnS heterointerface and an increased number of surface crystal defects in the ZnS shell layer were crucial for enhancing the photocatalytic performance of the ZnO-ZnS composite rods. A 3 h suldation process yielded a high degree of charge separation, which enabled these ZnO-ZnS composite rods to exhibit the highest photocatalytic activity for MB dye degradation.

Conflicts of interest
There are no conicts to declare.