Antimony sulphoiodide (SbSI), a narrow band-gap non-oxide ternary semiconductor with efficient photocatalytic activity

Abstract

In the context of the harvesting of solar photons, one dimensional semiconductors are attractive as they provide uninterrupted transport pathways for charge carriers, along the covalently bonded atomic chain direction. Moreover, the one-dimensional growth leads to a lower proportion of dangling bonds at the surfaces due to preferential growth in a particular direction. We report here a non-metal oxide semiconductor, antimony sulphoiodide (SbSI) which displays high optical absorption and a low and tunable band gap for visible light photocatalytic applications. Highly crystalline 1-D micro-rods of SbSI which eventually self-assemble into 3-D “urchin”-shaped structures are synthesized by using a simple solution method. The morphology of the SbSI is studied in terms of the dangling bonds at the surface planes. The results conclusively show that the SbSI has a lower proportion of dangling bonds at the surface. The electronic structure of SbSI, studied using density functional theory, displays a large static dielectric constant due to the ns² cation (Sb³⁺) which enhances the separation of electron and hole pairs effectively. The combination of these two features makes SbSI a promising material for visible photocatalytic degradation of organic pollutants in water, in spite of an overall low surface area (≈2.6 m² g⁻¹).

---

1. Introduction

Photocatalysis has been widely considered as an important chemical process to degrade organic pollutants infesting natural water bodies as well as for green energy production of gases such as hydrogen and methane.^1–3 The efficiency of a photocatalyst is mainly determined by the extent of recombination of the photo-generated charge carriers i.e. the electrons and holes.⁴ Thus, chemical design of efficient photocatalysts should focus on strategies resulting in fast and low rates of charge separation and recombination, respectively. Various strategies have been employed to enhance the charge separation at the photocatalyst surfaces. It has been proposed that reducing the surface states and setting up an internal electric field inside the photocatalyst may enhance the separation of the electron–hole pairs.^5–8 In the case of ferroelectric materials internal electric field (polarization) exists as a result of the polar non-centro symmetry, leading to separation of electrons and holes at the photocatalyst surface and enhancement in the photocatalytic activity.^9–11 The conventional oxide-based perovskite ferroelectrics such as strontium titanate (SrTiO₃), barium titanate (BaTiO₃) are inappropriate for photocatalysis as they are wide band-gap (E_g ≥ 3.2 eV) materials.¹² Recently, organometal halide perovskites viz. MAPbX₃ (MA – methyl ammonium iodide, X – halide) have been demonstrated as promising visible light absorbers in the realm of photovoltaics. However, MAPbX₃ are generally not stable on exposure to air or moisture and hence cannot be used as a photocatalysis for degradation of pollutants. There are only a few examples of non-oxide based inorganic ferroelectrics. GeTe, ZnCdTe, SbSI are some of the notable examples.^13–15 Among these, ferroelectric properties of the antimony sulphoiodide (SbSI) chalcohalide (T_Curie ≈ 295 K; ε = 50 000) a ternary V–VI–VII semiconductor has been extensively studied in the early 1960's.^16–18 An interesting physical parameter of the V–VI–VII group ferroelectrics (including SbSI), which until recently has not received due importance and is critical from the point of view of solar photon harvesting, is the low electronic energy band gap (E_g). The SbSI E_g equals 1.8 eV, which is lower compared to many ferroelectric oxides PbTiO₃, BaTiO₃.^19,20 The E_g of SbSI can be further tuned via substitution of S and I with other group VI and VII elements. Generally, the one dimensional growth of SbSI takes place by stacking of the ribbon-like structure along the atomic chain direction.²¹ This results in minimum number of dangling bonds at the surface and should also lead to minimization of recombination centers for charge carriers.²² Moreover the Sb³⁺ cations in the SbSI, having outermost ns² electron configuration, is expected to display large static dielectric constant by setting strong lattice polarization which will aid in the effective enhancement of charge separation.²³ Combination of favorable (one-dimensional) structure, large static dielectric constant and exceptionally high stability in aqueous medium, it is strongly felt that SbSI and in general the V–VI–VII materials should make highly efficient and cost-effective visible light photocatalysts.

We report here a simple and novel solution-based synthesis of self-assembled micron-sized sea “urchin” shaped SbSI with a high density of 1D micron-sized rods, which have a lower proportion of dangling bonds at the particles surfaces. The micron-sized sea “urchin” shaped SbSI display efficient visible light photodegradation of organic dyes in aqueous solutions with high degree of reproducibility. The high photocatalytic activity of SbSI is attributed to lower dangling bonds on the surface and cross band hybridization between p states of Sb³⁺ and I⁻ ions.

