Agro-Waste Derived Biomass Impregnated with TiO 2 as a Potential Adsorbent for Removal of As(III) from Water

: A novel type of adsorbent, TiO 2 impregnated pomegranate peels (PP@TiO 2 ) was successfully synthesized and its e ﬃ cacy was investigated based on the removal of As(III) from water. The adsorbent was characterized using Scanning Electron Microscopy (SEM), Energy Dispersive X-ray Spectrometer (EDS), X-ray Di ﬀ raction (XRD) analysis, and Fourier Transform Infrared (FTIR) Spectroscopy, to evaluate its morphology, elemental analysis, crystallinity, and functional groups, respectively. Batch experiments were conducted on PP@TiO 2 for As(III) adsorption to assess the adsorption isotherm, e ﬀ ect of pH, and adsorption kinetics. Characterization data suggested that TiO 2 was successfully impregnated on the biomass substrate. The equilibrium data better ﬁtted to the Langmuir isotherm model having a maximum adsorption capacity of 76.92 mg / g and better distribution coe ﬃ cients (K D ) in the order of ~10 3 mL / g. The highest percentage of adsorption was found at neutral pH. The adsorption kinetics followed the pseudo-2nd-order model. X-ray Photoelectron Spectroscopy (XPS) of the adsorption product exhibited that arsenic was present as As(III) and partially oxidized to As(V). PP@TiO 2 can work e ﬀ ectively in the presence of coexisting anions and could be regenerated and reused. Overall, these ﬁndings suggested that the as-prepared PP@TiO 2 could provide a better and e ﬃ cient alternative for the synergistic removal of As(III) from water.


Introduction
Safe water is the basic need for the existence of life. Due to the rapid industrialization, the world's supply of safe water is decreasing and the pollutant levels are rising steadily [1][2][3]. Arsenic (As) contaminated water is one of the most challenging environmental concerns because of its chronic toxicity and carcinogenicity (black foot disease; skin, liver, bladder, kidney, and lung cancer; stillbirths; and heart attacks) [4]. Naturally occurring As-contaminated groundwater has caused hazardous effects to hundreds of millions of population globally, especially in Bangladesh, India, Pakistan, There are voluminous research works available in the literature regarding the use of agro-waste derived biomass materials [3,10,18,20,21,28,31,42,44] and bare nano-TiO 2 [37][38][39][40] for sequestration of heavy metals including arsenic from aqueous medium. However, impregnation of TiO 2 into agro-waste derived biomass is still lacking though TiO 2 composite with other precursors like activated carbon [32], metal oxides [36], chitosan [45][46][47][48], graphene [54], and carbon nanotube [55] have been studied. To the best of our knowledge, no reports are available in the literature on the synthesis of TiO 2 impregnated pomegranate peels (PP@TiO 2 ) for the removal and oxidation of As(III) simultaneously. Because of such circumstances, PP@ TiO 2 is supposed to be a novel materials herein studied and its application as an adsorbent for removal of As(III) was proposed.
The objective of the present work was to fabricate TiO 2 impregnated pomegranate peels (PP@TiO 2 ) by using a modified sol-gel method and to investigate its adsorption behaviors of As(III) from water under different conditions through batch experiments. The adsorption characteristics of the PP@TiO 2 were analyzed by SEM, EDS, XRD, FTIR, and XPS. The adsorption isotherms, kinetics parameters, effect of pH, and effect of common coexisting anions on As(III) adsorption were also investigated and discussed. From the results, the plausible adsorption mechanisms had been proposed. A desorption study was conducted to investigate the regeneration and reusability of the adsorbent. With further development, this research may offer a cost-effective, environmentally friendly, and easy-handle treatment alternative for remediation of arsenic from aqueous medium.

