Synthesis, characterisation and application of imidazolium based ionic liquid modified montmorillonite sorbents for the removal of amaranth dye

Abstract

The removal of amaranth dye using montmorillonite modified with an ionic liquid (IL) was investigated. An ionic liquid (1-methyl, 3-decahexyl imidazolium) was synthesised, and characterised by nuclear magnetic resonance (¹H and ¹³C NMR), Fourier transform infrared (FT-IR) spectroscopy, high resolution mass spectroscopy (HR-MS) as well as thermal gravimetric analysis and differential thermal (TGA/DCS) analysis. The ionic liquid was thereafter used to modify sodium montmorillonite (Na⁺-Mt) to form a hydrophobic montmorillonite that has a positively charged surface, by adding IL in excess of the cation exchange capacity (CEC) of Na⁺-Mt. Energy dispersive X-ray analyser (EDX) was used to determine the chemical composition of Na⁺-Mt. Modified montmorillonite (Mt-IL) and Na⁺-Mt was characterised using FT-IR, scanning electron microscopy (SEM), X-ray diffraction (XRD) and Brunauer–Emmett–Teller (BET) method, and used in the adsorption study of an anionic amaranth dye. A higher CEC of the Mt-IL (112 meq per 100 g) than Na⁺-Mt (89 meq per 100 g) was observed. The effects of sorbent dosage, dye concentration, solution pH and contact time were investigated in order to determine optimal experimental conditions. Optimum adsorption was obtained at pH 2 and adsorption data was better described by the pseudo-second-order kinetics and Langmuir adsorption isotherm. A maximum adsorption capacity of 263.2 mg g⁻¹ was calculated. The thermodynamic parameters ΔG°, ΔH° and ΔS° were also investigated at temperatures of 313, 303 and 293 K. Column studies were performed on the modified material and the data generated was applied to the Thomas model where high column adsorption capacities of 393.7, 580.9 and 603.6 mg g⁻¹ at different concentrations were obtained.

---

1. Introduction

Montmorillonite has been extensively used for the adsorption of organic compounds.^1–3 Recently, research has focused on the modification of montmorillonite using a variety of methods and modifiers. Montmorillonite has been modified using dimethyl dihydrogenated tallow ammonium, ammonium and phosphonium surfactants.^4,5 Other modifiers include polypropylene on bentonite,⁶ hexadecyltrimethylammonium (HDTMA) on montmorillonite, montmorillonite intercalated with cetyltrimethylammonium octadecylamine and cetylpyridinium^7–9 among others.

These modifications have been based on the high cation exchange capacity (CEC) of montmorillonite, its swelling capacity and high surface area, all resulting in strong adsorption and adsorption capacities.^10,11 The aluminosilicate layer of montmorillonite has a ratio of 2 : 1 type structure, with an octahedral Al_2–3(OH)₆ layer.¹² The position of the aluminosilicate results in a weak interaction between adjacent lamellae (Si–O⋯O–Si). As a result, montmorillonite easily separates and coupled with its high cationic exchange capacity ensures favourable conditions for intercalation reactions. Si⁴⁺ substituted with Al³⁺ in the tetrahedral layer and Mg²⁺ substituted for Al³⁺ in the octahedral layer results in a net negative charge on the clay surfaces. The deficit in charge is balanced by the exchangeable Na⁺ or Ca²⁺ cation in the interlayer.¹⁰ If the inorganic cations of montmorillonite are replaced by large alkyl ammonium ions, the physical properties of the montmorillonite changes and the hydrophilic nature of montmorillonite is converted to a hydrophobic form (organo-montmorillonite).¹³ Organo-montmorillonite has been used for the adsorption of dye,^14,15 phenolic compounds and other aromatic compounds¹⁶ as well as pesticides.¹⁷