2. Experimental section

2.1 Solution synthesis of SbSI

SbSI is synthesized using a modified solution synthesis process, based on as reported in ref. 24. Briefly, 1 mmol of SbCl₃, 1 mmol of KI and 1 mmol of thioacetamide are taken in a three-neck round bottom flask along with 30 mL of glacial acetic acid. The solution was heated to 50 °C for a sufficient period of time until a clear yellow colored solution is formed. After this, the temperature of the solution is raised to 110 °C and the solution is refluxed with constant stirring for 2 h. A wine-red compound is formed which settles at the bottom of the round bottom flask. The reaction mixture is the cooled down to room temperature and acetic acid and unreacted chemicals are removed by centrifugation. Finally, the wine-red sediment is washed with absolute ethanol (5 times) and then dried at 70 °C under vacuum for 12 h.

2.2 Structural and electronic property characterization

The crystallographic phase analysis of SbSI sample is performed using powder X-ray diffraction (PANanyltical Empyrean X-ray diffractometer; Cu-K_α radiation, 1.5418 Å, nickel filter, step width = 0.026 Å, scan rate = 0.33 Å s⁻¹). The morphology of the sample is studied using scanning electron microscopy (Ultra55 FESEM Karl Zeiss). Chemical composition and elemental mapping of the sample is performed using with EDS facility of the FESEM. High resolution transmission electron microscopy (HRTEM) and selected area electron diffraction (SAED) are done using a JEOL-200 kV FETEM and the micrographic images are analysed using a Digital micrograph software. Diffuse reflectance spectra (DRS) are recorded using solid-state UV-visible spectrophotometer (PerkinElmer, Lambda 35). Using Micromeritics surface area analyzer model ASAP-2020, the nitrogen adsorption and desorption were recorded and the specific surface area was calculated by the Brunauer–Emmett–Teller (BET) method. The X-ray photoelectron spectroscopy (XPS) and ultraviolet valence photoelectron spectroscopy data are recorded by AXIS-Ultra using monochromatic Al-K_α radiation (225 W, 15 mA, 15 kV).

2.3 Theoretical calculation

The calculation of band structure and density of states (DOS) are done using CASTEP package in Materials Studio (Version 6.0). All calculations are carried out by using generalized gradient approximation of Perdew, Burke, and Ernzerhof (GGA-PBE) functionals. Monkhorst–Pack K-points (3 2 6) are used to set the Brillouin zone. A cut-off of 500 eV is used for both band structure and DOS calculations. Morphology of SbSI is simulated by the CSD-materials package in Mercury software.

2.4 Photocatalytic activity

The photocatalytic performance of the as-synthesized SbSI is probed via the room temperature photodegradation of methyl orange under visible light irradiation. In all experiments, the photocatalyst concentration is typically maintained at 1 g L⁻¹. Following the addition of photocatalyst to a known concentration of MO solution, the mixture is kept in the dark for 2 h under constant stirring. This is done to achieve equilibrium between the rates of adsorption and desorption of the dye on the surface of SbSI. Oriel Class3A solar simulator (1 sun, 1.5G AM) attached with a 400 nm cut-off filter is used to produce the visible light solar spectrum. At suitable chosen time intervals, aliquots are withdrawn from the solution and wrapped with aluminum foil to avoid further light exposure. The supernatant liquid is separated from the photocatalyst by centrifugation and degradation of MO is evaluated by recording the corresponding absorbance band of MO at 450 nm using UV-visible spectroscopy. Various reactive species of the MO photodegradation process is scavenged via the addition of suitable reagents. During the degradation kinetics, benzoquinone (BQ), ammonium oxalate (AO), sodium azide (AZ), isopropanol (IPA) are added to scavenge the reactive species superoxide, hole, singlet oxygen, and hydroxyl radicals respectively.

3. Results and discussion

The crystallographic structure and phase purity of the SbSI sample is analyzed using powder X-ray diffraction (PXRD) (Fig. 1a). PXRD pattern of the synthesized SbSI sample matches exactly with the simulated PXRD pattern for the space group Pnam (JCPDS no. 01-074-1195). No extra peaks are observed in the PXRD pattern which strongly confirms phase purity of the as-synthesized SbSI. The atomic positions and structural parameters of the as-synthesized SbSI sample can be obtained by using GSAS program. The Rietveld refinement of the SbSI XRD pattern is shown in Fig. S1† along with the calculated profile. The refined crystallographic parameters of SbSI are as shown in Table S1† (taken from the least square refinement values). Analysis of the crystal structure reveals a ribbon-like stacking along the c-axis [001] with a double chain of [(SbSI)_∞]₂ linked together by strong covalent bonding between Sb–S (Fig. 1b and c). In the ribbon stack framework, each chain is mainly formed from the distorted edge sharing pseudo-octahedral of two Sb–I bonds and three Sb–S bonds along with the lone pair electron of Sb (Fig. 1b upper). In a [100] and b [010] directions the ribbons are joined together by a weak van der Waals force.

