ZnO-ZnS Heterojunctions: A Potential Candidate for Optoelectronics Applications and Mineralization of Endocrine Disruptors in Direct Sunlight.

Simple solution combustion synthesis was adopted to synthesize ZnO-ZnS (ZSx) nanocomposites using zinc nitrate as an oxidant and a mixture of urea and thiourea as a fuel. A large thiourea/urea ratio leads to more ZnS in ZSx with heterojunctions between ZnS and ZnO and throughout the bulk; tunable ZnS crystallite size and textural properties are an added advantage. The amount of ZnS in ZSx can be varied by simply changing the thiourea content. Although ZnO and ZnS are wide band gap semiconductors, ZSx exhibits visible light absorption, at least up to 525 nm. This demonstrates an effective reduction of the optical band gap and substantial changes in its electronic structure. Raman spectroscopy, transmission electron microscopy, X-ray photoelectron spectroscopy, and secondary-ion mass spectrometry results show features due to ZnO and ZnS and confirm the composite nature with heterojunctions. The above mentioned observations demonstrate the multifunctional nature of ZSx. Bare ZSx exhibits a promising sunlight-driven photocatalytic activity for complete mineralization of endocrine disruptors such as 2,4-dichlorophenol and endosulphan. ZSx also exhibits photocurrent generation at no applied bias. Dye-sensitized solar cell performance evaluation with ZSx shows up to 4% efficiency and 48% incident photon conversion efficiency. Heterojunctions observed between ZnO and ZnS nanocrystallites in high-resolution transmission electron microscopy suggest the reason for effective separation of electron-hole pairs and their utilization.

Introduction

Band gap engineering of wide band gap semiconductor materials is one of the current research areas to harvest maximum visible light absorption[1,2a,2b] because of its potential applications in clean and renewable energy and pollution abatement. Among various semiconductor oxides, TiO2 and ZnO are accepted to be the best because of their abundance, chemical stability, nontoxicity, and the efficiency for degradation of hazardous pollutants; the holes produced by photon absorption in these materials exhibit high oxidizing ability.[3a,3b,4] Although TiO2 is considered the most important photocatalyst, ZnO is also a suitable alternative photocatalyst because of its similar band gap energy (3.37 eV), low cost, and higher electronic mobility than TiO2.[5,6] Moreover, in certain cases, ZnO has high quantum efficiency and photocatalytic activity than TiO2.[7−9] Nonetheless, because of significant photocorrosion, ZnO has not been explored as thoroughly as TiO2. The common form of Wurtzite ZnO is a well-known material with a direct band gap of 3.37 eV and employed as a photocatalyst.[3] In the past few decades, many results have been reported on the synthesis of nanoscale ZnS crystals because of their special properties.[10−13] However, ZnS absorbs only UV light and hence is less attractive for solar light harvesting. Indeed, the ZnO–ZnS (ZSx) core–shell nanostructure has been predicted to exhibit a band gap of 2.07 eV.[13b]

Various schemes have been adopted to increase the absorption of visible light, including doping of impurities, sensitization with quantum dots,[14] and formation of heterostructure semiconductor materials. Doping of anions and/or cations into the ZnO or ZnS lattice has been attempted by a number of research groups,[15−19] and this helps to reach many sophisticated applications, as its electronic structure and properties can be modified significantly. There are many reports on nitride and/or sulphide doping, which can reduce the band gap and lead to visible light absorption in the semiconductor materials.[19−23] Although nitride is the best candidate for p-type doping in ZnO, there are controversies, such as its not-so-easily reproducible nature, N-content, that still exist.[23,24] Mapa and Gopinath used simple solution combustion method (SCM) to introduce nitrogen in the ZnO lattice, but N 2p states were reported to be in the forbidden gap.[16] Introduction of cations into the lattice of metal oxide causes thermal instability and invariably introduces mid-gap energy states between the valence band (VB) and the conduction band (CB) which act as recombination centres.[25] Visible light absorption in the wide band gap semiconductor materials has been developed by solid solution materials.[26−30] A few groups have studied solid solutions, where a significant improvement in photocatalytic activity was reported with solar light due to red shift in the absorption edge and also modification of textural properties. In 2009, Mapa et al.[26] prepared (Zn1–Ga)(O1–N)(ZnO–GaN) solid solution by SCM; ZnO–GaN exhibits absorption in the visible light region because of formation of solid solution with a band gap reduction. The same group subsequently prepared (Zn1–In)(O1–N) (ZnO–InN)[30] and (Zn1–InGa)(O1–N)(ZnO–InGaN)[27] solid solutions by adopting the SCM. Later, one was demonstrated to show overall water splitting in the visible light and significant photoconversion efficiency. On similar lines, formation of ZSx was expected to change the electronic structure along with changes in electrical and optical properties; however, the large difference between the ionic sizes of O2– and S2– limits the solubility of ZnS in ZnO or vice versa, though they belong to the same structure group.[31] Also, the synthesis of ZnO–ZnS heterojunction remains a challenge.

In the present communication, we report on ZSx composites with heterojunction as an optoelectronic material for photocurrent generation, dye-sensitized solar cells (DSSC), and direct solar light-driven photocatalysis. ZSx was prepared by a simple SCM and without using any template in less than 10 min. It was also demonstrated to show high activity for degradation of endocrine disruptors (ECDs), such as endosulfan (ES) (6,7,8,9,10,10-hexachloro-1,5,5a,6,9,9a-hexahydro-6,9 methano 2,4,3-benzo dioxathiepine-3-oxide), 2,4-dichloro phenol (2,4-DCP), in direct sun light. The visible light photocatalytic activity under direct sun light has been explained on the basis of the synergetic electronic interaction of wurtzite ZnO and ZnS semiconductors as well as the hetreojunctions in ZSx composites. The present work is a part of ongoing efforts in our laboratories to understand solar energy harvesting/conversion aspects.[32−37]

