Separation of coii and sevi from a metal and glass industry wastes using graphene oxide-mangane oxid

2.1. Preparation of Nanocomposites

2.1.1. Synthesis of Graphene Oxide-Mangane Oxide nanocomposite Graphene oxide-mangane oxide was produced from graphite powder using the method proposed by Hummers [24]. 2 g of graphite was put to 100 mL of H2SO4 in an box containing ice. 7.0 g of KMnO4 was put to the mixture and stirred for 4 h at 37 °C. Deionized water was spreyed slowly to the luquid which its temperature is 90 °C. Then, 35 mL of 6% H 2O2 was added to the mixture; and the liquid was filtered with a 0.05 μm pore diameter Millipore membrane. This mixture was washed with 15% HCl and distilled water. Graphene oxide was obtained by using a sonicator. The solid phase was filtered and washed with distilled water. The solid phase was graphene oxide. It was dried and mixed in the deonized water at a concentration of 15 mg/L KMnO4 and is put to the graphen oxide at a volume ratio of 0.05%. This is defined as graphene oxide-mangane oxide nanocomposite. This mixture was put to the sonicator at a power of 89 W cm−1 for 30 min. Then, the mixture was filtered and washed with distilled water and methanol. The produced product is graphene oxide-mangane oxide nanocomposite. 2.1.2. Batch Adsorption Studies Adsorption experiments were performed in 150 mL glass flasks. 0.01 g graphen oxidemangane oxide was put into 40 mL Se(VI) or Co(II) containing wastewater. The pH was adjusted to 8.5 using 0.1 M NaOH. The flasks were shaken at 278 K at 250 rpm for 22 hours using a shaker. For adsorption kinetics; 0.1g graphene oxide-mangane oxide was shaken into 500 mL industry wastewater containing Se (VI) or Co(II) to determine the adsorption kinetics. 2.1.3. Physicochemical Analysis of Graphene Oxide-Mangane Oxide Nanocomposite The specific surface area and pore size distribution of the graphene oxide, mangane oxide and graphene oxide-mangane oxide nanocomposite were done using a BET surface area analyzer (ASAP2020, USA) by the N2 adsorption–desorption tests at 89 K. The surface properties of graphene oxide-mangane oxide was detected by a scanning electron microscope (SEM, S-3400 N, Japan) and a high-resolution transmission electron microscopy (HR-TEM, Tecnai G2 F30 STwin, USA) containing a field distribution apparatus at 197 kV. Scanning transmission electron microscopy (STEM) analysis was conducted to obtain the distribution of some metals into graphene oxide-mangane oxide. To analyze the organic and inorganic characteristics of the graphene oxide-mangane oxide; Fourier transform infrared spectroscopy (FT-IR) (Nicolet Model 810 FT-IR instrument-Nicolet analytical Instruments, WI) device was used. The crystallinity of the graphene oxide-mangane oxide was investigated by X-ray diffraction (XRD) with Kα radiation and scan rate of 0.04° s −1 PRO, USA). The IR Raman spectra of the samples were detected by Horiba JobinYvon, France at λ=532 nm laser power 1.7 mW, 100 x objective lens, 0.9 NA. The energy-dispersive X-ray spectroscopy (EDS) analysis was carried out on Bruker EDS QUANTAX. The thermal stability of the graphene oxide-mangane oxide was examined by thermogravimetric experiments (TGA) (Perkin–Elmer Diamond TG analyzer). Zetasizer Nano ZS 90 was used to measure the zeta potential of graphene oxide-mangane oxide. X-ray photoelectron spectroscopy (XPS, Kratos AXIS Ultra DLD, and Germany) was used to detect the situation of the adsorbed material onto graphene oxide-mangane oxide. The high-resolution Cu 2p3/2 peak was fitted by a curve-fitting program (XPS-peak4.1). The surface area (SBET) was calculated by usng the Brunauer–Emmett–Teller (BET) equation. The total hollow volume (Vp) was detected at P/P0 = 0.98 and the adsorption mean hollow size (4 V/A) was determined by BET. Micropore volume

(V micro) and the external area (Sexternal) were calculated by the t-method using an suitable standard.

