A mechanistic approach for superoxide radicals and singlet oxygen mediated enhanced photocatalytic dye degradation by selenium doped ZnS nanoparticles

Abstract

Structural modification of semiconductor type nanoparticles by doping with metals and non-metals renders versatility of the material. Here, we report the synthesis of <4 nm sized nanocrystalline selenium doping in ZnS nanoparticles (Se-doped ZnS NPs) by chemical precipitation method, and discuss the mechanism involved in drastic improvement in its photocatalytic activity as compared to pristine ZnS nanoparticles. An array of characterization techniques, including X-ray diffraction (XRD), high-resolution transmission electron microscopy (HRTEM), diffuse reflectance spectroscopy (DRS), Fourier-transform infrared spectroscopy (FT-IR), zeta potential and BET surface area measurements were performed to confirm the synthesis of stable Se-doped ZnS NPs with large surface area. Selenium doping not only led to lattice distortion in the ZnS lattice but it also resulted in band gap narrowing due to incorporation of defect sites within the forbidden gap, supported by fluorescence emission spectroscopy studies. These modified properties in Se-doped ZnS NPs were important for enhanced sunlight induced photocatalytic activity, demonstrated via degradation of trypan blue dye. Compared to the pristine ZnS nanoparticles, the kinetic study revealed a three-fold increase in the efficiency of the dye degradation and about seven-fold increase in the rate of dye degradation by Se-doped ZnS NPs. The mechanism of the dye degradation is studied in terms of scavenging of generated reactive oxygen species (ROS) in the reaction medium. The p-benzoquinone and sodium azide assay confirmed that the dye degradation by Se-doped ZnS was due to production of superoxide radicals and singlet oxygen species, respectively, which were generated due to interaction of photoexcited conduction band electrons with molecular oxygen at the surface of the nanoparticles. The proposed mechanism is supported by the theoretical calculation of the band structure of Se-doped ZnS nanoparticles.

---

1. Introduction

Wastewater effluents from textile industries are a severe threat to the environment and human health and hence demand suitable remediation technology.^1,2 In this regard, semiconductor type nanomaterials have rapidly evolved as efficient heterogenous photocatalysts for wastewater treatment.^3–7 Semiconductor nanomaterials as photocatalysts for dye degradation are not only economically favorable and technologically sustainable, but also environmentally more compatible than conventional oxidation by bleaching or by advanced oxidation process (AOP) based on Fenton chemistry using a UV light source and strong oxidizing agents such as ozone, or by electron beam irradiation.^8–10 The photoinduced reaction in a semiconductor nanoparticle is initiated by absorption of light energy greater than or equal to the band gap of the material to generate valence band holes (h⁺) and conduction band electrons (e⁻), which could either recombine or transport to the surface of the nanoparticles.¹¹ Typically, photocatalytic dye degradation by semiconductor type nanomaterials is an oxidation process where the free or trapped holes oxidize water molecules directly to produce hydroxyl radicals or undergo indirect oxidation by ROS (e.g., singlet oxygen, superoxide radicals), which are formed due to reduction of molecular oxygen by free or trapped electrons.^12,13 Various experimental studies including electron spin resonance spectroscopy (ESR), laser induced fluorescence spectroscopy, and use of fluorescent probes are reported for establishing the mechanism of ROS generation and subsequent dye degradation.^14–16 Most of these studies are based on the usages of various metal oxide nanoparticles, in the presence and absence of co-catalysts such as H₂O₂ or S₂O₈²⁻.^17–19 The most important feature regarding dye degradation is to achieve a faster rate of degradation especially for industrial dyes which comprise of complex structures containing one or more azo (–N=N–) chromogenic groups which are stable against conventional methods of degradation.²⁰

It may be remarked here that there exists an enormous scope for exploring materials other than metal oxides as photocatalysts for wastewater treatment, and particularly to explore dye degradation.^21,22 In this regard, ZnS nanoparticles are investigated as an efficient photocatalyst for degradation of organic pollutants in wastewater treatment.²³ However, the major disadvantage of using ZnS NPs as photocatalyst is attributed to its wide band gap (3.68 eV for bulk or higher for NPs). This would therefore require UV light irradiation and hence imposes a self-limiting constraint for utilizing the solar energy spectrum. However, conjugation of polymer–ZnS complex and doping of ZnS nanocrystals with Cu, Ni and Pb was reported to exhibit photocatalytic dye degradation using visible light.^24,25 The other challenge for ZnS NPs has been attributed to its route of synthesis, where chemical precipitation method is recommended to be followed by heat treatment under a nitrogen atmosphere.²⁶

Here we report the mechanism of photocatalytic dye degradation by Se-doped ZnS nanoparticles. Selenium was chosen in this study for its favorable photoelectric conversion property.^27–29 In addition, it is non-toxic and therefore suitable for environmental applications.³⁰ The mechanism of enhanced ability of photocatalytic degradation of trypan blue (a model azo dye) has been discussed in the light of structural changes in ZnS due to Se doping and ROS generation in aqueous medium when exposed to sunlight.

2. Materials and methods

2.1. Chemicals

Zinc acetate, sodium sulfide (Na₂S), mercaptoacetic acid (MAA) or thioglycolic acid (HSCH₂COOH), selenium dioxide (SeO₂), trypan blue dye, terephthalic acid (C₆H₄(COOH))₂, ethanol (C₂H₅OH), sodium azide (NaN₃), p-benzoquinone, dimethyl sulfoxide (DMSO), hydrogen peroxide (H₂O₂) and concentrated hydrochloric acid (HCl) were procured from Himedia Labs Pvt. Ltd. India. All reagents and precursors used in this study were of analytical grade and were used without further purification. All aqueous solutions used in this study were prepared with de-ionized water (Millipore).

