Partial Oxidation Strategy to Synthesize WS2/WO3 Heterostructure with Enhanced Adsorption Performance for Organic Dyes: Synthesis, Modelling, and Mechanism

In this work, a facile oxidation strategy was developed to prepare novel tungsten disulfide/tungsten trioxide (WS2/WO3) heterostructures for adsorbing organic dyes efficiently by combining the hydrophilic property of WO3 and the superior dye affinity of WS2. The structural and elemental properties of the synthesized hybrid materials were systematically investigated, and the results demonstrated the retained flower-like morphology of the primitive WS2 and the successful introduction of WO3. Furthermore, surface properties such as a superior hydrophilicity and negative-charged potential were also demonstrated by a water contact angle characterization combined with a Zeta potential analysis. The performance of the obtained WS2/WO3 hybrid materials for removing Rhodamine B (RhB) from wastewater was evaluated. The results showed that the maximum adsorption capacity of the newly synthesized material could reach 237.1 mg/g. Besides, the adsorption isotherms were also simulated by a statistical physics monolayer model, which revealed the non-horizontal orientation of adsorbates and endothermic physical interaction. Finally, the adsorption mechanism and the recyclability revealed that the partial oxidation strategy could contribute to a higher adsorption capacity by modulating the surface properties and could be applied as a highly efficient strategy to design other transition metal dichalcogenides (TMDs) heterostructures for removing organic dyes from wastewater.


Introduction
In the past several decades, it was estimated that millions of tons of dyes were produced annually, and amount of dyes were discarded into water as waste [1][2][3]. These organic dyes, which are toxic, non-biodegradable, and even carcinogenic, had posed serious threats to human health and marine organisms [4]. Consequently, it is urgent to develop low-cost and efficient methods to remove these organic dyes from wastewater. Various technologies have been developed and employed for treating wastewater, including membrane filtration, electrochemical decolorization, photocatalytic degradation, chemical oxidation/reduction, biodegradation, and adsorption [2,[5][6][7][8][9][10][11][12][13][14][15]. Among them, adsorption is considered a promising candidate to eliminate dyes from industrial wastewater due to the low capital (0.894 g) was dissolved in NMP (70 mL), and this suspension was stirred for 10 min to obtain a dark brown solution. Next, sublimed sulfur was added into the mixture accompanied with an ultrasonic treatment for 20 min until a homogenous solution was formed. Subsequently, the solution was transferred into a 100 mL PolyPhenyLene-lined stainless steel autoclave and was heated at 240 °C for 20 h. After the solvothermal process, the black precipitates were collected and dried under vacuum at 100 °C for 12 h. The black powder was WS2 and was thereafter annealed at 850 °C for 2 h under argon atmosphere to increase the crystallinity.
After the synthesis of WS2, the solids were treated in a furnace at 400 °C for n (n = 5, 30, and 60) min under oxygen flow (100 mL/min) to obtain a series of partially oxidized products which were denoted as WSO-n. Finally, pure WO3 was obtained with an extending duration of 4 h.

Material Characterizations
X-ray diffraction patterns of the obtained materials were recorded by an X-ray powder diffractometer (D/max 2500, Rigaku, Tokyo, Japan) with Cu Kα radiation (λ = 1.5418 Å). A thermal gravimetric analysis (TGA) was performed under oxygen flow on a NETZSCH TG 209F3 analyzer (Bavaria, Germany). Morphological information was recorded by scanning electron microscopy (Nanosem 430, FEI, Eindhoven, Netherlands) and transmission electron microscopy (Tecnai G 2 F20, FEI, Eindhoven, Netherlands). Elements contents and their chemical states were analyzed by X-ray photoelectron spectroscopy (ESCALAB 250XI, Al Kα, Thermo, MA, USA) which was calibrated by contaminant carbon at 284.8 eV. Zeta potentials were analyzed by Malvern Zetasizer Nano Series (Nano ZSP, Malvern, Worcestershire, UK) in the pH range of 10 to 2. The Fourier transform infrared (FTIR) spectra with the wavelength range of 4000-400 cm −1 were obtained using a Bruker Alpha spectrometer (Regensburg, Germany). Water contact angles were measured on a Dataphysics OCA contact angle meter.

Batch Adsorption Experiments
All the following adsorption experiments except the adsorption isotherms were conducted at 30 °C. The kinetic studies of different obtained materials toward the absorption of RhB were investigated via the following procedure: 20 mg of adsorbents were put into a 40 mL RhB solution (200 mg/L), after which the mixture was placed in a dark place and stirred vigorously. The mixture was sampled at predetermined time intervals, and then the solid absorbent was separated from the solution through centrifugation. The concentration of the remaining RhB dye on solution was measured via a UV-vis spectrophotometer (Hitachi U4100 spectrophotometer, Kyoto, Japan), at the maximum adsorption wavelength of RhB (554 nm). The adsorption amount (mg/g) of RhB was expressed as follows: , (1) where and (mg/L) represent the concentrations of remaining RhB at times 0 and (min), respectively.
(L) is the volume of RhB solution, and (g) is the mass of the adsorbent. The adsorption isotherms of RhB on WSO-5 were conducted at 30, 40, and 50 °C with a solution pH of 6. Specifically, the adsorption experiments were performed by putting 20 mg of the as-prepared materials into a 40 mL RhB solution with different concentrations (100-400 mg/L) After the synthesis of WS 2 , the solids were treated in a furnace at 400 • C for n (n = 5, 30, and 60) min under oxygen flow (100 mL/min) to obtain a series of partially oxidized products which were denoted as WSO-n. Finally, pure WO 3 was obtained with an extending duration of 4 h.

