Optimization, equilibrium and kinetic studies of Zn²⁺ and Ni²⁺ adsorption from aqueous solutions using composite adsorbent

The increased concentrations of heavy metals in the environment are a global problem. The World Health Organization (WHO) is concerned about drinking water that contains heavy metals like aluminum, manganese, chromium, cobalt, iron, nickel, zinc, copper, mercury, lead, and cadmium (WHO 2011). A part of that, Zn²⁺ and Ni²⁺ are among the most highly toxic heavy metals. Even at extremely low concentration, Ni²⁺ presents an environmental threat and causes cancer of lungs and nasal sinus (Ahmedna et al. 2004). On the other hand, Zn²⁺ is important for humans in small quantities, but it can affect health when the prescribed limit is exceeded. The World Health Organization (WHO 2011) has limited the concentrations of Zn²⁺ and Ni²⁺ to 3 mg/L and 0.07 mg/L, respectively. Most Zn²⁺ and Ni²⁺ enters the environment from various industrial effluents including cadmium–nickel batteries, lead and cadmium ores, purifying zinc, coal burning and burning of wastes, steel production, phosphate fertilizers, mining, alloy, pigments, and stabilizers (Low & Lee 1991).

There are many techniques available for heavy metals' removal such as adsorption reverse osmosis, solvent extraction, ion exchange and precipitation. Among these, adsorption using activated carbon is a well-known technique for heavy metals' removal, but the high cost of this technique limits its large-scale application in developing countries. In addition, many other adsorbents have been developed from different industrial by-products for heavy metal removal. However, there is still the need to investigate the capability of low cost and efficient adsorbents, especially for the adsorption of Zn²⁺ and Ni²⁺ from aqueous solutions (Zwain et al. 2014).

Apart from producing heavy metal-containing wastewater, industrial activities also leave behind a great deal of solid waste in ash form. Some naturally occurring agriculture by-products may serve as low cost adsorbents for heavy metal removal because they are widely available in large quantities, represent unused resources, and are environmentally friendly. However, the adsorption capacity of adsorbents made of agricultural by-products is usually less than synthetic adsorbents, but these materials may offer cost-effective alternative technique for heavy metals' removal from water and wastewater (Zwain 2012; Kim & Lee 2015).

Many attempts have been made to utilize waste materials as an alternative adsorbent, especially waste-derived siliceous materials such as rice husk ash (RHA), palm oil fuel ash (PFA), and coal fly ash (CFA). Although RHA, PFA, and CFA have been investigated as adsorbents for pollutants' removal, most of the studies only use one type of ash (with or without modification) and other studies have converted these ashes into activated carbon before use as an adsorbent. Furthermore, no work has been conducted to synthesize a composite adsorbent out of these waste-derived siliceous materials.

In addition, an effective method was required to synthesize the composite RHA/PFA/CFA adsorbent. A previous study by Adam & Chua (2004) has shown that RHA adsorbents could be synthesized using sol-gel method in the presence of aluminum ion. They found that this silica-incorporated aluminum is a very promising adsorbent for palmitic acid adsorption. Metal oxides contained in RHA might be essential in the adsorption of this palmitic acid. Therefore, the objective of this research is to investigate the removal of Zn²⁺ and Ni²⁺ from aqueous solutions using composite RHA/PFA/CFA adsorbent prepared by sol-gel method. Several operating parameters, including adsorbent dosage, contact time, and pH were studied during batch adsorption experiment. The study also investigates the adsorption isotherms and adsorption kinetics.

Three types of ash (i.e., RHA, PFA, and CFA) were collected from burning industrial fuels and used in the adsorbent preparation. RHA was obtained from Kilang Beras & Minyak Sin Guan Hup Sdn. Bhd., Pulau Pinang, Malaysia. PFA was supplied directly by United Oil Palm Mill, Pulau Pinang, Malaysia. CFA was collected from Stesen Janakuasa Sultan Azlan Shah, Manjung, Perak, Malaysia. Prior to use, the ashes were sieved to a fine particle size of less than 63 μm and then dried in an oven overnight at 110 °C.