Ionic liquids (ILs) have recently been used for the modification of montmorillonite¹⁸ and nanoparticles.¹⁹ There is presently limited information on the adsorption of organics (including dye) using IL modified materials. To the best of our knowledge the only reported studies are for the adsorption of reactive red-120 (RR) and 4-(2-pyridylazo) resorcinol (PAR) from aqueous solution by Fe₃O₄ magnetic nanoparticles using ionic liquid as a modifier which was recently studied by Absalan and co workers²⁰ and who reported a significant increase in adsorption capacity. Also, adsorption of folic acid, riboflavin, and ascorbic acid from aqueous samples by Fe₃O⁺ magnetic nanoparticles using ionic liquid as modifier was also investigated²¹ but no reports have been reported on IL modified montmorillonite. This lack of research on IL modified montmorillonite for the adsorption of dyes has prompted our investigation into the adsorption behaviour of dyes on IL modified montmorillonite.

They (ILs) are low-melting-point salts and form liquids that consist only of cations and anions. Ionic liquids are often considered as green solvents capable of replacing traditional organic solvents,²² composing of an organic cation and an inorganic anion.²³ The anion mainly controls their water miscibility²⁴ whereas the length of the alkyl chains of the cations influences all the other properties²⁵ such as the hydrophobicity of the material.

In this work, cationic IL (1-methyl, 3-decahexzyl imidazolium) was synthesised (the choice of 16 carbon atoms was guided by the work of Beall²⁶) and used in modifying montmorillonite which was then used to adsorb anionic amaranth dye. Amaranth is a class of azo dye, which is toxic, and its removal from wastewaters is very important as its presence, even in very low concentration, affects aquatic life and invariably affects the food chain. To the best of our knowledge this is the first report of the use of ionic liquid modified montmorillonite for the adsorption of dye.

2. Experimental section

2.1 Materials

Montmorillonite K 10 powder (CAS number 1318-93-0) used in the present study was purchased from Sigma Aldrich. Its chemical formula is Na_2/3Si₈(Al_10/3Mg_2/3)O₂₀(OH)₄. Methyl-imidazole, 1-bromodecahexane, montmorillonite, NaCl and amaranth (CAS number 915-67-3) were purchased from Sigma Aldrich. Double distilled deionized water was used for all sample preparation.

2.2 Apparatus

¹H and ¹³C NMR spectra of the derivatives were recorded on an AVANCE DPX-400 (400 MHz) spectrophotometer (Bruker, U.S.A.) with tetramethylsilane (TMS) as the internal standard. High-resolution mass data were obtained using a Bruker micro TOF-Q II ESI instrument operating at ambient temperature. Thermal gravimetric analysis and differential thermal analysis (TGA-DTA/DSC) (SDT Q 600 V 20.9 Build 20 instrument) was used to measure changes in chemical and physical properties of the material. This was measured as a function of increasing temperature from ambient to 1000 °C (with a constant heating rate of 5 °C min⁻¹) under nitrogen atmosphere with a flow rate of 50 mL min⁻¹. FT-IR spectra of the IL – sodium montmorillonite (Na⁺-Mt) and sodium montmorillonite modified with 1-methyl, 3-decahexyl imidazolium (Mt-IL) samples were recorded on a Perkin Elmer 100 series FT-IR spectrometer. BET Tri-star II 3020.VI.03 was used to determine the surface area, pore-size, pore volume and pore size distribution. The material was degassed using (Micromeritics vacprep 061, sample degas system) at 90 °C for 1 h and was increased to 200 °C for 12 h. The sample was then analysed by Tri-star II 3020.VI.03 under nitrogen atmosphere at 77 K. The surface morphology of the material was examined by scanning electron microscopy (SEM) using the Leo 435 VP model. Chemical composition analysis of montmorillonite was determined using on a Leo 1450 Scanning Electron Microscope equipped with energy dispersive X-ray analyser (EDX). The interlayer spacing of Na⁺-Mt and Mt-IL was measured by X-ray powder diffraction (XRD, Shimadzu XRD-6000, Japan, Cu Kα radiation, 40 kV, 40 mA, Japan), and data were collected for 2θ ranging from 1.3° to 10° with the scan speed of 1° min⁻¹.

