Equilibrium, kinetics and thermodynamics of Cu(II) biosorption on Chinese chestnut shell pretreated with steam explosion

Heavy metals have been widely used in large quantities in many industries over the past few decades, which has inevitably brought about an increase in the discharge of metallic substances into adjacent water bodies. These substances can get into the human body and accumulate in the organs, causing serious health problems if ingested beyond the recommended concentrations. Therefore, metal-contaminated wastewater needs to be treated prior to its discharge into the environment.

Copper, an important heavy metal widely used in industry, agriculture and household products, is emitted into the environment from sources such as drainage discharge, mining, plating, paints, pigments, and the fertilizer industry (Malik et al. 2017). Intake of acute dosage of copper by humans can cause serious mucosal corrosion, nervous irritation, capillary damage, and hepatic and renal failure (Malik et al. 2017). Copper-loaded wastewater can be treated by ion exchange, precipitation, chemical oxidation, chemical reduction, solvent extraction, reverse osmosis, ultrafiltration, electrodialysis, and sorption (Yao et al. 2010). Biosorption, a subcategory of sorption, is based on the ability of biomolecules or biomass to adsorb and concentrate some ions or other molecules from aqueous solutions (Cataldo et al. 2018). It is considered a promising technology for heavy-metal removal from wastewater.

Some low-cost agricultural biomass, such as the shells from chestnuts, walnuts, hazelnuts, and almonds (Yao et al. 2010; Cataldo et al. 2018), have been given more attention recently thanks to their eco-friendly and economic characteristics. However, these agricultural wastes are generally poor in their sorption capacities for heavy metals and incapable of meeting the requirements of industrial applications. Therefore, they need to be modified or treated to overcome this disadvantage. Many chemicals have been employed for the modifications, including organic and mineral acids/bases, oxidants, and organic compounds (Vijayaraghavan & Balasubramanian 2015). Physical methods are generally very simple and inexpensive. Autoclaving, steaming, thermal drying, lyophilization, cutting and grinding have been used for biosorbent modification. Steam explosion, which originally was used in the pulping process, has been recognized as a cost-effective technology for pretreating lignocellulosic biomass. Chi & Chen (2009) showed that steam explosion affects the specific surface area of corn stalk and its structure, making it better for ethanol production. Steam explosion has also been used for chestnut shell pretreatment to produce fermentable sugars, and its effect on enzymatic saccharification was reported by Maurelli et al. (2013). However, there have been no reports of applying steam explosion to modifying biomass for biosorption.

In China, the yearly production of Chinese chestnuts (Castanea mollissima) is about 1.88 million tons (FAO 2017), and the shells are the waste remaining after production. Our previous work (Yao et al. 2010) revealed the potential use of Chinese chestnut shell for the sorptive removal of copper ions from aqueous solutions. Here, our goal is to examine the applicability of steam explosion to improving the sorption performance of Chinese chestnut shell for copper removal from water. The biosorption by the steam-exploded chestnut shell was characterized by equilibrium isotherms, kinetics, thermodynamics, and Fourier transform infrared spectroscopy (FT-IR). To evaluate the effects of steam explosion on sorption, the results were compared with previously determined results of Cu(II) biosorption by raw chestnut shell (Yao et al. 2010).

Chinese chestnuts were purchased from a market in Shenyang, China, and peeled by hand. The shells were steam exploded at 1.3 MPa (about 192 °C) for 5 min with a QBS-200B-type steam blasting machine (Bio Energy Co., Ltd, Hebi, China), washed thoroughly with distilled water, air-dried in an oven for 24 hours at 65 °C, crushed, and sieved through a sieve with a 0.38 mm mesh.

Stock copper solution (10,000 mg/L) was made by dissolving the desired quantity of Cu(NO₃)₂·3H₂O (AR grade) in pure water. The test solutions were made by diluting the stock solution to the required Cu(II) concentrations with pure water.