Adsorbent was prepared by sol-gel method (Adam et al. 2006). About 15 g of each ash (RHA, PFA, CFA) was mixed and stirred in nitric acid (HNO₃ 65%) for 24 h. The mixture was filtered, rinsed with distilled water until the pH of the rinse became constant, and then dried in an oven at 110 °C for 1 day. About 15 g of the acidified RHA/PFA/CFA adsorbent was dissolved in 250 mL of 6 M NaOH, stirred for 12 h, and then filtered to eliminate insoluble particles. Thereafter, the filtrant was titrated with 3 M HNO₃ that contained 10% (w/w) aluminum ion [Al (NO₃)₃.9H₂O in HNO₃]. When the pH reached 11.5, a black suspension was seen, then titration was carried out until the pH reached 5. Finally, the RHA/PFA/CFA composite adsorbent was kept for 6 days, then the forming soft gel was filtered and oven dried at 110 °C for 1 day.

In this study, the synthetic wastewater was prepared by dissolving 132 mg of ZnSO₄.7H₂O and 124 mg of Ni(NO₃)₂.6H₂O in 1 L of deionized water to achieve Zn²⁺ and Ni²⁺ concentrations of 30 ± 2 mg/L and 25 ± 2 mg/L, respectively. This synthetic wastewater represents Zn²⁺ and Ni²⁺ concentrations in metal plating industries wastewater (Bayat 2002). The elements of RHA, PFA, and CFA were analyzed by X-ray photoelectron spectrometer (XPS, PHI 1600). The properties of the composite RHA/PFA/CFA adsorbent were also investigated using Fourier transform infrared (FTIR), specific surface area and particle size distribution. The surface functional groups of the adsorbent were detected using FTIR spectroscopy (Perkin Elmer, spectrum 100 Series model). The spectra were recorded from 4,000 to 400 cm⁻¹. Particle size distribution and specific surface area of the composite RHA/PFA/CFA adsorbents were examined using Mastersizer 2000.

In this study, a composite RHA/PFA/CFA adsorbent was made by combining three different types of ash together. RHA contains 63% of SiO₂, 19% of C, and other metallic elements in minor quantities; PFA contains about 34% of SiO₂, 25% of C, 6% of Al₂O₃, 5% of CaO, and other amounts of metallic elements; and CFA contains about 31% of SiO₂, 24% of C, 11% of Al₂O₃, 10% of Fe₂O₃, 7% of CaO, and other metallic components. The characteristics of composite RHA/PFA/CFA adsorbent have been investigated using FTIR, particle size distribution and specific surface area. Figure 1 shows the corresponding FTIR spectrums of the composite RHA/PFA/CFA adsorbent. The FTIR absorption bands appearing at 465, 797, and 1,053 cm⁻¹ are assigned with Si–O groups, attributed to bending of Si–O, symmetric stretching vibration of the Si–O (quartz), and asymmetric stretching vibration of the Si–O, respectively (Muller et al. 2014). According to the preparation process, the peak at around 1,384 cm⁻¹ is attributed to the NO₃− group, while the peak at 1,635 cm⁻¹ is assigned to H–OH bending of water molecules trapped in the silica matrix. The absorption peak around 3,436 cm⁻¹ is associated with the O–H stretching vibration from the solid Si–OH and the HO–H vibration of the water molecules adsorbed on the silica surface (Adam & Chua 2004).

The composite RHA/PFA/CFA adsorbent was also analyzed for particle size distribution and surface area, and the results are presented in Figure 2. The composite RHA/PFA/CFA adsorbent had a bimodal particle size distribution, with a predominance of particles in the range of 1–91 μm (averaged at 25 μm), followed by particles in the range 240–630 μm (averaged at 450 μm). This might be due to the combination of different materials (i.e., RHA, PFA, and CFA), influenced by the preparation process using the sol-gel method, leading to large particles' breakup or variable growth mechanisms during the preparation method. Adsorbents with bimodal particle size distribution have a higher adsorption capacity than unimodal particle size distribution. Sim et al. (2014) reported that an adsorbent with bimodal structure showed a higher uptake capacity and faster adsorption rate of silver ion and silver nanoparticles. The bimodal particle size distribution of the composite RHA/PFA/CFA adsorbent structure enables enhanced capillary driven aqueous solution distribution without hindering the aqueous solution intake of the adsorbent core.