2.2. Synthesis of ZnS NPs and Se-doped ZnS NPs

Two batches of selenium doped ZnS NPs were prepared by chemical precipitation method, where 11 mg (X) or 22 mg (Y) of selenium dioxide were respectively mixed in 50 mL of 20 mM zinc acetate solution and stirred for 1 h at 60–80 °C, at pH 7. Similarly, 60 mM solution of Na₂S as a precipitating agent was prepared in deionized water. To this solution, 175 μL of 40 mM mercaptoacetic acid (MAA) was added as a capping agent. This mixture was homogenized by magnetic stirring and was added dropwise to the solution containing zinc acetate and selenium dioxide under stirring. First a brownish turbid solution was formed to which a 0.5 M solution of NaOH was added dropwise. This mixture was subjected to stirring until a light yellow precipitate of ZnS nanoparticles doped with Se was formed (Se-doped ZnS NPs). It may be suggested here that mercaptoacetic acid played the role of capping agent by forming S–S bonds with ZnS. The nanoparticles were recovered by centrifugation and washed with ethanol several times to remove impurities and then air-dried overnight. Similarly, pristine ZnS nanoparticles (ZnS NPs) were synthesized without adding selenium dioxide in the above protocol. The actual concentration of selenium and zinc in the batches of Se-doped ZnS NPs were analyzed by atomic absorption spectrometry (Perkin Elmer Analyst 800) after digesting Se-doped ZnS NPs with a mixture of concentrated nitric acid and hydrochloric acid (1 : 3 molar ratio). An accurately weighed 50 mg sample of Se-doped ZnS NPs was treated with 15 mL of the acid mixture and heated until brown fumes ceased to evolve. The digest was slowly cooled to room temperature and the volume was then made up to 1000 mL by serial dilution in a volumetric flask. Three sets of digestion of respective batches X and Y of Se-doped ZnS NPs were performed for independent triplicate analysis. The concentrations of selenium in the respective batches of the digested samples were determined in triplicate against the calibration plot obtained by analysing known concentrations of standard selenium (Merck). The concentration of Se in Se-doped ZnS NPs was calculated using the formula given below:

Se (wt%) = [mass of Se in acid digest (mg)/mass of Se-doped ZnS NPs digested (mg)] × 100

where, mass of Se in acid digest = Se concentration measured by AAS (say, α mg L⁻¹) × volume make up (i.e., 1000 mL).

2.3. Characterization techniques

The X-ray diffraction measurements of the Se-doped ZnS NPs and pristine ZnS NPs were recorded with a Bruker ARS D8 Advance diffractometer. The instrument was operated at 40 kV where graphite-monochromatized Cu-Kα X-rays (λ = 1.54 Å) was scanned at 2° min⁻¹ in a wide angle region from 10 to 80° on a 2θ scale. The morphology and particle size distribution of the nanoparticles were studied by high-resolution transmission electron microscopy (HRTEM) coupled with element analysis facility by energy dispersive X-ray (EDX) analysis. The TEM studies were carried out using an FEI Technai-G² microscope operated at an acceleration voltage of 200 kV. The samples for TEM were prepared by drop casting the nanoparticles dispersed in ethanol on to a 150 mesh copper grid. The Brunauer–Emmett–Teller (BET) specific surface area of the nanoparticles were measured by a Quantachrome Nova 2200 analyzer at liquid-nitrogen temperature using nitrogen adsorption–desorption method.

The optical band gap of the as-synthesized Se-doped ZnS NPs and pristine ZnS NPs were determined from their respective UV-visible absorption spectra (Shimadzu UV1800). The emission spectra of the nanoparticles were recorded by a Spectro-fluorimeter FLS 980 (HORIBA), in the wavelength range of 400–1500 nm, where the excitation wavelength was 680 nm. The Fourier-transform infrared (FT-IR) spectra were recorded on a Thermo NICOLET 6700 FTIR in KBr pellets. The zeta potential of the nanoparticles was recorded in aqueous medium by a Malvern Zetasizer Nano ZS90.

2.4. Photocatalytic degradation study

Trypan blue dye solution (15 mg L⁻¹) was chosen as a test dye for sunlight induced photocatalytic degradation study by the batches of as-synthesized pristine ZnS NPs and Se-doped ZnS NPs as photocatalysts. All photocatalytic degradation experiments were carried out at IIT Roorkee (29°51′ N; 77°53′ E) in April and May 2015. First, a batch of 0.8 mg mL⁻¹ photocatalyst was treated with 50 mL of the dye solution in the dark. 1 mL aliquots were collected after a specific time interval and the concentration of dye was measured by a UV-visible spectrophotometer (Shimadzu UV-1800). A 15% decrease in dye concentration was recorded and the adsorption–desorption equilibrium was attained within 60 min.

For studying photocatalytic dye degradation, the batches of dye solution treated with respective photocatalysts were first kept in dark for 60 min and then the experimental assembly was exposed to sunlight for 3 h. A kinetic study of dye degradation was performed by measuring the concentration of the dye solution at specific time intervals. The absorbance of the decolourized solution was measured at λ_max = 590 nm after centrifugation at 15 000 rpm at 25 °C for 5 min in Beckman Coulter™ Allegra TM X-22R centrifuge. Simultaneously a controlled kinetic experiment was conducted where the dye solution of same concentration, without treating with nanoparticles (as photocatalyst), was exposed to sunlight.

