Sorption of phenols on hexadecyltrimethylammonium- modified bentonite: Application of Polanyi−Manes potential theory

Abstract

This study investigated the sorption of phenol and 4-chlorophenol (4-CP) on natural bentonite modified with hexadecyltrimethylammonium (HDTMA) cation. The Freundlich, Langmuir, Dubinin−Radushkevich (DR), Sips, and Polanyi−Dubinin−Manes (PDM) models fitted the sorption data well (R² > 0.92). The Freundlich coefficient and the maximum sorbed amount of the Langmuir and PDM models of 4-CP were higher than phenol because of higher hydrophobicity (log K_ow = 2.39 for 4-CP and 1.46 for phenol). The PDM model that includes solubility and molar volume was highly useful in predicting the sorption of phenols having widely different hydrophobicity and solubility. The characteristic curves, the plot of sorbed volume (q_v) versus the sorption potential per molar volume (ε/V_m) of 4-CP and phenol were distinctly different although they have similar chemical compositions. The selectivity of 4-CP (3.72) was higher than that of phenol (0.27) in binary sorption systems. The sorbed volume (q_v) in the binary sorption was remarkably reduced and the characteristic curve had wider distribution owing to competition in pore-filling. The sorption behaviors were elucidated by partitioning and pore-filling mechanisms. Among the tested binary sorption models, the modified Langmuir competitive model was the best in the prediction of the binary sorption (R² > 0.98).

Keywords

HDTMA-bentonite partitioning phenols polanyi potential theory pore-filling sorption

Introduction

Bentonite, a swelling clay is a structurally good material for sorption¹ but not effective for hydrophobic organic compounds (HOCs) owing to inherent hydrophilicity. Modification of the bentonite by exchanging its hydrophilic interlayer to quaternary ammonium cation with long hydrocarbon chains was investigated to improve the sorption ability for the HOCs.² Many researchers^1–6 have studied hexadecyltrimethylammonium (HDTMA), having long-chain (C > 12) quaternary ammonium. The HDTMA-modified organic matter in bentonite (HDTMA-bentonite) can be 10 to 30 times more effective in removing organic contaminants from water than organic matter in natural soils.^2,7

The sorption of HOCs on organically modified clays (organoclays) was explained by linear (partitioning) and nonlinear (adsorption) isotherms.^2,6,8–13 Several studies^14–17 have reported on the sorption of phenols and HOCs on HDTMA-clays. Generally, sorption isotherm models such as Freundlich, Langmuir, DR, Sips, and PDM models have been to predict single sorption. Marsal et al.¹⁸ and Houhoune et al.¹⁹ summarized the characteristics of HDTMA modified bentonite in different surfactant concentrations. Alkaram et al.²⁰ investigated the characteristics of the organoclay (HDTMA- or phenyltrimethylammonium (PTMA)-modified bentonite) and analyzed sorption behavior by Langmuir and Freundlich models. Rawajfih and Nsour²¹ analyzed partition and adsorption contributions to the total sorption amounts of phenol and chlorinated phenols onto HDTMA-bentonite.

The solubility-normalized models were reported for their successful predictions of the nonlinear sorption of phenols.^7,10,11,13 The Freundlich model normalized by solubility has been used for single sorption of organic contaminants (e.g., BTEX, PAHs, and PCBs).^7,11 Carmo et al.¹⁰ also used a unit equivalent Freundlich model for naphthalene and phenanthrene sorption. Shin and Song² used the Freundlich model normalized by solubility to describe the sorption characteristics of phenols onto HDTMA-modified montmorillonite; linear (partitioning) isotherm satisfying Henry’s law in low (C→0) and non-linear adsorption in the high concentration region (C→S_w), respectively. The ideal adsorbed solution theory (IAST) linked to the Freundlich model normalized by solubility resulted in a good performance for fitting binary sorption data of phenols. However, the use of other solubility-normalized models such as DR and PDM paid little attention to predict the binary sorption of phenols.

A few studies have reported on using the DR model to explain the sorption mechanisms (physical versus chemical sorption) for the sorption of 4-chlorophenol (4-CP)²² and herbicides²³ on tetrabutylammonium (DEDMAM)-montmorillonite. Fuller et al.²⁴ applied the DR model, which is a simple form of PDM model, for nonionic organics (benzene, CCl₄, TCE, and 1,2-DCB) to tetraalkylammonium-bentonites. The mechanistic sorption behaviors (partitioning and microporous adsorption) was analyzed by the characteristic curve of the PDM model. The PDM model includes the aqueous concentration normalized by the solubility of organic compounds in the water based on the Polanyi potential theory.^10,25 The potential theory was useful in finding the correlation of sorption isotherms for the classes of structurally similar organic compounds.²⁶ A normalizing factor to describe the sorbate-sorbent interaction forces was proposed to concurrently correlate the sorption isotherms of several organic compounds.^11,27,28

In this work, we used natural bentonite modified by the HDTMA cation to the 50% of its cation exchange capacity (CEC) (50% HDTMA-bentonite) as a sorbent. Single and binary sorption experiments were conducted using of phenol and 4-CP as sorbates. Single sorption data were fitted by the sorption isotherm models such as Freundlich, Langmuir, DR, Sips, and PDM models. Binary sorption data were analyzed by the Sheindorf−Rebhun−Sheintuch (SRS), Murali−Aylmore (MA), Langmuir competitive model (LCM), modified Langmuir competitive model (MLCM), P−factor model, and the IAST linked to the single sorption isotherm models.

Material and methods

Materials

Aqueous solutions of phenol (99.0−100.5%, Sigma−Aldrich, Munich, Germany) and 4-CP (4-CP, ≥99%, Sigma−Aldrich, Munich, Germany) were prepared by dissolving in distilled and deionized (DDI) water. The solution pH was adjusted to 6.0 using 10 mM phosphate buffer containing 8.8 mM potassium phosphate monobasic (KH₂PO₄, 99.5%, Kanto Chemical, Tokyo, Japan) and 1.2 mM potassium phosphate dibasic (K₂HPO₄, 98–101%, Yakuri Pure Chemical, Kyoto, Japan). The physicochemical properties of 4-CP and phenol are summarized in Table 1.

Table 1.

Physicochemical properties of phenols used.

