ZrMOX Particles for Enhanced Removal of Methyl Orange from Wastewater: Preparation, Characterization, and Adsorption Study

The wide application of organic dyes in many industries has brought challenges to the effective treatment of organic wastewater. In this study, a series of ZrMOX (M: Fe, Co, Ni, Cu) particles were prepared by the coprecipitation method to adsorb methyl orange (MO) in aqueous solution. The adsorbents were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FT-IR), and N2 adsorption-desorption. The selected adsorbent with the best adsorption performance was ZrFeOX with a molar ratio of 1 : 1 and calcination temperature of 573 K; the maximum adsorption capacity was 138.95 mg·g-1. The adsorption behavior of MO onto the adsorbent was studied as a function of contact time, initial concentrations, adsorption temperature, and pH conditions. The analysis results showed that pseudo-second-order, Elovich, and Langmuir models were suitable to describe the adsorption behavior of MO on the adsorbent. In addition, regeneration experiments presented that the MO removal rate reached over 96% after repeated recycling for 5 times. The adsorbent developed in this work is not only simple to prepare and low cost but also green and energy-saving, which can make some contributions to environmental governance.


Introduction
Dye is a very important part of human production and life [1]. Up to now, more than 100 000 types have been discovered or invented, and over 7 000 tons of dyes are being manufactured and used yearly world widely [2]. There are about 100 tons of dye wastewater discharged into the water body every year in the process of using dyes [3]. MO belongs to the type of azo dye; it is mainly used as an acid-base indicator and biological dye. Due to the existence of the −N=N− group and the benzene ring, MO and other azo dyes are highly toxic [4,5]. They not only cause various diseases in humans and animals but also cause harm to the environment. Therefore, it is necessary to remove MO and other dyes before the dye wastewater is discharged.
Many efforts have been made for eliminating dyes from polluted water, such as advanced oxidation technology [6], extraction [7], photocatalysis [8], and adsorption [5]. Among those, compared with the electrochemical oxidation energy consumption, extraction required more additives; photocatalysis is limited by the intensity of light. Adsorption has the advantages of mild conditions and does not need high temperature. More importantly, adsorption is a convenient, low-cost method for its extremely well-sourced adsorbent materials [9,10].
In recent years, various adsorbent materials were discovered to the removal of dyes from wastewater. Active carbon (AC) [11][12][13] and modified AC [14][15][16] were the largest type, because they showed a high specific surface area up to 2 000 m 2 ·g -1 . MOF [17] such as UiO-66-NH 2 showed excellent adsorption ability to methyl orange and methylene blue, and the maximum adsorption capacities were as high as 148.4 and 549.6 mg·g -1 , respectively. Metal composite oxides [9,18] and some natural products [19][20][21] also showed excellent adsorption performance for dye absorption. However, there are many problems in adsorbents such as deficient adsorption capacity, poor regeneration performance, and loss of active components. Zirconium has good ability to accept lone pair electrons, and various coordination compounds have stable spatial structure [22,23]. Composite oxide-based zirconium represented favorable stability against most solution systems with different acid-base properties and temperatures. The adsorbents possessed larger adsorption capacity; good stability and good reusability are optimum.
In this work, an anionic dye MO was selected as the target contaminant molecule. A series of metal composite oxides based on zirconium were prepared by the coprecipitation method as adsorbents. The materials were characterized by N 2 adsorption-desorption, XRD, SEM, and FT-IR. Then, the effects of contact time, adsorption temperature, pH, and initial concentration on the adsorption of adsorbents were evaluated through batch experiments. The adsorption kinetics, isotherm, and thermodynamics of the experimental data were analyzed. In addition, the regeneration performance of the adsorbent was evaluated.    China, 160 rpm), and each batch was controlled no less than 24 h. All concentrations of MO solutions were analyzed by a dual-beam ultraviolet-visible spectrophotometer (PERSEE, TU-1900, Beijing, China). Before determination, the standard curve was plotted by the testing of standard samples, and the correlation coefficient reached 0.99. All experimental data were repeated twice in order to reduce errors. The adsorption capacity q t (mg·g -1 ) and equilibrium adsorption capacity q e (mg·g -1 ) were used to describe the performance of MO adsorption, which were calculated through Equations (1) and (2), respectively [25]:

Experimental
where c 0 , c t , and c e (mg·L -1 ) are the MO concentrations of the moment 0, t (min), and equilibrium, respectively. v and m are the volume of MO solutions and the mass of adsorbents, respectively.