Results and Discussion

Structural and Textural Properties

X-ray Diffraction Studies

The structural aspects of all ZSx catalysts have been explored by PXRD (powder X-ray diffraction) measurements. Figure a,b shows a full and enlarged (21–38°) XRD pattern of the ZSx materials, respectively. Urea/thiourea employed for ZSx preparation varied from 100:0, 75:25, 50:50, and 25:75 to 0:100, and the resulting materials are designated as Z1, ZS1, ZS2, ZS3, and ZS4, respectively. Details of characterization methods are given in S1. It reveals that the as-prepared materials exhibit a hexagonal wurtzite structure and its diffraction pattern changes with the urea/thiourea ratio in the preparation mixture. For the comparison purposes, XRD patterns of ZnO, ZnS, and ZnSO4 and JCPDS patterns of ZnSO3 and Zn3O(SO4)2 were also included. XRD patterns of pure ZnO and ZnS show a hexagonal wurtzite structure, whereas ZnSO4 is of orthorhombic structure, and Zn3O(SO4)2 shows a monoclinic structure.[18] The XRD pattern of Z1 is very similar to that of ZnO, except for lattice contraction because of the introduction of N into the lattice.[16] The ionic radius of S2– (1.84 Å) is higher than that of O2– (1.4 Å);[18] further, with variation in the ratio of urea to thiourea, average combustion temperature and amount of sulfur introduced also change as well. In the case of ZS1, the urea to thiourea ratio is high at 3 and it leads to complete and fast combustion with high combustion temperature because urea is known to be a good fuel.[38] Hence, under this combustion condition, a significant amount of sulfur gets oxidized to sulfite, oxy sulfate, and/or sulfate of zinc. A 100% intense peak for ZS1 was indexed to the hexagonal ZnO structure with a space group of P63mc. Low intense peaks observed between 21° and 31° (Figure b) are attributed to a mixture of oxy sulfate and sulfate of zinc (Zn3O(SO4)2 and ZnSO4), with the former dominant on ZS1. Very broad and low intensity ZnS features could be seen on ZS1 at 28.7°, corresponding to the (002) facet of ZnS,[4] indicating the oxidation of surface ZnS to sulfate-related features (Figure b).

Figure 1

(a) PXRD pattern of ZSx materials prepared with different ratios of urea, along with ZnO and ZnS. Z1 is prepared with urea alone. (b) Enlarged XRD pattern of ZSx materials for clarity and compared with Zn3O(SO4)2, ZnSO4 and ZnSO3. A dash-dot line indicates a shift in ZnO features (101) to a higher angle and a lattice contraction. * and $ indicate ZnS and Si features, respectively. Dotted and dashed lines are a guide to the eye.

In the case of ZS2, peak intensity for ZnSO4, Zn3O(SO4)2 was suppressed enormously, and the ZnS features begin to appear (dashed arrows in Figure b). With increasing thiourea content in the preparation mixture, the reaction atmosphere changes predominantly toward sulfides along with ammonia. This is partially due to the presence of no oxygen in thiourea compared to urea and hence, enough amount of oxygen was not available for combustion. The lack of oxygen in thiourea does not allow the oxidation of sulfide and significant the ZnS phase is retained in the product. In other words, the material preparation furnace chamber becomes filled increasingly with in situ-produced reductants, such as H2S and ammonia, with very less oxygen, to interact with nascent ZnO clusters in a short span of time (≤1 min) under combustion conditions. This is the critical condition which leads to ZSx. Gas product analysis by IR confirms the formation of H2S and NH3. Unlike ZS1 and ZS2, ZnS and predominant ZnO features were observed for ZS3 and without any sulfate feature. Nonetheless, in ZS4, equally strong diffraction features of both ZnO and ZnS are found together without any other impurities. Thiourea (100%) was used as the fuel for ZS4, which is likely the reason for incomplete oxidation of ZnO (x/y ≈ 2) clusters and favors ZnS formation, under SCM conditions, attesting the successful use of thiourea as a source of S. Sulfide species are not oxidized, which helps to form a large amount of the hexagonal ZnS wurtzite phase. Therefore, distinct and strong ZnS features are observed in the XRD pattern along with ZnO. It is reasonable to conclude that a large ionic radius difference between S2– (1.84 Å) and O2– (1.4 Å) generates the ZnS phase along with ZnO in ZS2, ZS3, and ZS4. It is speculated that the structural similarity of ZnS and ZnO helps to form the heterojunctions in a 1:1 ratio; however, this needs careful verification. Not only the diffraction pattern, but also the relative intensity of the diffraction lines of as-prepared materials varied with the changing thiourea content in the reaction mixture. With no or low thiourea content, good crystallinity of Z1 and ZS1 can be observed, as they show very sharp peaks. However, with increasing thiourea content (ZS2) in the preparation mixture, crystallinity of the material decreases, and FWHM of the peaks increases, with the significant peak broadening because of smaller crystallite sizes of ZnS. However, this trend reverses with further increase in the thiourea content, and the crystallite size of ZnS increases, as shown in Table . Indeed, this observation demonstrates the possibility of fine tuning the ZnS crystallite size as well as its amount in ZSx. A shift in ZnO diffraction features to a higher angle from Z1 to ZS4 suggests a minor contraction of lattice parameters.

Table 1

Physicochemical Properties of ZSx Materials

ZSx compositiona (code)	U/TU mole ratio (%)b	surface area (m2/g)	ZnS size (nm)c
Zn1.04O0.84N0.15 (Z1)	100:0	1
ZnO1.02N0.08S0.11 (ZS1)	75:25	5	9
ZnO0.92N0.05S0.17 (ZS2)	50:50	8	16
ZnO0.65N0.04S0.32 (ZS3)	25:75	15	22
ZnO0.51N0.04S0.47 (ZS4)	0:100	20	50 (20)d

Material composition measured from energy-dispersive X-ray analysis.

The fuel (urea + thiourea)/Zn(NO3)2 molar ratio is maintained as 1. U/TU (urea/thiourea) percent mole ratio varied is given.

Crystallite size was calculated from the Scherer method.

Crystallite size was measured after calcination at 650 °C.

Thermal stability of ZSx was explored with XRD after ex situ calcination of ZS4 and ZS1 at different temperatures for 4 h in air, and the results are shown in Figure S1a,b, respectively. No significant change in the intensity of ZnS and ZnO features was observed at ≤400 °C. However, the intensity of ZnS peaks decreases, and a simultaneous increase in the intensity of ZnO features occurs, especially for the materials calcined at ≥500 °C. The above mentioned changes are attributed to the systematic oxidation of ZnS to ZnO under air atmosphere with an increase in calcination temperature. To have a better understanding of thermal treatment, XRD patterns of ZS4 were recorded in situ at temperatures between 480 and 700 °C, and the results are depicted in Figure . Peaks due to ZnS (ZnO) decrease (increases) in intensity as the temperature increases >480 °C, with the occurrence of a simultaneous broadening (narrowing) demonstrating a reduction (increase) in the crystallite size. Maximum changes in intensity and peak broadening of ZnS features occur between 560 and 650 °C, suggesting that the size reduction is relatively fast in the above mentioned temperature regime; however, a longer heating time at lower temperatures or heating in dilute oxygen atmosphere is expected to generate similar changes in a controlled manner. ZS4 at 700 °C shows only a trace amount of ZnS features, suggesting an almost complete oxidation to ZnO. An important underlying point in the above mentioned experiment is the controlled size reduction of ZnS between 400 and 650 °C, which could be utilized to prepare a controlled ZnS particle size in the ZSx matrix. The above mentioned bottom-up approach is likely to control the crystallite size better than other methods, especially in the solid state. A simultaneous increase in the intensity of ZnO features further supports the oxidation of ZnS to ZnO.