Parameter	Phenol	4-CP
M.W.	78.0	128.56
Solubility (mmol/L)	879.82	186.82
log K_ow	1.46	2.39
pK_a at 25 ^oC	9.99	9.41
UV wavelength (nm)	272.3	283.2
Molar volume (V_m) (mL/mol)	87.863	99.814
Molecular dimension (nm)	0.43−0.57*	0.43–0.63**
Molecular volume (nm³)	0.162*	0.178**

log K_ow: Octanol-water partition coefficient.

*Lorenc-Grabowska⁵⁰; **Lorenc-Grabowska et al.⁵¹

The natural bentonite used in this study was obtained from Gampo, Kyungpook Province, Korea. The CEC analyzed for CEC using Rhoades method²⁹ was 94 meq per 100 g bentonite. The HDTMA chloride cationic surfactant solution (25 wt%, Sigma−Aldrich, Munich, Germany) was used as organic modifier.

Preparation of HDTMA-bentonite

The impurities in bentonite was washed with distilled water for several times at room temperature. The bentonite suspension was filtered through a 0.45 μm membrane filter paper (ϕ = 110 mm, Advantec, Tokyo, Japan), and impurities in the filtrate was examined using a UV-Vis spectrophotometer (8453, Agilent, Santa Clara, CA, USA). The washed bentonite was dried in an oven for 24 hours at 60 °C and then kept in an amber bottle.³⁰

The schematic diagram for preparation of HDTMA-bentonite is showed as Figure 1. The HDTMA-bentonite was prepared by cation exchange sorption. To prepare HDTMA-bentonite, 30 g of the washed bentonite was added into a 2-L baffled beaker containing 1.0 L of 4,500 mg/L HDTMA solution (≈50% of the CEC of the bentonite) and stirred for 24 h at 250 rpm. The modified bentonite was collected by filtering through the membrane filter, washed with DDI water and dried in an oven at 60 °C for a day. The HDTMA-bentonite was sieved through a US standard No. 200 mesh (75 µm) and kept in an amber bottle.²⁶ The interlayer spacing of the washed bentonite and the HDTMA-bentonite were determined by X-ray diffraction (XRD, X'pert Pro MRD, Malvern PANalytical, Malvern, Almelo, The Netherlands) with a Cu Kα source and analyzed by the Bragg equation (nλ = 2dsinθ). As the cationic surfactant was inserted into the interlayer spaces, the interlayer spacing increases.³¹ The interlayer spacing of HDTMA-bentonite (18.51Å) was larger than that of bentonite (15.04Å). The organic carbon content (f_oc) of the 50% HDTMA-bentonite measured by elemental analysis was 5.62 (w/w%). The physicochemical characteristics of HDTMA-bentonite are summarized in Table 2. The spectra of the HDTMA-bentonite after HDTMA modification were analyzed using Fourier transform infrared spectroscopy (FT-IR) (Spectrum GX & AutoIMAGE, PerkinElmer, Waltham, MA, USA). The bands at 2850 cm⁻¹ and 2920 cm⁻¹ are assigned to the stretching vibration band of −CH₂ (Figure 2). Similar result was also reported by Yaming et al.³² Zhang et al.³³ reported that more than 99% of HDTMA were sorbed on Na-montmorillonite and less than 1.0% of the total adsorption were desorbed. Lee et al.³⁴ also reported that the desorbed concentration of HDTMA was less than 2.0% of the total adsorbed mass. Therefore, the adsorbed HDTMA is not easily desorbing from the modified clay.

Figure 1.

The schematic diagram of preparation of HDTMA-bentonite.

Table 2.

Physicochemical characteristics of HDTMA-bentonite used.

Parameter	HDTMA-bentonite
pH	7.5
pH_pzc	7.9
Basal spacing (Å)	18.51
Organic carbon content (f_oc, %)	5.62
Pore size (Å)	156.7
Pore volume (cm³/g)	0.01
Surface area (m²/g)	2.65

Figure 2.

FT-IR spectrums of bentonite and HDTMA-bentonite.

Sorption onto HDTMA-bentonite

For single sorption, 0.5 g of HDTMA-bentonite was added into 50 mL Erlenmeyer flasks capped with Teflon-faced silicone septa before adding 20 mL of chemical solutions; 0.78 to 77.8 mmol/L for 4-CP and 1.06 to 106.3 mmol/L for phenol, respectively. The flasks were shaken using an orbital shaker for 24 h at 150 rpm. After mixing, the supernatant was separated by centrifugation for 30 min at 2,000 rpm. The aqueous-phase concentration at equilibrium was determined by the spectrophotometer. The solid-phase concentration at equilibrium (q, mmol/g) was calculated by:

q = \frac{(C_{0} - C) W}{W}

(1)

where C₀ is the initial aqueous-phase concentration (mmol/L), C is the aqueous-phase concentration at equilibrium (mmol/L), V is the volume of chemical solution (L), and W is the sorbent weight (g). Sorption experiments onto the 50% HDTMA-bentonite were run in triplicate.

For the binary sorption experiment, binary solution was prepared with mixing 4-CP and phenol at equal molar ratio. After sorption, the equilibrium concentrations of the binary solutions were also determined by the spectrophotometer.

Sorption models

Single sorption mode

The single sorption data were fitted by Freundlich model:

q = K_{F} C^{N_{F}}

(2)

where K_F is the Freundlich coefficient [

(mmol / kg) / {(mmol / L)}^{N_{F}}

] and N_F is the Freundlich exponent (–), respectively.

The Langmuir model was also fitted to the single sorption data:

q = \frac{q_{m L} b_{L} C}{1 + b_{L} C}

(3)

where q_mL is the maximum sorption capacity (mmol/g) and b_L is the site energy factor (L/mmol), respectively.

