Role of functional groups on protonated de-oiled soybean involved in triclosan biosorption from aqueous solution

Abstract

Protonated de-oiled soybean (PDOS) as a low-cost and effective biosorbent was used for the biosorption of triclosan (TCS) from aqueous solution. The contribution of functional groups such as amine, carboxyl, sulfhydryl, phosphate and lipids on PDOS to TCS biosorption was evaluated by means of chemical modification. The results showed that sulfhydryl groups gave the largest contribution, followed by lipids and amine groups, while the contribution of carboxyl groups could not be confirmed. FTIR spectra offered help to identify the introduced functional group changes. Based on the results, it was proposed that the TCS biosorption onto PDOS was determined by hydrophobic interaction and hydrogen bonding rather than electrostatic interaction. The TCS biosorption process depended much on initial pH and the optimum pH value was 4.0. Kinetic studies demonstrated that a pseudo-second-order model was more applicable for description of the biosorption of TCS onto PDOS; both intra-particle diffusion and boundary layer diffusion might affect the biosorption rate. The Dubinnin–Radushkevich (D–R) isotherm agreed well with the biosorption equilibrium data over the whole concentration range, and the maximum biosorption capacity was 124.55 mg g⁻¹, revealing that biosorption via PDOS could be used as an effective method for TCS removal from aqueous solution.

---

1. Introduction

More attention has been paid to the concern of unintentional human exposure owing to the continuous detection of trace concentrations of pharmaceuticals and personal care products (PPCPs) in wastewater treatment plants (WTTPs).^1,2 Triclosan (5-chloro-2-(2,4-dichlorophenoxy)phenol, TCS), a member of PPCPs, is a synthesized, broad-spectrum antimicrobial agent widely used in many personal care products. Nowadays, TCS is ubiquitous in consumer care products such as toothpaste, mouthwash, soaps, as well as in household cleaners like detergents and even in cosmetics and pharmaceuticals at concentrations up to 0.3%.^3,4 After discharge of these PPCPs during normal use into domestic sewage, up to 96% of TCS may reach the environment due to their incomplete removal in WWTPs or direct discharge of wastewater without treatment.^5,6 TCS restrains bacterial growth by blocking lipid biosynthesis via inhibition of the enzyme carrying proteins.^7,8 Although it has not shown potential risks to mammals, the occurrence of TCS in waters exhibits its toxicity to water living organisms such as fish, Daphnia magna and certain algae species (i.e., Scenedesmus subspicatus).^4,9 Even worse, TCS displays possible bioaccumulation and further environmental persistence because of its special chemical properties.¹⁰ Therefore, it is urgent to eliminate TCS from water environment with useful ways.

Various physicochemical and biochemical treatment processes have been employed to remove TCS from drinking water and wastewater. Singer et al. reported that approximately 79% of TCS could be removed through biological wastewater treatment processes, but Ying et al. demonstrated that biological degradation of TCS was relatively slow with a biodegradation half-life of weeks or months under aerobic and anaerobic conditions.^2,4 Although chemical oxidation treatments could degrade TCS, great concern was paid to its toxic degradation products such as 2,8-dichlorodibenzo-dioxin, 4,5-dichloro-[1,1-biphenyl]-2,2-diol and 2,4-dichlorophenol.¹¹ This problem also troubles photocatalytic degradation and restricts its practical application. Previous work supports that biosorption plays the main part in the removal of TCS.^12–14 With the advantages of low cost, ease of operation, and free from secondary pollution, biosorption can be regarded as an ideal and efficient method to remove TCS from drinking water and wastewater.

Moreover, the exploitation of low-cost and high effective biosorbents has become a hotspot in biosorption. A diversity of biological materials have been exploited as biosorbents, such as bacteria,¹⁵ fungi,¹⁶ algae,¹⁷ industrial wastes,¹⁸ and agricultural wastes.¹⁹ Among of them, agricultural waste materials have been proved to be effective biosorbents for the removal of contaminants from wastewater, owing to their unique chemical composition, availability in abundance, and low capital cost.²⁰ De-oiled soybean, a waste material processed out from oil industry, is usually used as animal feedstuff owing to their lower nutrients value. The U.S. department of agriculture (USDA) announced that the production of soybean oil of China had reached around 14.46 million tons in 2015, and which will keep an increase trend in future. Every 191 gram of soybean oil, can produce 1000 gram de-oiled soybean.²¹ It means that about 75.71 million tons of de-oiled soybean were produced in 2015 in China. However, not all of de-oiled soybean were made full use, which may cause solid wastes. Recently, de-oiled soybean has been used as a potential biosorbent to remove dyes and heavy metals.^22–24 As far as we know, there are no studies on its TCS biosorption performance and mechanism.

In the present work, de-oiled soybean was utilized to prepare low-cost and efficient protonated de-oiled soybean (PDOS) for TCS removal from aqueous solution. This study aimed to probe into the biosorption mechanism between PDOS and TCS. The roles of amine, carboxyl, sulfhydryl, phosphate and lipids functional groups present on PDOS in the TCS biosorption were studied by chemical modification. The Fourier transform infrared (FTIR) analysis was conducted to supply the information about the changes of functional groups on PDOS before and after both biosorption and chemical modification. The effect of pH on TCS biosorption was assessed. Furthermore, the kinetics, isotherms, and thermodynamics were investigated.