Figure 2

Powder XRD patterns of ZS4 recorded in situ between 480 and 700 °C show a systematic decrease in the crystallite size of ZnS. An arrow mark indicates the formation of a new phase and it is yet to be identified.

Surface Area Analysis

Nitrogen adsorption–desorption isotherms have been measured to investigate the textural properties of porous ZSx materials. Figure a shows the N2 adsorption–desorption isotherm results obtained at 77 K and analyzed by the Brunauer–Emmett–Teller method for surface area and Barrett–Joyner–Halenda (BJH) method for pore size distribution in Figure b; surface area and porosity have been calculated, and the results are given in Table . All materials show a type III adsorption–desorption isotherm with a H3 hysteresis loop that demonstrates the presence of mesopores and macropores.[39] Predominantly, mesopores are observed for ZS1 with a surface area of 5 m2/g (Table ). An increase in the amount of thiourea increases the surface area to 8 and 15 m2/g for ZS2 and ZS3, respectively, with marginal changes in the pore size and pore volume (not shown). In the case of ZS4, the surface area increases to 20 m2/g because of formation of more pores, and it is fully supported by microscopy results.

Figure 3

(a) N2 adsorption–desorption isotherms measured at 77 K and (b) BJH-pore size distribution pattern for ZS1, ZS2, ZS3, and ZS4 materials.

Thermogravimetric Analysis

Figure depicts the thermal analysis of (a) ZS4 in air and N2 and (b) ZS1 and ZS2 in air atmosphere and compared with that of ZnS and ZnSO4·7H2O (in b), respectively. ZS1, ZS2, and ZS4 shows about 2–3% weight loss up to 150 °C because of the elimination of water and adsorbed components. Between 150 and 570 °C, the ZS4 material shows hardly any weight loss, and a sharp weight gain of 2.5% was observed between 570 and 680 °C. Weight loss begins >680 °C and continues up to 850 °C. On comparison of ZS4 results with those of ZnS in air, it can be directly inferred that the weight gain between 570 and 680 °C is mainly due to the oxidation of ZnS to a sulfate/sulfite-like compound. Thermal analysis of a ZS4 material in an N2 atmosphere shows no weight gain between 500 and 650 °C, and a rather minor weight loss was observed between 600 and 700 °C, which clearly suggest that any oxidation is fully prevented in an N2 atmosphere. The thermogravimetry (TG) result in air atmosphere is supported by the XRD of a calcined material at different temperatures; however, no sulfate features observed in XRD (Figure ) are attributed to the decomposition of sulfate species under in situ heating conditions. Besides, this also suggests that the reaction temperature under SCM conditions might be between 600 and 680 °C for sulfate formation in the cases of ZS1 and ZS2, and lower than 570 °C for ZS3 and ZS4. An exothermic peak observed in the differential thermal analysis (DTA) curve of ZS4 at 690 °C similar to that of ZnS at 660 °C supports the oxidation of sulfide. A significant shift in the temperature of the above mentioned peak for ZS4 indicated a relatively better thermal stability for ZSx composites compared to bulk ZnS. However, a shift of +30 °C observed in the DTA exotherm for ZS4 is likely due to the distribution of smaller ZnS clusters in the well-protected ZnO host lattice. It is well-known in the literature[27,30] that the protected clusters, like the above mentioned, exhibit better thermal stability. Weight loss observed between 680 and 850 °C is due to decomposition of sulfate and/or oxysulfates of zinc to ZnO and the distribution of smaller clusters of ZnS in well-protected ZnO, and it is in good correspondence with ZnSO4·7H2O results (Figure b). No significant weight loss occurs above mentioned 850 °C and supports a complete removal of sulfur from ZS4. The weight loss pattern observed for ZS1/ZS2 materials is quite different in air atmosphere compared to that for ZS4.

Figure 4

Thermogravimetric analysis of (a) ZS4 and ZnS and (b) ZS1, ZS2, and ZnSO4·7H2O carried out in air (and nitrogen for ZS4) atmosphere at a heating rate of 10 °C/min. DTA of respective materials carried out in air atmosphere is given in the corresponding insets.

TG–DTA results of ZnSO4·7H2O given in Figure b is worth comparing with those of ZS1/ZS2. For a pure ZnSO4·7H2O compound, there is an initial weight loss of about 42% up to 250 °C in two phases; this is due to the loss of water of crystallization. About 5–6% weight loss observed up to 270 °C on ZS1/ZS2 is mainly due to adsorbed water. A sharp weight loss of 29% observed between 700 and 880 °C for ZnSO4·7H2O is attributed to the decomposition of sulfate to oxide. Indeed, the weight loss observed between 600 and 850 °C (Figure a) is due to the decomposition of ZnSO4 to ZnO. The loss of SO2/SO3 leads to weight loss in TG. It is to be noted the endothermic nature of ZnSO4 decomposition, in contrast to the exothermic oxidation of ZnS to ZnSO4 followed by decomposition. In a similar manner, ZS1 and ZS2 show a weight loss of 13 and 7%, respectively, between 550 and 850 °C. Low weight loss with ZS2 hints at a low sulfate and high ZnS content.

SEM and HRTEM

The morphology and the textural properties of ZSx materials were explored using scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HRTEM). From the SEM images of as-prepared ZSx materials shown in Figure , it is clear that the materials exhibit pores of the order of 1–10 μm, and they are macroporous in nature. The porous nature of the materials was well-supported by surface area analysis. It is also important to compare the morphology and textural properties of Z1 to ZSx materials. Z1 shows a ordered hexagonal pyramid without any pore.[16] Indeed, the well-ordered microcrystals observed in Figure a for Z1 highlights the compact micro crystals of ZnO1–N; an enlarged image of a single hexagonal particle from the top view is shown in the Figure a inset. Further, the surface area of Z1 is also very low at 1 m2/g, which underscores the above mentioned point. ZS1 shows a porous structure with a needle-shaped particle, highlighting the role of thiourea in changing the surface and particle morphology.

Figure 5

SEM images of (a) Z1, (b) ZS1, (c) ZS3, and (d) ZS4 and HRTEM images of (e) ZS3 and (f) ZS4. Transparent white patches (indicated by blue squares) observed on ZS1 are due to sulfate/sulfite. One-dimensional needle-shaped particles observed on ZS1 change to V-shaped morphology on ZS4. Z1 shows triangle and hexagonal prisms in panel (a). Inset in panel (e) shows line profile analysis to show a progressive change in the d value from ZnS to ZnO measured in the bottom-right side. The number of heterojunctions is evident in ZS3 and ZS4 given in panels (e,f).