The DR model is different from the Langmuir model because it assumes heterogenous surface with different sorption potential. The DR model was used to characterize whether sorption is physical or chemical:^22,23

q = q_{m D} \exp (- β ε^{2})

(4)

where q_mD is the saturation sorption capacity (mmol/g), β is a constant (mol²/J²) related to the mean free energy of sorption per mole of the sorbate, and ε (J/mol) is the Polanyi potential (= RT ln(1 + 1/C)); R is the ideal gas constant (J/mol⋅K), and T is the absolute temperature (K). The mean free energy, E (J/mol), can be calculated by:^22,23

E = \frac{1}{\sqrt{2 β}}

(5)

where the mean free energy (E) value can be used to differentiate the sorption mechanisms. When the E value is in the range of 8 to 16 kJ/mol, the sorption occurs by ion-exchange. The E value less than 8 kJ/mol indicates the physical sorption process, whereas E value greater than 16 kJ/mol, the sorption is chemical.^22,23

The Sips model is defined as:³⁵

q = \frac{q_{m S} {(b_{S} C)}^{N_{S}}}{1 + {(b_{S} C)}^{N_{S}}}

(6)

where q_mS is the maximum sorption capacity (mmol/g), b_S (L/mmol) is Sips equilibrium constant, and N_S (–) is the Sips model exponent. If N_S = 1, the Sips model is equal to the Langmuir model. The Sips model becomes the Freundlich model in the low concentration region (C → 0):

q = q_{m S} {(b_{S} C)}^{N_{S}} = q_{m S} b_{S}^{N_{S}} C^{N_{S}} = K_{F S} C^{N_{S}}

(7)

where

K_{F S} = q_{m S} b_{S}^{N_{S}}

[

(mmol / g) / {(mmol / L)}^{N_{S}}

]. The model reduces to monolayer coverage in the high concentration region (C → ∞):

q = \frac{q_{m S} {(b_{S} C)}^{N_{S}}}{{(b_{S} C)}^{N_{S}}} = q_{m S}

(8)

The PDM model has been used to describes nonlinear sorption in highly meso- and microporous solids hypothesizing a pore-filling mechanism:^11,12,24

q = q_{m} \exp {- α {(\frac{ε}{V_{m}})}^{β}} = q_{m} \exp {- α {(\frac{R T ln(S / C)}{V_{m}})}^{β}}

(9)

where q_m is the maximum sorption capacity (mmol/g) of the PDM model, ε is effective adsorption potential (Polanyi potential, = RT ln(S/C), J/mol), S is the aqueous solubility of the solute (mmol/L), and V_m is the molar volume of the solute (mL/mol). α (mL/J) ^β and β are the fitting parameters.

The sorption isotherm model was fitted using TableCurve 2 D® (Version 5.01, SYSTAT Software, Inc.).

Binary sorption model

The SRS, MA, P–factor, LCM, MLCM, and the IAST linked to single sorption models were used to predict binary sorption.

The SRS postulates that the Freundlich model represents single sorption³⁶ assuming an exponential sorption energy distribution for the solute. The SRS model to predict the binary sorption is given as:

q_{i} = K_{F_{i}} C_{i} {(\sum_{j = 1}^{N} α_{i j} C_{j})}^{N_{F_{i}} - 1}

(10)

where q_i is solid-phase concentrations and C_i is aqueous-phase concentrations in the binary sorption, respectively. The subscripts i and j indicates the two solutes 4-CP and phenol and N is the total number of competing phenols. The parameters K_Fi and N_Fi are the Freundlich parameters obtained from single sorption of solute i. The α _ij is the competition coefficient between the two solutes i and j.

The MA model proposed by Murali and Aylmore³⁷ was to fit the binary sorption data:

q_{i} = \frac{K_{F_{i}} C_{i}^{N_{F_{i}} + 1}}{\sum_{j = 1}^{N} a_{i j} C_{j}}

(11)

where a_ij is the parameter that accounts for the competition between the solutes i and j.

The LCM has been typically used to analyze binary sorption behaviors:^38–40

q_{i} = \frac{q_{m L, i} b_{L, i} C_{i}}{1 + \sum_{j = 1}^{N} b_{L, j} C_{j}}

(12)

where q_mL,i (mmol/g) and b_L,i (L/mmol) are the Langmuir model parameters obtained from single sorption.

In the P−factor model, a lumped capacity factor P_i is defined to compare and correlate the maximum sorption capacity of single and binary sorption:^41,42

P_{i} = \frac{q_{m L, i}}{q_{m L, i}^{*}}

(13)

Where q_mL,i is the maximum sorption capacity of solute i in a single sorption, while $q_{m L, i}^{*}$ is the maximum sorption capacity of solute i in a binary component. The P−factor model based the Langmuir model is expressed as:

q_{i} = \frac{1}{P_{i}} \frac{b_{L, i} q_{m L, i} C_{i}}{1 + b_{L, i} C_{i}}

(14)

The MLCM is defined as:^42,43

q_{i} = \frac{q_{m L, i} b_{L, i} (C_{i} / η)}{1 + \sum_{j = 1}^{2} b_{L . j} (C_{j} / η)}

(15)

where q_mL,i and b_L,i are the Langmuir model parameters for solute i in single sorption system (Table 3). The constant interaction factor, η, can be estimated from the binary sorption data.

Table 3.

Sorption model parameters for single sorption of 4-CP and phenol onto HDTMA-bentonite.

Freundlich model
Solute	K _F	N_F (−)	–	R²	SSE
4-CP	0.2420 ± 0.0275	0.4178 ± 0.0344		0.9520	0.1300
Phenol	0.1137 ± 0.0189	0.4866 ± 0.0426		0.9498	0.0870

Langmuir model
Solute	q_mL (mmol/g)	b_L (L/mmol)	–	R²	SSE
4-CP	1.2866 ± 0.0330	0.1296 ± 0.0113		0.9923	0.0268
Phenol	1.1596 ± 0.0343	0.0429 ± 0.0036		0.9942	0.0130

DR model
Solute	q_mD (mmol/g)	β (mol²/J²)	E (KJ/mol)	R²	SSE

4-CP	0.9473 ± 0.0381	1.227E-6 ± 2.941E-7	0.638	0.9234	0.2668
Phenol	0.7992 ± 0.0284	1.403E-5 ± 2.827E-6	0.189	0.9489	0.1139

Sips model
Solute	q_mS (mmol/g)	b_S (L/mmol)	N_S (−)	R²	SSE
4-CP	1.3369 ± 0.0771	0.1175 ± 0.0188	0.9393 ± 0.0772	0.9926	0.0259
Phenol	1.0257 ± 0.0375	0.0558 ± 0.0045	1.2142 ± 0.0738	0.9963	0.0081

Polanyi-Manes model
Solute	q_m (mmol/g)	α [(mL/J) ^β ]	β (−)	R²	SSE
4-CP	1.2167 ± 0.0389	2.065E-5 ± 1.675E-5	2.3667 ± 0.1707	0.9938	0.0215
Phenol	1.0538 ± 0.0391	4.433E-7 ± 2.680E-7	3.0443 ± 0.2118	0.9963	0.0082

Unit: K_F = [ ${(mmol/g)/(mmol/L)}^{N_{F}}$ ], α = (mL/J) ^β .