2.5. Reactive oxygen species (ROS) assay

Interaction of semiconductor nanoparticles, including ZnS in aqueous medium tends to generate reactive oxygen species (ROS), e.g., hydroxyl radicals (OH˙), superoxide radicals (O₂˙⁻) and singlet oxygen (¹O₂). The tentative generation of ROS were studied by the following methods:

2.5.1. Terephthalic acid assay. The generation of hydroxyl radicals due to interaction of pristine and Se-doped ZnS NPs in aqueous medium under the exposure to sunlight was determined by terephthalic acid assay.³⁰ Batches of as-prepared nanoparticles (0.8 mg mL⁻¹) were dispersed in 2 mM terephthalic acid solution prepared in NaOH solution and were sonicated for 10 min and then exposed to sunlight for 40 min. The hydroxyl radicals, if generated in the medium would form a fluorescent product with terephthalic acid. After centrifuging the reaction mixture, the fluorescence emission spectrum of the product was measured at an excitation wavelength of λ = 315 nm. A control batch of dye solution without treating with nanoparticles was simultaneously exposed to sunlight to confirm that the dye degradation was not due to photolysis. A second control experiment was also performed in dark condition in the presence of the nanoparticles to assess the extent of adsorption of dye on the nanoparticles.

2.5.2. Superoxide radical, singlet oxygen and hydroxyl radical scavenging. The generation of superoxide radicals (O₂˙⁻), singlet oxygen (¹O₂) and hydroxyl radicals (OH˙) in the reaction medium were determined by treating with their respective quenchers. In this regard, p-benzoquinone was used as a superoxide radical specific quencher;³¹ sodium azide as singlet oxygen specific quencher;³² and dimethyl sulfoxide as hydroxyl radical quencher.³³ If the dye degradation is triggered due to generation of any of these ROS then the photocatalytic dye degradation would be inhibited in the presence of their respective quenchers.

To assess the role of singlet oxygen the batches of trypan blue dye solution (concentration of 15 mg L⁻¹) were mixed with 10, 50, 100 and 200 mM sodium azide, respectively. These batches were then treated with Se-doped ZnS NPs (0.8 mg mL⁻¹), respectively. A control experiment was run without adding sodium azide. After equilibrating the reaction mixture in the dark followed by exposure to sunlight, the kinetics of dye degradation inhibition was studied by monitoring the intensity of the UV-visible absorption spectrum of the dye. Similarly, the generation of hydroxyl radicals, if any, in the reaction medium was determined by treating with dimethyl sulfoxide (DMSO), as a quencher. The methodology for DMSO based hydroxyl radical scavenging was the same as that described for the sodium azide assay. However, the methodology for p-benzoquinone based scavenging of superoxide radicals was different due to poor solubility of p-benzoquinone in water. First, the commercially procured p-benzoquinone was recrystallized from ethanol. The freshly prepared yellow coloured crystals of p-benzoquinone were then dissolved in ethanol (98%), which was mixed with the trypan blue solution prepared in deionized water so that the ethanol : water ratio was 1 : 9. 0.8 mg ml⁻¹ of Se-doped ZnS NPs was added to the dye solution mixed with p-benzoquinone and equilibrated in the dark for 60 min followed by exposure to sunlight for 3 h. A control experiment was performed where the dye solution in ethanol–water was monitored without treating with Se-doped ZnS NPs. Since ethanol evaporates faster than water it was necessary to maintain the ethanol : water ratio in the dye solution throughout the period of the kinetic study by adding appropriate volumes of ethanol. The inhibition of dye degradation due to addition of p-benzoquinone was determined from a kinetic study as described above.

3. Results and discussion

3.1. Characterization of Se-doped ZnS NPs

The exact concentrations of selenium in the X and Y batches of Se-doped ZnS NPs measured from atomic absorption spectrometry analysis were determined to be 2.60 ± 0.12 wt% and 3.50 ± 0.10 wt%, of the as-prepared batches of Se-doped ZnS nanoparticles, respectively. The Se concentrations corresponded to mean ± standard deviation of triplicate analysis. The X-ray diffraction patterns of the pristine ZnS NPs and Se-doped ZnS NPs are shown in Fig. 1. The characteristic Bragg reflections for the pristine ZnS NPs were located at 28.69, 47.68, 56.39, 76.93 and 88.75°, corresponding to (111), (220), (311), (400) and (331) planes, respectively, of the cubic crystalline structure of ZnS, and matched with JCPDS file number 05-0566. The XRD pattern of Se-doped ZnS NPs revealed three important features, viz. (a) shift in the XRD peaks to higher 2θ values, i.e., 28.99° (111), 47.92° (220) and 56.65° (311) planes; (b) decreasing order of crystallinity as evident from the decrease in the intensities of the respective diffraction peaks; and (c) broadening of the peaks. The peak shift to higher 2θ value is attributable to the doping effect, particularly due to substitution with smaller sized Se⁴⁺ dopants (0.050 nm) in the lattice sites of ZnS where the ionic radius of Zn²⁺ is 0.060 nm.^34,35 Such a situation would result in the decrease of the lattice parameters of ZnS and consequently lead to lattice distortion. The effect of lattice strain is reflected from the broadening of the XRD peaks for the batch of 2.6 wt% Se-doped ZnS NPs, and further broadening was observed for the batch doped with 3.5 wt% Se-doped ZnS NPs (Fig. 1). The observed broadening of the XRD peaks could also be due to a decrease in the crystallite size.³⁶ The sizes of the as-synthesized ZnS and 2.6 wt% Se-doped ZnS NPs were measured by TEM and results are presented in Fig. 2a and b, respectively. The particle sizes were observed in the range of 2–7 nm with average size of about 3.9 nm (inset of Fig. 2b). Such small particle sizes measured by TEM were corroborated with large BET surface areas, i.e., 168.4 m² g⁻¹ for pristine ZnS NPs and 172.6 m² g⁻¹ for 2.6 wt% Se-doped ZnS NPs. The crystalline nature of the Se-doped ZnS NPs was confirmed from the high-resolution TEM measurement which revealed lattice fringes separated by 0.311 nm, shown in Fig. 2c and the interatomic spacing was also confirmed in the profile plot (inset of Fig. 2c). The interatomic layer spacings measured in our study were consistent with those reported in literature.³⁷

Fig. 1 X-Ray diffraction (XRD) pattern of pristine ZnS NPs, 2.6 wt% Se-doped ZnS NPs and 3.5 wt% Se-doped ZnS NPs.