Needle-shaped one-dimensional particles observed predominantly are attributed to ZS1 and ZS2, and transparent white patches observed on ZS1 (at top left and bottom right above the scale bar on Figure b) are attributed to sulfate-/sulfite-related compounds on the surface. On moving from ZS1 to ZS4, the above mentioned sulfate-related compounds decrease in content, and the particle morphology also changes drastically from a one-dimensional needle to two-dimensional V-shaped arrays with an edge size of about 1.5 μm at an angle of 72°. Indeed, the surface area also increases from 5 to 20 m2/g from ZS1 to ZS4, respectively, indicating an increased porosity from ZS1 to ZS4. EDX analysis was carried out to measure the material composition as well as to find out the extent of uniform distribution over a large surface area (1 mm2). Representative EDX results are given in Figure S2 for ZS2 and ZS4, and they support uniform distribution of all elements.

HRTEM results of ZS3 and ZS4 are shown in Figure e,f, respectively. Distinct d-spaces observed (0.29 and 0.26 nm) correspond to ZnS(101) and ZnO(002) facets, respectively, in both panels. Careful analysis reveals the abundant availability of heterojunctions between ZnO and ZnS and confirms the formation of ZnO–ZnS heterojunctions. Indeed, many such heterojunctions observed in several samples, prepared in different batches hints at the reliability of the method and measurements. Line profile analysis shown in the Figure e inset shows a clear change in the d value from ZnS to ZnO. The central rectangular box shows a gradual change in ZnS to ZnO. In fact, this indicates a possibility of the presence of solid solution to some extent. Thiourea and urea dissociate to generate in situ ammonia and H2S gaseous products at 500 °C, which helps to form ZSx composites with ZnO–ZnS heterojunctions.

Optical Absorption

It is well-known that anion doping affects light absorption characteristics of any oxide material,[19] and ZnO is not an exception. Diffuse reflectance UV–visible spectra of ZSx composites are shown in Figure a and compared with reference materials [ZnS, ZnO, and a physical mixture (ZnS–ZnO = 7:3)]. It is to be noted that ZnS and ZnO show absorption cutoff around 340 and 375 nm, respectively. However, the absorption cutoff is extended, at least, to 470 nm for all ZSx materials; ZS4 shows absorption up to 550 nm. A red shift is observed with increasing ZnS content from ZS1 to ZS4, and it is attributed to the modified chemical environment because of the formation of the heterojunction of ZnS and ZnO. Further ZnS crystallite sizes measured from XRD (Figures and 2) suggest an increasing size from ZS1 to ZS4 which is in good agreement with the above mentioned red shift.

Figure 6

(a) Diffuse reflectance UV–vis spectra of the ZSx materials along with ZnO, ZnS, and a physical mixture of ZnS and ZnO (7:3). A digital photograph is shown in the inset of panel (a) for the color associated with ZSx. (b) Diffuse reflectance UV–vis spectra of ZS4 (inset ZS1) materials calcined at 550 and 800 °C and ZnO.

This is further confirmed from the absorption spectrum recorded for a 7:3 physical mixture of ZnS–ZnO, which exhibits absorption due to ZnS up to 340 nm and up to 375 nm due to ZnO, as expected. A physical mixture of any ratio of ZnS and ZnO does not exhibit absorption >375 nm supporting ZSx with heterojunctions between ZnS and ZnO. Band gap energies calculated from the diffuse reflectance spectroscopy results for ZSx show a prominent decrease with increasing ZnS content. The band gap calculated from the absorption spectral data for ZS4 is 2.67 eV, suggesting a significant band gap reduction and changes in the electronic structure of ZSx composites compared to the pure components. However, a careful analysis of absorption by ZS3 and ZS4 reveals that the former absorbs more visible light with a considerably high absorption coefficient, especially between 400 and 500 nm. The band gap diminution of the ZSx composites has been attributed to the formation of hybrid bands/orbitals of ZnO and ZnS at the heterojunctions or interfaces. It is expected that S 3p and O 2p are likely to overlap in the VB, and the extent of overlap increases from ZS1 to ZS4 because of the increasing extent of heterojunctions. Sulfur exists as an anion (S2–) in ZS3 and ZS4, and the top of VB might be predominantly composed of S 3p energy levels, as they have higher orbital energies than O 2p levels. Thus, in the VB of ZSx, S 3p and O 2p bands overlap and broaden the VB than in ZnO or ZnS. It is expected that the CB of ZS1-ZS4 is formed by Zn 3d as well as S 3d orbitals. Our experimental results are in good agreement with DFT predictions by Schrier et al.[13b] for the ZnO–ZnS planar superlattice with a band gap of 2.31 eV. ZS1 and ZS4 calcined at high temperatures have been subjected to optical absorption studies, and the results are shown in Figure b. They exhibit a systematic decrease in the visible light absorption at calcination temperatures >550 °C.

X-ray Photoelectron Spectroscopy

Core-level spectra of S 2p and N 1s are compared in Figure , and the results from Zn 2p3/2, O 1s, and survey spectra are given in the Supporting Information (Figure S3). It can be seen from Figure a that the S 2p core level displays two different peaks at binding energy (BE) around 168.5 and 161.5 eV. A comparison of the observed BE to that of reported values for different sulfur compounds demonstrates that the oxidation state of S at 161.5 eV is due to S2– as in typical sulfides, and at 168.5 eV is due to S6+ as in sulfate or oxysulfate.[13] The peak at 161.5 eV corresponds to ZnS and demonstrates the formation of ZnS in ZSx. On ZS1 and ZS2, a characteristic, but broad, peak for sulfide was found with low intensity, along with a majority of sulfate on the surface. The sulfide feature increases in intensity from ZS1 to ZS4 and directly supports the observation of XRD. The above mentioned observation supports the oxidation of sulfide species on the surface to sulfate under preparation conditions. Nonetheless, sulfate formation is mostly restricted to the surface on ZS3 and ZS4, and no feature corresponding to sulfate has been observed in any of the bulk characterizations, such as XRD, Raman, and secondary-ion mass spectrometry (SIMS). It is clear from Figure a that with increasing thiourea content in the preparation mixture, the total sulfur content on the surface increases, which was demonstrated in the elemental mapping of ZS4 (Figure S2f). This can be clearly seen in the prepared material (in ZS3 and ZS4), where a prominent peak for sulfide and some sulfate was observed. Because of combustion conditions, some sulfate formation on the surface cannot be avoided; however, it can be removed by simple washing with water.

Figure 7

X-ray photoelectron spectroscopy spectra obtained from (a) S 2p and (b) N 1s core levels of ZSx.