Material Characterizations
X-ray diffraction patterns of the obtained materials were recorded by an X-ray powder diffractometer (D/max 2500, Rigaku, Tokyo, Japan) with Cu Kα radiation (λ = 1.5418 Å). A thermal gravimetric analysis (TGA) was performed under oxygen flow on a NETZSCH TG 209F3 analyzer (Bavaria, Germany). Morphological information was recorded by scanning electron microscopy (Nanosem 430, FEI, Eindhoven, Netherlands) and transmission electron microscopy (Tecnai G 2 F20, FEI, Eindhoven, Netherlands). Elements contents and their chemical states were analyzed by X-ray photoelectron spectroscopy (ESCALAB 250XI, Al Kα, Thermo, MA, USA) which was calibrated by contaminant carbon at 284.8 eV. Zeta potentials were analyzed by Malvern Zetasizer Nano Series (Nano ZSP, Malvern, Worcestershire, UK) in the pH range of 10 to 2. The Fourier transform infrared (FTIR) spectra with the wavelength range of 4000-400 cm −1 were obtained using a Bruker Alpha spectrometer (Regensburg, Germany). Water contact angles were measured on a Dataphysics OCA contact angle meter.

Batch Adsorption Experiments
All the following adsorption experiments except the adsorption isotherms were conducted at 30 • C. The kinetic studies of different obtained materials toward the absorption of RhB were investigated via the following procedure: 20 mg of adsorbents were put into a 40 mL RhB solution (200 mg/L), after which the mixture was placed in a dark place and stirred vigorously. The mixture was sampled at predetermined time intervals, and then the solid absorbent was separated from the solution through centrifugation. The concentration of the remaining RhB dye on solution was measured via a UV-vis spectrophotometer (Hitachi U4100 spectrophotometer, Kyoto, Japan), at the maximum adsorption wavelength of RhB (554 nm). The adsorption amount Q t (mg/g) of RhB was expressed as follows: where C 0 and C t (mg/L) represent the concentrations of remaining RhB at times 0 and t (min), respectively. V (L) is the volume of RhB solution, and m (g) is the mass of the adsorbent. The adsorption isotherms of RhB on WSO-5 were conducted at 30, 40, and 50 • C with a solution pH of 6. Specifically, the adsorption experiments were performed by putting 20 mg of the as-prepared materials into a 40 mL RhB solution with different concentrations (100-400 mg/L) under a constant Nanomaterials 2020, 10, 278 4 of 18 stirring rate of 200 rpm. After adsorption for 5 h to reach the equilibrium condition, the solid phase was separated from the liquid phase by centrifugation, and the equilibrium concentration of RhB was measured. The relative adsorption capacity of RhB was calculated by the following formula: where Q e (mg/g) represents the equilibrium adsorption capacity of RhB, and C e (mg/L) is the concentration of RhB at equilibrium status. The effect of the pH was explored by adding 20 mg adsorbent into a 40 mL RhB solution (200 mg/L) at a series of pH levels (10 to 2). The required pH values were adjusted by adding a negligible volume of 0.1 M HNO 3 and NaOH solutions. Furthermore, the effect of the adsorbent dosage was also studied by conducting a 40 mL RhB solution (200 mg/L) with a varying mass of WSO-5 from 10 mg to 60 mg for 5 h. The concentration of the remaining RhB in the solution was measured with the aforementioned method.

Characterizations
The microstructures and morphologies of WS 2 , WSO-5, WSO-30, WSO-60, and WO 3 were characterized by SEM and TEM, and the results are shown in Figures 2 and 3. The SEM images in Figure 2a,b demonstrate that the synthesized WS 2 has a uniform 3D flower-like morphology, which consists of intercrossed nanoflakes with a thickness of several nanometers. According to the thermal-gravimetric analysis (TGA) result ( Figure S1), WS 2 starts to lose weight at about 360 • C. Thus, WS 2 microflowers were determined to be calcined at 400 • C for the partial oxidation of WS 2 to WO 3 . The obtained WS 2 /WO 3 heterostructures were also characterized by SEM (Figure 2c-e). It can be clearly seen that the well-defined flower-like structures crumble gradually with an increase in the calcination time. As displayed in Figure 2c,d, the flower-like microstructures in WSO-5 and WSO-30 are retained, while there is almost no flake morphology in WSO-60. Finally, the three-dimensional structures were entirely destroyed into aggregating particles in WO 3 after the thoroughly oxidation of WS 2 into WO 3 (Figure 2f). under a constant stirring rate of 200 rpm. After adsorption for 5 h to reach the equilibrium condition, the solid phase was separated from the liquid phase by centrifugation, and the equilibrium concentration of RhB was measured. The relative adsorption capacity of RhB was calculated by the following formula: , where (mg/g) represents the equilibrium adsorption capacity of RhB, and (mg/L) is the concentration of RhB at equilibrium status.
The effect of the pH was explored by adding 20 mg adsorbent into a 40 mL RhB solution (200 mg/L) at a series of pH levels (10 to 2). The required pH values were adjusted by adding a negligible volume of 0.1 M HNO3 and NaOH solutions. Furthermore, the effect of the adsorbent dosage was also studied by conducting a 40 mL RhB solution (200 mg/L) with a varying mass of WSO-5 from 10 mg to 60 mg for 5 h. The concentration of the remaining RhB in the solution was measured with the aforementioned method.