Regeneration of Adsorbents.
Regeneration study was carried out as follows. 0.5 g of used adsorbent was put into 100 mL of 0.1 M NaOH solution stirring for 1 h. The adsorbent was washed to neutrality and centrifuged out. Then, it was put into 0.001 M HCl solution containing 100 mL and stirred for 1 h, after which centrifugalization and lavation were performed. Adsorption experiment was carried out after the adsorbent was dried at 353 K for 8 h. Removal rate of MO was calculated as follows: 2.5. Characterization of Adsorbents. Morphological features of all adsorbents in this work were examined via scanning electron microscopy (SEM) using a JSM-6700F equipment (JEOL, Japan). X-ray diffraction analysis of the adsorbents was performed with a Bruker D 8 Advance (Bruker-AXS, Germany) equipped with a copper anode (Cu Kα = 1:5406 Å). Adsorbents were scanned from 5°to 90°with a step rate of 1°s -1 . The results of XRD were compared with the Powder Diffraction File (PDF) database.
0.095 0.061 0.079 0.155 * S mic , S mes , V mic , and V mes : the surface area of micropore and mesopore and the volume of micropore and mesopore, respectively.

Adsorption Science & Technology
FTIR spectra of the adsorbents were acquired using a Vertex 70 spectrometer (Bruker, Germany). N 2 adsorptiondesorption isotherms were investigated on Autosorb-iQ-MP-C (Quantachrome, America). All adsorbents were outgassed at 573 K for 4 h. The specific surface area (S BET ) and total pore volume (V t ) were calculated by the Brunauer-Emmett-Teller (BET) method, and the pore size distribution was calculated by the DFT theory.

Results and Discussion
3.1. Effect of the Type of ZrMO X . A batch of Zr/Cu, Zr/Co, Zr/Ni, and Zr/Fe oxides with the molar ratio of 1 : 1 was prepared for equilibrium adsorption experiment. The results are shown in Figure 1(a); the equilibrium adsorption capacity of the 1-573-ZFe was the largest. Surface area and pore volume are an important indicator that affects the adsorption capacity [26]. From Table 1, it can be seen that 1-573-ZFe has abundant surface area and pore volume in the four ZrMO X , which attributed to obtaining the highest equilibrium adsorption capacity among the four ZrMO X .

Investigation on the Optimal Zr/Fe Molar Ratio.
A series of ZrFeO X particles with different molar ratios were prepared as adsorbents to investigate the removal of MO. From Figure 1(b), it was obvious that the largest adsorption capacity was obtained at 1 : 1 of the Zr/Fe molar ratio. This situation may be caused by the synergistic effect between Zr and Fe [27]. When the Zr/Fe molar ratio was greater than 1, the synergistic effect gradually increases with the increase of Fe content, and the adsorption sites gradually increase, resulting in an increase in the equilibrium adsorption capacity. However, when the molar ratio was less than 1, iron oxides may coat the adsorption sites [28], resulting in a decrease in adsorption sites and further a reduction in the equilibrium adsorption capacity.

Effect of Calcination
Temperature. The calcination temperature has a significant effect on the formation [29] and absorption efficiency of adsorbents. The adsorption properties of 1-ZFe prepared after calcination at 473 K, 573 K, 673 K, and 773 K for MO were investigated, respectively; the result is shown in Figure 1(c). It can be seen from Figure 1(c) that when calcination temperature was between 473 K and 573 K, q e gradually increased as the temperature rose. q e was related to the crystallinity of 1-ZFe. As calcination temperature increased, the crystallinity of 1-ZFe gradually increased. The results of XRD characterization data can well prove this point of view. Then, with the increase of temperature, q e gradually drops to 20 mg·g -1 , and then, there was no significant change with the increase of temperature. When the temperature rose to 673 K, the decrease in q e may be due to the intense agglomeration of the metal oxides at high temperature, resulting in a sharp decrease in the number of adsorption sites.

Adsorption Science & Technology
when pH > 7. When pH met between 4 and 6, protonation of the active groups appeared on the surface of 1-573-ZFe, which increased the electron attraction with the C 14 H 14 N 3 SO 3 -. So it was conducive to adsorption of MO on the 1-573-ZFe at pH of 4~6. However, q e decreased when pH was less than 4 because of the unstable structure [30] of adsorbent under acidic conditions, which led to destruction or reduction of adsorption sites. Therefore, all adsorption experiments were carried out at the pH of 4~6 subsequently. Figure 2 shows the XRD spectrum of four 1-ZFe calcined at 473 K, 573 K, 673 K, and 773 K. It was shown that the adsorbents were amorphous. By comparing with the standard PDF card of the oxide of zirconium and iron, the peak appearing at about 32°may be contributed by Fe 2 O 3 and ZrO 2 (JCPDS card No. 40-1139 and 49-1746, respectively). When the calcination temperature reached 773 K, it can be seen that there were two peaks near 32°, which may be caused by the increase of crystallinity which reflected in the (1 1 3) and (1 1 6) planes of Fe 2 O 3 and the (1 2 0) and (1 1 1) planes of ZrO 2 . There was no information about the broad peak near 15°related to the oxides of zirconium and iron, which was consistent with the results reported by Dou et al. [31]. N 2 adsorption-desorption isotherms are shown in Figure 3(a), and the adsorption of N 2 on ZrMO X was type IV isotherm according to IUPAC [32]. There was a hysteresis loop at relative pressure (P/P 0 ) of 0.4~0.7, which indicated the presence of mesoporous structure. It can be seen the pore size distributions of ZrMO X were centered around 3 nm from Figure 3(b). It is clear from Table 1 that the BET-specific surface area and pore volume of Zr/Fe composite oxide were the largest among the four adsorbents. The rich specific surface area and pore volume provide sufficient locations for the adsorption of MO on ZrFeO X . Figure 4 shows the FT-IR spectrum of 1-573-ZFe. The absorption peaks at 3408 cm -1 and 1630 cm -1 were bending vibration and O-H tensile vibration of water molecules adsorbed by 1-573-ZFe, respectively [33]. The signals at 1385 cm -1 were attributed to the bending vibration of the Zr-OH groups [34]. The wide peak near 500 cm -1 was caused by the vibration of Zr-O and Fe-O bonds.