2.2. Analytical Procedure

2.2.1. Se(VI) Measurement Method 1 mL of glass industry and metal industry wastewater samples was separately digested with 5 mL HNO3 and H2O2 at an oven at 250°C for 20 min and at 250°C for 15 min. After cooling the samples were diluted with 1% HNO3 and analyzed by ICP-MS. The limit of detection for Se(VI) was 0.39x10− 2 mg/L. 2.2.2. Co (II) Measurement Method An atomic absorption spectrometer (PG-990, USA) with deuterium and cobalt hollow cathode lamps was utilized for detect the cobalt at a wavelength of 251 nm. The pH of 100mL samples having Co(II) at levels varying between 0.01 and 10 mg/L was adjusted to 2.00 using 3.0x10 −2 mol/L HCl. Then the samples were centrifuged and 0.9 mL of 2-2-Methylimidazole and 0,8 mL Tetrafluoroborate was added into the supernatants. They were stirred and then, 6.0 mL of hexafluoride containing 0.01 g sulfur, and 0.5 mol/L sodium hexafluorophosphate was added. The samples were centrifuged for 20 min at 3800 rpm to separate the solid and liquid phases [25].

2.3. Co(II) and Se(VI) Concentrations in the Glass and Raw Metal Industries

The Co(II) and Se(VI) concentrations concentrations in the raw metal concentrations were about 2.5 mg/L.

2.4. Statistical Analysis

Regression analysis is used to understand the independent variables are related to the dependent variable. Alpha (α) level is the significance of the Analysis of Variance (ANOVA) statistic. In the study α was accepted as 0.05. F value of the analysis was performed using MS Office 2010 Excell program.

3.1. Characterization of Graphene Oxide-Mangane Oxide Nanocomposite with SEM

Fig. 1 shows the micro-morphology of 5 mg/L graphene oxide-mangane oxide nanocomposite. Graphene oxide-mangane oxide exhibited a lamellar and a wrinkled morphology. The MnO was doped to the graphene oxide surrounding. A structure with a smooth surface was detected in graphene oxide. A deposition of MnO was detected in the upper layer of the graphene oxide-mangane oxide nanocomposite since the surface of the graphene oxide-mangane oxide is smooth.

3.2. FT-IR Spectra and IR Raman Spectra of Graphene Oxide-Mangane Oxide

The FT-IR spectra of graphene oxide-mangane oxide are given in Fig. 2a. Hydroxylic, carboxylic, and carbonylic groups were detected in the surface of graphene oxide-mangane oxide nanocomposite. The maximum peak observed at around 3250 cm−1. This verify the hydroxylic groups in graphene oxide The C–C peak was detected at 1600 cm−1. A C-O-C peak was detected at 1200 cm−1 bonding to carbonylic/carboxylic double C O with C–O in 1250 cm−1 and 1700 −1 −1 cm . The detection of an absorption peak at 501 cm describe the Mn-O vibration and showed that MnO was bounded to the surface of graphene oxide [26]. The IR–Raman spectra of graphene oxide and graphene oxide-mangane oxide are illustrated in Fig. 2b. The IR spectrum of graphene oxide showed characteristic peaks at 250, 400, 607, 1300 and 1700 cm−1. In the compairison of the two peak bands at graphene oxide and graphene oxidemangane oxide; the band spectrum of graphene oxide-mangane oxide is lower, showing the formation of the chemical bonds between graphene oxide and MnO.

Figure 2a. FTIR spectra of Graphene Oxide-mangane Oxide (GO-MO) and MnO

Figure 2b. IR Raman spectra of Graphene Oxide-mangane Oxide (GO-MO) and MnO

Figure 2. The FTIR spectra (a) and Raman spectra (b) of graphene oxide and graphene oxide-mangane oxide nanocomposte (GO-MO)

3.3. XRD Characterisation of Graphene Oxide-Mangane Oxide

Fig. 3 showed the XRD spectra of graphene oxide-mangane oxide. Graphene oxide-mangane oxide spectra illustrated that the intensity of graphene oxide peaks decreased. The XRD spectra of graphene oxide-mangane oxide samples showed the maximum specific peaks at 2θ = 12,2°, 18,5°, 37,11° and 65° and indicates the existance MnO at a Birnessite type. This shows the presence of MnO on graphene oxide-mangane oxide. All peaks have low intensity, showing the presence of very small crystallized particles. In the position at 2 θ ~ 12-13° and 36-38°, there are two obvious hint of graphene oxide peak (Fig. 3). They were poorly crystallized in the surface of graphene oxide-mangane oxide.