Adsorbent Characterization
The surface morphology of adsorbent was studied through SEM measurements. The SEM images (Figure 1a,b) illustrate the porous nature and irregular morphology of cross-linked pomegranate peels powder (PP) and PP@TiO 2 . The decrease in porosity in PP@TiO 2 (Figure 1b) as compared to PP (Figure 1a) is probably due to the occlusion of pores by TiO 2 particles. As shown, some particles of TiO 2 are attached to the surface of biomass. Therefore, it is justified that the impregnation of PP by TiO 2 has been done effectively. SEM image of As(III) adsorbed PP@TiO 2 (Figure 1c) shows that the surfaces of adsorbent are covered because of the adsorption of arsenic and forms layer on the surfaces. EDS spectra of the PP, PP@TiO 2 , and As(III) adsorbed PP@TiO 2 , are shown in Figure 1d-f, respectively. As seen in Figure 1d, the main peaks corresponding to C, O, and Si were observed for the PP sample. After impregnation with TiO 2 (PP@TiO 2 ), new peaks related to Ti were observed (Figure 1e), suggesting that the TiO 2 had been successfully impregnated onto biomass. As a result of arsenic adsorption, another peak belonging to arsenic was observed in the EDX spectra of As(III) adsorbed PP@TiO 2 ( Figure 1f). This verified that PP@TiO 2 have successfully adsorbed arsenic from aqueous solution. The adsorption product was termed as As(III) adsorbed PP@TiO 2 and abbreviated as As//PP@TiO 2 hereafter. Figure 2a shows the result of the XRD pattern of PP and PP@TiO 2 . There were no sharp peaks of PP indicating the amorphous nature. In the case of PP@TiO 2 , the diffraction peaks attributed to crystalline TiO 2 appear at 2θ = 25.  [56][57][58]. The confirmation of the functional groups on adsorbents was done using FTIR spectra as shown in Figure 2b. It was examined in the ATR mode of FTIR. In the case of PP, absorption bands observed at around 3346, 2915, 1735, and 1621 cm −1 are attributed to -OH, -CH 2 , -COO, and C=O stretching, respectively. The bands with the highest intensity located at 1023 cm −1 correspond to the C-O stretching of the hydroxyl functional group connected to the carbon atom. In the case of PP@TiO 2 , the absorption peaks between 420 and 700 cm −1 ascribed to the Ti-O vibration representing the interaction of TiO 2 with biomass [54,56]. This provides clear evidence about the impregnation of TiO 2 and the formation of the Ti-O-C bond. The intensities of bands of the aforementioned oxygen-containing functional groups of PP@TiO 2 were significantly decreased as compared to PP, indicating that loading of TiO 2 onto PP. After arsenic adsorption, at the FTIR spectra of As//PP@TiO 2 , a new band observed at 825 cm −1 , corresponding to As-O stretching vibrations. This confirms As(III) adsorption onto PP@TiO 2 . A similar phenomenon has been reported for arsenic adsorption onto nano scaled activated carbon modified by iron and manganese oxide [59].  Figure 2a shows the result of the XRD pattern of PP and PP@TiO2. There were no sharp peaks of PP indicating the amorphous nature. In the case of PP@TiO2, the diffraction peaks attributed to crystalline TiO2 appear at 2θ = 25.2°, 38.1°, 48°, 55°,63°,70°, and 75°. These peaks correspond to the (1 0 1), (0 0 4), (2 0 0), (1 0 5), (2 0 4), (2 2 0), and (2 1 5) planes, respectively. This confirms the successful impregnation of TiO2 in the crystalline anatase form in biomass. The peaks show good agreements with JCPDS, card No. 21-1272 of anatase phase of TiO2 [56][57][58]. The confirmation of the functional groups on adsorbents was done using FTIR spectra as shown in Figure 2b. It was examined in the ATR mode of FTIR. In the case of PP, absorption bands observed at around 3346, 2915, 1735, and 1621 cm −1 are attributed to -OH, -CH2, -COO, and C=O stretching, respectively. The bands with the highest intensity located at 1023 cm −1 correspond to the C-O stretching of the hydroxyl functional group connected to the carbon atom. In the case of PP@TiO2, the absorption peaks between 420 and 700 cm −1 ascribed to the Ti-O vibration representing the interaction of TiO2 with biomass [54,56]. This provides clear evidence about the impregnation of TiO2 and the formation of the Ti-O-C bond. The intensities of bands of the aforementioned oxygen-containing functional groups of PP@TiO2 were significantly decreased as compared to PP, indicating that loading of TiO2 onto PP. After arsenic adsorption, at the FTIR spectra of As//PP@TiO2, a new band observed at 825 cm −1 , corresponding to As-O stretching vibrations. This confirms As(III) adsorption onto PP@TiO2. A similar phenomenon has been reported for arsenic adsorption onto nano scaled activated carbon modified by iron and manganese oxide [59].

Adsorption Isotherm
To investigate the quantitative As(III) adsorption capacity of PP@TiO2, adsorption isotherm experiments were carried out using batch methods within a broad range of As(III) concentrations (V/m =1000 mL/g, pH~7). The initial and equilibrium concentrations of As(III) in solution are measured by inductively coupled plasma-mass (ICP-MS) spectroscopy. The obtained concentrations