Characterizations
The microstructures and morphologies of WS2, WSO-5, WSO-30, WSO-60, and WO3 were characterized by SEM and TEM, and the results are shown in Figures 2 and 3. The SEM images in Figure 2a,b demonstrate that the synthesized WS2 has a uniform 3D flower-like morphology, which consists of intercrossed nanoflakes with a thickness of several nanometers. According to the thermal-gravimetric analysis (TGA) result ( Figure S1), WS2 starts to lose weight at about 360 °C. Thus, WS2 microflowers were determined to be calcined at 400 °C for the partial oxidation of WS2 to WO3. The obtained WS2/WO3 heterostructures were also characterized by SEM (Figure 2c-e). It can be clearly seen that the well-defined flower-like structures crumble gradually with an increase in the calcination time. As displayed in Figure 2c,d, the flower-like microstructures in WSO-5 and WSO-30 are retained, while there is almost no flake morphology in WSO-60. Finally, the three-dimensional structures were entirely destroyed into aggregating particles in WO3 after the thoroughly oxidation of WS2 into WO3 (Figure 2f).   plane of WO3. The elemental mapping results of WSO-5 further confirm the uniform distribution of W, S, and O elements on the nanoflakes, suggesting that the WS2 and WO3 species are interconnected on the nanoscale rather than isolated from each other (Figure 3b inset). These polycrystalline structures are also confirmed by a Fast Fourier Transform (FFT) pattern in Figure 3b. Finally, the lattice fringes of WS2 disappear, and only the lattice fringes of WO3 remain (Figure 3c) when the composites are entirely transformed into WO3 particles. The crystalline information is demonstrated by XRD patterns (Figure 4). From the pattern of pure WS2, several broad peaks at 14.5°, 32.7°, 33.6°, 39.5°, and 58.4° are in good agreement with the (002), (100), (101), (103), and (110) planes of the hexagonal phase of WS2 (JCPDS Card No.08-0237) [29]. With the increase of the calcining time, the intensity of the (0 0 2) peak decreases gradually from WSO-5 to WOS-60. Simultaneously, the peaks at about 24.0° and 33.6°, which are ascribed to the (0 2 0) and (2 2 0) peaks of WO3, emerge and get strengthened. This phenomenon also suggests the successful construction of WS2/WO3 heterostructures. Moreover, pure WO3 was also analyzed, and it was found that WO3 possesses characteristic peaks at 23.  The microscopic structures of the obtained composites were further investigated using high-resolution transmission electron microscopy (HRTEM), and the results are displayed in Figure 3. Initially, the interlayer spacing of WS 2 nanosheets displayed in the Figure 3a, measured to be 0.616 nm, can be ascribed to the (0 0 2) plane of WS 2 [28]. For the sample with a short partial oxidation time of 5 min, a new lattice fringe with d = 0.375 nm appears beside the interlayer spacing of 0.616 nm according to Figure 3b, ascribed to the characteristic interplane distance of the (020) crystallographic plane of WO 3 . The elemental mapping results of WSO-5 further confirm the uniform distribution of W, S, and O elements on the nanoflakes, suggesting that the WS 2 and WO 3 species are interconnected on the nanoscale rather than isolated from each other (Figure 3b inset). These polycrystalline structures are also confirmed by a Fast Fourier Transform (FFT) pattern in Figure 3b. Finally, the lattice fringes of WS 2 disappear, and only the lattice fringes of WO 3 remain (Figure 3c) when the composites are entirely transformed into WO 3 particles.
The crystalline information is demonstrated by XRD patterns (Figure 4). From the pattern of pure WS 2 , several broad peaks at 14.5 • , 32.7 • , 33.6 • , 39.5 • , and 58.4 • are in good agreement with the (002), (100), (101), (103), and (110) planes of the hexagonal phase of WS 2 (JCPDS Card No.08-0237) [29]. With the increase of the calcining time, the intensity of the (0 0 2) peak decreases gradually from WSO-5 to WOS-60. Simultaneously, the peaks at about 24.0 • and 33.6 • , which are ascribed to the (0 2 0) and (2 2 0) peaks of WO 3 , emerge and get strengthened. This phenomenon also suggests the successful construction of WS 2 /WO 3 heterostructures. Moreover, pure WO 3 was also analyzed, and it was found that WO 3   For the purpose of examining the chemical compositions and electronic states of elements of WSO-n, X-ray photoelectron spectroscopy (XPS) was further utilized in this work. It can be seen from the XPS full-range spectra ( Figure 5a) that these obtained composites exhibit characteristic peaks of C, W, O, and S elements. The characteristic peak of C 1s situated at ~284.8 eV, ascribed to the contaminant carbon, was utilized to calibrate the XPS spectra. After the calibration, the corresponding narrow-range W 4f, S 2p, and O 1s spectra were discussed in detail. Figure 5b shows that the O 1s spectra of the WS2/WO3 heterostructures can be