The specific surface area was determined by the Brunauer-Emmett-Teller (BET) method with an ASAP2020 sorption apparatus (Micromeritics, USA). The point of zero charge (pH_pzc) was measured by the immersion technique reported by Nomanbhay & Palanisamy (2005). The scanning electron microscopy (SEM) analysis was conducted with a Hitachi S-3000N SEM. FT-IR spectra were recorded with a Bruker Tensor 27 FT-IR spectrometer, using the KBr disc technique. The samples were dried over P₂O₅ in a desiccator for 7 days prior to the FT-IR measurement.

After the sorption process, the mixtures were filtered through a microporous membrane with 1.2-μm pores and the copper concentrations were measured with a flame atomic sorption spectrophotometer (AA-100, Perkin-Elmer Inc., Foster City, USA).

All the sorption experiments were conducted three times, and the mean values are presented. The equilibrium and kinetic data fitted to the models were solved by linear regression using Microsoft^® Excel.

The pretreated and untreated chestnut shell were characterized to illustrate the effect of steam-explosion pretreatment on the properties of the biosorbent. Steam explosion can decrease the pH_pzc from 4.9 for the untreated biomass (Yao et al. 2010) to 3.8 for the modified biomass. This may be due to the increased acidity of the biosorbent (Jacquet et al. 2015), which was confirmed by FT-IR spectra. The surface morphology of the biomass with and without pretreatment was characterized using SEM and indicated that the untreated particle surface is smooth with some pores while the particle surface of the steam-exploded chestnut-shell particle is rough (Figure 1). The roughness of the particle is caused by breakage and destruction of the biomass during the rapid escape of steam from the internal pores. Generally, an enlarged surface area endows sorbents with higher sorption capacity. Steam-explosion treatment has been shown to expand the specific surface area of some lignocellulosic materials (Chi & Chen 2009); it increased the surface area of the chestnut shell from 14.44 m²/g to 29.27 m²/g.

The FT-IR spectra of the untreated chestnut shell (Figure 2(a)) and the steam-exploded chestnut shell before (Figure 2(b)) and after (Figure 2(c)) Cu(II) sorption exhibit similar absorption bands, indicating that steam explosion did not produce new functional group species on the chestnut shell. However, significant changes were found in the absorption bands around 3,440, 1,740 and 1,521 cm⁻¹ for stretches in OH, COOH and C = C of aromatic rings. The steam explosion made the bands at 3,440 and 1,521 cm⁻¹ stronger and changed the band at 1,740 cm⁻¹ from a shoulder to a small peak, suggesting more hydroxyls, carboxyls, and aromatic rings were exposed by the pretreatment. After the sorption of copper onto the steam-exploded chestnut shell, the band at 3,440 cm⁻¹ was weakened and the peak at 1,740 cm⁻¹ for the carboxyl shifted to a band at 1,695 cm⁻¹ for copper carboxylate, indicating hydroxyls and carboxyls played important roles in binding copper onto the biomass. These functional groups have also been found to take part in the sorption of copper onto the untreated chestnut shell (Yao et al. 2010).