Fig. 2 Transmission electron microscopy (TEM) images of (a) pristine ZnS NPs; (b) 2.6 wt% Se-doped ZnS NPs; (inset: size distribution histogram of 2.6 wt% Se-doped ZnS NPs); (c) high-resolution TEM (HRTEM) image of 2.6 wt% Se-doped ZnS NPs showing lattice fringes marked by the arrows; (inset: profile image of the lattice fringes).

The stabilization of Se-doped ZnS NPs of sizes in the range 2–7 nm was attributable to effective capping of the nanoparticles by mercaptoacetic acid. Under this scheme, the capping of ZnS NPs by mercaptoacetic acid would lead to formation of S–S bonds. This was confirmed from FT-IR measurement, where the S–H bond was detected in mercaptoacetic acid at 2566 cm⁻¹ while it was absent in the batches of pristine as well as Se-doped ZnS NPs (Fig. 3). In addition, the band observed at 1714 cm⁻¹ corresponding to the carboxylic acid group in mercaptoacetic acid reagent, shifted to 1642 cm⁻¹ for Se-doped ZnS NPs and 1638 cm⁻¹ for pristine ZnS NPs. This shift is implicative of the carboxylate group favoring electrostatic interaction between the positively charged ZnS NPs and negatively charged carboxylate ions. It may be argued that such electrostatic interaction rendered high stability of the as-synthesized nanoparticles. The colloidal stability of the as-synthesized Se-doped ZnS NPs in aqueous medium was also confirmed from the measured value of large zeta potential at neutral pH (ξ = −44.5 mV). The FT-IR band at 665 cm⁻¹ observed for pristine ZnS NPs and in Se-doped ZnS NPs can be attributed to Zn–S vibration (corresponding to sulfide).

Fig. 3 Fourier-transform infrared (FT-IR) spectra of (a) mercaptoacetic acid (MAA) used as capping agent; (b) 2.6 wt% Se-doped ZnS NPs capped with MAA; and (c) pristine ZnS NPs capped with MAA.

The band structure of pristine ZnS NPs and Se-doped ZnS NPs were determined from UV-visible absorption spectra (Fig. 4). The band gap of the samples was estimated from the absorption edge of a Tauc plot, using the Tauc expression, given below,³⁸

(αhν) = A(hν − E_g)ⁿ;

here α (absorption co-efficient) = (2.303A)/d, hν, A, d and E_g are the energy of the incident light, absorbance, path length and band gap, respectively. The value of exponent ‘n’ determines the nature of electronic transition. Assuming a direct transition (i.e., n = 1/2), the Tauc plot was obtained from (αhν)² vs. hν and the band gap energy was determined by extrapolating the linear region of (αhν)² = 0. The band gap of pristine ZnS NPs, 2.6 wt% and 3.5 wt% Se-doped ZnS NPs were determined as 3.84 eV, 3.47 eV and 3.54 eV, respectively. Notably the band gap for pristine ZnS NPs was larger than that of the bulk ZnS (i.e., 3.68 eV), and the increase in the band gap can be attributed to a quantum confinement effect in ZnS nanocrystals.³⁹

Fig. 4 Tauc plot obtained from UV-visible absorption spectroscopy of pristine ZnS NPs, 2.6 wt% and 3.5 wt% Se-doped ZnS NPs.

Though the sizes of the Se-doped ZnS NPs were similar to those of pristine ZnS NPs, the Se-doped ZnS NPs revealed narrowing of the band gap. This decrease in the band gap for Se-doped ZnS NPs can be attributed to structural defects corresponding to interstitial energy levels in the forbidden gap, which could be due to Zn vacancies, sulfur vacancies or cations at interstitial sites, as demonstrated in several cases of doped nanoparticles.^40–42 The structural defects in Se-doped ZnS was further evident from the fluorescence emission spectrum, shown in Fig. 5a. In addition to a sharp peak at 367 nm, a broad emission peak in the wavelength region of 380–550 nm was observed. The peak at 367 nm is attributable to a near band emission peak. However, the broad peak was associated with a number of small peaks on top of it, which indicated multiple emission peaks in that region. Suitable data deconvolution was performed which revealed emission peaks at 415, 456, 470, 483 and 493 nm (Fig. 5b), whose sum peaks fitted well with our experimental data. These emission peaks corresponded to interstates in the Se-doped ZnS lattice owing to structural defects. The corresponding energy levels were in the forbidden gap closer to the valence band (shown in Scheme 1). These intermediate energy levels might facilitate interfacial electron (e⁻) transfer from the conduction band and hence promote ROS generation and thereby facilitate photocatalytic activity.⁴³

Fig. 5 Fluorescence emission spectra of (a) pristine ZnS NPs and (b) 2.6 wt% Se-doped ZnS NPs.