The N 1s core level appears at around 399.5 ± 0.2 eV for all ZSx (Figure b) materials, suggesting the nature of nitrogen on the surface, especially that the charge density is similar to that of ammonia.[16] It is also surprising that the nature of N remains the same on all ZSx compositions and increasing ZnS content did not change the nature. However, no nitride phase such as Zn3N2 was found. Moreover, the amount of surface nitrogen decreases from Z1 to ZS4, and it is also in correspondence to the bulk N-content (Table ) estimated by EDX. We attribute this trend to the N doping in the ZnO lattice, and no significant role in photocatalysis is expected, which is in line with ZnO1–N.[16] The BE values for the Zn 2p3/2 region was found at 1021.9 ± 0.1 eV, which is in agreement with literature values reported for ZnO (1022 eV) and ZnS (1021.9 eV) (see the Supporting Information Figure S3).[16,26] The O 1s core level also shows a typical oxide feature for Z1 at 529.4 eV. Nonetheless, the main O 1s peak appears at 531.8 eV for all ZSx materials uniformly because of sulfate, whereas the ZnO feature was suppressed because of preferential sulfate formation.

Raman Spectra

Raman spectra of ZSx materials along with Z1, ZnO, and ZnS were measured, and the results are shown in Figure . Distinct Raman modes {E2, longitudinal optical [A1(LO)] and transverse optical [A1(TO)]} are observed for ZnO, ZnS, and Z1.[39] ZnS shows characteristic phonon modes with high intensity at 428 (2E2) and 711 cm–1 [A1/E1(2LO)], and two other features at 492 and 611 (2TO) cm–1. All the above mentioned four modes were observed for ZS1 to ZS4 materials. A strong and sharp peak observed at 433 cm–1 on all ZSx materials, 437 cm–1 on ZnO, and 429 cm–1 on ZnS is due to the typical E2 (high) mode, and this mode is characteristic of the wurtzite phase. Among the Raman modes, the E2 (high) at 437 cm–1 shows the strongest intensity in ZnO and on all ZSx materials, indicating the high quality of the material. The presence of the above mentioned mode in ZSx materials indicates that they exhibit the same local geometry. Nonetheless, it is to be noted that all ZSx materials exhibit the 2E2 mode at 433 cm–1 and exactly in between the E2 modes of ZnO and ZnS, indicating the strong interaction of ZnO and ZnS with the formation of the heterojunction nature of ZSx materials. Broad-and low-intensity features observed for ZS1 are likely due to sulfate species present in it; however, no new mode for sulfate species was observed. Indeed, it is surprising that first-order E2 and A1/E1(LO) phonon modes are not observed; rather second-order 2E2 and A1/E1(2LO) modes are observed with high intensity for ZnS as well as ZSx. In addition to the above, broad peaks are observed at 497 cm–1 on ZSx materials, which does not belong to the first- or second-order structure of neither ZnO nor ZnS; further, this mode is not observed for ZS1, suggesting that this is not due to sulfate-related species also. Although XRD does not show prominent ZnS features for ZS2, Raman results show the same, highlighting the formation of ZnS in ZS2. Sharp peaks at 507 and 582 cm–1 in Z1 are attributed to A1(LO) due to N-doping.[16,39]

Figure 8

Raman spectra of ZSx materials, compared with Z1, ZnS, and ZnO.

SIMS Analysis

SIMS analysis results obtained are shown in Figure . Initially, the recorded mass spectrum (not shown) demonstrated the presence of different species, namely, Zn, S, ZnO, ZnN, ZnS, N, and O with significant to high counts. Secondary-ion intensities of the first five species are shown in Figure , as a function of sputtering time or depth. No change in counts/intensity of any species with sputtering time emphasizes the introduction of S throughout the bulk of ZSx and the uniformity of the distribution throughout the bulk. Further, S-related fragments dominate over N-related fragments, indicating that the N content is relatively lower. This is further confirmed from the SIMS results from Z1, in which N and Zn–N exhibit higher intensity than Zn–N in ZSx.[16] Indeed, any species that exhibits 10 counts/s and less is not considered because of low signal/noise and poor reliability.

Figure 9

SIMS results display secondary-ion intensities measured as a function of sputtering depth or time for (a) ZS1, (b) ZS4, and (c) Z1.

It is fully surprising that a heavier element like S does compete with N, and preferential ZnS formation occurs under combustion conditions. Nonetheless, counts for all species were significantly different on Z1, ZS1, and ZS4, indicating that the material characteristic has changed significantly from each other. Higher count rates are observed for Zn–N than Zn–O fragments, despite ZnO being the host lattice in Z1, which is attributed to the different ionization capacity of emitted species, and this is strongly dependent on the local surface characteristics of the materials, known as the matrix effect.[40] Although the bulk N content in ZS1 is half that of Z1, very low Zn–N fragments observed on ZS1 reiterates that the ionization capacity varies for the same fragment from material to material, especially when the nature of the materials changed significantly. An important point to be mentioned here is that no other N- and S-related species, such as N2, NO, ZnNO, ZnSO, NH, SO (x = 1–4), and NS, are observed in SIMS, which suggests that the status of N and S in ZnO1–N and ZSx materials is none of the above. It is also surprising to note the absence of SO, in spite of sulphate/sulphite species observed in other methods. As indicated by other characterizations, the nature of sulfur changes increasingly toward sulfide from ZS1 to ZS4.

Chronoamperometry

Photo-electrochemical measurement was carried out using three electrode assemblies in a 0.5 M NaClO4 electrolyte to understand the light harvesting character. A photoresponsive material was casted on a fluorine-doped tin oxide plate and evaluated for photocurrent generation under illumination. Chronoamperometry measurement (at 0 V vs Ag/AgCl) was carried out to study the instantaneous photoresponse of the materials, and the results are shown in Figure . Two types of measurements were carried out, namely, AM 1.5 (100 mW/cm2) and with a 400 nm (92 mW/cm2) cutoff filter to follow the photocurrent change and efficiency of visible light absorption. Among all catalysts, ZS4 generates an at least 4-fold and 2-fold higher current than ZS1 or ZS2, and ZS3, respectively (Figure a) under irradiation conditions. It demonstrates the effective absorption of light by ZS4, and the large number of heterojunctions present helps achieve high current than other compositions; in fact, the high surface area and porosity associated with ZS4 helps this. No current was produced under dark (shutter close) conditions, confirms the photofunctional behavior of ZSx materials. The photocurrent-generated value shows less in the visible light region than AM 1.5 (Figure a) because of no UV light in the 400 nm filtered light (Figure b). This observation demonstrates the necessity to utilize the small amount UV light (4–5%) in the solar radiation (1 sun condition) to maximize solar harvesting. Further, 3.5-fold higher current generation observed for ZS4 than ZS3 underscores the effective absorption of visible light (Figure b) and charge separation by ZS4. High photocurrent generated from ZS4 is attributed to better necking between the particles, which is essential for electron conduction. The extent of necking is relatively higher with ZS4 than ZS3, and hence the former shows higher current. Because of high photocurrent generation, ZS4 and ZS3 show better quantum efficiency in DSSC measurement and high photocatalytic degradation of ECD than ZS1 or ZS2, which is described below.

