Removing Thallium (I) Ion from Aqueous Solutions Using Modified ZnO Nanopowder

In this study, the adsorption of thallium (I) ion as a dangerous pollutant from aqueous solution onto modified ZnO nanopowder as a fairly cheap adsorbent has been examined in batch mode. It was known that modification of the adsorbent was necessary to reach a significant adsorption percentage. The adsorbent used here was modified by sodium phosphate solution. The effect of experimental conditions such as initial pH of solution, contact time, adsorbent dosage, initial concentration of thallium and temperature is studied. The results showed the dependence of the adsorption percentage to these conditions specially its pH. The maximum adsorption percentage of Tl (I) ions at 25±1 o C was 92.8%. Freundlich isotherm model provided a better fit with the experimental data than Langmuir and Temkin isotherm models by high correlation. Separation factor, RL, values showed that modified ZnO nanopowder was favorable for the adsorption of Tl (I) ion. The negative value of ΔH 0 showed that Tl (I) sorption is an exothermic process and the negative value of ΔS 0

In recent years, nano materials have been also used as adsorbents due to theirs special properties. Their surface has a lot of unsaturated atoms that can easily attach to other atoms, thus they have a high capacity for the adsorption. The examples of nanocompounds for the adsorption of thallium ions are nano-Al2O3 [22], nano-TiO2 [23] and multiwalled carbon nanotubes [24].
In this paper, ZnO nanopowder was used as an adsorbent to remove Tl (I) ions from aqueous solutions, for the first time.
In nano materials, nano-ZnO is abundant and a cheap adsorbent. It was known that net charge of ZnO nanopowder in aqueous solution is positive, thus surface modification is essential to obtain negatively charged ZnO nanopowder [25]. We accomplished the modification by sodium phosphate solution to adsorb Tl (I) ions.

Apparatus
An Atomic Absorption Spectrometer Model AA220 (VARIAN Co., USA) to determine the concentration of Tl (I) ion and a 420A model pH meter (ORION Co., USA) was used to measure the pH of the solutions . A TDL80-2B model centrifugal machine (Shanghai Anting Scientific Instrument Co., China) was used to separate the adsorbent from the mixture. We also used Ultrasonic Bath (71020-DTH-E, Model 1510 DTH, 220 v, EMS Company) for agitation and to prevent aggregation.

Modification of adsorbent
For adsorbing position thallium ions onto the ZnO nanopowder, it should be modified by a suitable anion to obtain the negative surface charge. Definite amount of ZnO nanopowder (0.05 g) was suspended into 10 ml from the various compound solutions (the suitable compound was tri-sodium phosphate). The mixture was then placed in the ultrasonic bath for 60 minutes for agitation and to prevent aggregation. The modified ZnO nanopowder was then used for the adsorption experiments.

Adsorption experiments
The adsorption experiments were done in 100 mL flasks containing 50 mL of Tl (I) solution of known concentration (10 mg L -1 ) obtained from the dilution of 250 mg L -1 stock solutions. The initial pH of the solution was adjusted to optimum value (pH 6) with 0.1 M HCl or 0.1 M NaOH solution. The mixture was then added to each flask and it was placed in the ultrasonic bath for 60 min for agitation and to prevent aggregation. After completion, the sample was removed from the flask and was separated by centrifuging at 4000 rpm for 5 min. The obtained solution was analyzed for residual Tl (I) ion. Then, the adsorption percentage (% Adsorption) was obtained as where Ci and Cf are the initial and the final concentrations of Tl (I) ions in solution, respectively.
The amount of Tl (I) ions adsorbed per unit mass of adsorbent (qe) was obtained by following equation: J u n e 1 1 , 2 0 1 5 where Ci and Ce are initial and equilibrium concentrations Tl (I) ions (mg L -1 ), respectively and V is the volume of the solution (L) and m is the mass of the adsorbent (g).
The average absolute value of relative error, AARE, compares the predicted results with experimental data. This is defined as follows:  (3) in which N is the number of data points.