Fig. 1 (a) PXRD pattern of micron-sized “urchin” shaped SbSI. The pattern in red shows the reference pattern (JCPDS no. 01-074-1195). Schematic diagram of distorted edge sharing pseudo-octahedral single (b-upper), ribbon arrangement of SbSI viewing along c axis (b-lower) and viewing along b axis (c) [red sphere – Sb, black sphere – I, yellow sphere – sulfur].

Fig. 2a shows the absorption spectrum versus wavelength of SbSI (along with the solar radiance spectrum) as obtained from diffused reflectance spectroscopy. SbSI absorbs over a wide range in the visible region (≈400–620 nm; the absorption edge ≈650 nm) strongly suggesting it as a potential absorber material for solar photon harvesting applications. Optical band gap of the SbSI is evaluated from the diffused reflectance spectrum with the help of Kubelka–Munk postulation.²⁵ From the Tauc plots (cf. Fig. S2a and b†) the sample reveals both an indirect (=1.84 eV) and direct (=1.94 eV) band gaps by using the following equation: αhν = A(hν − E_g)^1/n, where parameters α, E_g, hν, are the absorption coefficient, band gap, photonic energy respectively and A is a constant. Evaluation of indirect and direct band gaps of SbSI is achieved via substitution of the value of exponent n with 1/2 and 2 respectively in the above equation. From the above calculations, we conclude that the SbSI is an indirect band gap semiconductor with the optical band gap of 1.84 eV. The difference between indirect and direct band gap is 0.1 eV, and is comparable with the reported value.²⁶ Indirect band gap materials are not generally preferred for solar photon harvesting applications however, the V–VI–VII group semiconductors are envisaged to be potentially useful for photovoltaic and photocatalysis applications as it displays a high absorption co-efficient in visible range region.²⁷

Fig. 2 (a) Solar spectrum and UV-visible absorbance spectrum of SbSI, (b) band structure of SbSI calculated by using GGA-PBE functionals. (c) Density of states for SbSI. s-States (black), p-states (red), total (green). Note that the band gap is underestimated in GGS-PBE functionals calculation.

The band structure of SbSI has been studied previously.^10,20 We study here the band structure of SbSI based on the refined crystallographic parameters using PBE functionals (Fig. 2b) in the range of −2.5 eV to 3.5 eV. Due to PBE functionals, which are used for the calculation, the calculated band gap in SbSI is indirect (E_g ≈ 1.53 eV) and is less than the experimental value (E_g ≈ 1.84 eV). The valence band maximum (VBM) and conduction band minimum (CBM) show dispersion in atomic chain direction (c-axis, G–Z direction in Fig. 2b) because of the strong covalent bonding in Sb–S and Sb–I. Partial and total density of states of SbSI are shown in the Fig. 2c.

The top of the conduction bands is ultimately dominated by the antimony 3p states and the valence bands are dominated by the p states of the anions, especially by iodine 5p rather than by the sulfur 2p states. The distance between the inter chain along b-axis is 4.0417 Å and is smaller than that along a-axis (4.5493 Å), leading to a strong coupling via large I–I ions (Fig. 1b). This interchain coupling between iodine ions along the b-axis, significantly increases the p orbital contribution in the valence band. From Fig. 2c, we observe that a significant cross band gap hybridization between Sb 6p and I 5p in both VBM and CBM leads to a mixed ionic–covalent character. This hybridization in semiconductor is known to generate a strong lattice polarization which in turn generates a large static dielectric constant. The large static dielectric constant will increase the carrier lifetime and diffusion length in the semiconductor by screening the charges (defects and impurities) effectively.²⁸ Same cross band gap cation–anion p orbital hybridization of VBM and CBM has also been observed in the systems with ns² cation such as BiSI and CH₃NH₃PbI₃, TlBr.^23,29,30 Though the SbSI shows a paraelectric (from Rietveld analysis) phase above room temperature, it is expected that the static dielectric constant will be high due to cross band gap hybridization which will enhance the photo conversion by screening the charges effectively.³⁰