Characterization of ZrMO X .
As shown in Figure 5, the morphology for 1-573-ZFe was observed by SEM. It can be seen from the figure that the size distribution of the material was wide and amorphous, and there have many particles smaller than 2 microns on the solid surface of the adsorbent. There was no obvious pore structure observed on the surface of the adsorbent. Combined with the characterization results of N 2 adsorption-desorption, it can be well proved that there was no macroporous structure in the adsorbent.
3.6. Adsorption Kinetics. The kinetic data was obtained under the conditions of 303 K and initial concentrations of 50 mg·L -1 and 200 mg·L -1 MO solutions. Several classical kinetic models were used in this work, such as pseudo-first-order (PFO), pseudo-second-order (PSO), and Elovich (ELO) models presented in the following (Equations (4)-(6)) [35]: where q e and c e are as defined above; k 1 (min -1 ) and k 2 (g·mg -1 ·min -1 ) are the rate constants of pseudo-first-order and pseudo-second-order, respectively; a (mg·g -1 ) is the parameter about the surface coverage and adsorption free energy in the chemisorption; b (mg·g -1 ·min -1 ) is the initial sorption rate. Fitting parameters of PFO, PSO, and ELO kinetic models for MO adsorption are shown in Figure 6 and Table 2. It can be seen from Figure 6 that no matter the initial concentration was 50 mg·L -1 or 200 mg·L -1 , there was a fast adsorption rate in the initial stage of adsorption, and the adsorption was almost completed within 300 minutes. Combined with the chart, the kinetic models of PSO and ELO are more consistent with experimental data than PSO. It can be explained that the rate-limiting step was chemical adsorption [36] on the heterogeneous interface of 1-573-ZFe [37].
In general, the Weber and Morris model (W-M) was often applied to fit the kinetic data in order to deduce the rate-limiting step [32], expressed as Equation (7) [38]: where k W−M−n (mg·g -1 min -0.5 , n is the number stage) is the rate constant of intraparticle diffusion kinetic model; I n (mg·g -1 ) is a proportional constant to the boundary layer thickness. It can be seen from the multisegment fitting diagram in Figure 6(d) that the adsorption process of MO on 1-573-ZFe has undergone two stages of film diffusion and pore diffusion; Qian et al. [39] adsorption of haloforms on granular activated carbon and Liu et al. [30] adsorption of MO on FeOCl-MoS 2 both observed similar phenomena. Surface adsorption occurs first in the first stage, and the MO in the solution rapidly diffuses and adsorbs to the surface of 1-573-ZFe, resulting in a larger value of k W−M−2 . The second stage was the intraparticle diffusion. MO diffuses into the interior of 1-573-ZFe and was adsorbed on the internal adsorption active sites. The intraparticle diffusion stage was a gradual process, which can be obtained from k W−M−2 in Table 2. Since the fitted line did not cross the origin, it indicated that the intraparticle diffusion was not the rate-limiting step of adsorption [35].