3.4. EDS Analysis

EDS analysis results showed that Mn mass ratio was detected as 17.40 Wt % in graphene oxide-mangane oxide while no pure Mn element was detected in graphene oxide. This showed that the combination is between graphene oxide and MnO (data not shown). Furthermore, the graphene oxide-mangane oxide sample contains oxygenophilic functinonal groups (data not shown). This result confirmed that MnO has been coated on the graphene oxide sheets. The results of the surface area and pore structure of graphene oxide is given in Table 1. Graphene oxide has the smallest surface area (32 m2/g) and pore volume (0.11 cm3/g), as reported by Li et al., (2013) [27]. The average pore width of graphene oxide is 17.3 nm.This showed that graphene oxide is a mesoporous material. The mesoporous morphology of graphene oxide and its small surface area can be explained by the accumulation of graphene oxide during drying process at 100 0 C. This can be explained by the van der Waals force between graphene oxide metal oxides. Table 1. Properties of graphene oxide sample: surface area (SBET), total pore volume (Vp), the average pore width (L) Sample

3.5. Zeta Potential of the Graphene Oxide-Mangane Oxide Nanocomposite

The variation of the zeta potentials of graphene oxide, mangane oxide and graphene oxidemangane oxide nanocomposite versus pH is illustrated in Fig. 4. The aforementioned nano-metal 34

oxides and graphene oxide- mangane oxide nanocomposite have enlarged their negative zeta potentials (≤15.8 mV) in the pH values varying between 2 and 9. This showed that the particles did not aggregate and were dispersed homogeneously in the wastewater. The negative zeta potential shows that the acidity of nanocomposite surface containing oxygenated functional groups as reported by Amir et al., (2017) [28].

Figure 4. Zeta potential variations versus pH of graphene oxide, mangane oxide and graphene oxide-mangane oxide nanocomposite

3.6. HR-TEM Images of Graphene Oxide, Mangane Oxide and Graphene Oxide-Mangane

Oxide Nanocomposite The HR-TEM image of graphene oxide-mangane oxide nanocomposite is illustrated Fig. 5. In this figure it was observed that the mangane oxide was well dispersed on the graphene oxide. It was showed that the particle size of mangane oxide was 6 nm. With doping of the graphene oxide; the graphene oxide-mangane oxide nanocomposite exhibited crystal morphology and the calculated lattice void was 0.188 nm. This occurred with the distribution of mangane oxide nanoparticles resulted in big BET surface area of graphene oxide-mangane oxide nanocomposite (356.33 m2/g), which is an order magnitude higher than that of graphene oxide (32 m2/g) and dispersed mangane oxide (30.98 m2/g). The BET properties of the graphene oxide and graphene oxide-mangane oxide nanocomposite are summarized in Table 2. The results agree with the findings of Jasinski et al., (2013) [29].

Figure 5. HR-TEM images of graphene oxide, mangane oxide and graphene oxide-mangane oxide nanocomposite Table 2. BET surface area of graphene oxide, mangane oxide and graphene oxide-mangane oxide nanocomposite graphene oxide 32

BET surface area (m2/g) mangane oxide graphene oxide-mangane oxide nanocomposite 30,98 356.33