2. Materials and methods

2.1. TCS solution and analysis

Triclosan (molecular formula = C₁₂H₇Cl₃O₂, HPLC grade) was purchased from Sigma-Aldrich Corporation. Its physicochemical properties and molecular structure were listed in Table 1. TCS stock solution was prepared by dissolving 0.5 g TCS powder into 1 L HPLC grade methanol, and stored at 4 °C without light. The sample solutions were prepared by mixing up TCS/methanol with the background solution containing 0.1 g L⁻¹ NaN₃ (to prevent bacterial growth) and 2 wt% NaCl (to increase ionic strength). The volume of methanol added to the sample solutions was controlled in 20 (v/v%) to minimize cosolvent effects. And the desired pH value of the sample solutions was adjusted with 0.05 M H₂SO₄ or 0.1 M NaOH solution, which was measured by using a pH-meter (WTW-340i, Germany).

Table 1 Physicochemical properties and structural formula of TCS

Molecular formula	Structural formula	Molecular weight	Solubility (aqueous)	Solubility (methanol)	pKa	lg Kow
C12H7Cl3O2		289.5 g mol^−1	4.62 mg L^−1	Very soluble	8.14	4.76

---

The concentration of TCS was determined by using a UV-vis spectrophotometer (UV-1200, Mapada, China) at 280 nm. Quantification of TCS was based on a calibration curve performed within 0.5–50 mg L⁻¹, with the correlation coefficient R² of 0.9978.

2.2. Preparation of biosorbent

De-oiled soybean used in experiments was obtained from an oil mill in Shijiazhuang, Hebei Province, China. De-oiled soybean was rinsed with distilled deionized water to remove dust and impurities, dried at 105 °C overnight in an oven (DHG9240A, Shanghai, China), and sieved to 300–600 μm. Then, the biomass was converted to PDOS by the following procedure according to Schiewer and Volesky.²⁵ The biomass was protonated in 0.05 M H₂SO₄ (10 g biomass L⁻¹) for 6 h, washed 10 times in the same volume of distilled deionized water to remove the excess of acid, and dried in an oven at 105 °C overnight. The resulting dried PDOS was reserved in a dry cabinet for future experiments.

2.3. Chemical modification of PDOS

To explore the contribution of various functional groups on PDOS such as amine, carboxyl, sulfhydryl, phosphate, and lipids in the biosorption of TCS, portions of PDOS were chemically treated in different ways as following for modification of functional groups, which was schematically illustrated in Fig. 1. Enough chemical treated biosorbents were prepared at a time and the biosorbents modified with the same method were well mixed to reduce errors produced during the chemical modification.

Fig. 1 Schematic of chemical modifications performed on PDOS.

Control: PDOS without further processing.

Type 1: 1 g PDOS with 20 mL formaldehyde (molecular formula = HCHO, analytical grade) and 40 mL formic acid (molecular formula = HCOOH, analytical grade) were suspended in a water bath shaker with a constant shaking speed of 125 rpm at 20 °C for 6 h, which was referred to as T1. This treatment will cause methylation of amines by the following reaction:²⁶

Type 2: 1 g PDOS with 65 mL anhydrous methanol (molecular formula = CH₃OH, analytical grade) and 0.6 mL concentrated hydrochloric acid (molecular formula = HCl, analytical grade) were agitated in a water bath shaker with a constant shaking speed of 125 rpm at 20 °C for 6 h, which was referred to as T2. It has been known that this treatment results in esterification of carboxylic acids by the following reaction:²⁷

Type 3: 1 g PDOS with 100 mL 0.001 M 2,2′-dithiopyridine (molecular formula = C₁₀H₈N₂S₂, analytical grade) was agitated in a water bath shaker with a constant shaking speed of 125 rpm at 20 °C for 6 h, which was referred to as T3. This treatment can generate a modification of available sulfhydryl groups.²⁸

Type 4: 1 g PDOS was heated with 40 mL of triethyl phosphate (molecular formula = (C₂H₅O)₃P, analytical grade) and 30 mL of nitromethane (molecular formula = CH₃NO₂, analytical grade) under reflux conditions for 6 h, which was referred to as T4. This treatment will cause esterification of phosphate groups present on biomass.²⁹

Type 5: 1 g PDOS was heated with 75 mL of acetone (molecular formula = CH₃COCH₃, analytical grade) under reflux conditions for 6 h, which was referred to as T5. This treatment can extract the lipid fractions from PDOS.³⁰

2.4. Biosorption experiments

Biosorption of TCS onto PDOS was studied with batch experiments in 50 mL glass stopper Erlenmeyer flasks agitated in a water bath shaker with a constant shaking speed of 125 rpm at 20 °C (but not applicable for isotherms experiments) for 12 h to reach equilibrium (the biosorption profile of TCS versus contact time at different initial TCS concentrations ranging from 20 to 50 mg L⁻¹ was shown in Fig. S1, in ESI†). All the biosorption tests were conducted twice, and the average values of biosorption capacity were adopted with the standard deviation of 0.0528–0.7482.

To investigate the effect of initial pH value on the biosorption capacity, a series of experiments were carried out over initial pH value range of 2.0 to 10.0 at 20 °C, and PDOS dosage was 2.0 g L⁻¹ and initial TCS concentration was fixed at 50 mg L⁻¹. The optimal pH value (4.0) was used for subsequent experiments.

To explore the role of different functional groups in the biosorption of TCS by PDOS, the comparison tests between PDOS and biosorbents experienced different chemical modification were carried out at 20 °C with initial TCS concentration of 50 mg L⁻¹, initial pH value of 4.0 and biosorbent dosage of 2.0 g L⁻¹.

The experiments to study kinetics of the biosorption process were performed at 20 °C, and initial TCS concentration was in the range of 20–50 mg L⁻¹ with initial pH value of 4.0 and PDOS dosage of 2.0 g L⁻¹.

The experiments to study isotherms of the biosorption process were conducted at temperatures of 20, 35, 50 °C respectively, and initial TCS concentration varied in the range of 20–50 mg L⁻¹ at initial pH value of 4.0 and PDOS dosage of 2.0 g L⁻¹.