Adsorption Isotherm
To investigate the quantitative As(III) adsorption capacity of PP@TiO 2 , adsorption isotherm experiments were carried out using batch methods within a broad range of As(III) concentrations (V/m =1000 mL/g, pH~7). The initial and equilibrium concentrations of As(III) in solution are measured by inductively coupled plasma-mass (ICP-MS) spectroscopy. The obtained concentrations were used to calculate the % adsorption (A%, from Equation (1)) and the adsorption capacity of PP@TiO 2 (q e , from Equation (2)).
where, V, m, C o , and C e indicate the volume of As(III) solution (mL), amount of PP@TiO 2 taken (g), initial concentration (mg/L), and equilibrium concentration (mg/L) respectively. The distribution coefficient (K D ) is used to express the chemical binding capacity of an adsorbent and is stated as the ratio of the concentration of arsenic adsorbed in per gram of the adsorbent and its concentration per mL in the aqueous phase. It was expressed according to Equation (3) [60].
The equilibrium data are presented in Figure 3a. Past studies have reported that adsorbents having a K D value of 10 3 mL/g are considered good adsorbent [61]. K D values for the adsorption of As(III) obtained are in the order of~10 3 mL/g(see Figure 4b), which indicates that As(III) binds efficiently to the adsorbent.     The equilibrium data were fitted using two binding models, Langmuir and Freundlich. The Langmuir model assumes the saturated monolayer adsorption on the homogenous and energetically identical adsorption sites, and no significant interaction between adsorbed species, which is expressed according to Equation (4) in the linear form [62] as: Ce where, qe (mg/g) is the amount of As(III) adsorbed at equilibrium concentration Ce (mg/L), qm (mg/g) is the maximum adsorption capacity of adsorbent, and b (L/mg) is the Langmuir equilibrium constant The equilibrium data were fitted using two binding models, Langmuir and Freundlich. The Langmuir model assumes the saturated monolayer adsorption on the homogenous and energetically identical adsorption sites, and no significant interaction between adsorbed species, which is expressed according to Equation (4) in the linear form [62] as: where, q e (mg/g) is the amount of As(III) adsorbed at equilibrium concentration C e (mg/L), q m (mg/g) is the maximum adsorption capacity of adsorbent, and b (L/mg) is the Langmuir equilibrium constant related to adsorption energy. The q m and b were determined from the slope and intercept of the Langmuir plot of C e versus C e /q e presented in Figure 3b. Freundlich model considers a heterogeneous sorption surface and sorption sites with a different energy, which can be expressed by Equation (5) in its linear form [63] as: where K F (mg/g) and n (L/mg) are Freundlich constants that indicate the adsorption capacity and the adsorption intensity, respectively. The two constants K F and n were calculated from the slope (1/n) and intercept (logK F ) of the Freundlich isotherm plot of logC e versus logq e presented in Figure 3c.
The evaluated values of the isotherm parameters of both models are listed in Table 1. It is obvious from the result that the PP@TiO 2 is a suitable adsorbent for the adsorption of As(III) from water. For good adsorbent, the value of Freundlich constant (n) ranged between 1 and 10 (i.e., 1/n < 1). It is evident from Table 1 that the value of 'n' was found to be 2.72, indicating a strong adsorbent-adsorbate interaction. This tells us that the result obtained from Freundlich isotherm certainly cannot be neglected. Thus, both Langmuir and Freundlich isotherms showed a good fit to the experimental equilibrium data. The value of correlation coefficients (R 2 ) shows that the Langmuir model is better fitted than the Freundlich isotherm model. A similar observation was reported by Chandra et al. [64] for arsenic adsorption on magnetite reduced graphene oxide and Shehzad et al. [9] for As(III) adsorption on ZrO 2 nanosheets. These results indicated that the adsorption of As(III) onto PP@TiO 2 was not governed by a single mechanism. The observed maximum As(III) adsorption capacity, q m , 76.92 mg/g compares well with previously reported materials for Arsenic adsorption, which are listed in Table 2. This comparison indicates that the adsorption capacity of PP@TiO 2 is higher than that of previously studied adsorbents [9,[16][17][18]21,37,42,44,[65][66][67][68][69][70]. Hence, the prepared PP@TiO 2 may be a promising material towards the removal of As(III) from water. The feature of Langmuir isotherm was further justified by the analysis of the dimensionless separation factor, R L , [71], which is given by Equation (6).
where b (L/mg) is Langmuir equilibrium constant and C o (mg/L) is the initial concentration of As(III). The value of R L maybe 0, 1, or >1 indicating irreversible, linear, and unfavorable adsorption respectively. From the experimental data for all the concentrations tested, the evaluated R L values were ranges from 0.06 to 0.76 (see Figure 3d). All of these values lie between 0 and 1 (0 < R L < 1), which indicates more favorable adsorption of As(III) onto PP@TiO 2 .