Scheme 1 Schematic diagram showing energy levels of the valence band, conduction band and interstates of the 2.6 wt% Se-doped ZnS NPs in NHE scale and its suitability for photocatalytic degradation of trypan blue dye.

3.2. Photocatalytic degradation

The large surface area of Se-doped ZnS NPs was favourable for photocatalytic reaction, which is evident from the kinetic studies of sunlight exposed photocatalytic degradation of trypan blue dye by pristine ZnS NPs and Se-doped ZnS NPs (Fig. 6). Notably, more than 99% dye degradation was achieved in 3 h for the batch treated with 2.6 wt% Se-doped ZnS NPs as compared to only 38% dye degradation for the batch treated with pristine ZnS NPs. The control experiment for the batch of dye solution treated with Se-doped ZnS NPs conducted in the dark condition did not reveal dye removal. This indicated that the adsorption of trypan blue dye on the nanoparticles was negligible. The unfavourable dye adsorption phenomenon can be attributed to the negatively charged surface of the nanoparticles at pH 7 (ξ = −44.5 mV). Similarly, another control experiment was performed to rule out the case of dye degradation due to photolysis, where degradation of dye only due to sunlight exposure was insignificant (Fig. 6). The negative results in both control experiments indicated that dye removal was neither due to its adsorption of dye on nanoparticles nor due to photolysis. It may be concluded that dye removal was due solely to photocatalytic dye degradation by Se-doped ZnS NPs. Further, it was noted that 60% of the dye was degraded in the first two hours, and the rate of dye degradation by Se-doped ZnS NPs was apparently much faster than other reported metal oxide nanoparticles, e.g., ZnO NPs as photocatalyst.⁴⁴ The kinetics of the photocatalytic degradation by pristine ZnS NPs and Se-doped ZnS NPs as photocatalyst was studied using the Langmuir–Hinshelwood model.^45,46

ln(C_o/C) = kt,

where ‘C_o’ is the initial concentration of the dye solution, ‘C’ is the dye concentration at time ‘t’ and ‘k’ is the pseudo-first order rate constant. From the kinetic studies it was deduced that the photocatalytic dye degradation followed first-order kinetics with rate constant k = 0.00316 min⁻¹ (R² = 0.987) for ZnS NPs and k = 0.02116 min⁻¹ (R² = 0.96) for Se-doped ZnS NPs. Notably, the degradation rate constant for Se-doped ZnS NPs was seven-fold greater than that of ZnS NPs and was comparable with dye degradation reported for batches of metal oxides assisted with strong oxidizing agents such as H₂O₂ as co-catalyst.¹⁹

Fig. 6 Sunlight induced kinetic studies of degradation of trypan blue dye treated with batches of pristine ZnS NPs; 2.6 wt% Se-doped ZnS NPs and control batch of dye solution without treating with any nanoparticles.

In order to evaluate the stability and re-usability of the Se-doped ZnS NPs as photocatalysts, three repetitive cycles of degradation process were performed, i.e., equilibration in dark followed by sunlight exposure and recovery of the photocatalyst for subsequent cycles. In all the three cycles, more than 98% degradation of trypan blue dye was achieved in 180 min of sunlight exposure and their respective rate constants (k) are given in Fig. 7. Notably, the rate constants of photocatalytic dye degradation determined in each cycle were similar to the test sample (Fig. 6). This indicated that Se-doped ZnS NPs can be efficiently used as a photocatalyst for dye degradation. It may be commented that the drastic improvement in the dye degradation exhibited by Se-doped ZnS NPs could be due to the high reduction potential of selenium, which is a favorable condition for extraction of photoexcited electrons from the conduction band and to facilitate generation of ROS.³⁰

Fig. 7 Sunlight induced kinetic study showing reusability of the photocatalyst for sustained efficiency in dye degradation performed for three successive cycles.

3.3. Mechanism of dye degradation

3.3.1. Role of reactive oxygen species (ROS). Light induced generation of free radicals due to the interaction of semiconductor type nanoparticles in aqueous medium is considered to be the key event towards photocatalytic dye degradation.^47,48 In our present study, the role of ROS generation by pristine and Se-doped ZnS NPs for photocatalytic dye degradation was confirmed from the terephthalic acid assay, which is selective to hydroxyl radicals.³⁰ A strong fluorescence emission peak was recorded at λ_em = 424 nm which was due to formation of an adduct between ROS and terephthalic acid (Fig. 8). Notably in the absence of sunlight, the characteristic peak at λ_em = 424 nm was negligible, which indicated that ROS production was insignificant in the dark condition. It was further noted from the terephthalic acid assay that the intensities of the λ_em = 424 nm emission peak was similar for both the batches treated with pristine and Se-doped ZnS NPs, though the rate of dye degradation was several fold higher for the batch treated with Se-doped ZnS NPs. This indicated that in addition to hydroxyl radicals, Se-doped ZnS NPs produced a different types of ROS. It also meant that a different mechanism was involved in the dye degradation by Se-doped ZnS NPs when compared to pristine ZnS NPs. It may be commented here that superoxide radicals and singlet oxygen are the other forms of reactive oxygen species that might be generated by metal oxide photocatalysts, and singlet oxygen is eventually one of the end products of superoxide radicals (O₂˙⁻).^11,12

Fig. 8 Terephthalic assay profile for hydroxyl radical generation in the presence of photocatalysts (pristine ZnS and 2.6 wt% Se-doped ZnS NPs), both in dark condition and under sunlight exposure.