The morphology of SbSI is probed using both scanning electron microscopy (SEM) and high resolution transmission electron microscopy (HRTEM). SEM (Fig. 3a, b and f) reveals a spherical morphology with large concentration of rods on the surface, closely resembling a sea “urchin”. The particle growth mechanism involves various intermediate stages, starting from needles/rods (1D) and then passing through formation of bundles, dumbbells and finally leading to urchin-like structures (3D) (cf. Fig. S3†). Each of the urchin-like structures are ≈10–20 μm in diameter and consists of one dimensional (1-D) SbSI rods. The HRTEM image (Fig. 3h) from Fig. 3g clearly show lattice fringes with a d-spacing of 0.39 nm, which is in good agreement with the (210) plane obtained from PXRD (and JCPDS no. 01-074-1195). Thus, the electron microscopy and PXRD confirm that the evolved structure of SbSI is highly crystalline. The elemental composition and distribution in an individual SbSI urchin are evaluated using X-ray energy dispersive spectroscopy (EDS) facility attached with the FESEM. The elemental mapping images reveal a uniform distribution of the elements viz. Sb, S, I elements in the individual SbSI urchin (Fig. 3c–e). The specific area of “urchin” shaped SbSI, obtained from BET analysis of the N₂ adsorption–desorption isotherms (cf. Fig. S4†), is approximately 2.56 m² g⁻¹.

Fig. 3 (a) Low magnification FESEM image of “urchin” shaped SbSI. (b) High magnification FESEM image of a single unit. (c–e) Elemental mapping using FESEM (c) Sb (red), (d) sulphur (green), (e) iodine (white) in single unit of SbSI. (f and g) FESEM and TEM image of rod of SbSI. (h) HRTEM of single rod SbSI.

When the plane (210) satisfies the following condition: hu + kv + lw = A; if A = 0 implies (hkl) planes ∥ [uvw] direction, one can easily conclude that the plane (210) is growing along the c-axis [001] (Fig. 3h). According to the Bravais–Friedel–Donnay–Harker (BFDH) law, the preferential growth of crystal face is approximately proportional to the inversion of inter-planar spacing (dhkl).^31–33 In the case of SbSI, the value of dhkl (l = 0) is much larger than dhkl (l ≠ 0) (cf. Table S2†). Therefore, growth rate along the c-axis proceeds faster than along the a and b axes because of the strong bonding interaction between the Sb–S in c-axis than the weak van der Waals force along the a and b axes. As the result, the SbSI grows in one-dimension rod shape (Fig. 3f and g) with (100), (010) planes expected to be side surface of the rods. When the (100), and (010) planes are terminated in SbSI rod, there are no dangling bonds at their surface (grain boundaries) because no bond breaking occurs between Sb–S and Sb–I bonds (Fig. 4a). At the same time (001) plane shows dangling bonds at the terminated surface because of bond breaking between Sb–S and Sb–I. The atoms having dangling bonds are displayed in green color in the Fig. 4b and c. Based on the BFDH morphology prediction, the simulated morphology of SbSI is shown in the Fig. 4d. From the Fig. 4e, the plane along the c-axis such as (110), and (020) are side surface of the SbSI rods and don't have dangling bonds at the surface. The planes (111) and (011) which are non-parallel to c-axis have dangling bonds at their surface as shown in Fig. 4f. From the experimental X-ray diffraction pattern, it can be concluded that the SbSI grows in (121) plane because of high intensity when compared to any other planes. In the orthorhombic space group, the (121) plane is composed by high inter-planar spacing of planes with hkl, l = 0 such as (110), (020), (120), (200) which are free of dangling bonds at the surface. Each SbSI rods in urchin shape particle contains nearly 72% of surface area (grain boundaries), without any dangling bonds (cf. Table S2†). It is well known that the dangling bonds introduce defeats at the surface of the particle and these become centers for recombination of photogenerated electrons and holes in light absorbing materials.²² It will affect the photoelectron conversation efficiency of light absorbers in photovoltaics and photocatalysis. Herein, we have synthesized the SbSI with low dangling bond surface and at a lower temperature. This is expected to be a highly efficient photocatalyst.

Fig. 4 (a–c) View of stacked [(SbSI)_∞]₂ ribbons along c axis (a), b axis (b) and a axis (c). Red dotted lines are representation of non-dangling bond planes. (d) Simulated morphology of SbSI by BFDH method. Planes ∥ c axis (red surface). Planes non parallel to c axis (green). Viewing the simulated morphology of SbSI along b axis (e) and c axis (f) [red sphere – Sb, black sphere – I, yellow sphere – sulfur, green sphere – dangling bond atoms].