The SRS, MA, LCM, P-factor, and MLCM were predicted by using the curve–fitting toolbox MATLAB^® (Version R2019a, Mathworks, Inc.). The estimated parameters are listed in Table 4.

Table 4.

Sorption capacity of organic pollutants by HDTMA-bentonite.

Pollutant	Sorbent	SorptionCapacity(mmol/g)	References
Phenol	Bentonite	0.0089	Banat et al.⁵⁰
Phenol		0.0026	Alkaram et al.²⁰
4-chlorophenol		0.0827	Koumanova and Peeva-Antova⁵¹
Phenol	HDTMA-bentonite	1.1596	This study
		0.0896	Alkaram et al.²⁰
		0.0105	Yapar et al.⁵²
		1.1157	Rawajfih and Nsour²¹
4-cholophenol		1.2866	This study
4-cholophenol		0.3368	Rawajfih and Nsour²¹
Chlorobenzene		0.2143	Messikh et al.⁵³
Trichloroethylene		0.0025	Cho et al.⁵⁴
2,4-dichlorophenol		0.6687	Ruan et al.⁵⁵
2,4-dichlorophenol		0.2718	Rawajfih and Nsour²¹
m-nitrophenol		0.3886	Koyuncu et al.⁵⁶
p-nitrophenol		0.2970	Koyuncu et al.⁵⁶
2,4,6-trichlorophenol		0.3779	Anirudhan and Ramachandran⁵⁷

We also employed the IAST proposed by Radke and Prausnitz,⁴⁴ to fit the binary sorption (4-CP and phenol) onto HDTMA–bentonite. The IAST calculations are described in detail elsewhere.^6,45,46

Results and discussion

Single sorption

The sorption of 4-CP and phenol was performed at pH 6 using HDTMA-modified bentonites as sorbents. Since the pK_a values of 4-CP (= 9.41) and phenol (= 9.99) were greater than the pH values at equilibrium (pH 6), neutral speciation is dominant (4-CP: 99.96%, phenol: 99.99%). The single sorption of 4-CP and phenol onto HDTMA-bentonite is depicted in Figure 3. The sorption of the phenols in the HDTMA-bentonites showed a type I isotherm with higher sorption affinity.^47,48 The type I isotherms are implicated strong interactions between sorbate and sorbent where the magnitude is well reflected by the hydrophobicity (log K_ow) (see Table 1). The strong sorption is mainly because of the dissolution (or partitioning) of the neutral speciation (pH < pKa) of phenols onto the pseudo-organic phase formed in the interlamellar spacing of the HDTMA-bentonite.^6,45,49 However, the sorption capability of HDTMA-bentonite is rather limited because it was prepared by cation-exchange adsorption HDTMA only up to 50% CEC of bare bentonite surface.

Figure 3.

Single sorption of 4-CP and phenol onto HDTMA-bentonite. Lines represent sorption models.

The affinity of 4-CP sorption was higher than that of phenol. Table 3 lists the fitted model parameters of the Freundlich, Langmuir, Sips, DR, and PDM models.

The experimental data were fitted by the Freundlich model well (R² > 0.950). The Freundlich coefficient, K_F, is an indicator of the sorption affinity of HDTMA-bentonite. The K_F values of 4-CP were higher than phenol due to higher hydrophobicity (i.e., higher log K_ow). The exponent (N_F) is indicative of the intensity of sorption; the N_F values at equilibrium were within 0.33 to 0.49, indicating highly non-linear and favorable sorption. This explains that the sorption affinity of phenols onto the sorbent increases with hydrophobicity (i.e., log K_ow values).

As listed in Table 3, the Langmuir model was better fitted (R² > 0.992) than the Freundlich model (R² > 0.950). The 4-CP had higher q_mL than phenol because of higher hydrophobicity (i.e., higher log K_ow) (see Table 1). The site energy factor, b_L, reflected sorption affinity. The b_L of 4-CP was also higher than that of phenol.

The DR model was also predicted the single sorption well (0.923 < R² < 0.950). The 4-CP had higher q_mD than that of phenol, the same pattern as in the Langmuir model. The q_mD in the DR model were slightly less than the q_mL in the Langmuir model (Table 3) owing to the difference in sorption mechanisms. The estimated mean free energy, E in equation (5) were less than 8 kJ/mol indicating that the phenols are sorbing on the HDTMA-bentonite mainly via physical sorption.

The sorption capacities of various pollutants onto HDTMA-bentonite were compared in Table 4. HDTMA-modification improves the sorption capacity of phenolic compounds. The sorption capacities of HDTMA–bentonite that used in this study were higher than those of HDTMA-bentonites in the literature.

The three-parameter models; Sips (R² > 0.993) and PDM (R² > 0.994) models also fitted the data well. The q_mS of 4-CP in the Sips model and q_m of 4-CP in the PDM model was also higher than that of phenol, consistent with q_mL in the Langmuir model and q_mD in the DR model. We attempted the dual-mode model,^11,58–60 including partitioning and adsorption; however, the sum of squares of errors was significantly reduced. Therefore, we focused on the PDM model only in this study.