Supernatant sample was taken from flasks at predetermined time intervals and then the residual TCS concentration was analyzed by using a UV-vis spectrophotometer (UV-1200, Mapada, China). The amount of TCS adsorbed by PDOS at equilibrium (q_e, mg g⁻¹) could be calculated according to the mass balance equation below:

q_e = (C₀ − C_e)V/m (3)

where C₀ and C_e are the initial and equilibrium concentration in mg L⁻¹, respectively; V is the volume of the solution in L; m is the mass of biosorbent used in g.

2.5. Characterization of biosorbent

The morphology of PDOS before and after biosorption was observed by a scanning electron microscope (SEM, S-4300, Hitacha, Japan). The changes of main chemical components for PDOS before and after TCS uptake were analyzed by an energy-dispersive X-ray spectrometer (EDS, Genesis XM2 60S, Edax, USA). Textural properties of PDOS were determined by the N₂ physisorption at 77 K in an adsorptive and desorptive apparatus (ASAP 2020, Micromeritics, USA) using the Brunauer–Emmet–Teller (BET) method. Before and after both biosorption and chemical modification, PDOS was characterized by a FTIR spectroscopy (Vertex 70, Bruker, Germany) in the wavenumber range 4000–400 cm⁻¹ at 4 cm⁻¹ spectral resolution to evaluate and confirm the functional groups on PDOS that might be bound up with the TCS.

3. Results and discussion

3.1. Characterization of biosorbent

The SEM micrographs of PDOS were given in Fig. 2. The surface of PDOS exhibited heterogeneity (Fig. 2a), parts of PDOS were intact (Fig. 2b), while parts were like morning glory (Fig. 2c). The parts like morning glory might provide a deeper penetration of TCS to the internal structure of PDOS. The EDX analysis of PDOS before and after TCS biosorption was shown in Fig. 3. Comparing the content of chlorine on PDOS, there was a distinct increase (from 0.06 wt% to 0.67 wt%) after TCS uptake, revealing that PDOS has realized a specific biosorption of TCS. The BET surface area, total pore volume, and average pore diameter of PDOS were 2.1560 m² g⁻¹, 0.0027 cm³ g⁻¹, and 8.0372 nm, respectively.

Fig. 2 SEM micrographs of PDOS.

Fig. 3 EDX analysis of PDOS before (a) and after TCS biosorption (b).

Fig. 4 depicted the FTIR spectra of TCS, PDOS before and after TCS biosorption over the range of 4000–400 cm⁻¹. There were a lot of absorption peaks, which revealed the existence of different functional groups on PDOS (Fig. 4(i)). The broad strong band observed at 3437.08 cm⁻¹ might be caused by the overlapping of stretching vibration of O–H in hydroxyl group and symmetric stretching vibration of N–H (amine I).³¹ The bands at 2960.61, 2931.26, 2877.68 cm⁻¹ represented –CH₂ symmetric stretching vibration, –CH₂ asymmetric stretching vibration and –CH₃ symmetric stretching vibration, respectively.³² The bands at 2360.78 and 2341.06 cm⁻¹ were due to the asymmetric stretching of CO₂ caused by background absorption. The medium strong band around 1642.41 cm⁻¹ might be a combination band of –C=O stretching vibration and N–H in-plane deformation (amine I).³³ The band at 1546.85 cm⁻¹ corresponded to a combination of the stretching vibration of C–N and deformation vibration of N–H (amine II).³⁴ The band at 1442.70 cm⁻¹ was attributed to the C–H asymmetric bending vibration in –CH₂ group.³² The band at 1413.77 cm⁻¹ might be due to the symmetrical stretching vibration of C=O in carboxylate and deformation vibration of O–H in alcohols.³² The band at 1232.46 cm⁻¹ was for the stretching vibration of C–N (amine III).³² The band at 1157.24 cm⁻¹ could be assigned to the stretching vibration of C–N in amine III and P–O–C links of the organic phosphate groups.³³ The band at 1082.52 cm⁻¹ was consistent with the stretching vibration of P–O in (C–PO₃²⁻) moiety.³⁵ The bands in finger region demonstrated the existence of sulfur or phosphate groups. Fig. 4(ii) gave the spectra of TCS, which was characterized at 3302.69 cm⁻¹ (phenol hydroxyl group), and 1595.86, 1500.47, 1468.10 cm⁻¹ (benzene rings).³⁶

Fig. 4 FTIR spectra of PDOS before and after TCS biosorption.

By comparing Fig. 4(i) and (iii), an obvious decrease of the band intensity for all bands was found after TCS biosorption. The band at 3437.08 cm⁻¹ became weak, and also a shift of band from 3437.08 to 3436.47 cm⁻¹. These changes might be caused by the split of N–H in amine I groups, and the integration of phenol hydroxyl group of TCS onto PDOS. The band intensity of 1642.41 (amine I), 1548.85 (amine II), 1157.24 and 1232.46 cm⁻¹ (amine III) all clearly decreased after TCS uptake, which revealed that there might be an interaction between TCS and the amine groups. Compared with the band at 1642.41 cm⁻¹, the bands at 1548.85, 1232.46 and 1157.24 cm⁻¹ did not shift, indicating that amide II and III might be not as important as amine I in the TCS biosorption. The band at 1413.77 cm⁻¹ shifted to 1407.98 cm⁻¹ and the band intensity decreased obviously, indicating that C=O in carboxylate and O–H in alcohols played an important role in the TCS biosorption. The band intensity at 1442.70 cm⁻¹ decreased but the band did not shift, which demonstrated that C–H in –CH₂ group was involved in the TCS biosorption onto PDOS. There was a slight shift from 1082.52 to 1081.51 cm⁻¹ as the band intensity had a clear decrease, suggesting that phosphate groups might play a part in the TCS biosorption. FTIR analysis revealed the main functional groups for TCS biosorption onto PDOS might be amine, hydroxyl, carboxyl, phosphate groups and either.