2.3 Preparation of ionic liquid (IL), 1-methyl, 3-decahexyl imidazolium

1-Methyl, 3-decahexyl imidazolium is a salt which consists of imidazolium (a five ring compound of two ammonium with one of them a quaternary ammonium) at the centre, a methyl group and decahexyl alkyl chain attached to the ammonium and the quaternary ammonium in the ring respectively. The method used by Dzyuba and Bartsch²⁷ was modified for the synthesis of the salt. An oil bath with a stirred flask containing equimolar amounts of N-imidazole and the primary alkyl bromide (bromodecahexane) was heated at 140 °C for 10–15 min and thereafter cooled. The solution was heated again in the oil-bath at 140 °C for another 10–15 min, followed by drying under vacuum at 100–120 °C to give the desired product, a highly viscous golden liquid that solidified at room temperature.

2.4 Preparation of sodium-montmorillonite (Na⁺-Mt) and modification with IL

Montmorillonite was soaked in deionised water for 24 h and thereafter centrifuged. The upper layer having very fine clay, was decanted and converted to sodium montmorillonite (Na⁺-Mt) using the method employed by Zaghouane-Boudiaf and Boutahala.¹⁰ Briefly, 10 g of montmorillonite was mixed with 1 M NaCl solution and stirred for 24 h. Thereafter, it was dialyzed in deionized water until it was free of chloride and dried at 80 °C. Na⁺-Mt was further passed through a 53 μm sieve allowing for the fine Na⁺-Mt particles to be collected. The cation exchange capacity (CEC) of Na⁺-Mt and Mt-IL were determined using the Gillman and Sumpter method,²⁸ with a CEC value of 89 meq per 100 g.

The Na⁺-Mt was then modified by adding IL in excess of the CEC of the clay to form Mt-IL. Newman¹² reported that clay can adsorb more than its CEC and in the work of Xu and Boyd,^29,30 addition of an organic cation greater than the CEC of clay, led to the adsorption of the organic cation on the external surface of the material through cation exchange and hydrophobic bonding, which generated positive charges on the surface. The mass of IL required was calculated using the following equation:³¹

m_c = f × CEC × X × GMW (1)

where m_c is the mass of organic cation required to achieve the desired fraction of CEC (g), f is the fraction of cation-exchange capacity satisfied by the organic cation (no unit), CEC is the cation exchange capacity of the clay (mmol g⁻¹ clay), X is the mass of clay (g) and GMW is the gram molecular weight of IL (g mol⁻¹).

The desired IL was slowly added to the desired mass of Na⁺-Mt dispersed in methanol. The mixture was stirred overnight and the modified montmorillonite was washed several times with double distilled deionized water and thereafter dried in an oven at 100 °C overnight.

2.5 Batch adsorption studies

Adsorption experiments were performed in batch mode to investigate the effects of various process parameters such as pH, adsorbent and initial dye concentrations, contact time, and temperature on Na⁺-Mt and Mt-IL. The effect of dose of the adsorbent was carried out with 0.005 g to 0.1 g in 30 mL of amaranth dye having a concentration of 50 mg L⁻¹; subsequent parameters (effect of concentration, pH, and temperature) were carried out using the same concentration and volume as in the case of dose. A blank was also determined under the same conditions but without the adsorbent. The concentration of the dye was always measured before and after the adsorption. All experiments were run for 24 h, to allow sufficient time for the adsorption to reach equilibrium conditions and were carried out in duplicate.