3.3.2. ROS scavenging studies. In order to get an insight of the mechanism of photocatalytic dye degradation by Se-doped ZnS NPs, specific ROS scavengers e.g., sodium azide as singlet oxygen scavenger, DMSO as hydroxyl radical scavenger and p-benzoquinone as superoxide radical scavenger were treated in the reaction medium comprising the dye solution and Se-doped ZnS NPs as photocatalyst. The kinetic studies of photocatalytic dye degradation in the presence of DMSO and sodium azide are shown in Fig. 9a. Compared to the batch treated with sodium azide the rate of inhibition of photocatalytic dye degradation for the batch treated with DMSO was significantly less. This indicated that generation of hydroxyl radicals in the reaction medium was not significant for Se-doped ZnS NPs. On the other hand the photocatalytic dye degradation was significantly inhibited for the batch treated with sodium azide. This is attributable to scavenging of singlet oxygen species from the reaction medium. In other words, this indirectly implied that singlet oxygen species were generated in the reaction medium by Se-doped ZnS NPs. It may be remarked here that singlet oxygen species are produced due to reaction of holes with superoxide radicals.¹¹ The tentative generation of superoxide radicals was studied using p-benzoquinone as a specific scavenger. Since p-benzoquinone is insoluble in water but soluble in ethanol the kinetic study of inhibition of dye degradation by p-benzoquinone was performed in an ethanol–water mixture. Compared to the control batch (i.e. without treating with p-benzoquinone), 1 mM solution of freshly recrystallized ethanolic solution p-benzoquinone was capable of inhibiting trypan blue dye degradation (Fig. 9b). This indirectly implied that superoxide radicals (O₂˙⁻) were generated in the reaction medium during the interaction with Se-doped ZnS NPs. In this study it was however not possible to directly compare the extent of generation of superoxide radicals and singlet oxygen species from the rate of inhibition of dye degradation by p-benzoquinone (superoxide radical scavengers) and that by sodium azide (singlet oxygen scavenger), respectively, as these studies were performed in different solvent systems. However, singlet oxygen species are formed from the superoxide radicals and hence it may be remarked that both superoxide radicals and singlet oxygen species were generated in the reaction medium and were responsible for dye degradation.

Fig. 9 Kinetic studies of inhibition of trypan blue dye degradation treated with 2.6 wt% Se-doped ZnS NPs as photocatalyst exposed to sunlight for batches (a) in the presence of sodium azide as singlet oxygen scavenger and DMSO as hydroxyl radical scavengers, and (b) in the presence of p-benzoquinone (BQ) as superoxide radical scavenger.

As the ROS generation study was based on scavenging action, it was necessary to confirm that sodium azide did not scavenge hydroxyl radicals. The hydroxyl radicals were generated in batches of dye solution (without treating with photocatalyst) by spiking suitable concentrations of H₂O₂ and exposing to UV light (λ = 360 nm). These batches were treated with sodium azide and DMSO, respectively. As expected, the batch of dye solution treated with sodium azide did not inhibit dye degradation while the batch treated with DMSO revealed inhibition of dye degradation (Fig. 10a). This clearly explained that sodium azide cannot scavenge hydroxyl radicals. The role of sodium azide in scavenging singlet oxygen species was also revealed from the increase in the rate of inhibition of dye solution by increasing sodium azide concentration, where the batch treated with 200 mM sodium azide completely inhibited dye degradation (Fig. 10b). It may therefore be concluded that the drastic inhibition of dye degradation observed in the test samples (shown in Fig. 9a) by sodium azide was due to scavenging of singlet oxygen species which were generated in the reaction medium due to interaction of Se-doped ZnS NPs with dye solution.

Fig. 10 (a) The consequences of mimicking hydroxyl radical generation towards dye degradation in the batches of trypan dye solution treated with DMSO (hydroxyl radical scavenger) and sodium azide (singlet oxygen scavenger); (b) effect of varying concentration of sodium azide on the rates of inhibition of dye degradation.

3.3.3. Band structure of Se-doped ZnS NPs and singlet oxygen generation. In order to account for the role of singlet oxygen towards photocatalytic dye degradation by Se-doped ZnS NPs, it is important to understand the electronic band structure and possible charge transport phenomenon in Se-doped ZnS NPs. The electronic structure of Se-doped ZnS NPs is given in Scheme 1 with an assumption that the structure was similar to ZnS NPs without Se doping. The electronic structure was characterized by the band gap energy (E_g) corresponding to the difference between the valence band (E_VB) and the conduction band (E_CB), each comprising of a high density of states. The conduction band edge minima (E_CB) and valence band maxima (E_VB) of the semiconductor-like materials of Se-doped ZnS NPs as photocatalysts could be empirically calculated using the following expressions with Nernstian correction for pH:^49,50

E_CB = χ(A_aB_b) – 1/2E_g + E_o;

and the pH corrected expression,

E_CB(pH) = E_CB + 0.059(PZZP − pH)

and

E_VB = E_CB + E_g:

and the pH corrected expression,

E_VB(pH) = E_VB + 0.059(PZZP − pH)