well fitted into two peaks at 530.94 eV and 531.94 eV, attributed to the O in WO3 and chemisorbed oxygen-containing materials, respectively [31,32]. As shown in Figure 5c, the two peaks of W 4f5/2 and W 4f7/2 of WS2 are situated at about 34.60 eV and 33.24 eV. After partial oxidation, new peaks at 37.99 eV and 35.96 eV appeared, which could be assigned to W 4f5/2 and W 4f7/2 in WO3 [33]. The characteristic peaks of both WS2 and WO3 can be clearly observed in WSO-n, which further confirms the successful synthesis of the aimed heterostructures. Ultimately, the S 2p spectra of these composites were characterized ( Figure 5d). There are two binding energy of S 2p located at 162.30 eV and 163.70 eV, which can be assigned to the presence of S 2− in WS2. The special peak present at 169.10 eV could be attributed to S 6+ due to the oxidation of divalent sulfide ions [34]. Significantly, it only existed in WS2/WO3 hybrids. The characteristic peaks of S decreased gradually with the oxidation process due to the enhanced transformation from WS2 to WO3. Overall, all the XPS spectra results mentioned above demonstrated a successful construction of WS2/WO3 heterostructures. For the purpose of examining the chemical compositions and electronic states of elements of WSO-n, X-ray photoelectron spectroscopy (XPS) was further utilized in this work. It can be seen from the XPS full-range spectra (Figure 5a) that these obtained composites exhibit characteristic peaks of C, W, O, and S elements. The characteristic peak of C 1s situated at~284.8 eV, ascribed to the contaminant carbon, was utilized to calibrate the XPS spectra. After the calibration, the corresponding narrow-range W 4f, S 2p, and O 1s spectra were discussed in detail. Figure 5b shows that the O 1s spectra of the WS 2 /WO 3 heterostructures can be well fitted into two peaks at 530.94 eV and 531.94 eV, attributed to the O in WO 3 and chemisorbed oxygen-containing materials, respectively [31,32]. As shown in Figure 5c, the two peaks of W 4f5/2 and W 4f7/2 of WS 2 are situated at about 34.60 eV and 33.24 eV. After partial oxidation, new peaks at 37.99 eV and 35.96 eV appeared, which could be assigned to W 4f5/2 and W 4f7/2 in WO 3 [33]. The characteristic peaks of both WS 2 and WO 3 can be clearly observed in WSO-n, which further confirms the successful synthesis of the aimed heterostructures. Ultimately, the S 2p spectra of these composites were characterized (Figure 5d). There are two binding energy of S 2p located at 162.30 eV and 163.70 eV, which can be assigned to the presence of S 2− in WS 2 . The special peak present at 169.10 eV could be attributed to S 6+ due to the oxidation of divalent sulfide ions [34]. Significantly, it only existed in WS 2 /WO 3 hybrids. The characteristic peaks of S decreased gradually with the oxidation process due to the enhanced transformation from WS 2 to WO 3 . Overall, all the XPS spectra results mentioned above demonstrated a successful construction of WS 2 /WO 3 heterostructures. Nanomaterials 2020, 10, 278 7 of 18 A contact angle analysis was adopted to assess the obtained materials' wettability, which is traditionally determined by the substrates' morphologies and chemical compositions [35,36]. Initially, the contact angle of pure WS2 was measured to be about 102° (Figure 6a). After oxidization, a sharp decrement of the contact angle values was observed with all values lower than 20° ( Figure  6b-e), indicating the hydrophilicity of WS2/WO3 hybrids, and WO3 was dramatically promoted with the introduction of oxygen atoms. The increasing wettability might result from the formation of hydrogen bonds which are introduced by strong-polar oxygen atoms interacting with water molecules [37,38]. In addition, the high-magnification SEM images of WS2/WO3 hybrids (Figure 2ce) shows that these heterostructures have rougher surface morphologies with voids and many highly water-permeable channels, which can further enhance the hydrophilicity. A contact angle analysis was adopted to assess the obtained materials' wettability, which is traditionally determined by the substrates' morphologies and chemical compositions [35,36]. Initially, the contact angle of pure WS 2 was measured to be about 102 • (Figure 6a). After oxidization, a sharp decrement of the contact angle values was observed with all values lower than 20 • (Figure 6b-e), indicating the hydrophilicity of WS 2 /WO 3 hybrids, and WO 3 was dramatically promoted with the introduction of oxygen atoms. The increasing wettability might result from the formation of hydrogen bonds which are introduced by strong-polar oxygen atoms interacting with water molecules [37,38]. In addition, the high-magnification SEM images of WS 2 /WO 3 hybrids (Figure 2c-e) shows that these heterostructures have rougher surface morphologies with voids and many highly water-permeable channels, which can further enhance the hydrophilicity.