In addition, the composite RHA/PFA/CFA adsorbent had a specific surface area of 77.4 m²/g. In comparison, the specific surface areas of RHA, PFA, and CFA prior to treatment were 11.35 m²/g (Fernandes et al. 2017), 1.775 m²/g (Megat Johari et al. 2012), and 1.1 m²/g (Xie et al. 2014), respectively. There is a notable difference in specific surface area between individual ashes and the composite RHA/PFA/CFA adsorbent, which might be due to the preparation using the sol-gel method and the combination of different materials. Specific surface area is defined as the sum of total area of particles including pores, and has been directly correlated with particle size and porosity. Smaller, highly porous particles have considerable surface area, while larger, non-porous ones have reduced surface area values (Fernandes et al. 2017).

By varying the adsorbent amounts from 2 to 10 g/L, the effect of the adsorbent amount was investigated. Table 1 provides the experimental data for Zn²⁺ and Ni²⁺ adsorption. For all experiments, the concentration of initial metal ions was fixed at 30 ± 2 mg/L of Zn²⁺ and 25 ± 2 mg/L of Ni²⁺. Figure 3 shows the adsorption of Zn²⁺ and Ni²⁺ ions increases rapidly with increasing the amount of RHA/PFA/CFA adsorbent. The major increase in Zn²⁺ and Ni²⁺ removal was noticed from 41 to 81% and 25 to 66% when the adsorbent amount was increased from 2 to 10 g/L, respectively. Increasing the adsorbent amount will increase the availability of surface area and binding sites on the adsorbent surface. In addition, RHA, PFA, and CFA contain carbon and silica based substances that attach the metal ion to each other in an aqueous solution (Dahlan et al. 2007, 2008; Lee et al. 2008).

The results showed that the removal efficiency for Zn²⁺ ions is higher than for Ni²⁺ ions. This might indicate that the surface of composite RHA/PFA/CFA adsorbent has a higher affinity for Zn²⁺ ions than for Ni²⁺ ions. This could be associated with the adsorbent pore size, which is more favorable for smaller atomic size of Zn²⁺ ions (139 pm) than Ni²⁺ ions (163 pm) (Enghag 2008). In addition, higher charge density on the surface of smaller sized Zn²⁺ ions could also be another reason for their higher removal efficiency compared to Ni²⁺ ions.

The results also showed that maximum removals of Zn²⁺ and Ni²⁺ were obtained at adsorbent amount of 10 g/L. After that, the adsorbent uptake efficiency reached equilibrium condition without further increase. Therefore, the next adsorption studies were conducted using adsorbent amount of 10 g/L. At high adsorbent amount, the available heavy metal concentration is insufficient to cover the exchangeable available sites on the adsorbent, resulting in low metal adsorption. Further, increased adsorbent amount leads to an interference between binding sites and may result in a low specific removal. The interactions between heavy metal ions become more essential when the adsorbent amount in the liquid phase is higher, as this may cause physical blockage of some adsorption sites, leading to a decreased adsorption efficiency. These interactions can result in electrostatic interferences, where the electrical surface charges on the closely packed particles diminish attractions between the surfaces of individual grains and adsorbed solutes (Malamis & Katsou 2013).

Table 1 shows the adsorption of Zn²⁺ and of Ni²⁺ onto RHA/PFA/CFA adsorbent while varying the contact time. Adsorption of Zn²⁺ and of Ni²⁺ is highly enhanced by increasing the contact times from 10 to 60 min. Figure 4 shows that the adsorption of Zn²⁺ and Ni²⁺ rapidly increased in the first 30 min, where about 77% of Zn²⁺ and 41% of Ni²⁺ was removed, after which, equilibrium was slowly reached. This could be due to the excessively adsorbent surface area available at the start of the Zn²⁺ and Ni²⁺ adsorption process (Bhattacharya et al. 2006). Similarly, Rashid et al. (2016) reported that the removal of Zn²⁺ and Ni²⁺ was faster in the beginning and slowed down until equilibrium was reached. They reported that the adsorption process occurred in two stages: initial shorter duration (fast stage), followed by longer second duration (slower stage); then it continued until equilibrium was achieved.

When the adsorbent surface area becomes limited, further increase in adsorption capacity is associated with the transport of adsorbate from the exterior to the interior sites of the adsorbent substances (Dahlan & Zwain 2013). Increment in contact time has increased the Zn²⁺ and Ni²⁺ adsorption, but it remains unchanged after reaching equilibrium in 40 min and 50 min for Zn²⁺ and Ni²⁺, respectively. Moreover, the maximum Zn²⁺ and Ni²⁺ removal efficiencies of 81% and 60%, respectively, were attained at contact time of 60 min. Likewise, Kara et al. (2017) mentioned that Zn²⁺ and Ni²⁺ adsorption reached equilibrium at 40 and 50 min, with adsorption capacity of 60.06 mg/g and 29.40 mg/g, respectively.