Modification of the adsorbent by various compounds
First, we modified ZnO nanopowder (as adsorbent) by various compounds and then, we studied the adsorption percentage of Tl (I) ion in them and we selected the adsorbent gained by higher adsorption. These compounds were: 1M NaOH, 2M NaOH, 0.1M HCl, 0.1M NaCl, 1% sodium citrate and 1% tri-sodium phosphate. As seen in Figure 1, the adsorbent modified by sodium phosphate gave the best adsorption. Upon modification, PO4 3ions attach to the adsorbent sites and subsequently to the adsorbent better than other ions, due to more aggregation of the negative charge. This favored Tl (I) ion adsorption onto the surfaces of the adsorbent.

Modification of the adsorbent by sodium phosphate solution
The adsorbent was modified by 1-15% w/v sodium phosphate solutions and adsorption process was done again. The results given in Figure 2 showed that the modified adsorbent by 5% w/v sodium phosphate solution is a more suitable adsorbent. Extra PO4 3ions disturb the desirable adsorption.

The effect of solution initial pH on adsorption percentage
The solution initial pH is the major parameter controlling sorption processes. For study this parameter, the solution initial pH was varied in the range of 5-11. At pH<5, ZnO nanopowder is dissolved in the solution. As seen in Figure 3, Tl (I) sorption on ZnO nanopowder is increased from pH=5 to pH=6 and then it is drastically decreased by increasing pH. The satisfactory Tl (I) sorption occurred at pH=6. When the pH increase, with increasing OH ions, they compete with PO4 3-ions for adsorption of the positive thallium ions and this deteriorate the adsorption process.

The effect of sorbent dosage
The adsorption percentage of Tl (I) ion onto the modified ZnO nanopowder affected the dosage of the sorbent ( Figure 5), while the other conditions maintained constant. The adsorption percentage increased with an increase in sorbent dosage from 0.03 g up to 0.1 g and then, it remains constant nearly. This may probably be due to increased adsorbent surface area and availability of more adsorption sites. The optimum sorbent dosage was 0.1 g.

The effect of temperature
Temperature is the other parameter that affect to the adsorption percentage. Thus, temperature was varied within the range of 25-60 0 C, while the other conditions were constant. As seen in Figure 6, the adsorption percentage decrease with an increase in temperature. It shows that the adsorption is an exothermic process. This agrees with the process enthalpy (ΔH 0 = -6.92 kJ mol -1 , Table 3). The suitable temperature was 25 o C.

Adsorption isotherms
In order to find the suitable model for the adsorption of Tl (I) ion onto modified ZnO nanopowder, the experimental data were correlated by Langmuir, Freundlich and Temkin models. The related linear equations are: where Ce (mg L -1 ) is the liquid phase concentration of Tl (I) ions at equilibrium; qe (mg g -1 ) is the amount of Tl (I) ions adsorbed per unit mass of adsorbent at equilibrium; KL (L mg -1 ) is the Langmuir isotherm constant; qm (mg g -1 ) is the maximum sorption capacity of Langmuir model; KF (mg 1-(1/n) L 1/n g -1 ) is the Freundlich constant, and n is the heterogeneity factor; B1 is a constant that it depend on the adsorption heat and KT (L g -1 ) is the Temkin isotherm constant. We correlated the adsorption data at different initial concentrations of Tl (I) ion in terms of the Langmuir isotherm (Eq. (4)).
The curve of 1/qe versus 1/Ce gave a straight line with a slope of 1/KLqm and intercept of 1/qm (Figure 8 a). We also examined the data according to the Freundlich isotherm (Eq. (5)). The plot of log qe versus log Ce gives a straight line with slope 1/n and intercept of log KF (Figure 8 b). We also studied the data based on the Temkin isotherm (Eq. (6)). The plot of qe against lnCe gave a straight line with a slope of B1 and intercept of B1lnKT (Figure 8 c). As it was seen, the experimental data have a better agreement with the Freundlich isotherm than those of Temkin and Langmuir because the values of regression coefficient and AARE % for the Freundlich isotherm are higher than Temkin and Langmuir. Table 1(a, b, c) shows the parameters of the Langmuir, Freundlich and Temkin models and theirs regression coefficients and AARE%.  Constant separation factor, RL, determines the essential characteristics of the Langmuir isotherm by Eq. (7) [26]: Where KL is the Langmuir constant and Ci is the initial concentration of Tl (I) ion in solution.
This factor illustrates the shape of the isotherm and the nature of the adsorption process as below: RL value Nature of the process The calculated RL values against initial Tl (I) concentration were shown in Figure 9. The value of RL in the range of 0-1 at all initial Tl (I) concentrations shows the favorable adsorption of Tl (I) ion.