The pH usually plays a crucial role in biosorption. It affects both the surface charge of biosorbents, and the degree of ionization and speciation of sorbates. The effect of the pH of the solution on Cu(II) removal was that the copper removal surged with increasing pH from 2.0 to 3.0, then gradually rose to the maximum of 98.2% at pH 5.0, and finally slightly decreased with the further increase of the pH value to 6.0 (Figure 3(a)). Therefore, pH 5.0 was selected as the optimum pH for the sorption process of Cu(II) onto the steam-exploded chestnut shell. At low pH values, more H⁺ is present in the solution, which competes with the Cu(II) ion for the sorption sites on the biosorbent. This may explain why the removal efficiency goes up with the pH increasing from 2.0 to 5.0. The effect of the solution pH on the Cu(II) sorption can also be explained based on the biosorbent's pH_pzc and the copper speciation. The pH_pzc of the steam-exploded chestnut shell was found to be 3.8. Consequently, the charges on the biosorbent surface are positive at a pH below 3.8 and negative at a pH over 3.8. At pH values ranging from 2.0 to 6.0 in the Cu(II) solution of 50 mg/L, Cu²⁺ and Cu(OH)⁺ are the dominant Cu(II) species, according to simulation results performed with the Visual Minteq v2.61 software. The biosorbent surface is protonated more pronouncedly at a low pH owing to the presence of concentrated protons in the solution. This condition is more unfavorable to copper capture because the positively charged biosorbent surface electrostatically repulses the metal cations. At the optimum pH value (pH 5.0), the surface is negatively charged and favorable to the sorption of the dominant species, Cu²⁺ and Cu(OH)⁺. To further explain the removal mechanism, all the final pHs after the sorption were measured, and the ΔpH (pH_final – pH_initial) values are shown in Figure 3(b). All the solution pHs went down after sorption (negative ΔpH), indicating that H⁺ was replaced from the phenolic hydroxyls and carboxyls on the biosorbent to the solution by Cu²⁺ and/or Cu(OH)⁺. Ion exchange was also involved in the sorption process of Cu(II) onto the untreated chestnut shell (Yao et al. 2010). This sorption mechanism was retained after steam-explosion pretreatment.

Sorption kinetics is a physicochemical way of evaluating the performance of a process. The pseudo-first-order, pseudo-second-order, and intra-particle diffusion equations were fitted to the kinetic data. In this study, equilibrium was attained within 300 min and the equilibrium times were the same for all tested concentrations. Therefore, sorption data for 2.5–300 min were fitted to the kinetic models.

The values of q_e, k₂, h_0, and R² are shown in Table 1. The plot of t/q_t versus t gives good straight lines for all the copper concentrations (Figure 4(b)), and all of the R² values are over 0.99 (Table 1). The theoretical values (q_e,cal) obtained from this equation are close to the experimental ones (q_e,exp). These signify that the sorption systems are a good fit for the pseudo-second-order kinetic model, on the basis of the assumption that the rate-limiting step might be chemisorption, involving electron sharing or exchange of sorbent with sorbate (Ho 2003). We found the Cu(II) sorption onto the untreated chestnut shell also obeyed the pseudo-second-order kinetic model (Yao et al. 2010). By comparison with the Cu(II) sorption onto the untreated chestnut shell (Yao et al. 2010), the steam-explosion pretreatment led to reductions of the sorption rate constant and the initial sorption rate (h₀), which is disadvantageous to a sorption process. Currently, there is no evidence to explain the decrease in sorption on steam-exploded biomass. As a thermo-mechanicochemical process, steam explosion opens lignocellulosic structures, modifies the chemical properties of the lignin and cellulose, and forms organic acids, etc. (Jacquet et al. 2015). Steam explosion made the chestnut shell more favorable to working as a biosorbent in both physical (rough surface and bigger surface area) and chemical (more binding functional groups) aspects. The contradiction between the sorption performance and the physicochemical properties is very interesting and worth further investigation.

The sorption isotherm shows the distribution behavior of sorbate between the liquid and the solid phases at equilibrium. The right isotherm model can be used to design the sorption process. Therefore, fitting the equilibrium data to different isotherms is a meaningful step in obtaining an applicable isotherm for design purposes. In this study, a sorption isotherm study was conducted using two famous isotherms, i.e., the Langmuir and Freundlich models.

The constants K_L and q_m at different temperatures were calculated from the plots of 1/q_e against 1/C_e, as shown in Figure 5, and their values, as well as the coefficient of determination (R²), are shown in Table 3. High R² values (>0.99, Table 3) indicate that the Langmuir model is a good fit for the biosorption of Cu(II) onto the chestnut shell pretreated by steam explosion.