3.2. Effect of pH on TCS biosorption

As an important variable factor, the effect of initial pH on the biosorption of TCS onto PDOS was explored in the initial pH range of 2.0–10.0 at 20 °C with a fixed biosorbent dosage of 2 g L⁻¹, and the results were enlisted in Fig. 5. The profile demonstrated that the biosorption capacity increased with an initial pH increase from 2.0 to 4.0, whereas it declined gradually when the pH further increased (>4.0). The optimum pH was 4.0, and it was selected for subsequent experiments.

Fig. 5 Effect of initial pH value on TCS biosorption by PDOS (C₀ = 50 mg L⁻¹, biosorbent dosage = 2.0 g L⁻¹, temperature = 20 °C).

A possible explanation could be that the potential effect of pH variation on adsorbent (PDOS) and adsorbate (TCS). Functional groups on PDOS whose pK_a are in the range of 8.0–10.0, 9.0–10.0, 3.5–5.0 and 6.5–11.0, corresponded to amine, hydroxyl, carboxyl and phosphate functional groups, respectively.^37–41 Therefore, except carboxyl, the other three functional groups existed as protonated form in acid and neutral conditions, whereas they became progressively negative-charged with increasing pH in alkaline conditions. On the other hand, TCS could be considered as a weak organic acid owing to its pK_a value is 8.14, implying TCS could exist in neutral form when the solution pH < 8.14 and in deprotonated form when the solution pH > 8.14. Consequently, electrostatic interaction was not the reason for TCS biosorption onto PDOS in acid and neutral conditions; further, the hydrogen bonding between the amine, hydroxyl, carboxyl functional groups of PDOS and the phenolic groups of TCS molecules may account for the TCS uptake onto PDOS in pH < 8.14 condition. As mentioned above, the pK_a value of carboxyl is 3.5–5.0, so the hydrogen bonding between the carboxyl functional groups of PDOS and the phenolic groups of TCS molecules was stronger when the pH value was less than 5.0, which might explain that the optimum pH was 4.0. At alkaline conditions (pH > 8.14), TCS molecules were in deprotonated form; meanwhile, the surface of PDOS became more negative-charged. Hence, the electrostatic repulsion between the surface of PDOS and negative-charged TCS inhibited the TCS biosorption onto PDOS. Besides, the hydrogen bonding weakened because of the deprotonation of the amine, carboxyl and hydroxyl groups on PDOS as well as the phenolic groups of TCS molecules. The net result of these two effects was the deterioration in performance as pH increased in alkaline conditions.

3.3. Contribution of functional groups on TCS biosorption by PDOS

To evaluate the effects of chemical treatment on the functional groups which might be involved in the TCS biosorption, the FTIR spectra of raw and chemically-modified PDOS were shown in Fig. 6. Fig. 6(ii) depicted the FTIR spectrum of PDOS that received the methylation of amine treatment using formaldehyde and formic acid (T1). Compared with Fig. 6(i), the bands at 3437.08 and 1546.85 cm⁻¹ significantly decreased in intensity and widened, suggesting a reduction of N–H bond amine groups. The treatment also resulted in that the bands at 1232.46 and 1157.24 cm⁻¹ disappeared and the band intensity at 1642.41 cm⁻¹ clearly decreased. Fig. 6(iii) revealed the FTIR spectrum of esterification of carboxyl groups (T2). The wider and lower bands at 1642.41 and 1413.77 cm⁻¹ indicated that carboxylic groups might be esterified. The FTIR spectrum of PDOS samples after treated with 2,2′-dithiopyridine (T3) was sketched out in Fig. 6(iv), which did not exhibit any obvious differences in the characteristic bands of sulfhydryl groups at 2600–2550 cm⁻¹. However, other bands were changed during the modification process, which suggested that there were conformational changes in the macromolecules of PDOS. The FTIR spectrum of T4 samples was shown in Fig. 6(v). The esterification of phosphate groups resulted in changes of relative bands at 1157.24 and 1082.52 cm⁻¹ whose intensity decreased significantly. Fig. 6(vi) illustrated the FTIR spectrum of PDOS treated with acetone (T5). The band at 2931.26 cm⁻¹ shifted to 2931.68 cm⁻¹ after acetone treatment, which may be caused by the partial elimination of –CH₂ group, implying that the lipid fractions were isolated from PDOS.

Fig. 6 FTIR spectra of PDOS before and after chemical modifications.

Fig. 7 disclosed the effect of different chemical modifications of functional groups on the biosorption capacity of PDOS. The TCS biosorption was obviously restrained when T1, T3 and T5 were conducted. A reduction of 12.67% in TCS biosorption was recorded when sulfhydryl groups present on PDOS was modified (T3), following by lipid extraction (T5) and methylation of amines (T1), where the reduction of the uptake capacity were found to be 9.72% and 8.70%, respectively. T4, that was esterification of phosphate groups, resulted in a slight increase of 4.74% in the TCS uptake. Besides, T2 (esterification of carboxyl) generated an unobservable reduction of 0.18% in TCS biosorption capacity. To the best of our knowledge, functional groups present on agricultural residues biomass and their participation in TCS biosorption have never been explored.

Fig. 7 Effects of different chemical modifications on the TCS biosorption by PDOS (C₀ = 50 mg L⁻¹, initial pH = 4.0, biosorbent dosage = 2.0 g L⁻¹, temperature = 20 °C).