The surface elemental composition and oxidation states of Sb, S, I in SbSI are verified by X-ray photoelectron spectroscopy (XPS). The overall survey spectrum (range: 0 to 1400 eV) of SbSI crystals is shown in (cf. Fig. S5†) and strongly confirms the existence of Sb, S, I elements. The high resolution XPS spectrum of the individual elements is fitted with Lorentzians to estimate the binding energy (Fig. 5a–c). The doublet peaks of Sb 3d_5/2 and Sb 3d_3/2 are observed at 528.5 eV and 537.7 eV respectively in the core spectrum of Sb (Fig. 5a). The separation between the peaks is 1.2 eV which confirms that the Sb is in the +3 oxidation state in SbSI and also matches with published reports.^34,35 Fig. 5b shows the core spectrum of I with peaks at 617.8 eV and 629.2 eV corresponding respectively to 3d_5/2 and 3d_3/2 of I. Fig. 5c shows the XPS spectrum of S with the doublet 2p peaks at 160.6 eV (2p_3/2) and 161.8 eV (2p_1/2). A difference of 1.2 eV between the two peaks is attributed to the coexistence of sulphide ions (S²⁻) in the SbSI samples. Thus, XPS confirms that the synthesized samples are phase pure.

Fig. 5 High resolution spectra of the individual elements corresponding to the Sb-3d (a), I-3d (b), S-2p (c) respectively. (d) UPS (He I: 21.21 eV) of SbSI. The red and blue circles in the UP spectra in (d) are enlarged and are shown below. (e) Secondary electron spectroscopy from UPS. (f) Valence band edges from UP spectra. The solid red lines in figures (e) and (f) are the fits to the experimental and denotes the secondary electron onset and valence band maximum (VBM) respectively.

From the ultraviolet photoelectron spectroscopy (UPS), relation between E_F, E_Vacuum, E_VBM (VBM: valence band maximum) edges are evaluated (Fig. 5d). The Fermi level of semiconductors and metals are evaluated by subtracting the secondary electron spectrum onset from the ultraviolet photon energy of 21.21 eV.^36,37 Based on this definition, the Fermi level of SbSI is obtained to be −4.61 eV with respect to the value of secondary electron onset (16.60 eV, Fig. 5e). Intersection of the best fits to the increasing and constant intensity of the plot in Fig. 5f provides an estimate of the difference between E_VBM and E_F. This is estimated to be 0.8 eV. The valence band maximum is located at around −5.41 eV (E = 0.91 eV in NHE) with respect to vacuum level. So, the conduction band using a band gap value of 1.84 eV, lies at around −3.57 eV (E = −0.93 eV in NHE) with respect to the vacuum level. The SbSI can show both n-type and p-type characteristics depending on the adopted synthesis method.³⁸ Here, the Fermi level of the as-synthesized SbSI lies 0.12 eV below the mid band gap position (−4.49 eV). Thus, the solution based synthesis protocol employed in this work produces an SbSI intrinsic semiconductor. The conduction band position of SbSI (−3.57 eV w.r.t. vacuum) is above the conduction band edge of most of the well-known metal oxides such as TiO₂ (–4.20 eV), ZnO (−4.20 eV) and SnO₂ (−4.5 eV) commonly used in the photovoltaics and photocatalysis.^39,40 Therefore, the SbSI has a very high potential in solar photon harvesting applications and coupled with its high stability which will be very beneficial for photocatalytic applications.

The photocatalytic performance of SbSI is investigated via photodegradation experiments of methyl orange (MO), an anionic dye, under simulated solar irradiation. Fig. 6a shows the concentration dependent photodegradation profiles of MO under simulated solar light. As a control, the photolysis of MO is also carried out under solar irradiance in the absence of the photocatalyst. Photodegradation of MO in the absence of photocatalyst is observed to be negligible. From this observation, it is clear that the MO dye itself plays no role in the photocatalytic process. Photocatalytic degradation of methylene blue (cationic), rhodamine B (zwitterionic) dyes using SbSI are also performed (cf. Fig. S6†). The photocatalytic activity methylene blue is very poor (only 18% at 20 min) compared to MO.

Fig. 6 (a) Fitted photodegradation profiles of MO at various concentrations (under simulated solar irradiance). (b) Photodegradation profiles of MO in the presence of scavenging agents such as benzoquinone (BQ), sodium azide (SA), ammonium oxalate (AO) and isopropanol (IPA). Figures (c) and (d) ln(C_o/C) versus time of MO degradation at various concentration at 20 min in the absence and presence of scavenging agents.