Adsorption Isotherms.
Adsorption isotherm is an important factor of adsorption performance. Adsorption isotherm demonstrates the relationship curve between q e and c e at a certain temperature. Langmuir and Freundlich isotherm models were used to fit the equilibrium data of the temperature at 303 K, 313 K, and 323 K. The equations can be expressed as Equations (8) and (9) [40]: where K L (L·mg -1 ) and K F (L·mg -1 ) are the constants of Langmuir and Freundlich adsorption isotherm models, respectively; q m (mg·g -1 ) relates to the theoretical maximum adsorption capacity; 1/n is the adsorption intensity factor; n depends on the property, intensity in the adsorption process, and distribution of active sites. As shown in Figure 7(a), it can be seen that q e increased with the increasing of temperature, and it can be inferred that the adsorption process was endothermic [41]. Not only that, but it can also be seen that when the initial concentration is 20~60 mg·L -1 , q e was not significantly affected by temperature and remains almost unchanged. It can be considered that this material has a wide temperature operating range at low MO concentrations. From Figures 7(b)-7(d) and Table 3, it can be seen that Langmuir adsorption isotherm was better fitted to the adsorption process. This showed that the adsorption behavior of MO on 1-573-ZFe was a monolayer adsorption [42], and it was calculated that the maximum adsorption capacity of MO on 1-573-ZFe surface was 143.68 mg·g -1 .
3.8. Adsorption Thermodynamics. In order to obtain the change of energy in the adsorption process, thermodynamic experiments were carried out at the temperature of 303 K, 313 K, and 323 K, respectively. The changes of thermodynamic parameters of Gibbs free energy (ΔG), enthalpy (ΔH ), and entropy (ΔS) were calculated by using the following: where K is the thermodynamic equilibrium constant; T (K) is the thermodynamic temperature; R (8.314 J·mol -1 ·K -1 ) is the molar gas constant. ΔH (J·mol -1 ) and ΔS (J·mol -1 ) were calculated based on the linear relationship between ln K and 1/T. The thermodynamic parameters are listed in Table 4. It can be seen from Table 4 that ΔG was between -10 and 0 kJ·mol -1 , indicating that the adsorption process under the experimental conditions was a spontaneous physical adsorption process [43,44]. Moreover, ΔG decreased with the increasing of temperature, which further verified that the increasing of temperature improved adsorption. The adsorption enthalpy of MO on 1-573-ZFe was positive, indicating that the adsorption process was endothermic. This may be because the interaction of MO molecules with the adsorption sites on the surface of 1-573-ZFe was weaker than that of water molecules [44]. The adsorption of MO on 1-573-ZFe required the desorption of water molecules firstly, so the adsorption took place based on absorption of heat. The change of adsorption entropy was positive, indicating that the spontaneous adsorption of MO on 1-573-ZFe was an entropy-controlled process [45], and the randomness of the solid-liquid interface increases [37].
3.9. Regeneration Performance. The regeneration of the adsorbent is an important indicator to judge whether it can be used as an industrial adsorbent. After the regeneration experiment was finished, 1-573-ZFe was directly used for readsorption of MO. The result is shown in Figure 8. It can be seen from the figure that although the removal rate of MO gradually decreases with the increase of the number of cycles, the results of the fifth cycle regeneration test show that the removal rate of MO can reach 96.37%. Even after the sixth time, the MO removal rate can still reach more than 80%. These results demonstrated that 1-573-ZFe not only has good regeneration stability but also has good application potential in environmental remediation.
In Table 5, the adsorption capacities of MO were compared between this work and other reported adsorbents. The adsorbents were often modified through some processes to increase their adsorption capacity for MO. By comparison, it can be seen that the 1-573-ZFe adsorbent reported in this work has good adsorption capacity. Moreover, the preparation method is simple, the economic cost is low, and the energy consumption is low. It is an adsorbent with good commercial application value.

Conclusion
In summary, among the four ZrMO X adsorbents prepared by the coprecipitation method, through step-by-step screening, 1-573-ZFe showed best adsorption performance when the pH was between 4 and 6; this may be due to their high BET surface area (255.754 m 2 ·g -1 ) and pore volume (0.218 cm 3 ·g -1 ). It can be seen from the XRD and SEM characterization results that 1-ZFe is an amorphous composite metal oxide with irregular structure. The experimental data of kinetics and PSO and ELO can describe the dynamic adsorption process of MO on 1-573-ZFe, indicating that the adsorption of MO on 1-573-ZFe was a chemical adsorption control process that occurs on a heterogeneous surface. When the MO concentration is 0~60 mg·L -1 , the temperature change has almost no effect on the equilibrium adsorption capacity, and the removal rate of MO can reach more than 94%. The isotherm model indicated that MO has a monolayer adsorption on the 1-573-ZFe surface. The thermodynamic parameter ΔG was between -10 and 0 kJ·mol -1 , indicating that the process was a spontaneous physical adsorption. Both ΔH and ΔS were positive, explaining that the adsorption process was endothermic and controlled by entropy. The removal rate of MO can be over 96% after 5 times of regeneration, which explained 1-573-ZFe has superior regeneration performance. Therefore, these results indicated that 1-573-ZFe present excellent prospects on the removal of organic pollutants from wastewater.

Data Availability
The data used to support the findings of this study are included within the article.

Conflicts of Interest
The authors declare that they have no conflicts of interest.

Authors' Contributions
Ms. Yulong Sun contributed equally to this work and should be considered co-first author.