The PDM model that includes the pore-filling concept correlated well with the 4–CP and phenol sorption with the highest R² (4-CP: 0.994; phenol: 0.996) and the lowest SSE values (4-CP: 0.02; phenol: 0.008) (Table 3). This indicates that pore-filling could be a critical mechanism in 4-CP and phenol sorption onto HDTMA-bentonite.^22,23 The q_m of phenol 4–CP was higher than that of phenol, although the solubility (S) of 4-CP (=186.8 mmol/L) is less than that of phenol (= 879.8 mmol/L). The molar volume (V_m) and molecular dimension of 4-CP (V_m = 99.8, dimension = 0.43–0.63 nm) is higher than that of phenol (V_m = 87.8, dimension = 0.43 − 0.57 nm) (Table 1), indicating sorption by the pore-filling mechanism. The q_m value of 4-CP was higher than that of phenol, the same pattern as the q_mS of the Sips model and q_mL of the Langmuir model. The fitting parameters β was 2.4 for 4-CP and 3.3 for phenol. Xia and Ball⁵⁸ reported β values ranging from 1.4 to 2.7 for sorption of nine different chlorobenzenes and PAH sorbates on aquitard solids. Fuller et al.²⁴ used a simplified PDM model by setting β = 2 for the sorption of chlorobenzene, CCl₄, TCE, and 1,2-DCB onto tetraalkylammonium-bentonites. Long et al.⁶¹ reported β values ranging from 0.93 to 1.505 for the sorption of naphthalene on nonpolar polymeric sorbents such as XAD-4, NAD 150, and NDA-1600. The β values were highly specific for different chemicals and sorbents.

The molar volume (V_m) and solubility of solute (S) are included in the PDM model, equation (9). Therefore, a sorption model such as the PDM would be suitable in predicting the sorption of organic compounds with widely different hydrophobicity (log K_ow), molar volume (V_m), and solubility (S), especially for binary competitive sorption.

Regarding the Polanyi theory, the plot (so called characteristic curve) of the sorbed volume (q_v) against the Polanyi potential normalized by the molar volume (ε/V_m), was used to describe the sorption process. In Figure 4, the ε/V_m value of phenol was higher than that of 4-CP because of a higher S/C and lower V_m for phenol. In the Polanyi theory, the solutes with similar compositions can be described by a single characteristic curve.²⁴ However, in this study, the characteristic curves of 4-CP and phenol were separated, although they have a similar chemical structure (Figure 4).

Figure 4.

Characteristic curves of 4-CP/phenol binary sorption onto HDTMA-bentonite.

Fuller et al.²⁴ used the correlation curve to explain the sorption mechanisms of HOCs on tetrabutylammonium–modified bentonite. The DR model, a simple form of the PDM model was potentially useful in predicting the sorption of dissimilar compounds to a sorbent because it could be estimated from the solubility-normalized isotherm of one solute. They observed that CCl₄ and TCE with similar compositions fall into the same curve, but not the benzene with different composition.

Xia and Ball⁵⁸ reported the unified correlation curves for sorption of benzene, chlorobenzene, dichlorobenzene, and trichlorobenzene on natural solids. Long et al.⁶¹ also observed that the characteristic curve (q_v versus ε/V_m) of naphthalene sorption on three different polymeric sorbents (XAD-4, NDA-150 and NDA-1600) had distinctly different curves, although the polymeric sorbents have the same polymer matrix. When the micropore volume was considered (V_micro), the characteristic curve of (q_v/V_micro versus ε/V_m) fall on a single curve for the sorption of naphthalene onto three polymeric sorbents. This explains that micropore-filling in the PDM model could also explain the sorption behaviors in HDTMA-bentonite. The molecular dimension of HOCs molecules (phenol = 0.43−0.57 nm, 4-CP = 0.43−0.63 nm) is less than the pore size of HDTMA-bentonite (15.7 nm), these molecules can be adsorbed by pore-filling.

It is generally accepted that the sorption of HOCs, including phenols onto organoclays, occurs by partitioning in the low and adsorption in the high concentration regions, respectively.^{6,7,12,45,62,63} Condon⁶⁴ showed that in the case of the standard deviation σ = 0.5, the PDM model becomes the Freundlich model (β = 1), and in the other case of activation energy E_a ≈ RT, the linear isotherm satisfying Henry’s law is obtained. When N_F = 1, the Freundlich isotherm becomes linear isotherm. The PDM model inherently includes linear (partitioning) and nonlinear adsorption (pore-filling) isotherms. Therefore, the PDM model is useful in explaining sorption mechanisms in organoclays, especially for the solutes with widely different solubility, molar volume (molecular dimension), and hydrophobicity.

Binary sorption

The binary sorption experiments were conducted for the 4-CP and phenol system, as shown in Figure 5. In comparison with single sorption, the sorbed amount in the binary sorption system was highly reduced. To determine the degree of reduction quantitatively, the Langmuir model was fitted to the binary sorption data of individual solute. Table 5 shows the q_mL values of the Langmuir in the single and binary sorption were compare as well as the percentage reduction of q_mL in the binary sorption. As the sorption affinity in the single sorption is stronger (i.e., 4-CP > phenol), the reduction in q_mL for the solute in the binary sorption system becomes less (i.e., 41.9% (4-CP) < 64.9% (phenol)). The competition between sorbates drives the weakly sorbed solute to be desorbed from the sorbed phase, thereby allowing the stronger (more hydrophobic) solute to occupy more sorption sites on the HDTMA-bentonite.

Figure 5.

Binary sorption of 4-CP/Phenol onto HDTMA-bentonite. Lines represent binary sorption models.

Table 5.

Model parameters for binary sorption of 4-CP (1) and phenol (2) onto HDTMA-bentonite.

SRS	α ₁₂	α ₂₁	R²	SSE
SRS	0.2896 ± 0.0426	5.5643 ± 0.3722	0.9303/0.9585	0.0574/0.0102
MA	a ₁₂	a ₂₁	R²	SSE
MA	0.1494 ± 0.0202	2.0279 ± 0.0982	0.9286/0.9641	0.0588/0.0089
CLM	−	−	R²	SSE
CLM			0.9176/0.9480	0.2395/0.0546
P−Factor	P _i	−	R²	SSE
P−Factor	1.7217/2.8477		0.9696/0.8910	0.0884/0.1144
MCLM	η₁	η₂	R²	SSE
MCLM	1.0050 ± 0.1050	1.509 ± 0.0595	0.9852/0.9847	0.0121/0.0038
IAST−Fr	−	−	R²	SSE
IAST−Fr			0.9617/0.9683	0.1114/0.0332
IAST−Lang	−	−	R²	SSE
IAST−Lang			0.9793/0.9890	0.0603/0.0116
IAST−Sips	−	−	R²	SSE
IAST−Sips			0.8232/0.7614	0.5142/0.2504