where, E_g is the band gap of Se–ZnS NPs, E_CB is the potential of the conduction band, E_VB is the potential of valence band, E_o is the scale factor taken as –4.50 eV, i.e., the energy of free electrons on the normal hydrogen electrode (NHE scale). The parameter χ(A_aB_b) corresponds to the absolute electronegativity of a semiconductor material type A_aB_b, which is calculated as the geometric mean of the electronegativity of the constituent atoms.⁵¹ In our calculation, the electronegativity values for Zn and S were taken as 4.45 and 6.22 eV, respectively. From the above expressions, the energy levels of the conduction band and the valence band were calculated as −1.17 and +2.29 eV, respectively. These energy levels are important parameters that decide the feasibility of charge transport of photoexcited electrons and holes to the surface of the nanoparticles and subsequent interaction with molecular oxygen and water, respectively, to produce the reactive oxygen species. The redox potentials for O₂/O₂˙⁻ and H₂O/OH˙ is given as −0.2 and +2.2 eV, respectively, in the NHE scale (shown in Scheme 1).⁵² Apparently in the case of Se-doped ZnS NPs, the charge transport is favourable for both photoexcited electrons from the conduction band to the redox couple of molecular oxygen to form superoxide radicals and holes from the valence band to the redox couple of H₂O to form hydroxyl radicals. However, our ROS scavenging studies revealed that only superoxide radicals and singlet oxygen were generated while generation of hydroxyl radicals was not evident. It may be remarked here that the holes generated in our system were utilized for converting superoxide radicals to singlet oxygen as given in Scheme 1.¹¹

Moreover the dye degradation efficiency of Se-doped ZnS NPs was several times higher than pristine ZnS NPs. The major difference between pristine ZnS NPs and Se-doped ZnS NPs corresponded to the structural defects in the doped NPs. The fluorescence emission spectrum of Se-doped ZnS clearly revealed the existence of energy levels in the forbidden gap of the nanocrystals of Se-doped ZnS owing to structural defects. Therefore the photoexcited electrons in the conduction band could either hop to the surface of the nanocrystals, trapped in the defect sites or re-combine with holes. Under such circumstances, the reduction of molecular oxygen at the surface of the nanoparticles by photoexcited electrons is more likely, which would facilitate enhanced photocatalytic dye degradation mediated by superoxide radicals and singlet oxygen. Further studies would be required to understand in more detail the importance of structure–activity relationships in semiconductor nanoparticles.

4. Conclusion

The doping of ZnS NPs with selenium resulted in narrowing of the band gap to the visible region. The two salient features of this study were (a) achievement of more than 99% sunlight induced photocatalytic dye degradation in 3 h by 2.6 wt% Se-doped ZnS as compared to 38% dye degradation by pristine ZnS NPs; (b) a seven-fold increase in the rate of dye degradation by Se-doped ZnS over pristine ZnS. Such a drastic improvement in the photocatalytic behaviour was attributed to consequences of doping with selenium leading to structural defects, e.g., formation of interstitial energy levels in the doped ZnS lattice. As a result the photoexcited electron–hole transport was altered and favoured, which impacted on the generation of the nature of reactive oxygen species (ROS). Specific ROS scavenger assay by sodium azide and p-benzoquinone confirmed that superoxide radicals and singlet oxygen, which were primarily generated in the reaction medium by Se-doped ZnS NPs, triggered the degradation of trypan blue. The experimental results were well supported by theoretical band structure calculations which suggested transportation of electrons from the conduction band to interact with oxygen molecules to produce superoxide radicals and singlet oxygen species.

Acknowledgements

One of the authors (S. T.) is thankful to Ministry of Human Resource Development (MHRD), Govt of India for awarding a Senior Research Fellowship. The authors are thankful for the instrumental facilities of the Institute Instrumentation Centre, Department of Chemistry and Centre of Nanotechnology, IIT Roorkee.