Figure 10

Photocurrent measurement under (a) AM1.5 (ZS1, ZS2, ZS3, and ZS4) and (b) 400 nm (ZS4 and ZS3).

Dye-Sensitized Solar Cells

Figure a and Table show the J–V curves and the photovoltaic (PV) performances of DSSC devices, respectively, based on the photoanodes using ZnO and ZSx composites under 1 sun irradiation (AM 1.5G solar simulation). Typical ZnO-based cells exhibit Voc—0.53 V, Jsc—4.95 mA/cm2, fill factor (FF)—59%, and η—1.53%. The PV performance linearly increases by varying the composition of sulfur from ZS1 to ZS4. ZS4 exhibits the best PV performance (Voc—0.67 V; Jsc—8.45 mA/cm2; FF—69%; and η—3.8%) than other ZSx materials and bare ZnO. This was further confirmed by the incident photon conversion efficiency (IPCE) spectra (see Figure S4 in the Supporting Information), wherein bare ZnO shows 30% conversion and ZS4 shows 48%. The SEM results reveal the porous nature and the high surface area for ZS4, which induce the photogenerated carriers to diffuse fast to the surface and react rapidly with the adsorbed molecules. More interestingly, the effective transport of charge carriers at the heterojunction between ZnO and ZnS helps to minimize the recombination which is proven by high FF (69%) for ZS4.

Figure 11

(a) J–V curves of (DSSCs) ZnO, ZS1, ZS2, ZS3, and ZS4. (b) Nyquist plots of all PV cells under AM 1.5G sun illumination (100 mW/cm2).

Table 2

PV Parameters of ZnO and ZSx Composites under AM 1.5G Sun Illumination (100 mW/cm2)

parametersa	ZnO	ZS1	ZS2	ZS3	ZS4
Voc (V)	0.53	0.51	0.64	0.62	0.67
Jsc (mA/cm2)	4.95	5.3	6.7	7.9	8.45
FF (%)	0.59	61	63	64	67
η (%)	1.54 ± 0.23	1.64 ± 0.27	2.7 ± 0.2	3.1 ± 0.24	3.8 ± 0.3
Rct (Ω)	43	39	33	28	23
Cμ (μF/cm2)	3.1	6.73	10.1	13.5	18.4
τr (ms)	1.33	2.62	3.31	3.78	4.1
IPCE (%)	30	32	36	38	48

Active photoanode area = 0.235 cm2. Thickness of material is 15 μm dipped into 0.5 mM of acetonitrile/tert-butanol (1:1) for 6 h.

Electrochemical impedance spectroscopy (EIS) is a powerful tool to investigate the internal resistance and the charge-transfer process in DSSC.[41] Charge-transfer resistance (Rct) of ZnO and the composites were obtained from Nyquist plots (Figure b) under light illumination (100 mW/cm2) at the photoelectrode/electrolyte interface. From the EIS analysis, Rct (Cμ) and electron lifetime (τr) were calculated and are summarized in Table . τr was calculated by the following equation; τr = Rct × Cchem; (Cchem = chemical capacitance). τr reflects the response time constant for recombination which is correlated with Voc. A ZS4-based photoanode exhibited the smallest charge-transfer resistance (23 Ω), with highest chemical capacitance (18 μF/cm2) and electron lifetime (4.1 ms), indicating faster charge transport and lower recombination across the interface of photoelectrode. Compared to ZS4, all other ZSx composites show higher Rct, and τr and smaller Cchem. This underscores the better charge separation characteristics associated with ZS4, which is expected to generally increase the solar harvesting performance.

Photocatalytic Degradation of 2,4-DCP in Direct Sun Light

2,4-DCP can be considered as a model compound for ECD, and its degradation was carried out as a model reaction to evaluate the photocatalytic oxidation ability of ZSx. Figure shows the UV–vis absorption spectrum recorded for aqueous 2,4-DCP solution that was irradiated under direct sun light for different time periods with ZS3 (Figure a) and ZS4 (Figure b). Two absorption bands (broad band below 240 and at 285 nm) are observed for 100% DCP or before the onset of degradation; the weaker band in 270–300 nm range corresponds to π–π* and n−π* transitions that are characteristic of benzene and its substituted derivatives.[42] Significant photocatalytic degradation is evident from the results observed in Figure by ZS3 and ZS4. A linear decrease in 2,4-DCP degradation is evident from the decrease in the band intensity of both peaks. In fact, ZS3 and ZS4 show a similar trend in activity. Although absorption intensity decreases in Figure , there is hardly any mineralization of fragmented components to CO2. Nonetheless, fragmented components can be mineralized relatively easier than the parent DCP.

Figure 12

Absorption spectra of 2,4-DCP solution recorded at different time intervals following photodegradation with (a) ZS3 and (b) ZS4 catalyst.

The activity measurements were also carried out with all ZSx materials in direct sunlight. Necessary control experiments were also carried out. Further, a UV lamp (λ = 240–400 nm; predominantly, 365 nm and 240–320 nm) or H2O2 (2 mL of 25% H2O2 in direct sunlight; see Figure S5) was employed to measure the 2,4-DCP degradation activity with ZS3. Figure compares the activity of all catalysts for 2,4-DCP degradation. Control experiments, such as pure ZnO and ZnS, show low activity (20–25% degradation) in 10 h. However, all ZSx catalysts show significant activity for 2,4-DCP degradation in direct sunlight. The bulk heterostructure of ZSx leads to absorption of visible light and shows better activity for 2,4-DCP degradation under direct sun light. The λ-max at 285 nm for the absorption of 2,4-DCP (Figure ) reduced as reaction time increased. Indeed, complete 2,4-DCP degradation occurs in about 10 h with ZS2, ZS3, and ZS4 catalysts under direct sun light as well as in UV light irradiation. This underscores the visible light activity associated with ZSx materials. The activity of all other catalysts follows the order given below: ZS3 > ZS4 > ZS2 > ZS1. High surface area and pore volume in ZS3 and ZS4 adsorb a large amount of reactants on the catalyst surface; because of highly crystallinity and small particle size, the photogenerated charge carriers diffuse fast to the surface of the catalyst and react rapidly with 2,4-DCP and hence an effective utilization of charge carrier for the oxidation.[31] In the presence of H2O2, the reaction was very fast, and 2,4-DCP degradation was completed in 3 h. It demonstrates the formation of the hydroxyl radical from photocleavage of H2O2 during the light irradiation which increases the oxidation of 2,4-DCP.[20] This reveals the importance of the hydroxyl radical for oxidation reaction and assists the reaction. Indeed, without a catalyst or in the absence of light, no activity was observed, which suggests that the degradation reaction occurs under a photocatalytic reaction path way. There is a finite difference in reactivity occurs between ZS3 and ZS4. ZS3 exhibits a small crystallite size (20 nm) with significantly high visible light absorption, which is highly required for the decomposition of ECD, whereas ZS4 shows a high crystallite size (50 nm) and a surface area with good necking between particles, which enhance the electron conduction and hence lead to high photocurrent generation. This underscores the necessity to have optical absorption, electrical conduction, and microstructural properties in a supplementing manner to achieve higher activity as well as photocurrent. The corresponding first-order kinetics plot and rate constant are shown in Figure S6, indicating the superior activity associated with ZS3 among ZSx catalysts. Indeed, either in the presence of UV light or sunlight, peroxide increases the rate constant effectively.