Adsorption Knetics
To explore the adsorption kinetics of these newly developed materials on RhB, the effect of the adsorption capacity over contact time was evaluated. As shown in Figure 7a, all the WS2/WO3 hybrid samples exhibited similar trends, with the sorption capacities increasing rapidly during the first 100 min, then gradually reaching equilibrium with a further increase of the adsorption time. Obviously, the pure WS2 showed a poor adsorption capacity of merely 16.4 mg/g, which can be ascribed to the hydrophobic surface property. After the introduction of oxygen atoms with a short duration time of 5 min, the adsorption capacity got a tremendous promotion (237.1 mg/g) due to the stronger hydrophilicity, as well as to the retention of the superior properties of WS2. However, with the further increment of the oxidation duration, the adsorption capacity gradually declined until reaching pure WO3, which exhibited a low adsorption performance of 30.1 mg/g.
The fitting results are displayed in Figure 7b-d, and the calculated kinetic parameters are summarized in Table S1. Compared with the pseudo-first-order model and Weber-Morris model, the pseudo-second-order model exhibits better correlation results for the experimental data, with a higher R 2 value (above 0.99).

Adsorption Knetics
To explore the adsorption kinetics of these newly developed materials on RhB, the effect of the adsorption capacity over contact time was evaluated. As shown in Figure 7a, all the WS 2 /WO 3 hybrid samples exhibited similar trends, with the sorption capacities increasing rapidly during the first 100 min, then gradually reaching equilibrium with a further increase of the adsorption time. Obviously, the pure WS 2 showed a poor adsorption capacity of merely 16.4 mg/g, which can be ascribed to the hydrophobic surface property. After the introduction of oxygen atoms with a short duration time of 5 min, the adsorption capacity got a tremendous promotion (237.1 mg/g) due to the stronger hydrophilicity, as well as to the retention of the superior properties of WS 2 . However, with the further increment of the oxidation duration, the adsorption capacity gradually declined until reaching pure WO 3 , which exhibited a low adsorption performance of 30.1 mg/g.
Three empirical kinetic models were adopted to conduct a further analysis toward the adsorption kinetics, namely, a pseudo-first-order model (Equation (3)), pseudo-second order model (Equation (4)), and Weber-Morris model (Equation (5)): where Q e (mg/g) and Q t (mg/g) represent the adsorption capacities of RhB onto samples at equilibrium and time t (min); c (mg/g) is a constant; k 1 (min −1 ), k 2 (g·mg −1 ·min −1 ), and k i (mg·g −1 ·min −1/2 ) are the rate constants of the pseudo-first order, pseudo-second order, and Weber-Morris models, respectively. The fitting results are displayed in Figure 7b-d, and the calculated kinetic parameters are summarized in Table S1. Compared with the pseudo-first-order model and Weber-Morris model, the pseudo-second-order model exhibits better correlation results for the experimental data, with a higher R 2 value (above 0.99).

Adsorption Isotherms
To elucidate the equilibrium states of RhB dye molecules on materials and in liquid surroundings, the adsorption isotherms were conducted [39]. In particular, WSO-5 was chosen as the model material due to its higher adsorption capacity. The adsorption isotherms of RhB on WSO-5 were carried out at gradient temperatures of 30, 40 and 50 °C. As shown in Figure 8, the adsorption isotherms exhibited conventional L-types with obvious platforms. The results showed that the increasing temperature in circumstance could facilitate the adsorption for WSO-5, suggesting an endothermic adsorption process. Furthermore, a straightforward statistical physics model was utilized to fit the adsorption isotherms of RhB. The mathematical equation of the model is presented as follows [40]: where represents the number of RhB molecules per site of adsorbent, (mg/g) is the density of the receptor sites, and ⁄ (mg/L) means the concentration at half saturation of the formed layer. According to the statistic physics model, the adsorption capacity at saturation is given by [41]: .
The expression of the adsorption energy , which contains the solubility (mg/L) of the adsorbates, can be expressed as follows [41,42]: The fitted model parameters were determined through nonlinear regressions of the experimental sorption isotherms ( Table 1). The values of are close to unity, indicating that this monolayer model could be used to elucidate the sorption mechanisms of the test adsorbents.