References

1. L. Pereira and M. Alves, Environmental Protection Strategies for Sustainable Development, ed. A. Malik and E. Grohmann, Springer, New York, 2012, vol. 14, pp. 111–162.
2. M. Perez, F. Torrades, X. Domenech and J. Peral, Water Res., 2002, 36, 2703–2710.
3. S. S. Boxi and S. Paria, RSC Adv., 2015, 5, 37657–37668.
4. P. V. R. K. Ramacharyulu, D. B. Nimbalkar, J. P. Kumar, G. K. Prasad and S.-C. Ke, RSC Adv., 2015, 5, 37096–37101.
5. M. M. Khin, A. S. Nair, V. J. Babu, R. Murugana and S. Ramakrishna, Energy Environ. Sci., 2012, 5, 8075–8109.
6. X. Hu, G. Li and J. C. Yu, Langmuir, 2010, 26(5), 3031–3039.
7. R. K. Dutta, B. P. Nenavathu, M. K. Gangishetty and A. V. R. Reddy, J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng., 2013, 48, 871–878.
8. S. Guo, G. Zhang and J. Wang, J. Colloid Interface Sci., 2014, 433, 1–8.
9. G. Pliego, J. A. Zazo, J. A. Casas and J. J. Rodriguez, J. Hazard. Mater., 2013, 252–253, 180–185.
10. J. Paul, K. P. Rawat, K. S. S. Sarma and S. Sabhrawal, Appl. Radiat. Isot., 2011, 69, 982–987.
11. T. Tachikawa and T. Majima, Langmuir, 2009, 25(14), 7791–7802.
12. R. K. Dutta, B. P. Nenavathu, M. K. Gangishetty and A. V. R. Reddy, Colloids Surf., B, 2012, 94, 143–150.
13. H. Jia, W. He, W. G. Wamer, X. Han, B. Zhang, S. Zhang, Z. Zheng, Y. Xiang and J.-J. Yin, J. Phys. Chem. C, 2014, 118, 21447–21456.
14. T. Tachikawa, M. Fujitsuka and T. Majima, J. Phys. Chem. C, 2007, 111, 5259–5275.
15. A. Lipovsky, Z. Tzitrinovich, H. Friedmann, G. Applerot, A. Gedanken and R. Lubart, J. Phys. Chem. C, 2009, 113(36), 15997–16001.
16. Y. Murakami, E. Kenji, A. Y. Nosaka and Y. Nosaka, J. Phys. Chem. B, 2006, 110, 16808–16811.
17. S. G. Kumar and K. S. R. K. Rao, RSC Adv., 2015, 5, 3306–3351.
18. S. H. Shen, T. Chan, W. Yeong, J. C. Juan and C. Y. Teh, J. Chem. Technol. Biotechnol., 2011, 86, 1130–1158.
19. A. Sharma and R. K. Dutta, RSC Adv., 2015, 5, 43815–43823.
20. M. R. Hoffmann, S. T. Martin, W. Choi and D. W. Bahnemann, Chem. Rev., 1995, 95, 69–96.
21. D. Das and R. K. Dutta, J. Colloid Interface Sci., 2015, 457, 339–344.
22. Z. Xiong, L. L. Zhang, J. Ma and X. S. Zhao, Chem. Commun., 2010, 46, 6099–6101.
23. M. Sharma, T. Jain, S. Singh and O. P. Pandey, Sol. Energy, 2012, 86, 626–633.
24. R. K. Dutta, U. C. Bind, J. B. M. Krishna, A. K. Sinha and G. S. Taki, J. Radioanal. Nucl. Chem., 2014, 302, 819–824.
25. S. Joicy, R. Saravanan, D. Prabhu, N. Ponpandian and P. Thangadurai, RSC Adv., 2014, 4, 44592–44599.
26. A. Kudo and M. Sekizawa, Chem. Commun., 2000, 1371–1372.
27. K. Wang, Y. Shi, H. Zhang, Y. Xing, Q. Dong and T. Ma, Phys. Chem. Chem. Phys., 2014, 16, 23316–23319.
28. N. T. Hahn, A. J. E. Rettie, S. K. Beal, R. R. Fullon and C. B. Mullins, J. Phys. Chem. C, 2012, 116, 24878–24886.
29. J. Qian, K. Jiang, J. Huang, Q. Liu, L. Yang and Y. Song, Angew. Chem., 2012, 124, 10497–10500.
30. B. P. Nenavathu, A. V. R. K. Rao, A. Goyal, A. Kapoor and R. K. Dutta, Appl. Catal., A, 2013, 459, 106–113.
31. E. D. Kerver, I. M. C. Vogels, K. S. Bosch, H. V. S. Rova, R. J. M. Vandenmunckhof and W. Frederiks, Histochem. J., 1997, 29, 229–237.
32. J. R. Harbour and S. L. Issler, J. Am. Chem. Soc., 1982, 104(3), 903–905.
33. M. Sahni and B. R. Locke, Ind. Eng. Chem. Res., 2006, 45, 5819–5825.
34. R. Dong, Q. Ye, L. Kuang, X. Lu, Y. Zhang, X. Zhang, G. Tan, Y. Wen and F. Wang, ACS Appl. Mater. Interfaces, 2013, 5, 9508–9516.
35. M. Kurian and C. Kunjachan, International Nano Letters, 2014, 4, 73–80.
36. R. J. Wiglusz, A. Bednarkiewicz and W. Strek, Inorg. Chem., 2012, 51, 1180–1186.
37. Y. P. Zhu, J. Li, T. Y. Ma, Y. P. Liu, G. Du and Z. Y. Yuan, J. Mater. Chem. A, 2014, 2, 1093–1101.
38. A. Beleanu, M. Mondeshki, Q. Juan, F. Casper, C. Felser and F. Porcher, J. Phys. D: Appl. Phys., 2011, 44, 475302.
39. A. P. Alivisatos, Science, 1996, 271(5251), 933–937.
40. Z. Fang, S. Weng, X. Ye, W. Feng, Z. Zheng, M. Lu, S. Lin, X. Fu and P. Liu, ACS Appl. Mater. Interfaces, 2015, 7, 13915–13924.
41. D. Mocatta, G. Cohen, J. Schattner, O. Millo, E. Rabani and U. Banin, Science, 2011, 332, 77–81.
42. W. Zhu, X. Qiu, V. Iancu, X. Q. Chen, H. Pan, W. Wang, N. M. Dimitrijevic, T. Rajh, H. M. Meyer III, M. P. Paranthaman, G. M. Stocks, H. H. Weitering, B. Gu, G. Eres and Z. Zhang, Phys. Rev. Lett., 2009, 103, 226401–226404.
43. R. K. Dutta, B. P. Nenavathu and M. K. Gangishetty, J. Photochem. Photobiol., B, 2013, 126, 105–111.
44. R. K. Dutta, B. P. Nenavathu and S. Talukdar, Colloids Surf., B, 2014, 114, 218–224.
45. Y. S. Ho, J. C. Y. Ng and G. McKay, Sep. Purif. Rev., 2000, 29(2), 189–232.
46. J. Li, F. Sun, K. Gu, T. Wu, W. Zhai, W. Li and S. Huang, Appl. Catal., A, 2011, 406, 51–58.
47. D. Channei, B. Inceesungvorn, N. Wetchakun, S. Ukritnukun, A. Nattestad, J. Chen and S. Phanichphant, Nature, 2014, 4, 5757.
48. D. Beydoun, R. Amal, G. Low and S. McEvoy, J. Nanopart. Res., 1999, 1, 439–458.
49. R. R. Reddy, Y. N. Ahammed, K. R. Gopal and D. V. Raghuram, Opt. Mater., 1998, 10, 95–100.
50. Y. Xu and M. A. A. Schoonen, Am. Mineral., 2000, 85, 543–556.
51. R. G. Pearson, Inorg. Chem., 1988, 27, 734–740.
52. Y. Li, W. Zhang, J. Niu and Y. Chen, ACS Nano, 2012, 6, 5164–5173.

---