Adsorption Thermodynamics
Thermodynamic analysis helps us obtain useful information on energetic changes during adsorption process. The change in free energy (ΔG 0 ) can be calculated as: Where R (kJ mol -1 K -1 ) is the gas constant, T is the absolute temperature, and K0 (L mol -1 ) is the thermodynamics equilibrium constant. The equilibrium constant (K0) can be defined as [27]:  Tables 2 and 3, respectively. The negative values of ΔG 0 indicate that the Tl (I) adsorption process onto modified ZnO nanopowder could occur spontaneously. Generally speaking, the heat released during physical adsorption has the same order of magnitude for the heat of condensation, i.e., 2.1-20.9 kJ mol -1 [28], while the heat of chemisorption changes in the range of 80-200 kJ mol -1 [29]. Therefore, it seems that Tl (I) adsorption onto modified ZnO nanopowder could be attributed to a physical sorption process. The negative value of ΔH 0 shows that Tl (I) sorption is an exothermic process. The low and negative value of ΔS 0 may state that the entropy have not changed significantly during the adsorption but there is a little decrease of randomness at the solid-solution interface during sorption.

Comparing nano-adsorbents
In this study, several nano-adsorbents have been compared and results were shown in Table 4. It can be seen that the highest values of the maximum adsorption capacity, qm, are related to the modified ZnO nanopowder. Among the nanoadsorbents, nano-ZnO is an abundant and cheap adsorbent. It seems that, among the nano-adsorbents, the modified ZnO nanopowder is the most suitable nano-adsorbent for removing Tl (I) ions from aqueous solutions. ) Reference nano-Al2O3 6.28 [22] nano-TiO2 4.09 [23] multiwalled carbon nanotubes 0.658 [24] nano-ZnO 6.95 current work

CONCLUSION
In this study, we showed that ZnO nanopowder, after modification with sodium phosphate solution, can be used as an adsorbent to adsorb Tl (I) ion from aqueous solution. We showed that 5% w/v sodium phosphate solution gave the better adsorption percentage. Experimental data indicated that the adsorption percentage was depended on parameters such as initial pH of solution, contact time, dosage of adsorbent, temperature and the initial concentration of thallium. The optimum conditions were: pH 6; contact time, 60 min; adsorbent dosage, 0.1 g; initial concentration of thallium, 50 mg L -1 and temperature, 25 o C. Under these conditions, the maximum adsorption percentage of Tl (I) ion on modified ZnO nanopowder obtained was 92.8%. Freundlich isotherm model provided a better fit with the experimental data than Langmuir and Temkin isotherm models by high correlation. The RL values showed that the modified ZnO nanopowder was preferred for the adsorption of Tl (I) ion. The negative value of ΔH 0 shows that Tl (I) adsorption is an exothermic process. The low and negative value of ΔS 0 may state that the entropy was not changed significantly during the adsorption but there was a little decrease in randomness at the solid-solution interface during sorption. Among several nano-adsorbents, nano-ZnO is a more suitable adsorbent because it is abundant and cheap and its maximum adsorption capacity is higher than others. J u n e 1 1 , 2 0 1 5