Adsorption Isotherms
To elucidate the equilibrium states of RhB dye molecules on materials and in liquid surroundings, the adsorption isotherms were conducted [39]. In particular, WSO-5 was chosen as the model material due to its higher adsorption capacity. The adsorption isotherms of RhB on WSO-5 were carried out at gradient temperatures of 30, 40 and 50 • C. As shown in Figure 8, the adsorption isotherms exhibited conventional L-types with obvious platforms. The results showed that the increasing temperature in circumstance could facilitate the adsorption for WSO-5, suggesting an endothermic adsorption process. Furthermore, a straightforward statistical physics model was utilized to fit the adsorption isotherms of RhB. The mathematical equation of the model is presented as follows [40]: where n represents the number of RhB molecules per site of adsorbent, D m (mg/g) is the density of the receptor sites, and c 1/2 (mg/L) means the concentration at half saturation of the formed layer. According to the statistic physics model, the adsorption capacity at saturation Q sat is given by [41]: The expression of the adsorption energy E, which contains the solubility c s (mg/L) of the adsorbates, can be expressed as follows [41,42]: E = RTln(c s /c 1/2 ).
The fitted model parameters were determined through nonlinear regressions of the experimental sorption isotherms ( Table 1). The values of R 2 are close to unity, indicating that this monolayer model could be used to elucidate the sorption mechanisms of the test adsorbents.  The parameter gives information about the orientation states of the adsorbates on the adsorbent surface. Traditionally, two situations are taken into consideration: if 1 , the adsorbates would interact with the multifunctional groups of adsorbents (horizontal orientation); if 1, more than one adsorbate molecules would be anchored in one receptor site on the adsorbent surface (non-horizontal orientation) [43]. The effects of the temperature on this parameter, displayed in Figure 9a, show that the values of in all tested temperature are above unity, varying from 1.84 to 2.92. In consequence, this evidence implies a multi-molecular adsorption process of RhB molecules with an oblique orientation on WS2/WO3 [40].
Moreover, the effect of the temperature on the receptor sites' density was also interpreted, and the results are shown in Figure 9b. This decreasing tendency is related to the increasing number of molecules captured per site, leading to a tendency of aggregation which would hinder the accessibility of the RhB molecules to the adsorption receptor sites [42,44].
The correlation of the parameters ( and ) of the monolayer can be examined by the variation of the adsorption capacity on the temperature. As shown in Figure 9c, the effect of the temperature on the parameter indicates that the increment of temperature results in an increasing value, which further verifies an endothermic process of RhB adsorption. The adsorption energy was calculated to further clarify the interactions between the RhB molecules and the adsorbent surface. The values were calculated via Equation (8), and the results that evolve with the tested temperature are displayed in Figure 9d. The results confirm again that the adsorption is an endothermic process, which matches the analysis conclusion of well. Furthermore, the results show that the adsorption is a physisorption process within the investigated temperatures, since all values are lower than 10 kJ/mol [42].  The parameter n gives information about the orientation states of the adsorbates on the adsorbent surface. Traditionally, two situations are taken into consideration: if n < 1, the adsorbates would interact with the multifunctional groups of adsorbents (horizontal orientation); if n ≥ 1, more than one adsorbate molecules would be anchored in one receptor site on the adsorbent surface (non-horizontal orientation) [43]. The effects of the temperature on this parameter, displayed in Figure 9a, show that the values of n in all tested temperature are above unity, varying from 1.84 to 2.92. In consequence, this evidence implies a multi-molecular adsorption process of RhB molecules with an oblique orientation on WS 2 /WO 3 [40].
Moreover, the effect of the temperature on the receptor sites' density D m was also interpreted, and the results are shown in Figure 9b. This decreasing tendency is related to the increasing number of molecules captured per site, leading to a tendency of aggregation which would hinder the accessibility of the RhB molecules to the adsorption receptor sites [42,44].
The correlation of the parameters (n and D m ) of the monolayer can be examined by the variation of the adsorption capacity on the temperature. As shown in Figure 9c, the effect of the temperature on the parameter Q sat indicates that the increment of temperature results in an increasing Q sat value, which further verifies an endothermic process of RhB adsorption.
The adsorption energy E was calculated to further clarify the interactions between the RhB molecules and the adsorbent surface. The E values were calculated via Equation (8), and the results that E evolve with the tested temperature are displayed in Figure 9d. The results confirm again that the adsorption is an endothermic process, which matches the analysis conclusion of Q sat well. Furthermore, the results show that the adsorption is a physisorption process within the investigated temperatures, since all E values are lower than 10 kJ/mol [42].

Effect of Adsorbent Dosage
The cost of an adsorbent in practical application depends, to a large extent, on its specific dosage to achieve the wastewater treatment standards. In this work, the correlation of the adsorbent dosage and the removal efficiency of RhB was explored ( Figure 10). It can be clearly observed in Figure 10a that the removal efficiency of RhB increases from about 24% to almost 95% as the dosage increases from 0.125 to 1.5 g/L. However, there is no evident growth anymore when the dosage exceeds 1.0 g/L. The reason for this might be that more available binding sites are provided to capture RhB molecules in solution with the increased content of WS2/WO3 composites. To reduce the expense of sewage disposal, an optimal WS2/WO3 dosage of 1 g/L is suggested for the efficient elimination of RhB, with an initial concentration of 200 mg/L at pH 6.0.
Meanwhile, the mass-based adsorption amount of RhB dye gradually decreased with the increment of solid content (Figure 10b). This can be explained by the fact that WS2/WO3 microflowers dispersed well in solution and that the surface active sites were highly available to form a stable complex with RhB molecules at low dosages. However, higher solid dosages would cause the collision and aggregation of the WS2/WO3 adsorbent, resulting in a decrease in the density of the active sites. In addition, under a vigorous stirring rate (200 rpm), the inter-collision between WS2/WO3 at a higher solid content would also cause the desorption of some weak-linked RhB molecules and ultimately reduce the adsorption amount [22,45].