TCS is relatively hydrophobic, with log K_ow of 4.76 at neutral pH and 25 °C. The lipid content of biosorbent decreased when lipid extraction with acetone was carried out (T5), which decreased the hydrophobic interaction between PDOS and TCS molecules, and resulted in a reduction of 9.72% for TCS biosorption capacity. Besides, esterification treatment of phosphate groups with triethyl phosphate and nitromethane (T4) yielded an increase of 4.74% in TCS uptake. Given this, it could be inferred that lipids present on PDOS played a role in the TCS biosorption while the participation of phosphate groups was insignificant to the TCS biosorption onto PDOS. On the other hand, TCS molecule could be seen as a metachlorine and para-o-dichlophenoxy substituted phenol, and it has been confirmed that para-o-dichlophenoxy substitute enabled the hydrogen atoms of phenol functional groups more positive.⁴² Thus, TCS would readily form a hydrogen bonding with the amine, carboxyl and sulfhydryl groups of PDOS. In the current study, methylation of amines (T1), esterification of carboxylic acids (T2) and modification of sulfhydryl groups (T3) weaken the hydrogen bonding between PDOS and TCS molecules. Compared with the control, the biosorption capacity of sulfhydryl-modified (T3) and amine-concealed (T1) biosorbents suffered a degradation of 12.67% and 8.70%, respectively, which demonstrated that sulfhydryl groups made more contribution to the TCS biosorption than amine groups. It was worth noting that the esterification of carboxylic acids (T2) increased the lipid context of PDOS, and it also broke the hydrogen bonding between carboxyl functional groups on PDOS and phenolic groups of TCS molecules. The net result of these two opposing effects led to a very small decrease of 0.18% in biosorption capacity. On the other hand, considering the effect of pH on TCS biosorption onto PDOS, it could be inferred that carboxyl functional groups played an important role in TCS biosorption process. Therefore, the participation of carboxyl groups seemed to be significant to the TCS biosorption by PDOS, but the extent of its contribution to the biosorption could not be confirmed. Above all, amine, carboxyl, sulfhydryl and lipids functional groups might play pivotal roles in the biosorption process. As mentioned in Section 3.2, electrostatic interaction was not responsible for the TCS biosorption in acid and neutral conditions. Thus, we proposed that hydrophobic interaction and hydrogen bonding are the two principal mechanisms controlling the biosorption of TCS onto PDOS.

3.4. Biosorption kinetic study

3.4.1. Pseudo-first-order and pseudo second-order kinetic model. Eqn (4) is the pseudo-first-order rate expression of Lagergren model:⁴³

lg(q_e − q_t) = lg q_e − k₁t/2.303 (4)

where q_t is the TCS uptake at time t in mg g⁻¹; k₁ is the pseudo-first-order rate constant for the biosorption process in min⁻¹.

The pseudo-second-order kinetic model put forward by Ho and Mckay has been applied to evaluate the rate constants of TCS by different biosorbents in recent studies.^18,44 The model is represented as eqn (5):⁴⁵

t/q_t = 1/k₂q_e² + t/q_e (5)

where k₂ is the pseudo-second-order rate constant for the biosorption process in [g (mg min)⁻¹].

The pseudo-first-order and pseudo-second-order kinetic plots were illustrated in Fig. 8, and Table 2 gave the rate constants and correlation coefficients. Pseudo-second-order model fitted the experimental data better than the pseudo-first-order model at the overall biosorption time (0–720 min). And the following discussion from three aspects would explain the reasons. Firstly, the correlation coefficients of the pseudo-second-order model were over the range of 0.9967–0.9996, while the correlation coefficients of the pseudo-first-order model were all less than 0.9394. Secondly, the gaps between the theoretical values q_e,cal and the experimental values q_e,exp of the pseudo-second-order model were much smaller than those of the pseudo-first-order model. Thirdly, it has been reported in many studies that the pseudo-first-order equation of the Lagergren model describes the experimental data well just in the previous 20–30 min in biosorption process.⁴⁶ In conclusion, the pseudo-second-order model might work better for describing the experimental data of TCS biosorption onto PDOS, which implied that the controlling step of the overall rate was the chemisorption mechanism instead of mass transfer.⁴⁵

Fig. 8 Kinetic models of TCS biosorption onto PDOS at different initial concentrations (C₀ = 20, 30, 40, and 50 mg L⁻¹, initial pH = 4.0, biosorbent dosage = 2.0 g L⁻¹, temperature = 20 °C), (a) pseudo-first-order model; (b) pseudo-second-order model; (c) intra-particle diffusion model.

Table 2 Kinetic constants of TCS biosorption onto PDOS at different concentrations (C₀ = 20, 30, 40, and 50 mg L⁻¹, initial pH = 4.0, biosorbent dosage = 2.0 g L⁻¹, temperature = 20 °C)

C0 (mg L^−1)	20	30	40	50
qe,exp (mg g^−1)	2.01	5.35	9.42	13.66
Pseudo-first-order kinetic model
k1 (min^−1)	0.0048	0.0041	0.0044	0.0048
qe,cal (mg g^−1)	1.24	3.61	5.89	7.43
R^2	0.9071	0.9394	0.9190	0.9022
Pseudo-second-order kinetic model
k2 [g (mg min) ^−1]	0.0099	0.0039	0.0026	0.0025
qe,cal (mg g^−1)	2.08	5.33	9.47	13.76
R^2	0.9967	0.9967	0.9993	0.9996
IPD model
kint [mg (g min^1/2)]	0.1421	0.3482	0.5955	0.8048
I (mg g^−1)	0.0938	0.1148	0.6992	2.4268
R^2	0.9339	0.9822	0.9809	0.9671

---

3.4.2. Intra-particle diffusion model. The intra-particle diffusion model (IPD) proposed by Webber and Morris is one of the most widely used models for studying the rate limiting steps of biosorption.⁴⁷ In this study, the IPD model was employed to study the rate-determining step in biosorption of TCS onto PDOS, and it can be presented as follows:⁴⁸

q_t = k_intt^0.5 + I (6)

where k_int is the intra-particle diffusion rate constant in mg (g min^1/2)⁻¹; and I is the intercept in mg g⁻¹.