The Freundlich model−based binary sorption models, SRS (0.930 < R² < 0.959) and MA (0.929 < R² < 0.964) models, predicted the binary sorption well. The Langmuir model–based models, LCM (0.918 < R² < 0.948), P-factor model (0.891 < R² < 0.970), and MLCM (R² = 0.985), also showed good prediction. In the SRS model analysis, α₂₁ (5.564) was greater than α₁₂ (0.290), indicating that the presence of the competing solute affects the 4-CP (solute 1) more than the phenol (solute 2). The same pattern was observed in the MA model; a₂₁ (2.028) was greater than a₁₂ (0.149), elucidating that more hydrophobic solute (4-CP) affect less hydrophobic solute (phenol) in the binary sorption. The P-factor value of phenol (2.848) was greater than that of 4-CP (1.722); that is, phenol is more affected than 4-CP. In the MLCM analysis, the interaction factor (η) values of phenol (η₁ = 1.509) were greater than that of 4-CP (η₂ = 1.005). Again, 4-CP suppresses the phenol sorption in the binary sorption. Among the tested models, MLCM was the best predicting model regarding R² values (R² = 0.985). The selectivity of phenols in binary sorption system was calculated at C_initial = 77.8 mmol/L using equations (16) and (17). The selectivity of 4-CP (K_d,4-CP/K_d,phenol = 3.72) was higher than that of phenol (K_d,phenol/K_d,4-CP = 0.27) in binary sorption systems.

Selectivity of 4-CP =_{K_{d, 4 -CP}} / K_{d, phenol}

(16)

Selectivity of phenol = K_{d, phenol} /_{K_{d, 4 -CP}}

(17)

The IAST predictions for the binary sorption are also presented in Figure 5. The parameters of Freundlich, Langmuir, and Sips models (Table 3) obtained from single sorption were used in the predictive IAST. The R² values, IAST−Fr (R² > 0.962), IAST−Lang (R² > 0.979) and IAST−Sips: R² > 0.761) showed that the IAST were favorable in their predictions of 4-CP/phenol binary sorption (Table 4).

The characteristic curve for the binary sorption also exhibited a distinctly separated curve (Figure 6). The comparison of the characteristic curve of the single (Figure 4) and binary sorption (Figure 6) showed that the maximum sorbed volume (q_vm) was remarkably reduced for both 4-CP (from 0.12 to 0.07 mL/g) and phenol (from 0.09 to 0.03 mL/g), and the q_v curve exhibited wider distributions in ε/V_m than those in single sorption, explaining the competition effect in pore-filling. The steric hindrance because of the differences in the molecular dimension (phenol = 0.43−0.57 nm, 4-CP = 0.43−0.63 nm) and molecular volume (phenol = 0.162 nm³, 4CP = 0.178 nm³) could also be critical in binary competitive sorption.

Figure 6.

Characteristic curves of 4-CP/phenol binary sorption onto HDTMA-bentonite.

Conclusions

This study investigated the sorption mechanisms of phenol and 4-CP onto HDTMA-bentonite by various single and binary sorption isotherm models. The single sorption isotherms were fitted well by the Freundlich, Langmuir, DR, Sips, and PDM models (0.92 < R² < 0.99). The 4-CP had higher the K_F of Freundlich model and q_mL of Langmuir model than phenol, mainly because of its higher hydrophobicity (log K_ow). Because the neutral speciation of 4-CP and phenol are dominant at the working solution pH 6, the sorption occurs because of partitioning (or dissolution) into the pseudo–organic medium. In the sorption of the phenols with widely different hydrophobicity (log K_ow), solubility (S), and molar volume (V_m), a sorption model such as the PDM model that includes solubility and molar volume would be more appropriate, especially for binary competitive sorption. Although phenol has higher solubility (S) but less molar volume (V_m) than 4–CP, phenol had less sorption capacity than 4-CP. According to the Polanyi theory, the characteristic curves of phenols onto HDTMA-bentonite were significantly separated, although they have similar chemical compositions. The steric hindrance related to the size and shape of phenol and 4-CP molecules (Table 1) could also induce differences in sorption affinity.

In the binary sorption, SRS, MA, LCM, MLCM, and IAST linked with single sorption models fitted positively to the binary sorption (0.76 < R² < 0.99). The higher the interaction coefficients (α in SRS, a in MA and η in MCLM) of 4-CP are less affected by presence of phenol in binary sorption. Compared to single sorption, the q_v,m of pharmaceuticals was reduced (94% for 4-CP and 67% for phenol), and the characteristic curve exhibited a wide distribution because of the competition in pore-filling. In summary, the sorption of phenols occurs by i) partitioning onto the pseudo-organic phase formed by HDTMA modification and ii) nonlinear adsorption (pore-filling) mechanism. The developed model can be used to predict multi-component sorption of contaminants (including hydrophobic organic compounds, dyes, heavy metals and pharmaceuticals and personal care products) in wastewater treatment and also fate and transport modeling in groundwater.

Footnotes

Acknowledgments

This work was supported by Korea Environment Industry & Technology Institute (KEITI) through Subsurface Environmental Management (SEM) Projects, funded by Korea Ministry of Environment (MOE) (2019002480005).

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Won Sik Shin

Jiyeon Choi finished PhD at School of Architecture, Civil, Environmental and Energy Engineering in Kyungpook National University, Korea. She is a postdoctoral research associate in Kyungpook National University.

Dong-Ik Song received PhD in Department of Chemical Engineering, Purdue University, USA. He is a Professor at Department of Chemical Engineering, Kyungpook National University, Korea. His research interests are sorption mechanism and modeling.

Won Sik Shin received PhD in Civil and Environmental Engineering, Louisiana State University, USA. He is a Professor at School of Architecture, Civil, Environmental and Energy Engineering, Kyungpook National University, Korea. His research is concerned with biological treatment(MBBR), sorption, organoclays, modeling, oxidation, soil and groundwater contamination, remediation technology, toxicity and sediment. He has published 65 international and 46 national journals.