Effect of pH and As(III) Adsorption Mechanism
The % adsorption of As(III) and distribution coefficient, K D onto PP@TiO 2 evaluated in the pH range of 2-12 are presented in Figure 4a,b, respectively, which show that the effective pH range for As(III) adsorption is from 6 to 9. A maximum K D value of 3.21 × 10 3 mL/g with the highest As(III) adsorption capacity of 81.4% is observed at pH 7. On increasing pH of the solution from 2 to 7, As(III) adsorption increased and then decreased beyond the pH 7. At lower pH values, As (III) exists only as neutral H 3 AsO 3 species, and the surfaces of PP@TiO 2 are protonated. Thus, binding of As (III) on positively charged adsorbent is not favorable [20]. The adsorption of As(III) species took place via both electrostatic attraction and ligand exchange of the hydroxyl ions or water molecules contained in the coordination spheres of impregnated TiO 2 from the biomass surface. At higher than 10 pH values, Catalysts 2020, 10, 1125 8 of 18 the predominant As(III) species H 2 AsO 3 -, HAsO 3 2and AsO 3 3are repelled by the negatively charged adsorbent surface and therefore lower adsorption of As(III) has been obtained [21]. The decrease in As(III) adsorption can be attributed to the combined effect of electrostatic repulsion and the competition between the hydroxyl ions, and anionic arsenic species for adsorption sites. This is in agreement with the observation reported for As(III) adsorption from aqueous solution using iron oxide/nano-porous carbon magnetic composite [16]. Mechanism of As(III) adsorption using PP@TiO 2 can be described as follows. Impregnation with TiO 2 results in the generation of additional adsorption sites on the exterior surface of PP@TiO 2 . It is inferred that during the impregnation of tetravalent Ti, neutralization of all the four positive charges of Ti by the functional groups (-OH, -COOH) present in biomass is difficult because of steric hindrance. Therefore, one or two positive charges of tetravalent Ti bind with a carboxylic acid or other functional groups in pomegranate peel biomass, and the rest are neutralized with hydroxyl ion. Thus, TiO 2 in aqueous solution undergoes a hydration reaction to give hydrated titanium oxide, Polymer ≡ Ti(OH) n (H 2 O) 6-n (Equation (7)). At near neutral as well as weakly alkaline conditions i.e., at pH 6 to 9, As(III) exists as anionic species H 2 AsO 3 and HAsO 3 2as well as neutral H 3 AsO 3 molecule [20]. The adsorption of As(III) on PP@TiO 2 may be considered as a ligand exchange mechanism by the substitution of hydroxyl anions or neutral H 2 O molecules. Thus, the undermentioned mechanism may be responsible for the sorption of anionic species (Equation (8)) and neutral species (Equation (9)) of As(III). Polymer Polymer Besides, As(III) could be partially oxidized to less toxic and more easily adsorbed As(V) by the simple interaction with TiO 2 on the surface of the adsorbent, in batch experiments performed under room conditions [37]. The oxidized As(V) species, H 2 AsO 4 would be adsorbed according to A similar mechanism has also been reported by Biswas et al., [20] for the removal of arsenic using Zr(IV) loaded orange waste gel. Joshi et al., [16] also reported a ligand exchange mechanism for As(III) onto Fe 3 O 4 /sugarcane bagasse activated carbon composite.

Adsorption Kinetics
To understand the dynamics of adsorption, the time variation in the amount of adsorption of As(III) on PP@TiO 2 was measured. Figure 5a shows the equilibrium concentration and adsorption percentage of As(III) as a function of contact time. Figure 5b exhibits the adsorption capacity (q t ) of As(III) with contact time on PP@TiO 2 . The adsorption for As(III) reaches equilibrium within~6 h. Therefore, 24 h was chosen as an optimum agitation period to achieve the complete adsorption of As(III) on PP@TiO 2 in the subsequent adsorption experiments. The kinetic data were fitted with pseudo-1st-order and pseudo-2nd-order models, which are expressed respectively [72] as Equations (11) and (12) below. log(q e − q t ) = log q e − k 1 2.303 t (11) Catalysts 2020, 10, 1125 9 of 18 where q e and q t (mg/g) are the amount of As(III) ions adsorbed per unit mass of PP@TiO 2 at equilibrium and time t, respectively, and k 1 ,(min -1 ) and k 2 (g mg −1 min −1 ) are the rate constants of pseudo-1st-order and pseudo-2nd-order adsorption kinetics model. The pseudo-1st-order kinetic plot of log(q e − q t ) versus t (Figure 5c) gives a linear relationship from which k 1 and q e can be measured from the slope and intercept, respectively. Similarly, the pseudo-2nd-order kinetic plot t/q t versus t (Figure 5d) gives a linear relationship from which k 2 and q e can be measured from the intercept and slope, respectively. The kinetic parameters of As(III) adsorption evaluated based on both equations and R 2 are listed in Table 3. The pseudo-2nd-order kinetics is suitable for As(III) adsorption due to the higher value of R 2 (0.999), indicating the adsorption is chemisorption.
As(III) on PP@TiO2 in the subsequent adsorption experiments. The kinetic data were fitted with pseudo-1st-order and pseudo-2nd-order models, which are expressed respectively [72] as Equations (11) and (12)  where qe and qt (mg/g) are the amount of As(III) ions adsorbed per unit mass of PP@TiO2 at equilibrium and time t, respectively, and k1,(min -1 ) and k2 (g mg −1 min −1 ) are the rate constants of pseudo-1st-order and pseudo-2nd-order adsorption kinetics model. The pseudo-1st-order kinetic plot of log(qe − qt) versus t (Figure 5c) gives a linear relationship from which k1 and qe can be measured from the slope and intercept, respectively. Similarly, the pseudo-2nd-order kinetic plot t/qt versus t (Figure 5d) gives a linear relationship from which k2 and qe can be measured from the intercept and slope, respectively. The kinetic parameters of As(III) adsorption evaluated based on both equations and R 2 are listed in Table 3. The pseudo-2nd-order kinetics is suitable for As(III) adsorption due to the higher value of R 2 (0.999), indicating the adsorption is chemisorption.