The multi-linearity plots of q_t against t^0.5 for the biosorption of TCS onto PDOS at various initial TCS concentrations were shown in Fig. 8c. All the plots showed the similar features: the plots consisted of three separate segments—an initial portion attributed to the diffusion of TCS from the solution to the external surface of PDOS through the boundary layer diffusion, a linear portion due to the intraparticle diffusion and a plateau region due to final equilibrium step. The rate constants of IPD model from the linear portion were enlisted in Table 2. The correlation coefficients R² varied from 0.9399 to 0.9822, confirming that IPD model fitted the experimental data well. The linear portions didn't pass through the origin, revealing that the rate determining factors include not only intra-particle diffusion but also boundary layer diffusion. All the intercepts (I) were positive, indicating that there was a rapid biosorption in a short period of time. Besides, I was higher at a higher TCS concentration supporting that the larger the intercept, the greater the boundary layer effect.

3.5. Biosorption equilibrium study

The Langmuir isotherm predicts the formation of a monolayer biosorption onto the biosorbent surface with a fixed number of identical energy sites,⁴⁹ and it can be written as the following eqn (7):

q_e = Q⁰bC_e/(1 + bC_e) (7)

The linear form of Langmuir isotherm can be represented by eqn (8):

q_e = Q⁰ − q_e/bC_e (8)

where Q⁰ is the maximum amount TCS adsorbed under a given condition in mg g⁻¹; b is the Langmuir constant related to the affinity of the binding sites in L mg⁻¹.

The Temkin isotherm model is based on the assumption that the heat of biosorption decreases linearly owing to biosorbent–adsorbate interactions,⁵⁰ and it can be expressed in the following form:

q_e = RT ln(AC_e)/b (9)

The linear form of Temkin isotherm is written as the following eqn (10):

q_e = B ln A + B ln C_e (10)

where B = RT/b, B is the isotherm constant related to the heat of biosorption; A is the equilibrium binding constant corresponding to the maximum binding energy in L mg⁻¹; R [8.314 J (mol K)⁻¹] is the universal gas constant; T is temperature in K.

Based on the hypothesis that the mechanism for biosorption in micropores is that of pore-filling rather than layer-by-layer surface coverage, D–R isotherm was widely used to fit the equilibrium data for purpose of understanding the biosorption type.⁵¹ And it can be expressed in the following form:

q_e = Q⁰ exp(−K_DR[RT ln(1 + 1/C_e)]²) (11)

ε = RT ln(1 + 1/C_e) (12)

The linear form of D–R isotherm by logarithmic transfer of both sides is generally expressed as follow:

ln q_e = ln Q⁰ − K_DRε² (13)

where ε is polanyi potential; K_DR is the constant related to the mean free energy of biosorption in (mol kJ⁻¹)²; E is the mean free energy of biosorption in kJ mol⁻¹, and it can be calculated from the value of K_DR using eqn (14):

The linear plots for the three different isotherms were depicted in Fig. 9, and the isotherm constants were enlisted in Table 3. For Langmuir isotherm, the values of Q⁰ and b were negative, implying Langmuir isotherm wasn't appropriate for describing the equilibrium data. Similar results were also reported for the biosorption of Acid Red 19 onto slag and fly ash.⁵² The correlation coefficients of Temkin isotherm model ranged from 0.8540 to 0.9704, indicating that Temkin isotherm model did not match with the experimental data satisfactorily. As for D–R isotherm, the values of R² were 0.9997, 0.9902 and 0.9994 for 20, 35 and 50 °C, respectively. Q⁰ decreased with increasing temperature, which was consistent with the variation of E. Therefore, D–R isotherm could describe the equilibrium data of the TCS biosorption by PDOS well. It's widely accepted that the value of E is between 8 and 16 kJ mol⁻¹ suggesting that the biosorption process could be explained by ion-exchange, while for the value of E is less than 8 kJ mol⁻¹, the biosorption process might be physical in nature.⁵³ Further from Table 3, the values of E were in the range of 0.0332–0.0448 kJ mol⁻¹, corresponding to a physisorption rather than ion-exchange. What's more, the value of E decreased with increasing temperature indicated that the biosorption process was an exothermic biosorption.

Fig. 9 Isotherm models of TCS biosorption on PDOS at different temperatures (C₀ = 20, 30, 40, and 50 mg L⁻¹, initial pH = 4.0, biosorbent dosage = 2.0 g L⁻¹, temperature = 20, 35 and 50 °C), (a) Langmuir isotherm model; (b) Temkin isotherm model; (c) D–R isotherm model.

Table 3 Isotherm parameters for the biosorption of TCS onto PDOS at different temperatures (C₀ = 20, 30, 40, and 50 mg L⁻¹, initial pH = 4.0, biosorbent dosage = 2.0 g L⁻¹, temperature = 20, 35 and 50 °C)

Temperature (°C)	20	35	50
Langmuir
Q^0 (mg g^−1)	−1.17	−0.49	−0.35
b (L mg^−1)	−0.0370	−0.0300	−0.0238
R^2	0.9950	0.9973	0.9886
Temkin
A (L mg^−1)	0.0567	0.0468	0.0441
B	34.8890	20.6090	8.5603
R^2	0.9704	0.8540	0.8960
D–R
Q^0 (mg g^−1)	124.55	82.76	27.45
KDR (mol kJ^−1)^2	249.52	403.25	454.19
E (kJ mol^−1)	0.0448	0.0352	0.0332
R^2	0.9997	0.9902	0.9994

---

In terms of the correlation coefficients and other constants calculated from these three isotherm models, D–R isotherm gave a better description than Langmuir and Temkin isotherm. Table 4 summarized the Q⁰ values of different adsorbents used for the TCS removal in recent literatures.^{11–14,18,54} Q⁰ in this study was 124.55 mg g⁻¹ gained from D–R isotherm model at 20 °C, which was lower than carbon nanotubes, but much higher than activated carbon, kaolinite, montmorillonite, modified zeolites and membranes, suggesting its potential application in TCS removal.