References

Sternik

Gładysz-Płaska

Grabias

, et al. A thermal, sorptive and spectral study of HDTMA-bentonite loaded with uranyl phosphate. J Therm Anal Calorim 2017; 129: 1277–1289.

Shin

Song

D-I.

Solubility-normalized Fruendlich isotherm for the prediction of sorption of phenols in HDTMA modified montmorillonite. Geosci J 2005; 9: 249–259.

Smith

Jaffe

PR.

Benzene transport through landfill liners containing organophilic bentonite. J Environ Eng 1994; 120: 1557–1559.

Smith

Winquist

AS.

Permeability of earthen liners containing organobentonite to water and two organic liquids. Environ Sci Technol 1996; 30: 3089–3093.

Slimani

Ahlafi

Moussout

, et al. Adsorption of hexavalent chromium and phenol onto bentonite modified with hexadacyltrimethylammonium bromide (HDTMABr). J Adv Chem 2014; 8: 1601–1611.

Song

D-I

Choi

Shin

WS.

The modified song isotherm model: application to multisolute sorption of phenols in organoclays using the ideal adsorbed solution theory. Environ Technol 2021; 42: 1591–1602

Lee

Mortland

Boyd

, et al. Shape-selective adsorption of aromatic molecules from water by tetramethylammonium-smectite. J Chem Soc Faraday Trans I 1989; 85: 2953–2962.

Huh

Song

Jeon

YW.

Dual-mode sorption model for single- and multisolute sorption onto organoclays. Sep Sci Technol 1999; 34: 571–586.

Huh

Song

Jeon

YW.

Sorption of phenol and alkylphenols from aqueous solution onto organically modified montmorillonite and applications of dual-mode sorption model. Sep Sci Technol 2000; 35: 243–259.

10.

Carmo

Hundal

Thompson

ML.

Sorption of hydrophobic organic compounds by soil materials: application of unit equivalent Freundlich coefficients. Environ Sci Technol 2000; 34: 4363–4369.

11.

Allen-King

Grathwohl

Ball

WP.

New modeling paradigms for the sorption of hydrophobic organic chemicals to heterogeneous carbonaceous matter in soils, sediments, and rocks. Adv Water Res 2002; 25: 985–1016.

12.

Kleineidam

Schüth

Grathwohl

Solubility-normalized combined adsorption-partitioning sorption isotherms for organic pollutants. Environ Sci Technol 2002; 36: 4689–4697.

13.

Ran

Xing

Suresh

, et al. Importance of adsorption (hole-filling) mechanism for hydrophobic organic contaminants on an aquifer kerogen isolate. Environ Sci Technol 2004; 38: 4340–4348.

14.

Boyd

Mortland

Chiou

CT.

Sorption characteristics of organic compounds on hexadecyltrimethylammonium-smectite. Soil Sci Soc Am J 1988; 52: 652–657.

15.

Karapanagioti

Sabatini

Bowman

RS.

Partitioning of hydrophobic organic chemicals (HOC) into anionic and cationic surfactant-modified sorbents. Water Res 2005; 39: 699–709.

16.

Yildiz

Gur

Adsorption of phenol and chlorophenols on pure and modified sepiolite. J Serb Chem Soc 2007; 72: 467–474.

17.

Dong

Chen

, et al. Adsorption of bisphenol a from water by surfactant-modified zeolite. J Colloid Interface Sci 2010; 348: 585–590.

18.

Marsal

Bautista

Ribosa

, et al. Adsorption of polyphenols in wastewater by organo-bentonites. Appl Clay Sci 2009; 44: 151–155.

19.

Houhoune

Nibou

Chegrouche

, et al. Behaviour of modified hexadecyltrimethylammonium bromide bentonite toward uranium species. J Environ Chem Eng 2016; 4: 3459–3467.

20.

Rawajfih

Nsour

Characteristics of phenol and chlorinated phenols sorption onto surfactant-modified bentonite. J Colloid Interface Sci 2006; 298: 39–49.

21.

Alkaram

Mukhlis

Al-Dujaili

AH.

The removal of phenol from aqueous solutions by adsorption using surfactant-modified bentonite and kaolinite. J Hazard Mater 2009; 169: 324–332.

22.

Akçay

Yurdakoç

The characterization of prepared organomontmorillonite (DEDMAM) and sorption of phenoxyalkanoic acid herbicides from aqueous solution. J Colloid Interface Sci 2006; 296: 428–433.

23.

Akçay

Characterization and adsorption properties of tetrabutylammonium montmorillonite (TBAM) clay: thermodynamic and kinetic calculations. J Colloid Interface Sci 2006; 296: 16–21.

24.

Fuller

Smith

Burns

SE.

Sorption of nonionic organic solutes from water to tetraalkylammonium bentonites: mechanistic considerations and application of the Polanyi–Manes potential theory. J Colloid Interface Sci 2007; 313: 405–413.

25.

Chiou

Kile

Rutherford

, et al. Sorption of selected organic compounds from water to a peat soil and its humic-acid and humin fractions: potential sources of the sorption nonlinearity. Environ Sci Technol 2000; 34: 1254–1258.

26.

Crittenden

Sanongraj

Bulloch

, et al. Correlation of aqueous-phase adsorption isotherms. Environ Sci Technol 1999; 33: 2926–2933.

27.

Long

, et al. Polanyi-based models for the adsorption of naphthalene from aqueous solutions onto nonpolar polymeric adsorbents. J Colloid Interface Sci 2008; 319: 12–18.

28.

Long

, et al. Removal of benzene and methyl ethyl ketone vapor: comparison of hypercrosslinked polymeric adsorbent with activated carbon. J Hazard Mater 2012; 203–204: 251–256.

29.

Rhoades

JD.

Cation exchange capacity. In: Methods of soil analysis, Part II, chemical and microbiological properties. 2nd ed. Madison: American Society of Agronomy, 1982, pp.149–157.

30.

Kim

J-H

Shin

Song

D-I

, et al. Multi-step competitive sorption and desorption of chlorophenols in surfactant modified montmorillonite. Water Air Soil Pollut 2005; 166: 367–380.

31.

Mallavarapu

Naidu

Preparation, characterization of surfactants modified clay minerals and nitrate adsorption. Appl Clay Sci 2010; 48: 92–96.

32.