X-ray Photoelectron Spectroscopy (XPS) Studies
The surface composition and chemical states of the PP@TiO 2 after As(III) adsorption experiments were further analyzed by the XPS. The wide scan spectrum indicated signals from Ti, O, C, and As in the sorption product (Figure 6a). The high-resolution XPS scan of Ti 2p (Figure 6b) showed two broad peaks centered at~458.6 and 467.3 eV representing the Ti 2p 3/2 and Ti 2p 1/2 peaks of Ti 4+ , respectively [56]. Additional peaks at about 457.25 eV and 463.27 are attributed to the Ti 2p 3/2 and Ti 2p 1/2 peaks of Ti 3+ , respectively [73]. The Ti 2p spectrum mainly appears in Ti 4+ state with a small amount of Ti 3+ .
The occurrence of Ti 3+ is due to oxygen deficiency in the lattice of TiO 2 [73]. This indicated a reduction of Ti 4+ to Ti 3+ during the partial oxidation of adsorbed As(III). The high-resolution XPS scan of As 3d (Figure 6c) was composed of two peaks indicating the presence of both As 3+ (B.E. = 44.32 eV) and As 5+ (B.E. = 46.24 eV) with a qualitative ratio of 75.7% and 24.3%, respectively. Hence, the partial oxidation of As(III) by impregnated TiO 2 was verified by XPS using As 3d spectrum [74]. The high-resolution XPS scan of O 1s (Figure 6d) was composed of peaks at~528.4, 529.9, and 531.8 eV are associated with Ti-O, O-As, and O-H bonds, respectively.

X-ray Photoelectron Spectroscopy (XPS) Studies
The surface composition and chemical states of the PP@TiO2 after As(III) adsorption experiments were further analyzed by the XPS. The wide scan spectrum indicated signals from Ti, O, C, and As in the sorption product (Figure 6a). The high-resolution XPS scan of Ti 2p (Figure 6b) showed two broad peaks centered at ~458.6 and 467.3 eV representing the Ti 2p3/2 and Ti 2p1/2 peaks of Ti 4+ , respectively [56]. Additional peaks at about 457.25 eV and 463.27 are attributed to the Ti 2p3/2 and Ti 2p1/2 peaks of Ti 3+ , respectively [73]. The Ti 2p spectrum mainly appears in Ti 4+ state with a small amount of Ti 3+ . The occurrence of Ti 3+ is due to oxygen deficiency in the lattice of TiO2 [73]. This indicated a reduction of Ti 4+ to Ti 3+ during the partial oxidation of adsorbed As(III). The highresolution XPS scan of As 3d (Figure 6c) was composed of two peaks indicating the presence of both As 3+ (B.E. = 44.32 eV) and As 5+ (B.E. = 46.24 eV) with a qualitative ratio of 75.7% and 24.3%, respectively. Hence, the partial oxidation of As(III) by impregnated TiO2 was verified by XPS using As 3d spectrum [74]. The high-resolution XPS scan of O 1s (Figure 6d) was composed of peaks at ~528.4, 529.9, and 531.8 eV are associated with Ti-O, O-As, and O-H bonds, respectively.  Thus, XPS results revealed that As(III) is adsorbed onto PP@TiO 2 and is partially oxidized to As(V). Similar behavior was reported by Tamayo et al. [19] for As(III) removal from aqueous solution on calcium titanate nanoparticles. The partial oxidation of As(III) to As(V) could be due to the presence of physio-sorbed oxygen and surface hydroxyl groups. All results revealed that PP@TiO 2 , on one hand, can remediate As(III) from water, and on the other hand, it can oxidize some of the As(III) to As(V) at the surface of the adsorbent simultaneously. Previous studies have reported that the degree of photocatalytic oxidation of As(III) to As(V) under the influence of TiO 2 photocatalyst highly dependent on various operation factors such as pH, dissolved oxygen, initial As(III) concentration, a wavelength of light, the intensity of light, TiO 2 loading, calcination temperature, properties of TiO 2 (bandgap, particle size), presence of coexisting ions and batch reactor configuration [32,75]. Bang and Patel [37] reported that around 10% of As(III) is oxidized to less toxic and more easily adsorbed As(V) by the simple interaction with TiO 2 , in batch experiments performed under room conditions. Yao et al. [32] reported that 0.80 mg/L of As(III) could be completely oxidized to As(V) within 30 min in the presence of 3.0 g/L TiO 2 /ACF (composite of TiO 2 and activated carbon fiber) in the irradiation of UV-light. They observed that the presence of silicate and phosphate ions significantly decreased the oxidation of As(III). Past studies have reported the complete oxidation of As(III) to As(V) by TiO 2 in the presence of UV light and oxygen [30,34]. The As(III) adsorption and photo-oxidation capability of biomaterial/TiO 2 composite increases as n-TiO 2 loading [53]. Nevertheless, additional research is required to investigate detailed As(III) oxidation characteristics of PP@TiO 2 .