Figure 13

Photocatalytic degradation activity for 2,4-DCP with ZSx under direct sun light and UV light.

Photocatalytic Degradation of ES

ES is a toxic chemical and was employed as a pesticide around the globe in the past without realizing the harmful effects on human beings. Till date, an ES-infested agricultural land could not be used in many countries. ES exists as a mixture of α and β isomers, and its consumption through farm products or inhalation creates genetic problems in human beings and animals.[43] A small amount (2 μg/mL) of ES is harmful for male fertility and affects the endocrine system. Because of harmful effects, biodegradation of ES was carried out by enzymes[44] at a specific pH. It requires several days to degrade ES into its fragments and involves several steps to grow the microorganism in a suitable environment.[45] It is further limited by low degradation efficiency with high toxic metabolites, and hence a complete mineralization by cheap alternative methods, such as photocatalysis, is required.[46] Thomas et al., reported photocatalytic activity studies of Ag- and Au-incorporated TiO2 for degradation of ES under solar light irradiation.[46b,46c]

There are few reports on photocatalytic degradation of ES under UV light irradiation.[46e−46i] Sivagami et al.[46f] studied the photocatalytic activity of TiO2 for the degradation of ES predominantly under UV light (254 nm). Similarly, Tapia-Orozco and Vázquez employed photoactive TiO2 films for ES degradation in UV light.[46g] Very limited studies were carried out so far, and complete mineralization is yet to be realized in a sustainable manner. In the present study, photocatalytic degradation of aqueous ES was carried out under direct sun light irradiation using ZS3 and ZS4 without using any other component, such as noble metal co-catalyst or peroxide, and the results are shown in Figure . Decomposition of ES was indicated by UV–vis absorption and confirmed by gas chromatography–mass spectrometry (GC–MS) analysis. A very similar pattern of photocatalytic degradation of ES as a function of irradiation time is shown by ZS4 (Figure a) and ZS3 (Figure b). Virgin ES shows an absorption band at 215–225 nm due to π–π* transition. However, two new absorption peaks appear at higher wavelengths as a function of direct sunlight irradiation time; new absorption peaks appeared at 243 and 282 nm corresponding to π–π* and n−π* transition from one of the fragments or the decomposed product attesting the decomposition of ES into other products. The absorption feature around 200 nm keeps increasing with increasing illumination time, hinting the decomposition of ES into several products or small fragments, which contribute to the high intensity at 200 nm. The absorption feature due to ES at 215 nm disappears gradually. This observation demonstrates the formation of a new product increased with irradiation time. Nonetheless, it was not possible to confirm the complete degradation of ES and the nature of fragments or products because of decomposition from absorption spectral studies. To explore more on this aspect, GC–MS analysis was carried out for pure ES and products or fragments that remain in solution after 60 h of continuous solar irradiation over eight days with the ZS4 catalyst.

Figure 14

Absorption spectra of aqueous ES recorded at different time intervals following photodegradation with (a) ZS4 and (b) ZS3.

GC–MS analysis shows two isomer peaks of ES α and β at retention times (R) 16.4 and 17.6 min, as shown in Figure a (and Figure S7a). Corresponding mass spectra for ES α and β are given in Figure S7b,c. After 60 h of sunlight irradiation, no peak corresponding to α and β ES was observed in GC (Figure b); however, many new features are observed, which is attributed to the various fragments from the decomposition of ES, and the corresponding mass spectrum of each new fragment is given in Figure S8 (1–10). All possible products are listed in Table S1. R = 3.378 min (Figure S8-1) shows mass fragmentations at m/z 118 (C9H10+), 103 (C8H7+), 91 (C7H7+), 77 (C6H5+), 44 (C2H4O+), and 32 (CH4O+); chemical composition of this fragment is C10H12O, and the probability for the same is 97.9%. Figure S8-2 shows a typical mass fragmentation of α-methylbenzyl alcohol from the feature at R = 4.128 min; the relative mass fragmentation observed at m/zs 122 (M+=C8H10O+), 120 (C8H8O+), 107 (M+–CH3), 105 (C7H5O+), 79 (C6H7+), 77 (C6H5+), and 44 (C2H4O+) supports this. It is well-studied that benzene derivatives give a phenylium ion peak at m/z 77 and observed that the probability of α-methylbenzyl alcohol is 50.1% by the NIST library.[45]Figure S8-3 shows the MS corresponding to R = 4.225 min, and it exactly matches with the m/z patterns of acetophenone with a probability of 70%.[45] The observed mass fragments are m/z = 120 (M+), 105 (M–CH3CO+; C7H5O+), 77 (C6H5+), and 43 (CH3CO+). The mass spectrum of the next feature that appeared at R = 4.42 min (Figure S8-4) matches with that of 2-phenyl-2-propenol with a probability of 92%; this feature shows a parent peak at m/z 136 (M+=C9H12O+), and the following fragment peaks appeared at m/z 121 (M+–CH3), 118 M+–H2O), 103 (C8H7+), 91 (C7H8+), 77 (C6H5+), and 43 (CH3CO+). Last two features are the major products observed in the GC pattern because of ES degradation. Observation of these products attests dechlorination of ES in the first step followed by formation of acetophenone or 2-phenyl-2-propenol. Indeed, this is an important observation because dechlorination of ES makes the resulting products to a simple organic molecule that can be photocatalytically degraded easily. Further, it removes the harmful nature that leads to endocrine disruption.

Figure 15

GC results obtained for (a) ES and (b) ES fragments after 60 h of irradiation under direct sunlight with a ZS4 catalyst. Corresponding mass spectral patterns and identity are given in Figure S8 and Table S1, respectively.