The rhodamine dye showed a slightly better performance (61% at 20 min) however, no further degradation is observed when monitored for longer times. So, we focus our attention only on MO degradation. The photodegradation efficiency (=C/C_o × 100; C: concentration at time = t; C_o: initial concentration of photocatalyst at 20 minute) is 97%, 84%, 74%, 61% for 20 ppm, 30 ppm, 40 ppm, 50 ppm respectively. Fig. 6c and d shows a linear variation of ln(C_o/C) with time which suggests a pseudo-first order kinetics. From Fig. 6c, the rate constants of MO for 20, 30, 40, 50 ppm is obtained as 0.19, 0.09, 0.06, and 0.05 min⁻¹ respectively. The decline in the degradation rate and efficiency with increase in the dye concentration is attributed to the shielding of the photocatalyst by the MO dye in solution. At high MO concentrations, the MO shields the SbSI micro-particles from the irradiating light thus, limiting the quantum of light absorbed by the SbSI. As a consequence, the amount of reactive species generated by the photocatalyst are limited thus, leading to decline in the photocatalytic performance.⁴¹ From the experimental observation and theoretical studies, we conclude that the lower dangling bond on the surface of the 1-D SbSI rods in the 3-D urchin shapes decreases the electron–hole recombination which might be the reason for high photocatalytic activity of SbSI on the organic pollutants. The large static dielectric generation in SbSI also enhances the photocatalytic activity by effectively screening the charge carriers.

In order to enhance the activity of photocatalyst, it is very much essential to figure out the mechanism of organic pollutant degradation. The extent of organic pollutant degradation depends completely on various reactive species such as electron, hole, superoxide radical (˙O₂⁻), hydroxyl radical (˙OH⁻), which are involved in the photocatalytic process.^42,43 The photodegradation of MO by SbSI are performed also in the presence of following scavenging agents: benzoquinone (superoxide radical), ammonium oxalate (hole), isopropanol (hydroxyl radicals), sodium azide (singlet oxygen) (Fig. 6b). When benzoquinone is introduced in the reaction beaker, efficiency reduced from 97% (no scavengers) to 17%. Similarly, k decreases from 0.19 min⁻¹ (no scavengers) to 0.01 min⁻¹ in the presence of BQ (Fig. 6d). This is due to the quenching of ˙O₂⁻ generated by SbSI. This reveals that degradation of MO may be largely due to ˙O₂⁻. The energy of the photo generated electrons in the conduction band (E = −0.93 eV in NHE) of SbSI is more negative than the redox potential of O₂/˙O₂⁻ (E = −0.33 eV versus NHE) and it is highly feasible to produce ˙O₂⁻ from singlet oxygen which are absorbed by the surface of the photocatalyst.⁴⁴ Addition of isopropanol, the scavenging agent for ˙OH⁻, leads to decrease in degradation efficiency from 97% (no scavengers) to 86% (k = 0.11 min⁻¹ for IPA). This suggests that the ˙OH⁻ do not play a significant role, especially in comparison to ˙O₂⁻. It is well known that the ˙OH⁻ can be generated from H₂O and OH⁻. The energy required to produce hydroxyl radicals from H₂O and OH⁻ are 2.72 eV and 2.38 eV respectively (w.r.t. NHE).⁴⁴ On the other hand, the energy of holes in the valence band of SbSI is E = 0.91 eV in NHE. Thus, the energy differences are significantly large to produce ˙OH⁻ from H₂O and OH⁻. There is another possibility to generate ˙OH⁻ from the ˙O₂⁻ in water via the photogenerated electrons.⁴⁵ As the amount of ˙OH⁻ is considerably less compared to the ˙O₂⁻, the role of ˙OH⁻ in the photodegradation of MO is limited. To reveal the role of the photogenerated hole in the photocatalytic degradation process of MO by SbSI, ammonium oxalate is added as a hole quencher to the MO solution. The photodegradation efficiency of MO decreased to 54% with the addition of the hole scavenging agent (k = 0.04 min⁻¹ for AO). This suggests that holes also play a vital role in the photocatalytic degradation of MO along with ˙O₂⁻ radicals. It has been widely proposed that photogenerated reactive species ˙O₂⁻ are generated from singlet oxygen dissolved in the solution or absorbed on the photocatalyst surface. Hence, to examine the role of singlet oxygen on photodegradation of MO, sodium azide is added to the MO solution. This would ensure absorption of singlet oxygen and sustenance of photodegradation under anoxic condition. The results show decline in the rate constant of photodegradation from 97% to 21% (k = 0.014 min⁻¹ for SA). From the above results, it can be concluded that ˙O₂⁻ and hole are primary reactive species in the MO degradation process. The ˙O₂⁻ is generated by injection of electron from the conduction band to the singlet oxygen. The photodegradation pathway of the MO dye by SbSI is depicted in Fig. 7a. Thus, the findings from the present investigations of the reaction mechanism of MO degradation using various scavengers agree will with that proposed in ref. 46 and 47.