Effect of Adsorbent Dosage
The cost of an adsorbent in practical application depends, to a large extent, on its specific dosage to achieve the wastewater treatment standards. In this work, the correlation of the adsorbent dosage and the removal efficiency of RhB was explored ( Figure 10). It can be clearly observed in Figure 10a that the removal efficiency of RhB increases from about 24% to almost 95% as the dosage increases from 0.125 to 1.5 g/L. However, there is no evident growth anymore when the dosage exceeds 1.0 g/L. The reason for this might be that more available binding sites are provided to capture RhB molecules in solution with the increased content of WS 2 /WO 3 composites. To reduce the expense of sewage disposal, an optimal WS 2 /WO 3 dosage of 1 g/L is suggested for the efficient elimination of RhB, with an initial concentration of 200 mg/L at pH 6.0.

Effect of Solution pH
Usually, the solution pH is a significant factor that exerts great influence on the adsorption of dyes onto the adsorbents. Therefore, the effect of the pH on RhB adsorption for the WS2/WO3 hybrid was investigated, and the results are exhibited in Figure 11a. The equilibrium adsorption amount (mg/g) exhibited an increasing trend when raising pH values from 2 to 6 and then sharply decreased when the initial pH values exceeded 7. This trend suggests that WS2/WO3 exhibits an excellent Meanwhile, the mass-based adsorption amount of RhB dye gradually decreased with the increment of solid content (Figure 10b). This can be explained by the fact that WS 2 /WO 3 microflowers dispersed well in solution and that the surface active sites were highly available to form a stable complex with RhB molecules at low dosages. However, higher solid dosages would cause the collision and aggregation of the WS 2 /WO 3 adsorbent, resulting in a decrease in the density of the active sites. In addition, under a vigorous stirring rate (200 rpm), the inter-collision between WS2/WO3 at a higher solid content would also cause the desorption of some weak-linked RhB molecules and ultimately reduce the adsorption amount [22,45].

Effect of Solution pH
Usually, the solution pH is a significant factor that exerts great influence on the adsorption of dyes onto the adsorbents. Therefore, the effect of the pH on RhB adsorption for the WS 2 /WO 3 hybrid was investigated, and the results are exhibited in Figure 11a. The equilibrium adsorption amount Q e (mg/g) exhibited an increasing trend when raising pH values from 2 to 6 and then sharply decreased when the initial pH values exceeded 7. This trend suggests that WS 2 /WO 3 exhibits an excellent performance for RhB adsorption under neutral or slightly acidic circumstances, which can be interpreted by the correspondingly extraordinary negative Zeta potentials of the WS 2 /WO 3 materials displayed in this pH range (pH ≥ 4) (Figure 11b) that facilitate the attraction of RhB dye molecules.

Adsorption Mechanism
The adsorption mechanism of RhB dye onto WS2/WO3 heterostructures was further elucidated by considering all the results discussed above. As shown in Figure 12a, the electronegative sulfur and oxygen atoms in the polar active sites of the WS2/WO3 heterostructures would connect with the RhB ± molecules by electrostatic attraction, subsequently inducing the adsorption process. According to the fitting results of the monolayer model mentioned before, RhB ± molecules exhibit a non-horizontal orientation on the WS2/WO3 surface. Specifically, this is primarily conducted via the RhB ± molecules aggregating layer-by-layer on the active sites in a head-to-tail way, since this zwitterionic form has positive and negative sites simultaneously [22]. The adsorption of RhB molecules onto materials can be detected through a comparison of the FTIR spectroscopy before and after adsorption toward RhB (Figure 12b). The peak at about 580 cm −1 in the spectra of WS2/WO3 is assigned to O-W-O vibration. Meanwhile, the new peaks at about 1590 and 1340 cm −1 belonging to the characteristic peaks of in-ring C-C and C-H stretching vibrations in the aromatic ring of the RhB molecule could be detected after the adsorption of dye (Figure 12c) [26,49]. However, there are no The interrelation of the existing forms of RhB in solution and the adsorption capacity were further studied. It has been reported that RhB molecules exist in three forms in solution, including the cationic, anionic, and zwitterionic forms (Figure 11c) [46]. In an acidic environment, the RhB molecule is ionized and exists predominantly in its cationic form (RhB + ). However, the competitive protons introduced by the acidic environment exert a positive charge effect on the WS 2 /WO 3 surface, which is reflected by the Zeta potential analysis (Figure 11b). Therefore, under these circumstances it is difficult for WS 2 /WO 3 to come into contact with RhB + molecules, owing to the electrostatic repulsion. With gradually increased pH values, the fraction of the cationic form (RhB + ) decreased, while the zwitterionic form (RhB ± ) increased simultaneously. These zwitterionic RhB molecules, with both C=N + and COO − groups, would aggregate on the active sites, resulting in a higher adsorption capacity [47,48]. When the pH rose up to a higher value above 7, the anionic form (RhB − ) increased, and the adsorption capacity of WS 2 /WO 3 significantly decreased owing to the electrostatic repulsion between negatively charged RhBand negative WS 2 /WO 3 .