Effect of Common Coexisting Anions
The effect of common coexisting anions on As(III) adsorption from aqueous solution was examined considering various concentrations of chloride (Cl − ), sulphate (SO4 2− ), and phosphate (PO4 3− ). As shown in Figure 7, compared to the blank (no coexisting ions), an increase in the concentration of coexisting ions leads to a decrease in As(III) adsorption efficiency for all the tested ions. Minimal interference was observed in the presence of Cl − , while the stronger interference was observed for PO4 3− that reaches % adsorption loss up to 27.9% at its highest concentration (100 mg/L). Hence, phosphate is the strongest competitor with arsenic ions to binding sites on PP@TiO 2 . This can be attributed to the similar oxyanion structure of both phosphate and arsenic anions [76].
properties of TiO2 (bandgap, particle size), presence of coexisting ions and batch reactor configuration [32,75]. Bang and Patel [37] reported that around 10% of As(III) is oxidized to less toxic and more easily adsorbed As(V) by the simple interaction with TiO2, in batch experiments performed under room conditions. Yao et al. [32] reported that 0.80 mg/L of As(III) could be completely oxidized to As(V) within 30 min in the presence of 3.0 g/L TiO2/ACF (composite of TiO2 and activated carbon fiber) in the irradiation of UV-light. They observed that the presence of silicate and phosphate ions significantly decreased the oxidation of As(III). Past studies have reported the complete oxidation of As(III) to As(V) by TiO2 in the presence of UV light and oxygen [30,34]. The As(III) adsorption and photo-oxidation capability of biomaterial/TiO2 composite increases as n-TiO2 loading [53]. Nevertheless, additional research is required to investigate detailed As(III) oxidation characteristics of PP@TiO2.

Effect of Common Coexisting Anions
The effect of common coexisting anions on As(III) adsorption from aqueous solution was examined considering various concentrations of chloride (Cl − ), sulphate (SO4 2− ), and phosphate (PO4 3− ). As shown in Figure 7, compared to the blank (no coexisting ions), an increase in the concentration of coexisting ions leads to a decrease in As(III) adsorption efficiency for all the tested ions. Minimal interference was observed in the presence of Cl − , while the stronger interference was observed for PO4 3− that reaches % adsorption loss up to 27.9% at its highest concentration (100 mg/L). Hence, phosphate is the strongest competitor with arsenic ions to binding sites on PP@TiO2. This can be attributed to the similar oxyanion structure of both phosphate and arsenic anions [76].

Desorption Study and Reusability of PP@TiO 2
Desorption of As(III) from As//PP@TiO 2 is important to assess the recovery of toxic arsenic and regeneration of spent adsorbents for repeated use. Desorption tests were carried out by batch method using the adsorbent loaded with As(III) immediately after the adsorption processes. As shown in Figure 4, As(III) is efficiently adsorbed on PP@TiO 2 at pH range 6-9 but it is poorly adsorbed at higher pH (>9), indicating that desorption of adsorbed As(III) can be easily achieved by employing the weakly alkaline solution. Consequently, the desorption test was carried out with different concentrations of NaOH ranging from 0.01-0.5 M as presented in Figure 8a. The % desorption (% D) of As(III) was calculated using Equation (13) as follows [77].
where, A amount and D amount are the adsorbed and desorbed amounts (mg) of As(III) ions, respectively. The desorption of arsenic increases from 29.1% to 97.8% by increasing the concentration of NaOH from 0.01 to 0.1 M and remains constant with a further increase in concentration. Hence, a 0.1 M NaOH solution was found to be a suitable desorbing agent for the regeneration of spent adsorbent. Hydroxyl ions released from the NaOH solution might have replaced the adsorbed arsenic anions onto the PP@TiO 2 surface. The desorption of arsenic species is believed to take place by the ligand substitution mechanism by the replacement of adsorbed arsenic species by hydroxyl ions in alkali solution as shown below (Equation (14)).

% D = Damount
Aamount × 100 (13) where, Aamount and Damount are the adsorbed and desorbed amounts (mg) of As(III) ions, respectively. The desorption of arsenic increases from 29.1% to 97.8% by increasing the concentration of NaOH from 0.01 to 0.1 M and remains constant with a further increase in concentration. Hence, a 0.1 M NaOH solution was found to be a suitable desorbing agent for the regeneration of spent adsorbent. Hydroxyl ions released from the NaOH solution might have replaced the adsorbed arsenic anions onto the PP@TiO2 surface. The desorption of arsenic species is believed to take place by the ligand substitution mechanism by the replacement of adsorbed arsenic species by hydroxyl ions in alkali solution as shown below (Equation (14)).
Polymer ≡Ti(OH)n-1 (H2AsO3)(H2O)6-n + OH - Polymer ≡Ti(OH)n (H2O)6-n + H2AsO3 -, To evaluate the reusability of the PP@TiO2, it was subjected to four series of adsorption/desorption cycles. Figure 8b shows that the initial adsorption capacity of As(III) was decreased by 14.5% after four cycles. This result indicated that the PP@TiO2 exhibited excellent stability and reusability. In alkaline solution, the surface of the adsorbent becomes negatively charged and As(III) anions are displaced by OH -, leaving the sorption sites before the succeeding adsorption cycle without affecting the chemical stability of the adsorbent. The high desorption efficiency of arsenic from the PP@TiO2 created the low-cost removal technique since both adsorbent and arsenic were regenerated and recycled efficaciously. To evaluate the reusability of the PP@TiO 2 , it was subjected to four series of adsorption/desorption cycles. Figure 8b shows that the initial adsorption capacity of As(III) was decreased by 14.5% after four cycles. This result indicated that the PP@TiO 2 exhibited excellent stability and reusability. In alkaline solution, the surface of the adsorbent becomes negatively charged and As(III) anions are displaced by OH -, leaving the sorption sites before the succeeding adsorption cycle without affecting the chemical stability of the adsorbent. The high desorption efficiency of arsenic from the PP@TiO 2 created the low-cost removal technique since both adsorbent and arsenic were regenerated and recycled efficaciously.