The K_L values for sorption both onto the steam-exploded (Table 3) and onto the untreated (Yao et al. 2010) chestnut shell decreased with the increasing sorption temperature. At the same temperatures, the K_L values for sorption onto the steam-exploded shell are lower than for untreated shell. With increasing sorption temperature, q_m increases for copper sorption onto the steam-exploded biosorbent (Table 3) while it decreases for the untreated biosorbent (Yao et al. 2010). Below 313 K, the pretreatment impaired the Cu(II) sorption capacity. Under the experimental conditions, the maximum q_m value is 17.54 mg/g (Table 3) for the sorption onto the steam-exploded chestnut shell at 333 K and 12.56 mg/g (Yao et al. 2010) for sorption onto the untreated chestnut shell at 293 K. Therefore, the steam-exploded shell is preferable for the sorption of copper ions at higher temperatures, while the untreated shell is more effective at lower temperatures. The steam-exploded chestnut shell shows higher copper-sorption capacity than many biosorbents from various agricultural by-products, such as date stones, palm-tree waste (Belala et al. 2014), peanut husk (Li et al. 2007), wheat bran (Renu & Singh 2018), wheat straw (Muhamad et al. 2010), hazelnut shell, apricot stone (Šoštarić et al. 2015) and corn cob (Petrović et al. 2015) . Therefore, steam-exploded chestnut shell is promising for Cu(II) removal from wastewater.

Both the Langmuir and the Freundlich isotherms had R² values over 0.9 and the values derived from the former are closer to unity (Table 3), implying that the Langmuir model is more appropriate for characterizing Cu(II) sorption onto the steam-exploded chestnut shell than the Freundlich model. This confirms the monolayer coverage process of copper onto the steam-exploded chestnut shell according to the Langmuir model.

The values of ΔH° and ΔS° were obtained from the plot of ΔG° against T, as shown in Figure 6. The calculated thermodynamic parameters are given in Table 4. The positive value of ΔG° indicates that steam-exploded chestnut shell is unable to capture Cu(II) spontaneously. The negative value of ΔH° indicates that the sorption is exothermic, thereby demonstrating that the product is energetically stable (Khan & Singh 1987). The negative value of ΔS° suggests that the process is entropy driven and indicates that the Cu(II) sorption does not significantly alter the internal structure of the biosorbent. The Cu(II) sorption onto the untreated chestnut shell is an exothermic process driven by enthalpy, and it has negative ΔG° values which show it is a feasible and spontaneous process (Yao et al. 2010).

The steam-explosion pretreatment reduced the sorption rate constant and the initial sorption rate of the pseudo-second-order kinetic model. The Cu(II) sorption capacity was enhanced at higher temperatures. Additionally, the sorption process was changed from a spontaneous process driven by enthalpy to a nonspontaneous one driven by entropy. It increased the surface area of the chestnut shell and exposed more hydroxyl and carboxyl groups, which are binding sites for copper ions. Compared with untreated chestnut shell, steam-exploded shell is preferable for sorbing copper ions from higher-concentration solutions at higher temperatures. Steam explosion is a simple and inexpensive physical process. The pretreatment improved some aspects of the biosorbent while impairing others. Unfortunately, this is a pilot study without a systematic investigation into the sorption response to the pretreatment parameters such as temperature, pressure and time. The negative aspects might be ameliorated by optimizing the steam-explosion conditions, which should be further explored.

This research was supported by the National Natural Science Foundation of China (No. 31770673, 21167016 and 51268052). The atomic absorption spectrometer, scanning electron microscope, and FT-IR spectrometer were supplied by the Valuable Equipment Sharing Platform of Southwest Forestry University. Appreciation is expressed to Dr Ian M. Scott, a scientist from the London Research and Development Centre of Agriculture and Agri-Food Canada, for reviewing the manuscript.

The authors declare no conflict of interest.