Adsorption Mechanism
The adsorption mechanism of RhB dye onto WS 2 /WO 3 heterostructures was further elucidated by considering all the results discussed above. As shown in Figure 12a, the electronegative sulfur and oxygen atoms in the polar active sites of the WS 2 /WO 3 heterostructures would connect with the RhB ± molecules by electrostatic attraction, subsequently inducing the adsorption process. According to the fitting results of the monolayer model mentioned before, RhB ± molecules exhibit a non-horizontal orientation on the WS 2 /WO 3 surface. Specifically, this is primarily conducted via the RhB ± molecules aggregating layer-by-layer on the active sites in a head-to-tail way, since this zwitterionic form has positive and negative sites simultaneously [22]. The adsorption of RhB molecules onto materials can be detected through a comparison of the FTIR spectroscopy before and after adsorption toward RhB (Figure 12b). The peak at about 580 cm −1 in the spectra of WS 2 /WO 3 is assigned to O-W-O vibration. Meanwhile, the new peaks at about 1590 and 1340 cm −1 belonging to the characteristic peaks of in-ring C-C and C-H stretching vibrations in the aromatic ring of the RhB molecule could be detected after the adsorption of dye (Figure 12c) [26,49]. However, there are no obvious dye characteristic peaks in the spectra of WS 2 and WO 3 , which indicates that few RhB molecules are adsorbed onto these two materials. Besides, no characteristic peaks of RhB were observed after the desorption process, which means that electrostatic attraction is responsible for the RhB adsorption by WS 2 /WO 3 . On the other hand, the XRD patterns in Figure 12d also prove that the adsorption process didn't change the crystal structure of WSO-5. The characteristic peaks are well retained after adsorption, and there are no obvious peaks shifts, which demonstrates that the adsorption is a complete physical behavior that occurs primarily through electrostatic attraction.

Recyclability
For the practical application of an adsorbent, a quick adsorption rate and high adsorption capacity are of great significance, and the reusability of the adsorbent should be taken into consideration as well. After the adsorption experiments, the adsorbent was washed with ethanol as the eluting agent and reused for the next adsorption process. According to Figure 13, the adsorption

Recyclability
For the practical application of an adsorbent, a quick adsorption rate and high adsorption capacity are of great significance, and the reusability of the adsorbent should be taken into consideration as well. After the adsorption experiments, the adsorbent was washed with ethanol as the eluting agent and reused for the next adsorption process. According to Figure 13, the adsorption performance of WSO-5 toward RhB dye slightly decreased after the first cycle, while remaining almost constant in the following cycles. These results demonstrate that WS 2 /WO 3 has a good stability and can be regenerated and reused efficiently for the treatment of waste water contaminated with RhB dye. Figure 12. (a) The adsorption mechanism of WSO-5 microflowers toward RhB; (b) the FTIR spectra of WS2, WS2/WO3, and WO3 before and after the adsorption of RhB dye; (c) The enlarged image of WSO-5 of the zone marked in (b); and (d) the XRD patterns of WSO-5 before and after the adsorption of RhB dye.

Recyclability
For the practical application of an adsorbent, a quick adsorption rate and high adsorption capacity are of great significance, and the reusability of the adsorbent should be taken into consideration as well. After the adsorption experiments, the adsorbent was washed with ethanol as the eluting agent and reused for the next adsorption process. According to Figure 13, the adsorption performance of WSO-5 toward RhB dye slightly decreased after the first cycle, while remaining almost constant in the following cycles. These results demonstrate that WS2/WO3 has a good stability and can be regenerated and reused efficiently for the treatment of waste water contaminated with RhB dye.

Comparison with Other Adsorption Materials
To sum up, in this study the WS 2 /WO 3 hybrids were first synthesized and utilized to investigate their dye adsorption performance. The comparison of the mass-based adsorption performance between them and some other reported TMD absorbents is presented in Table 2. It can be clearly seen that the WS 2 /WO 3 hybrids, particularly WSO-5, exhibit a higher adsorption capacity and relatively medium adsorption time. Because of the multiple advantages, such as a simple synthesis process, high removal efficiency, and good reusability, the WS 2 /WO 3 heterostructure could be a promising candidate for the elimination and separation of organic dyes from wastewater.

Discussion
In this work, a new type of WS 2 /WO 3 heterostructure, which could be used as excellent absorbents, was developed and fabricated by coupling a solvothermal synthesis method with a partial oxidation strategy. Different characterization results confirmed the successful fabrication of WS 2 /WO 3 heterostructures, which possess more negative surface charges and which highly promoted hydrophilicity after a partial oxidation. The synthesized heterostructures were employed for the decontamination of dyes, and the results indicate that WSO-5 has an enhanced adsorption capacity of 237.1 mg/g, compared with pure WS 2 and WO 3 , which exhibited a limited capacity of 16.4 and 30.4 mg/g, respectively. Besides, three models were applied to simulate the kinetics in the adsorption process, and the results prove that the pseudo-second-order model gives better fitting results. Furthermore, the adsorption isotherms were also conducted by a precise monolayer adsorption model to reveal the interaction between the adsorbent and adsorbate, demonstrating a non-horizontal orientation of the RhB molecules and an endothermic physisorption process. Finally, the recyclability study showed that WS 2 /WO 3 can be easily regenerated through a simple ethanol washing and that the adsorption performance was well-maintained after five cycles. These results indicate that WS 2 /WO 3 can be a promising adsorbent for the removal of RhB dye from wastewater. Overall, the partial oxidation strategy can introduce more negative surface charges and enhance hydrophilicity, which gives an indication for the surficial modification of other TMD materials and their application in the elimination of dyes.