Materials
Analytical grade chemicals were used without additional purification. The waste biomass of pomegranate peels employed in this study was collected from a local juice vendor, Kathmandu, Nepal. All the solutions in the study were prepared using deionized water. Sodium arsenite (NaAsO 2 ) (Sigma-Aldrich, New Delhi, India), titanium(IV) n-butoxide (Samchun pure chemical, Seoul, Korea) were used. A standard stock solution of As(III) (1000 mg/L) was prepared by dissolving an exact amount of NaAsO 2 in deionized water and stored in the refrigerator for further use.

Synthesis of TiO 2 Impregnated Pomegranate Peels (PP@TiO 2 )
The pomegranate peel waste was kindly donated by a local juice vendor, Kathmandu, Nepal. Pomegranate peels were first washed thoroughly with distilled water. It was then oven-dried at 70-80 • C for 48 h. The dried mass was grounded and sieved through (150 µm) mesh copper sieve. To avoid the dissolution of the adsorbent in aqueous solutions, polyphenol and polysaccharide hydroxyl groups in the biomass were cross-linked through condensation reaction using concentrated sulfuric acid as a dehydrating agent. The cross-linked pomegranate peels were prepared as described in Paudyal et al., [78]. 15 g of this biomass powder was mixed with 30 mL of concentrated H 2 SO 4 (96%) in a round bottom flask and the mixture was agitated for 24 h at 100 • C followed by cooling at room temperature. The black product was then neutralized with sodium bicarbonate and reconditioned by stirring again in 1 M HCl solution. It was washed several times with distilled water till neutrality, after which it was dried overnight in a convection oven at 70 • C. The pomegranate peels powder obtained in this way is abbreviated as PP.
The PP@TiO 2 was prepared using a previously reported modified sol-gel method [73,[79][80][81]. About 10 mL of titanium(IV) n-butoxide was mixed with 15 mL of ethanol. After stirring for 30 min at room temperature, 2 g of powdered PP was added to it. Then, 15 mL of a solution of 1:1:1 ethanol, deionized water, and acetic acid were added dropwise from the burette to the magnetically stirred mixture. The whole content was further stirred for 4h until sol was obtained. The sol was dried overnight at 80 • C after aging at 40 • C for 2 h. Finally, the dried gel was ground into a powder and calcined at 400 • C in a muffle furnace for 2 h. The resulting PP@TiO 2 was stored in a plastic container before being used in the characterization and adsorption experiment.

Characterizations
To investigate the morphology and elemental composition of adsorbents, SEM images and EDS were carried out on a JEOL JSM-6700F instrument (Jeol Ltd., Tokyo, Japan) equipped with X-ray Energy Dispersive Spectrometer. The crystallinity of PP@TiO 2 samples was investigated by XRD, Rigaku Ultima IV X-ray diffractometer (Rigaku Co., Tokyo, Japan)with Cu Kα (λ = 1.54056 Å) radiation. The surface functional groups of samples were analyzed using FTIR spectroscopy on IR AFFINITY-1 Shimadzu (Shimadzu, Kyoto, Japan) spectrometer in the wavenumber range of 400-4000 cm −1 .

Batch Adsorption Studies
The adsorption studies for As(III) were carried out to examine the adsorption behavior on the PP@TiO 2 using the batch method (V: m~1000 mL/g). 25 mg (dry weight) of PP@TiO 2 was taken in different 50 mL conical flasks containing 25 mL of As (III) solution in the concentration range 10-600 mg/L and shaken for 24 h at 25 • C to attain equilibrium. The pH of the As(III) solution was maintained with 0.1M HCl and 0.1 M NaOH solutions. After adsorption, the PP@TiO 2 was separated from the mixture by filtration and dried at 70 • C. The initial and equilibrium concentration of arsenic was analyzed. From the equilibrium concentration, the sorption isotherms and maximum adsorption capacity were evaluated. Two isotherms model, Langmuir, and Freundlich were used to describe the As(III) sorption behavior.
Effect of pH: The influence of pH of the solution on the sorption of arsenic was investigated over the pH range 2-13. 25 mg of PP@TiO 2 was shaken for 24 h in 25 mL solution of 20-25 mg/L As(III) with varying pH solutions, prepared by using 0.1M HCl and 0.1M NaOH solutions. The resulting solution was analyzed for the equilibrium concentration.
Adsorption Kinetics: Kinetic studies of As(III) adsorption on to PP@TiO 2 were carried out until it reached equilibrium. For each experiment, 25 mg of the PP@TiO 2 was taken in a 50 mL conical flask containing 25 mL of~30 mg/L As(III) aqueous solution (V/m = 1000 mL/g) at pH 7, and the content was shaken well at room temperature. Adsorption experiments of various contact times were performed. After a predetermined time interval, the sample of each flask was immediately filtered and the equilibrium concentration of As(III) was analyzed.
Effects of Common Co-existing Ions: Natural water contains several coexisting ions, which may compete with As(III) for available active sorption sites and affect the As(III) adsorption efficiency of PP@TiO 2 . The effect of common coexisting anions in an aqueous system such as Cl − , SO 4 2−, and PO 4 3− on the adsorption of As(III) was examined by adding various concentrations (10, 50, and 100 mg/L) of each coexisting ions into 25 mg/L As(III) solution as a binary mixture. To prepare various concentrations of coexisting ion solutions, different potassium salts (KCl, K 2 SO 4 , and KH 2 PO 4 , 99% pure, Merk, Kenilworth, NJ, USA) were used. Additionally, the blank sample was made similarly without adding coexisting anion. The pH of all the solutions was maintained to 7.0 by using 0.1M HNO3 and 0.1M NaOH. A defined amount (1 g/L) of PP@TiO2 was added and the mixtures were shaken for 24 h at 25 • C to attain equilibrium. After filtration, the remaining concentration of As(III) was analyzed.