A minor feature observed at R = 5.65 min can be attributed to octanol from the mass spectrum (Figure S8-5) with a chemical composition of M+=C8H18O+ exhibiting major mass fragments at m/z 111 (C8H15+), 97 (C7H13+), 83 (C6H11+), 69(C5H9+), 55 (C4H7+), 44 (C3H8+), and 32 (CH4O+). Another minor feature at R = 8.2 min shows mass fragments (Figure S8-6) at m/z 111 (C8H15+), 97 (C7H13+), 83 (C6H11+), 69 (C5H9+), 55 (C4H7+), 44 (C3H8+), and 32 (CH3OH+is likely due to adduct formation with methanol used as a reaction solvent). Yet another minor feature at R = 9.6 min (Figure S8-7) predicted to be a norbornadiene derivative with chemical composition C8H7Cl3 (M+ = 208) exhibiting mass fragmentations at 206 (M+–2), 191 (C7H2Cl3), 91 (C7H7+), 44 (C3H8+), 40, and 32. A partially dechlorinated product from ES leads to norbornadiene; very low intensity highlights the fast decomposition of them as well as fast dechlorination as a whole. A fragment observed at R = 10.7 min (Figure S8-8) is attributed to diethyl phthalate, as the observed MS pattern is in good agreement with that.[47]

A feature observed at R = 17.15 min shows mass fragments at 307 (C9H6Cl5O), 207 (C8H6Cl3+), 44 (C2H4O+), and 32 (CH4O+) (Figure S8-9); from the above mentioned fragments, it has been predicted that the parent chemical composition is C9H6Cl6O (M+ = 342), and it is likely to be ES–ether. GC results show a peak at R = 17.4 min (Figures S8-10), and the corresponding MS results show a typical MS pattern of ES–diol; C9H8Cl6O2 with a molecular mass of 360 with a probability of 84.7%. The corresponding mass values appeared at m/z 325 (C9H8Cl5+O2), 307 (C9H4Cl5+), 277 (C8H8Cl3), 207 (C8H6Cl2O2+), 170 (C8H5ClO2+), 44 (C2H4O+), and 32 (CH4O+). Other unidentified products (trace amounts) were also observed on GC–MS. The above mentioned discussions are consistent with the UV–visible absorption spectrum of degraded products; the observed peaks at 243 and 282 nm correspond to π–π* and n−π* transitions of benzene derivatives attached with heteroatoms such as −OH or keto groups. From the GC–MS and UV–visible results, it demonstrates that ES was degraded through ES–diol, ES–ether, α-methylbenzyl alcohol, and acetophenone pathway; it was further degraded into smaller fragments. Importantly, the above mentioned observations highlight the visible light activity associated with ZSx catalysts. Indeed, without a catalyst or in the absence of light source, no ES degradation was observed. A semiquantitative calculation from the GC chromatogram recorded before and after the 60 h irradiation demonstrates that, at least 80% ES was completely mineralized to CO2 through various fragments. Some of the corollary experiments carried out for complete degradation of acetophenone or 2-phenyl-2-propanol with ZS4 (not shown) in sunlight occurs in 16–20 h with 50 ppm concentration.

ZnO–ZnS heterojunctions present in ZSx is the prime reason for the visible light absorption and better activity. The heterojunctions observed between ZnS and ZnO nanocrystallites help in electron–hole pair separation, and facile mobility of charge carriers allows for better photocatalytic degradation of ES. The porous structure and high surface area induce the photogenerated carriers to diffuse fast to the surface and react rapidly with the adsorbed molecules and hence an effective utilization of charge carriers for the degradation of ES.

Conclusion

ZnO–ZnS (ZSx) composites with heterojunctions between the constituent components were prepared by the SCM and characterized by a variety of physicochemical, spectral, and microscopy measurements. ZSx were indexed to the wurtzite structure; however, with an increase in the thiourea amount, an increase in the ZnS content was also observed. An important point to be noted is the increasing crystallite size of ZnS with increasing thiourea content in the reactant mixture. Heat treatment in air between 450 and 650 °C decreases the ZnS crystallite size, offering a method to fine tune the particle size. The present method of preparation and the tuning of the ZnS crystallite size in the solid state are important, and it is not available at present in the solid state. Further, high thermal stability of the present materials makes them flexible for several applications, including catalysis.

The porous structure with needle- or V-shaped particles with significant macropore contribution underscores its potential in sensing and catalysis applications. Raman, SIMS, TG, and XRD studies demonstrate the direct bonding between Zn and S as in typical ZnS. Optical absorption studies demonstrate a significant decrease in the band gap down to 2.3 eV for ZSx compared to wide band gap for parent materials (ZnO–ZnS). The band gap diminution of ZSx has been attributed to the formation of hybrid orbitals in the VB and CB. It is expected that S 3p and O 2p are likely to overlap in the VB of ZSx at and around interfaces and heterojunctions, and the extent of overlap increases from ZS1 to ZS4. Sulphur exists as an S2– anion in ZSx, and the top of the VB might be predominantly composed of S 3p bands, as they have higher orbital energies than O 2p levels. Thus, S 3p and O 2p bands overlap and broaden the VB compared to ZnO or ZnS. The above mentioned studies affirm the nature of ZSx with heterojunctions between ZnO and ZnS and ZSx as a potential candidate for PV application.

Photocurrent generation under 1 sun condition and heterojunctions observed between ZnO and ZnS demonstrates the effective sunlight absorption and separation of charge carriers. Probably, this is the first report to demonstrate photocurrent generation under 1 sun condition by combining two wide band gap semiconductors into a composite with heterojunctions. Photocatalytic activity studies of ZSx (for complete mineralization of 2,4-DCP and ES ECDs in direct sunlight) demonstrate its potential to develop efficient visible light active photocatalysts. We also suggest that ZSx composites might offer a possible solution to eradicate the EDC contamination in the large amount of farm lands in many countries. A simple mixing of ZSx with the soil in the presence of water and sunlight would mineralize EDC in about a month’s time, and it is very likely to make the land cultivable again. However, systematic and thorough pilot plant studies will be required to evaluate the feasibility of this option. These applications are depicted pictorially in Figure .

Figure 16

Possible applications with ZSx composites are depicted. (Right) ES-/ECD-infested land could be made fertile again and (left) possible to generate power. Both applications are with direct sunlight.

Experimental Section

Synthesis of ZnO–ZnS Composites

All chemicals employed were of analytical grade and used as such. Zn(NO3)2·6H2O (Merck) as a zinc precursor and urea and thiourea (Loba Chemie) as fuels were used. SCM[30] was adopted to make ZSx, which is a simple process that requires short time and a laboratory furnace. In a typical synthesis procedure, an equimolar (0.04 mol) amount Zn(NO3)2·6H2O and fuel, (but with a different ratio of urea and thiourea, varied from 100% urea to 100% thiourea) were taken in a 250 mL beaker with 10 mL of distilled water.

After thorough solubilization of reactants, the solution was introduced into a preheated furnace at 500 °C. Water evaporation takes place for up to 2 min, followed by ignition of the reactant mixture, yielding a yellow solid. With 100% urea, brown-colored ZnO1–N was obtained.[16] The solid product was collected after the completion of the combustion process. Urea/thiourea was varied from 100:0, 75:25, 50:50, and 25:75 to 0:100, and the resulting materials are designated as Z1, ZS1, ZS2, ZS3, and ZS4, respectively. All the above mentioned materials were thoroughly characterized, and the characterization methods and instruments are described in the Supporting Information (S1).