Yaming

Mingliang

Zhipeng

, et al. Organic modification of bentonite and its application for perrhenate (an analogue of pertechnetate) removal from aqueous solution. J Taiwan Inst Chem E 2016; 62: 104–111.

33.

Zhang

Sparks

Scrivner

NC.

Sorption and desorption of quaternary amine cations on clays. Environ Sci Technol 1993; 27: 1625–1631.

34.

Lee

J-J

Choi

Park

J-W.

Simultaneous sorption of lead and chlorobenzene by organobentonite. Chemosphere 2002; 49: 1309–1315.

35.

Carvajal-Bernal

Gomez-Granados

Giraldo

, et al. Application of the sips model to the calculation of maximum adsorption capacity and immersion enthalpy of phenol aqueous solutions on activated carbons. Eur J Chem 2017; 8: 112–117.

36.

Sheindorf

Rebhun

Sheintuch

A Freundlich-type multicomponent isotherm. J Colloid Interface Sci 1981; 79: 136–142.

37.

Murali

Aylmore

LAG.

Competitive adsorption during solute transport in soils: 1. Mathematical models. Soil Sci 1983; 135: 143–150.

38.

Sağ

Kutsal

Fully competitive biosorption of chromium(VI) and iron(III) ions from binary metal mixtures by R. arrhizus: use of the competitive Langmuir model. Process Biochem 1996; 31: 573–585.

39.

Siitonen

Sainio

Kaspereit

Theoretical analysis of steady state recycling chromatography with solvent removal. Sep Purif Technol 2011; 78: 21–32.

40.

Toth

Torocsik

Tombacz

Competitive adsorption of phenol and 3-chlorophenol on purified MWCNTs. J Colloid Interface Sci 2012; 387: 244–249.

41.

McKay

Al-Duri

Simplified model for the equilibrium adsorption of dyes from mixtures using activated carbon. Chem Eng Process 1987; 22: 145–156.

42.

Choy

KKH

Porter

McKay

Langmuir isotherm models applied to the multicomponent sorption of acid dyes from effluent onto activated carbon. J Chem Eng Data 2000; 45: 575–584.

43.

Valderrama

Barios

Farran

, et al. Evaluating binary sorption of phenol/aniline from aqueous solutions onto granular activated carbon and hypercrosslinked polymeric resin (MN200). Water Air Soil Pollut 2010; 210: 421–434.

44.

Radke

Prausnitz

JM.

Thermodynamics of multisolute adsorption from dilute liquid solutions. AIChE J 1972; 18: 761–768.

45.

Song

Shin

WS.

Three-parameter empirical isotherm model: its application to sorption onto organoclays. Environ Sci Technol 2005; 39: 1138–1143.

46.

Shin

WS.

Competitive sorption of anionic and cationic dyes onto cetylpyridinium-modified montmorillonite. J Environ Sci Heal A 2008; 43: 1459–1470.

47.

Parida

Dash

Patel

, et al. Adsorption of organic molecules on silica surface. Adv Colloid Interface Sci 2006; 121: 77–110.

48.

DD.

Adsorption analysis: equilibria and kinetics. London: Imperial College Press, 1998.

49.

Sanghi

Bhattacharya

Review on decolorisation of aqueous dye solutions by low cost adsorbents. Color Technol 2002; 118: 256–269.

50.

Banat

Al-Bashir

Al-Asheh

, et al. Adsorption of phenol by bentonite. Environ Pollut 2000; 107: 391–398.

51.

Koumanova

Peeva-Antova

Adsorption of p-chlorophenol from aqueous solutions on bentonite and perlie. J Hazard Mater 2002; 90: 229–234.

52.

Yapar

Özbudak

Dias

, et al. Effect of adsorbent concentration to the adsorption of phenol on hexadecyl trimethyl ammonium-bentonite. J Hazard Mater 2005; 121: 135–139.

53.

Messikh

Bougdah

Bousba

, et al. Modeling the adsorption of chlorobenzene on modified bentonite using an artificial neural network. Curr Opin Green Sustain Chem 2020; 3: 100026.

54.

Cho

H-H

Lee

Hwang

S-J

, et al. Iron and organo-bentonite for the reduction and sorption of trichloroethylene. Chemosphere 2005; 58: 103–108.

55.

Ruan

Liu

Chang

C-Y

, et al. Preparation of organobentonite by a novel semidry-method and its adsorption of 2,4-dichlorophenol from aqueous solution. Int Biodeterior Biodegradation 2014; 95: 212–218.

56.

Koyuncu

Yildiz

Salgin

, et al. Adsorption of o-, m- and p-nitrophenols onto organically modified bentonites. J Hazard Mater 2011; 185: 1332–1339.

57.

Anirudhan

Ramachandran

Removal of 2,4,6-trichlorophenol from water and petroleum refinery industry effluents by surfactant-modified bentonite. J Water Process Eng 2014; 1: 46–53.

58.

Xia

Ball

WP.

Adsorption-partitioning uptake of nine low-polarity organic chemicals on a natural sorbent. Environ Sci Technol 1999; 33: 262–269.

59.

Xia

Pignatello

JJ.

Detailed sorption isotherms of polar and apolar compounds in a high-organic soil. Environ Sci Technol 2001; 35: 84–96.

60.

Nguyen

Sabbah

Ball

WP.

Sorption nonlinearity for organic contaminants with diesel soot: method development and isotherm interpretation. Environ Sci Technol 2004; 38: 3595–3603.

61.

Long

, et al. Adsorption of naphthalene onto macroporous and hypercrosslinked polymeric adsorbent: effect of pore structure of adsorbents on thermodynamic and kinetic properties. Colloid Surface A 2009; 333: 150–155.

62.

Zhu

Chen

Sorption behavior of p-Nitrophenol on the interface between anion-cation organobentonite and water. Environ Sci Technol 2000; 34: 2997–3002.

63.

Zhu

Chen

Shen

Sorption of phenol, p-nitrophenol, and aniline to dual-cation organobentonites from water. Environ Sci Technol 2000; 34: 468–475.

64.

Condon

JB.

Equivalency of the Dubinin–Polany equations and the QM based sorption isotherm equation. B. Simulations of heterogeneous surfaces. Micropor Mesopor Mater 2000; 38: 377–383.