The pH of solution highly affects the adsorption process due to the speciation of adsorbate species, degree of ionization, and changes of charge on adsorbent surface. Figure 5 shows the effect of pH level (range 3–9) on adsorption of Zn²⁺ and Ni²⁺ by RHA/PFA/CFA adsorbent. It was found that the adsorption process was strongly dependent on the pH level presented in the media. When the pH increased from 3 to 7, there was a sharp increase in adsorption process from 60% to 96% and from 23% to 59% for Zn²⁺ and Ni²⁺, respectively. In pH from 7 to 9, no change was observed in the removal of zinc due to the fact that most of the Zn²⁺ ions were removed earlier. In contrast, the Ni²⁺ adsorption further increased from 59% to 83% when pH level increased from 7 to 9, however, the Zn²⁺ adsorption was higher than Ni²⁺. Thus, pH 7 was taken into account as the optimum pH level for the following studies.

At different pH values, the difference in adsorption of Zn²⁺ and Ni²⁺ by RHA/PFA/CFA adsorbent could be due to different surface charges resulting from different adsorbent surface and different degrees of solute speciation ionization (Rashid et al. 2016). At extremely acidic and basic pH, low adsorption of metal ions could be due to sorbate lyophobic behavior. The functional groups are a key factor for metal binding, which could be affected by pH through protonation/deprotonation, leading to a decrease/increase in the attraction of charged metal ions (Ullah et al. 2013). In addition, at low level of pH, there is practically less removal of metal ions due to the high electrostatic repulsion caused by high H⁺ ion concentration on the surface sites. When the pH is increasing, the H⁺ ion concentration on the adsorption site is reduced, resulting in decreased electrostatic repulsion, thus leading to an improvement of metal ions adsorption (Aklil et al. 2004). On the other hand, at higher levels of pH, OH⁻ competes for Zn²⁺ and Ni²⁺ with the active sites on the surface of adsorbent and formation of the precipitate of Zn(OH)₂ and Ni(OH)₂ occurs and contributes to the removal process (Kalyani et al. 2003). To prevent metal ion precipitation, the rest of the following experiments were conducted at levels of pH not higher than 7.

Langmuir isotherm model is used to theoretically estimate the maximum metal removal that cannot be obtained in experimental studies. From this Langmuir isotherm model study, the maximum adsorption capacities (Q_o) were 21.74 and 17.85 mg/g for Zn²⁺ and Ni²⁺, respectively. As compared to the actual adsorption capacities of 6.6 mg Zn²⁺/g and 3.2 mg Ni²⁺/g (measured at adsorbent amount of 2 g/L, contact time of 60 min, shaker rate of 110 rpm, neutral pH of 6, and initial Zn²⁺ and Ni²⁺ concentration of 30 ± 2 mg/L and 25 ± 2 mg/L, respectively), those values (Q_o) show that the RHA/PFA/CFA adsorbent did not reach its equilibrium condition. Hence, less amount of adsorbent could be used efficiently to remove these concentrations of metals (Vakili et al. 2016). The difference in adsorption capacities might be due to the different interaction mechanisms associated with different heavy metal ions. The affinity constant (b) for Zn²⁺ (0.022) was much higher than that for Ni²⁺ (0.011). Thus, the RHA/PFA/CFA adsorbent showed high affinity for the removal of Zn²⁺ (Table 2).

K_F represents the distribution coefficient of Zn²⁺ and Ni²⁺ adsorbed by RHA/PFA/CFA adsorbent at equilibrium concentration. The slope 1/n is ranged between 0 and 1.0 which measures the surface heterogeneity or adsorption intensity. As the value of the slope approaches zero, the adsorption surface become more heterogeneous. On the other hand, the adsorption surface become more homogenous when the slope gets closer to 1.0. The data of 1/n together with the coefficient of determination are presented in Table 2. In this study, the values of 1/n were 0.79 and 0.87 for Zn²⁺ and Ni²⁺, respectively. This result indicates that the surface of RHA/PFA/CFA adsorbent is less heterogeneous. In general, the Freundlich equation is an empirical analysis for very uneven adsorbent surface that can adsorb single adsorbate at a fixed range of concentration.