Table 4 Maximum adsorption capacity (Q⁰) for TCS removed from aqueous solutions by various adsorbents

Adsorbents	Q^0 (mg g^−1)	pH	T (K)	Reference
Montmorillonite	3.30	3.0	298	14
Montmorillonite	8.60	7.0	298	54
Kaolinite	22.02	3.0	298	14
Cetylpyridinium bromide modified zeolites	46.95	6.7	298	18
Activated carbon	70.42	6.0	298	14
Electrospun fibrous membranes	99.00	6.0	298	11
Protonated de-oiled soybean	124.55	4.0	293	Present work
Multi-walled carbon nanotubes	166.83	3.0	298	12
Multi-walled carbon nanotubes	434.70	7.0	296	13
Single-walled carbon nanotube	558.20	7.0	296	13

---

3.6. Thermodynamic analysis

Thermodynamic parameters such as Gibbs free energy change (ΔG°, kJ mol⁻¹), enthalpy change (ΔH°, kJ mol⁻¹) and entropy change [ΔS°, J (mol K)⁻¹] were evaluated by the following equations:

K_p = C_ae/C_e (15)

ΔG° = −RT ln K_p (16)

ln K_p = ΔS°/R − ΔH°/RT (17)

where C_ae is the concentration of TCS adsorbed in the biosorbent at equilibrium in mg L⁻¹; K_p is the distribution coefficient. The values of ΔH° and ΔS° can be calculated from the slope and intercept of the linear plot of ln K_p versus 1/T.

The values of ΔG°, ΔH°, ΔS° were listed in Table 5. ΔG° increased with temperature increasing from 20 to 50 °C, implying that a reduction in spontaneous nature of TCS biosorption onto PDOS. The negative values of ΔH° for all the temperatures showed that the biosorption process was exothermic in nature, which was consistent with the experimental observations. The value of ΔS° was −172.27 [J (mol K)⁻¹], confirming that a distinct increase in randomness of system.

Table 5 Thermodynamic parameters of TCS biosorption by PDOS

T (K)	ΔG° (kJ mol^−1)	ΔH° (kJ mol^−1)	ΔS° [J (mol K)^−1]	R^2
293	1.50	−49.04	−172.27	0.9979
308	3.71
323	6.06

---

4. Conclusion

PDOS can be used as a low-cost and effective biosorbent for the TCS removal with a high capacity of 124.55 mg g⁻¹ gained from D–R isotherm. TCS uptake was strongly affected by pH (optimum pH was 4.0) and electrostatic interaction wasn't the one controlling the TCS biosorption in acid and neutral condition. Through chemical modification to block functional groups on PDOS, five functional groups capable of biosorption were identified: amine, carboxyl, sulfhydryl, phosphate, and lipids groups. The results demonstrated that sulfhydryl groups gave the largest contribution to the biosorption of TCS onto PDOS, followed by lipids and amine groups. The distinct decrease of biosorption capacity for sulfhydryl-modified and amine-methylated biosorbents indicated that TCS biosorption onto PDOS might be primarily controlled by the hydrogen bonding between sulfhydryl, amine groups present on PDOS and the phenolic groups of TCS molecules. Lipid extraction also led to a decrease in biosorption capacity, implying that the role of the hydrophobic interaction between lipids on PDOS and TCS molecular should not be ignored. The participation of carboxyl groups seemed to be significant to the TCS biosorption by PDOS, but the extent of its contribution to the biosorption could not be confirmed. Phosphate groups contributed little to TCS biosorption.

Acknowledgements

We would like to thank the National Natural Science Foundation of China (51078007, 51378027, 51578015), the Natural Science Foundation of Beijing (8162010) and Beijing Talent Foundation of BJUT (2013-JH-L06) for the financial supports of this study.