3.7. XPS Analysis of Graphene Oxide, Mangane Oxide and Graphene Oxide-Mangane

Oxide Nanocomposite XPS was performed to detect the variations in the structure of the graphene oxide-mangane oxide nanocomposite during adsorption of the Se (VI) and Co(II). Fig. 6a shows extensively the XPS spectrum of the generated graphene oxide-mangane oxide nanocomposite. This figure shows the O 1s XPS spectra of mangane oxide-graphene oxide nanocomposite contained MnO2. In this nanocomposite; Mn, O, and C elements are detected. The XPS spectra (Fig. 6b) shows two peaks located at 640.9 and 651.9 eV. This originated with the presence of Mn 2p 3/2 and Mn 2p1/2 of Mn4+. The maximum spectra exhibited similarities with the studies reported for MnO 2, indicating mangane is oxidated to Mn4+ [30]. The O 1s peak generated from two oxygenated groups. One of them is O2− spectra of MnO2 at a binding energy of 549.6 eV. The other one is an OH − peak of manganese hydroxide. All these functional groups of graphene oxide were binded at a binding energy of 530.2 eV as reported by Pan et al., 2016 ( Fig. 6c) [31]. The binding energy for Mn 2p of Graphene oxide M2 increased during the adsorption of Se(VI) and Co(II) (Fig. 6c). This can be explained by to the formation of covalent Mn-O-Se or Mn-O-Co bonds by decreasing of extranuclear electron density around of the Mn as reported by Park et al., (2016) (data not shown) [30].

Figure 6c. Figure 6. XPS spectra of a) narrow-scan O1s in graphene oxide-mangane oxide nanocomposite b) narrow-scan Mn 2p in graphene oxide-mangane oxide nanocomposite c) O2− and OH− peaks in graphene oxide-mangane oxide nanocomposite

3.8. Scanning Transmission Electron Microscopy (STEM) Analysis of Mangane Oxide,

Graphene Oxide and GO-MnO Nanocomposite In the STEM image, GO and MnO can be easily distinguished in Fig. 7a. It seems that the white is MnO only dispersed around the surface of GO (black particles). The small MnO nanocrystals dispersed around the GO-MnO nanoparticles form an oxide-on-oxide nanostructure. The grey ones are the GO-MnO Nanoparticles. As shown in Fig. 7b, the lattice spacing of a MnO nanocrystal was measured to be 1.47 nm, while the lattice spacing of the oxide nanocrystal adjacent to the GO nanoparticle was measured to be 1.49 nm, consisting to the (101) lattice plane of tetragonal spinel MnO (111). This image clearly shows the formation of the oxide-on-oxide nanostructures through close particle-on-particle contact between two oxides in this composite. As depicted from Fig. 7c, the formation of the special nanostructure may result from the epitaxial growth of MnO along the exposed (111) lattice plane of the GO. Since graphene oxide having a planer hexagonal structure, the spinel MnO can be rationalized to preferentially deposit on the surfaces of graphene through epitaxial growth of MnO as reported by Chou et al., (2014) [32].

Figure 7c. Figure 7. STEM picture of Mangane Oxide, Graphene Oxide and Mangane Oxide doped Graphene Oxide(a,b,c)

3.9. Termal Stability of Graphene Oxide, Mangane Oxide and Graphene Oxide-Mangane

Oxide Nanocomposite- TGA analyss Thermal stability of the graphene oxide-mangane oxide nanocomposite was investigated by TGA (Fig. 8). The samples showed three-step weight losses. Weight loss up to 98 °C temperature can be explained by the losses of water moieties bounded on the graphene oxide layers. Weight losses was detected at a temperature of 189 °C as reported by She et al., (2015) [33]. Lowering of the weight above 280 °C can be attributed to the thermal splitting of the carbon in the graphene oxide. Increasing of the thickness of the mangane oxide cause to the growth of the mass moieties (data not shown).This, is consistent with recent reports explaining that mangane oxide is thermally in equilibrium up to 1500 °C temperature [34].