Fig. 7 (a) Schematic diagram of the photodegradation of methyl orange using SbSI. Pathways (I and II) are more favorable than III. (b) MO photodegradation by “urchin”-shaped SbSI over five consecutive cycles. Values inside the parenthesis represents the photodegradation efficiency in each cycle.

Recyclability experiments are performed by repeated photodegradation of MO over several cycles as shown in Fig. 7b. No drastic changes in the photodegradation profiles are observed between the first 5 cycles. PXRD of the recovered samples following each cycle (cf. Fig. S7†) show no impurity peaks and the crystallinity of the SbSI is well retained. No notable changes in the chemical composition and morphology of the single SbSI microstructure are observed from FESEM and EDS mapping (cf. Fig. S8†).

4. Conclusions

In summary, we have demonstrated here a ternary semiconductor viz. SbSI as a prospective water stable photocatalyst. The highly efficient degradation activity on the organic pollutant are accounted by the formation of lower dangling bond on the surface as well as by the large static dielectric constant generated by the cross band hybridization between the ns² (Sb³⁺) cation and iodine ion. The attractive electronic properties of SbSI viz. the narrow E_g, position of the conduction band edge (which is above that of most metal oxides), low dangling bonds on the surface makes it also a potential candidate as light harvesters in photovoltaics. Stability under aqueous conditions additionally makes it a promising material for photocatalytic production of hydrogen and fine chemical synthesis.

Acknowledgements

Authors acknowledge the CSIR, New Delhi for financial support and CENSE, IISc. Bengaluru and SSCU, IISc. Bengaluru for infrastructural support.