The isotherm constants K_T and B₁ can be obtained from the q_e versus ln C_e plotting. K_T is the equilibrium binding constant (L/mg) related to the maximum binding energy and constant B₁ is correlated to the adsorption heat. Figure 8 shows the plotted Temkin isotherm with the parameter values listed in Table 2.

Considering the coefficient of determination (R²) presented in Table 2, the adsorption isotherm study using the RHA/PFA/CFA adsorbent is explained by the Freundlich model followed by Temkin and Langmuir models. Additionally, the Freundlich model (Figure 7) better fits the experimental data than Temkin (Figure 8) and Langmuir models (Figure 6). The predicted equilibrium capacities of Zn²⁺ and Ni²⁺ onto RHA/PFA/CFA adsorbent using the Freundlich isotherm agrees precisely with the adsorption capacities of this experimental study. This suggests multilayer adsorption of Zn²⁺ and Ni²⁺ on less heterogeneous surfaces, and assumes that all of the active sites of the adsorbent participated in the adsorption processes.

The parameters (k₁, k₂) of this kinetic study are calculated and tabulated in Table 3. Based on the results presented in Figures 9 and 10, at initial concentrations of 30 ± 2 mg/L of Zn²⁺ and 25 ± 2 mg/L of Ni²⁺, the adsorption kinetics of Ni²⁺ seems to obey both kinetic models, while adsorption kinetics of Zn²⁺ obeys only the pseudo-second-order kinetic model. In addition, from the coefficient of determination (R²) values, pseudo-second-order model fits the data better than pseudo-first-order model for both Zn²⁺ and Ni²⁺ adsorption. Pseudo-second-order kinetic model assumes that chemisorption process may be the rate-limiting step. In chemisorption, the metal ions bind to the adsorbent surface, where a chemical (usually covalent) bond is formed, and look for sites that maximize their coordination number with the adsorbent surface (Atkins 1995).

The adsorption capacities of RHA/PFA/CFA adsorbent removing Zn²⁺ and Ni²⁺ have been calculated, compared with different adsorbents given in the literature, and shown in Table 4. From Table 4, several studies using a single solid waste material (in the form of ash with/without modification) have been conducted to examine the adsorption of Zn²⁺ and Ni²⁺ using adsorbent prepared from rice husk ash (Bhattacharya et al. 2006), palm oil ash (Chu & Hashim 2003), and fly ash (Bayat 2002). In this work, the RHA/PFA/CFA adsorbent prepared by sol-gel method had a relatively high adsorption capacity (i.e., 21.74 mg/g and 17.85 mg/g for Zn²⁺ and Ni²⁺, respectively), as compared to adsorbent prepared using a single solid waste material (with/without modification). The relatively high adsorption capacity shown by RHA/PFA/CFA adsorbent might be due to the interaction of silica between these solid waste materials to form more complex reactive species (Zwain & Dahlan 2012) that is responsible for Zn²⁺ and Ni²⁺ adsorption. On the other hand, among the three types of solid waste material which were prepared individually, rice husk ash had the highest Zn²⁺ and Ni²⁺ adsorption capacity. From this comparison, it shows that RHA/PFA/CFA adsorbent was more promising and efficient for Zn²⁺ and Ni²⁺ removal from aqueous solution.

This study confirmed that composite RHA/PFA/CFA adsorbent modified by the sol-gel method was an excellent adsorbent for removal of Zn²⁺ and Ni²⁺ from synthetic wastewater. The maximum adsorption capacity was observed at 10 g/L of adsorbent amount, 60 min of contact time, and pH 7 which gave 81% and 61% removal efficiency for Zn²⁺ and Ni²⁺, respectively. An increase in adsorbent amount and contact time leads to an increase in adsorption efficiency. A decrease in pH level causes a major decrease in the adsorption capacity of Zn²⁺ and Ni²⁺. The adsorption isotherm was found to best fit the Freundlich isotherm model for both Zn²⁺ and Ni²⁺. The maximum adsorption capacity (q_max) calculated was 21.74 mg/g and 17.85 mg/g for Zn²⁺ and Ni²⁺, respectively. In addition, it was found that the pseudo-second-order model successfully explained the adsorption kinetics of Ni²⁺ and Zn²⁺, as compared to the pseudo-first-order.

The authors are grateful for the Short Term Grant (A/C. 60310014) and Iconic Grant Scheme (A/C. 1001/CKT/870023) financial support received from Universiti Sains Malaysia.