Figure 8. TGA lines of graphene oxide-mangane oxide nanocomposite

3.10. Effect of Increasing Graphene Oxide-Mangane Oxide Nanocomposite Concentrations

on the Adsorption Yields of Se(VI) and Co(II) The studies were performed at 2 mg/L graphene oxide-mangane oxide nanocomposite concentration since in the previous study the maximum adsorption yields of Se(VI) and Co(II) were obtained with this adsorbent concentration at a room temperature of 21 oC after 25 min (data not shown). As the graphene oxide-mangane oxide composite concentrations were elevated from 1 mg/L up to 4 mg/L the the adsorption yields of Se(VI) increased from 57% up to 98% (Table 3). Further increase of composite concentration to 5, 6 and 7 mg/L did not affect the Se(VI) adsorption yields. The adsorption yields remained maximum at a 4 mg/L graphene oxidemangane oxide composite at 21 oC. Similar results was found for Co(II). The maximum adsorption yields for Co (II) was obtained at 4 mg/L graphene oxide-mangane oxide composite concentration (Table 3). Increasing of the composite concentration did not affect the adsorption yield of Co(II). These results showed that the adsorption yields of both metals depend to the nanocomposite concentration up to 5 mg/L. This can be explained with elevated ion exchange capability of the graphene oxide-mangane oxide nanocomposite. The results showed that Co(II) and Se(VI) adsoption yields increases up to an optimum concentration in which the removal yield has no change with the graphene oxide-mangane oxide composite concentration as reported by Somiya et al., (1988) and Matias et al., (2015) [35], [36]. As known, the equilibrium concentration of graphene oxide-mangane oxide nanocomposite lowered with elevating nanocomposite concentrations. At high nanocomposite concentration the unit adsorption decrease due to adsorption sites remaining unsaturated during the adsorption reaction whereas the number of sites available for adsorption site increases by increasing the adsorbent dose as reported by Somiya et al., (1988) and Matias et al., (2015) [35], [36].

Table 3. Effects of increasing graphene oxide-mangane oxide nanocomposite concentrations on the adsorption yields of Se(VI) and Co(II) levels at 21 oC after 25 min at increasing graphene oxide-mangane oxide composite concentration Graphene oxidemangane oxide Concentration (mg/L)

A multiple linear relationship between maximum adsorption yields of Se(VI) and Co(II) and graphene oxide-mangane oxid nanocomposite concentrations up to 5 mg/L graphene oxidemangane oxide nanocomposite was obtained (R=0.93) and this regression was significant (ANOVA p=0.005<α (0.05) and F=1.28). Further increase of graphene oxide-mangane oxide nanocomposite concentration did not affect both Co(II) and Se(VI) adsorptions.

3.11. Effects of Contact Time on the Adsorption Capacities of Co(II) and Se(VI)

The studies were performed at 4 mg/L graphene oxide-mangane oxide composite concentration to determine the effects of adsorption time on the adsorption yields of Se(VI) and Co(II) at 21 oC. Table 4 shows the effect of contact time on the adsorption capacities of Se(VI) and Co(II) by 4 mg/ L graphene oxide-mangane oxide nanocomposite. Table 4. Effects of contacting time on the adsorption capacities of Co(II) and Se(VI) at 4 mg/ L graphene oxide-mangane oxide nanocomposite at 21 oC Graphene Contacting OxideTime for Mangane Oxide Adsorp Concent tion (min) ration (mg/L) 4 20 4 30 4 50 4 60 4 70 4 80

Se(VI) Co(II) adsorption Concen capacity tration (mg/g) (mg/ L) 289 270 259 239 200 160

The adsorption capacity of nanocomposite increases with time and attain equilibrium within 20 min for both metals. The equilibrium time was dependent up to 30 min adsorption. Then, the increase of contacting time did not affect the adsorption capacities of both metals. It was found that Se(VI) adsorption capacity (289 mg/g) was higher than Co(II) adsorption capacity (256 mg/g) after 20 min adsorption/contacting time at 4 mg/L nano graphene oxide-mangane oxide nanocomposite. The metal absorption versus time are simultaneously elevated the saturation

gradient. This can be explained by the possible monolayer bounding of metal ions on the surrounding of the adsorbent up to a contacting time of 20 min. Later, the lower adsorption rate at high contact times is due to a decrease in number of vacant sites of nanocomposite and metal concentrations. The decreased adsorption rate, particularly, toward at high contact times, indicates the possible monolayer formation of metals on the adsorbent surface. This may be attributed to the lack of available active sites required for further uptake after attaining the equilibrium. A multiple linear relationship between maximum adsorption yields of Se(VI) and Co(II) and adsorption times up to 30 min was obtained (R=0.93) and this regression was significant (ANOVA p=0.003<α (0.05) and F=1.06). Further increase of contacting time did not affect both Co(II) and Se(VI) adsorption yields.