References

1. G. M. Bruce, R. C. Pleus and S. A. Snyder, Environ. Sci. Technol., 2010, 44, 5619–5626.
2. G. Ying and R. S. Kookana, Environ. Int., 2007, 33, 199–205.
3. P. J. Vikesland, E. M. Fiss, K. R. Wigginton, K. McNeill and W. A. Arnold, Environ. Sci. Technol., 2013, 6764–6772.
4. H. Singer, S. Müller, C. Tixier and L. Pillonel, Environ. Sci. Technol., 2002, 36, 4998–5004.
5. C. T. Anger, C. Sueper, D. J. Blumentritt, K. McNeill, D. R. Engstrom and W. A. Arnold, Environ. Sci. Technol., 2013, 47, 1833–1843.
6. J. Zhao, G. Ying, Y. Liu, F. Chen, J. Yang and L. Wang, J. Hazard. Mater., 2010, 179, 215–222.
7. L. M. McMurry, M. Oethinger and S. B. Levy, Nature, 1998, 394, 531–532.
8. N. Surolia and A. Surolia, Nat. Med., 2001, 7, 167–173.
9. D. R. Orvos, D. J. Versteeg, J. Inauen, M. Capdevielle, A. Rothenstein and V. Cunningham, Environ. Toxicol. Chem., 2002, 21, 1338–1349.
10. G. Dhillon, S. Kaur, R. Pulicharla, S. Brar, M. Cledón, M. Verma and R. Surampalli, Int. J. Environ. Res. Public Health, 2015, 12, 5657–5684.
11. J. Xu, J. Niu, X. Zhang, J. Liu, G. Cao and X. Kong, Emerg. Contam., 2015, 1, 25–32.
12. S. Zhou, Y. Shao, N. Gao, J. Deng and C. Tan, Clean: Soil, Air, Water, 2013, 41, 539–547.
13. H. Cho, H. Huang and K. Schwab, Langmuir, 2011, 27, 12960–12967.
14. S. K. Behera, S. Oh and H. Park, J. Hazard. Mater., 2010, 179, 684–691.
15. S. Ravikumar, I. Ganesh, I. Yoo and S. H. Hong, Process Biochem., 2012, 47, 758–765.
16. A. Javaid, R. Bajwa, U. Shafique and J. Anwar, Biomass Bioenergy, 2011, 35, 1675–1682.
17. Z. S. Birungi and E. M. N. Chirwa, J. Hazard. Mater., 2015, 299, 67–77.
18. C. Lei, Y. Hu and M. He, Chem. Eng. J., 2013, 219, 361–370.
19. E. Nakkeeran, N. Saranya, M. S. Giri Nandagopal, A. Santhiagu and N. Selvaraju, Int. J. Phytorem., 2016, 18, 812–821.
20. D. Sud, G. Mahajan and M. Kaur, Bioresour. Technol., 2008, 99, 6017–6027.
21. R. Dalgaard, J. Schmidt, N. Halberg, P. Christensen, M. Thrane and W. A. Pengue, Int. J. Life Cycle Assess., 2007, 13, 240–254.
22. V. K. Gupta, A. Mittal, L. Krishnan and J. Mittal, J. Colloid Interface Sci., 2006, 293, 16–26.
23. A. Mittal, J. Mittal, A. Malviya and V. K. Gupta, J. Colloid Interface Sci., 2010, 344, 497–507.
24. N. Daneshvar, D. Salari and S. Aber, J. Hazard. Mater., 2002, 94, 49–61.
25. S. Silke and V. Bohumil, Environ. Sci. Technol., 1995, 12, 3049–3058.
26. A. Kapoor and T. Viraraghavan, Bioresour. Technol., 1997, 61, 221–227.
27. L. R. Drake, S. Lin, G. D. Rayson and P. J. Jackson, Environ. Sci. Technol., 1996, 30, 110–114.
28. K. Parvathi and R. Nagendran, World J. Microbiol. Biotechnol., 2008, 24, 2865–2870.
29. A. Markowska, J. Olejnik and J. Michalski, Chem. Ber., 1975, 108, 2589–2592.
30. J. M. Tobin, D. G. Cooper and R. J. Neufeld, Enzyme Microb. Technol., 1990, 12, 591–595.
31. J. Gao, Q. Zhang, K. Su, R. Chen and Y. Peng, J. Hazard. Mater., 2010, 174, 215–225.
32. H. Schulz and M. Baranska, Vib. Spectrosc., 2007, 43, 13–25.
33. J. Gao, J. Wang, X. Wu, Q. Zhang and Y. Peng, Fresenius Environ. Bull., 2011, 20, 51–62.
34. M. Ghasemi, S. Mashhadi, M. Asif, I. Tyagi, S. Agarwal and V. K. Gupta, J. Mol. Liq., 2016, 213, 317–325.
35. G. Naja, C. Mustin, B. Volesky and J. Berthelin, Water Res., 2005, 39, 579–588.
36. L. Qian, Y. Guan and H. Xiao, Int. J. Pharm., 2008, 357, 244–251.
37. M. Tsezos, H. Eccles and S. Hunt, Immobilization of ions by biosorption, Ellis Horwood, Chichester, UK, 1986.
38. S. Schiewer and B. Volesky, Environ. Microbe-Met. Interact., 2000, 14, 329–362.
39. J. Buffle, Complexation reactions in aquatic systems. An analytical approach, John Wiley and Sons, New York, USA, 1988.
40. R. E. Martinez, D. S. Smith, E. Kulczycki and F. G. Ferris, J. Colloid Interface Sci., 2002, 253, 130–139.
41. X. Lopez, M. Schaefer, A. Dejaegere and M. Karplus, J. Am. Chem. Soc., 2002, 124, 5010–5018.
42. G. Zhang, M. Sun, Y. Liu, H. Liu, J. Qu and J. Li, Langmuir, 2015, 31, 1820–1827.
43. S. Lagergren, K. Sven. Vetenskapsakad. Handl., 1898, 24, 1–39.
44. X. Zhu, Y. Liu, G. Luo, F. Qian, S. Zhang and J. Chen, Environ. Sci. Technol., 2014, 48, 5840–5848.
45. Y. S. Ho and G. McKay, Process Biochem., 1999, 34, 451–465.
46. Y. S. Ho and G. McKay, Process Saf. Environ. Prot., 1998, 76, 332–340.
47. G. F. Malash and M. I. El-Khaiary, Chem. Eng. J., 2010, 163, 256–263.
48. J. C. Morris and W. J. Weber, ASCE J. Sanit. Eng. Div., 1963, 83, 31–59.
49. I. Langmuir, J. Am. Chem. Soc., 1918, 40, 1361–1403.
50. D. Temkin, Acta Physicochim. URSS, 1934, 1, 36–52.
51. M. Saeed, A. Rusheed and N. Ahmed, J. Radioanal. Nucl. Chem., 1996, 211, 283–292.
52. A. Andreadakis, K. R. Ramakrishna and T. Viraraghavan, Water Sci. Technol., 1997, 36, 189–196.
53. M. Mahramanlioglu, I. Kizilcikli and I. O. Bicer, J. Fluorine Chem., 2002, 115, 41–47.
54. B. Liu, J. Lu, Y. Xie, B. Yang, X. Wang and R. Sun, J. Colloid Interface Sci., 2014, 418, 311–316.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra06702f

---