Notes and references

1. S. Malato, P. Fernandez-Ibanez, M. I. Maldonado, J. Blanco and W. Gernjak, Catal. Today, 2009, 147, 1.
2. F. E. Osterloh, Chem. Soc. Rev., 2013, 42, 2294.
3. F. Sastre, A. V. Puga, L. Liu, A. Corma and H. García, J. Am. Chem. Soc., 2014, 136, 6798.
4. H. Wang, L. Zhang, Z. Chen, J. Hu, S. Li, Z. Wang, J. Liu and X. Wang, Chem. Soc. Rev., 2014, 43, 5234.
5. S. Bai, J. Jiang, Q. Zhang and Y. Xiong, Chem. Soc. Rev., 2015, 44, 2893.
6. Z. Zhang and J. T. Yates, Chem. Rev., 2012, 112, 5520.
7. C. R. Bowen, H. A. Kim, P. M. Weaver and S. Dunn, Energy Environ. Sci., 2014, 7, 25.
8. C. Acar, I. Dincer and C. Zamfirescu, Int. J. Energy Res., 2014, 38, 1903.
9. L. Li, P. A. Salvador and G. S. Rohrer, Nanoscale, 2014, 6, 24.
10. K. T. Butler, J. M. Frost and A. Walsh, Energy Environ. Sci., 2015, 8, 838.
11. M. A. Khan, M. A. Nadeem and H. Idriss, Surf. Sci. Rep., 2016, 71, 1.
12. P. Kanhere and Z. Chen, Molecules, 2014, 19, 19995.
13. R. E. Brandt, V. Stevanović, D. S. Ginley and T. Buonassisi, MRS Commun., 2015, 5, 265.
14. J. Varghese, R. W. Whatmore and J. D. Holmes, J. Mater. Chem. C, 2013, 1, 2618.
15. L. Liang, X. Kang, Y. Sang and H. Liu, Adv. Sci., 2016, 3, 1500358.
16. E. Fatuzzo, R. Nitsche, G. Harbeke, W. Ruppel, H. Roetschi and W. Merz, J. Phys. Rev., 1962, 127, 2036.
17. D. Berlincourt, R. Nitsche, W. J. Merz and H. Jaffe, Appl. Phys. Lett., 1964, 4, 61.
18. J. Varghese, C. O'Regan, N. Deepak, R. W. Whatmore and J. D. Holmes, Chem. Mater., 2012, 24, 3279.
19. S. Piskunov, E. Heifets, R. I. Eglitis and G. Borstel, Comput. Mater. Sci., 2004, 29, 165.
20. K. T. Butler, S. McKechnie, P. Azarhoosh, M. Schilfgaarde, D. O. Scanlon and A. Walsh, Appl. Phys. Lett., 2016, 108, 112103.
21. I. Cho, B. K. Min, S. W. Joo and Y. Sohn, Mater. Lett., 2012, 86, 132.
22. Y. Zhou, L. Wang, S. Chen, S. Qin, X. Liu, J. Chen, D. J. Xue, M. Luo, Y. Cao and Y. Cheng, Nat. Photonics, 2015, 9, 409.
23. H. Shi, W. Ming and M. H. Du, Phys. Rev. B: Condens. Matter Mater. Phys., 2016, 93, 104108.
24. G. Chen, W. Li, Y. Yu and Q. Yang, RSC Adv., 2015, 5, 21859.
25. P. Kubelka and F. Munk, Z. Tech. Phys., 1931, 12, 593.
26. S. A. Park, M. Y. Kim, J. Y. Lim, B. S. Park, J. D. Koh and W. T. Kim, Phys. Status Solidi B, 1995, 187, 253.
27. C. Tablero, J. Alloys Compd., 2016, 678, 18.
28. M. H. Du, J. Mater. Chem. A, 2014, 2, 9091.
29. M. H. Du and D. J. Singh, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 81, 144114.
30. D. Amoroso and S. Picozzi, Phys. Rev. B, 2016, 93, 214106.
31. A. Bravais, Etudes Crystallographiques, Academie des Sciences, Paris, 1913.
32. R. Docherty, G. Clydesdale, K. J. Roberts and P. Bennema, J. Phys. D: Appl. Phys., 1991, 24, 89–99.
33. J. D. H. Donnay and D. Harker, Am. Mineral., 1937, 22, 446–467.
34. M. Nowak, E. Talik, P. Szperlich and D. Stróż, Appl. Surf. Sci., 2009, 255, 7689.
35. M. Nowak, A. Nowrot, P. Szperlich, M. Jesionek, M. Kepińska, A. Starczewska, K. Mistewicz, D. Stróż, J. Szala, T. Rzychoń, E. Talik and R. Wrzalik, Sens. Actuators, A, 2014, 210, 119.
36. W. Chun, A. Ishikawa, H. Fujisawa, T. Takata, J. N. Kondo, M. Hara, M. Kawai, Y. Matsumoto and K. Domen, J. Phys. Chem. B, 2003, 107, 1798.
37. M. Ghaffari, H. Hunag, O. K. Tan and M. Shannon, CrystEngComm, 2012, 14, 7487.
38. M. Nowak, in Nanowire Science and Technology, ed. N. Lupu, InTech, Croatia, 2010, ch. 13.
39. E. Miller, Y. Zhao, C. Mercado, S. Saha, J. M. Luther, K. Zhu, V. Stevanovic, C. L. Perkins and J. van de Lagemaat, Phys. Chem. Chem. Phys., 2014, 16, 22122.
40. J. L. White, M. F. Baruch, J. E. Pander, Y. Hu, I. C. Fortmeyer, J. E. Park, T. Zhang, K. Liao, J. Gu, Y. Yan, T. W. Shaw, E. Abelev and A. B. Bocarsly, Chem. Rev., 2015, 115, 12888.
41. C. Wang, Q. Zhu, C. Gu, X. Luo, C. Yu and M. Wu, RSC Adv., 2016, 6, 85852.
42. Y. Guo, Y. Zhang, N. Tian and H. Huang, ACS Sustainable Chem. Eng., 2016, 4, 4003.
43. W. W. Lee, C.-S. Lu, C.-W. Chuang, Y.-J. Chen, J.-Y. Fu, C.-W. Siao and C.-C. Chen, RSC Adv., 2015, 5, 23450.
44. M. Yan, Y. Wu, Y. Yan, X. Yan, F. Zhu, Y. Hua and W. Shi, ACS Sustainable Chem. Eng., 2016, 4, 757.
45. L. Sun, X. Wu, M. Meng, X. Zhu and P. K. Chu, J. Phys. Chem. C, 2014, 118, 28063.
46. H. P. Zhao, Y. F. Zhang, G. F. Li, F. Tian, H. Tang and R. Chen, RSC Adv., 2016, 6, 7772.
47. W. Li, D. Li, Y. Lin, P. Wang, W. Chen, X. Fu and Y. Shao, J. Phys. Chem. C, 2012, 116, 3552.

---

Footnote

† Electronic supplementary information (ESI) available: Supplementary Fig. S1–S8 (Rietveld analysis, Tauc plots, BET, XPS, dye degradation, SEM images for growth mechanism, XRD and EDS imaging after photocatalysis process), Tables S1 and S2. See DOI: 10.1039/c6ra23750a

---