X-ray Photoelectron Spectroscopy (XPS) Studies
To elucidate chemical composition and arsenic speciation, the XPS investigation of sorption product was carried out on Kratos Axis Ultra DLD spectrometer (Kratos Analyticals, Manchester, UK) with a monochromatic Al Kα X-ray source (150 W). Samples were mounted using double-sided adhesive tape, and binding energies were referenced to the C 1s binding energy of adventitious carbon taken to be 286.4 eV. For the survey of elements, the wide scans were taken at pass energy of 80 eV, and scans of photoelectron peaks were taken at pass energy of 40 eV.

Desorption Study and Reusability of PP@TiO 2
For the desorption study, As(III) adsorbed PP@TiO 2 was prepared by shaking 250 mg of PP@TiO 2 with 250 mL of As(III) solution (25 mg/L) at pH 7 for 24 h at room temperature. After filtration, the filtrate was examined for residual As(III) concentration. From initial and residual concentration, the adsorbed amount of As(III) was evaluated. The filter cake was dried. The dried As(III) adsorbed PP@TiO 2 was used for the elution test. The adsorbed arsenic was effectively eluted by the alkali solution. To identify the optimum concentration of alkali, 25 mg of exhausted PP@TiO 2 was shaken with 25 mL of various concentrations of NaOH solution ranging from 0.01 to 0.5 M batch-wise, similar to adsorption studies. The sample of each flask was filtered and the filtrate was analyzed for the desorbed amount of arsenic. The % desorption was calculated. Four cycles of adsorption/desorption of As(III) were carried out to investigate the reusability of PP@TiO 2 . The PP@TiO 2 samples after the adsorption process were stirred with 0.1M NaOH solution followed by filtration, washing with deionized water, and drying. The regenerated PP@TiO 2 was again used for the adsorption experiments to evaluate reusability up to four times.

Analysis of Arsenic Concentration
For all the above experiments, the initial and equilibrium concentration of arsenic in the solution was analyzed by Inductively Coupled Plasma-Mass Spectrometer (ICP-MS) (Agilent Technologies, 7900 ICP-MS, Santa Clara, CA, USA) with argon plasma, and internal standards of germanium and scandium. For the calibration of the instrument, a repeating As(III) standard and blank were analyzed every 10 samples. Before analysis, all samples were acidified with 1% HNO 3 and analyzed within 48 h.

Conclusions
In summary, we herein synthesized PP@TiO 2 using a modified sol-gel method, which was then applied as an efficient adsorbent for removal of As(III) from water. Based on the results of characterizations, it was reasonable to conclude the effective loading of TiO 2 into biomass and successful adsorption of As(III) onto PP@TiO 2 . The maximum adsorption of As(III) on PP@TiO 2 was observed at neutral pH with a high distribution coefficient of 4.0 × 10 3 mL/g. The equilibrium data were better followed by the Langmuir-model, which indicated that the monolayer adsorption of As(III) occurred on a homogenous surface of the PP@TiO 2 . The maximum adsorption capacity for As(III) was evaluated as 76.92 mg/g, which was found to be higher than many other agro-waste based bio-adsorbents and bare TiO 2 reported in the literature. Kinetics data showed that the adsorption followed the pseudo-second-order model, suggesting the process might be chemisorptions with higher initial adsorption rates. The XPS analysis of PP@TiO 2 after the As(III) adsorption process revealed that the adsorbed species was present as As(III) and partially oxidized to As(V). Moreover, the PP@TiO 2 can work effectively in the presence of various coexisting anions, and the exhausted PP@TiO 2 can be easily separated for regeneration and reuse. PP@TiO 2 is a safe substance for human health and a feasible low-cost material. Thus, PP@TiO 2 investigated in this study can be considered an efficient, cost-effective, environmentally friendly, and reusable adsorbent for removing As(III) from water.

Conflicts of Interest:
The authors declare no conflict of interest.