3.12. Effect of pH on the Adsorption Yields of Co(II) and Se(VI)

The effect of pH on the Co(II) and Se(VI) adsorption at 4 mg/L graphene oxide-mangane oxide nanocomposite at 21oC at 20 min contacting time for pHs between 2.2 and 10.0 is presented in Table 5. It can be found that the adsorption yields increased with pH for both metals. The uptake of Co(II) and Se(VI) by graphene oxide-mangane oxide nanocomposite increased as the pH increased from 2.0 to 6.1. At higher pH values (8.0-8.9) the adsorption efficiency increased for both metals. As the pH of the samples elevated to >9.3; Co(II) started to accumulated in the solution. Therefore, the elevated mass of adsorption at pH = 8.9 may be a synergistic effect of both adsorption and accumulation on the monolayer surrounding of the graphene oxide-mangane oxide nanocomposite. It is considered that graphene oxide-mangane oxide had a maximum adsorption capacity at a pH of 8.9. Therefore, the optimum pH for maximum Co(II) and Se(VI) adsorption is 8.9. Similarly, Se(VI) adsorption on the graphene oxide-mangane oxide composite tends to elevate with the increasing of pH. This can be explained by the positively charges occured at low pHs. Therefore, the adsorption of Se(VI) at lower pH values resulted from an increased repulsion between the more positively charged Se species and positively charged surface sites. Furthermore, at lower pH, H+ ions were competitive with Se(VI) ions to binding on the surface of the nanocomposite [37]. The adsorption efficiencies of Se(VI) and Co(II) increased from 43% to 87% and up to 98% as the pH were increased from 2 to 4 and up to 8.9. In our study adsorption efficiency was directly dependent to the pH. For the graphene oxide-mangane oxide nanocomposite, the adsorption at pH above 9.5 shows a reduction due to the formation of hydroxylic functional groups of cobalt and selenium such as Co(OH) 2 ) and (Se(OH)2 ) [38]. Table 5. Effect of pH on the adsorption efficiency of Se(VI) and Co(II) at 4 mg/L graphene oxide-mangane oxide nanocomposite after 20 min adsorption time Graphene pH Oxide-Mangane Oxide Concentration (mg/L) 4 2.2 4 4.5 4 6.2 4 8.9 4 10.0

Se(VI) Se(VI) Co(II) Removal adsorption Concen efficiency capacity tration (%) (mg/g) (mg/ L) 56 69 72 98 87

Co(II) Co(II) Removal adsorpt efficiency ion capacity (%) (mg/g) 53 67 70 98 80

3.13. Adsorption Properties and Kinetics of Se(VI) and Co(II) with Constant Graphene

Oxide-Mangane Oxide Nanocomposite The adsorption of Se(VI) and Co(II) onto 4 mg/L graphene oxide-mangane oxide nanocomposite reached equilibrium within 20 min at 21 oC at pH=8.9. The adsoption kinetic data for both 1.5 mg/L Se(VI) and Co(II) were described by the pseudo-first order kinetic models for graphene oxide-mangane oxide, respectively (Fig. 9). The Se(VI) and Co(II) adsorption on the surface of graphene oxide-mangane oxide before reach to equilibrium can be explained by two steps: at the beginning an initial quik phase (0–4 min) followed by a slow phase (4–6 min). The low phase was originated from the pore diffusion within graphene oxide-mangane oxide as reported by Wan et al., (2018) [39].

Figure 9a. Adsorption kinetic of 1.5 mg/L Se(VI) (a) onto 4 mg/L

Figure 9b. Adsorption kinetics of 1.5 mg/L Co(II) (b) onto 4 mg/L

Figure 9. Adsorption kinetics of 1.5 mg/L Se(VI) (a) and 1.5 mg/L Co(II) (b) onto 4 mg/L graphene oxide-mangane oxide at pH = 8, 9, and at a temperature of 21 oC.

3.14. Recoveries of Se(VI) and Co(II)

The adsorption efficiencies in both metals were not reduced significantly after four sequential utilization of 4 mg/L graphene oxide-mangane oxid nanocomposite within 20 min contacting time (Table 6). In this research, the adsorption yields of Se(VI) and Co(II) were 99% and 98% in the first utilization. After four sequential utilization of the same graphene oxide-mangane oxide nanocomposite the adsorption yields of Se(VI) and Co(II) decreased slightly from 90 and to 86 %, respectively. Table 6. Sequential treatment of Se(VI) and Co(II) with 4 mg/L graphene oxide-mangane oxide nanocomposite concentration, after 20 min adsorption time at pH 8.9 and at a temperature of 21oC.