Adsorption kinetics were determined by analysing the adsorptive uptake of amaranth dye from its aqueous solution of 20 mg L⁻¹, 50 mg L⁻¹, 100 mg L⁻¹ and 200 mg L⁻¹ concentrations at different times. The equilibrium data for developing the isotherms were obtained at different temperatures (20, 30, and 40 °C) by mixing 0.05 g of Mt-IL with amaranth for 36 h. After equilibrium, the aqueous phase concentration of dye in the solution was determined by a double beam UV-Vis-NIR spectrophotometer (Shimadzu Model UV 3600, Japan) at λ_max 520 nm. The equilibrium adsorption capacity of Mt-IL and Na⁺-IL were calculated from the following relationship:

where q_e is the equilibrium adsorption capacity in (mg g⁻¹), C₀ and C_e are the dye concentrations at initial and at equilibrium respectively, in mg L⁻¹, V is the volume of solution in (L), and W is the weight of adsorbent in (g).

Adsorption data were fitted to Freundlich,³² Langmuir,³³ Temkin and Dubinin–Raduskevich,³⁴ isotherms. Linear equations of all the applied models are:

Temkin model: q_e = B In A + B In C_e (5)

Dubinin–Raduskevich model: In q_e = In q_m − β∑² (6)

where C_e is the equilibrium dye concentration in the solution in mg L⁻¹, b is the Langmuir adsorption constant (L mg⁻¹), and q_m is the theoretical maximum adsorption capacity (mg g⁻¹). K_f (L mg⁻¹) and n are Freundlich isotherm constants indicating the capacity and intensity of the adsorption, respectively. A is the equilibrium binding constant (L mg⁻¹) and B is the heat of adsorption. β is the Dubinin–Raduskevich model constant (mol² kJ⁻²) related to the mean free energy of adsorption per mole of the adsorbate and ∑ is the polanyi potential. E is the mean free energy of adsorption (kJ mol⁻¹).

In order to determine the mechanism and potential rate-controlling steps involved in the process of adsorption, pseudo first order,³⁵ pseudo second order.³⁶ The pseudo first order kinetic model equation is:

In(q_e − q_t) = ln q_e − k₁t (9)

where q_t is the amount of adsorption at time t in (mg g⁻¹), k₁ is the rate constant of the equation in (min⁻¹), and q_e is the amount of adsorption equilibrium in (mg g⁻¹). The adsorption rate constant (k₁) can be determined experimentally by plotting ln(q_e − q_t) versus t.

The pseudo second order kinetic model is expressed as:

where k₂ is the equilibrium rate constant of pseudo second order adsorption in (g mg⁻¹ min⁻¹). The values of k₂ and q_e can be determined from the slope and intercept of the plot t/q_t versus t, respectively.

The data was also fitted into the Weber–Morris model, to evaluate the possibility of intra-particle diffusion as the rate limiting step; it is given by:³⁷

q_t = k_idt^1/2 + C (11)

where k_id (mg g⁻¹ h^−1/2) is the intra-particle diffusion rate constant.

2.6 Fixed bed column studies

Fixed bed column studies were conducted using glass columns of 1.5 cm internal diameter and 55 cm length. The column was packed with the modified montmorillonite between two supporting layers of glass wool to prevent the adsorbent floating from the outlet. The mass of the adsorbent and the flow rate (3.2 mL min⁻¹) were kept constant while the influent concentration was varied from 50 to 200 ppm. The data was fitted into the Thomas model³⁸ in order to predict the adsorptive curve of breakthrough in a fixed-bed.

The Thomas expression is represented by eqn (12):

where k_Th is the Thomas rate constant (mL min⁻¹ mg⁻¹), q_o is the equilibrium uptake of the dye per g of the adsorbent (mg g⁻¹), x represents the mass of adsorbent in the column (g), V_eff is the effluent volume (mL), C₀ and C_t is the dye influent concentration and effluent concentration at time t (mg L⁻¹) respectively, and v is the flow rate (mL min⁻¹). The value of C_t/C₀ is the ratio of effluent and influent of the dye concentrations. The value of time t (min) is t = V_eff/v. The kinetic coefficient k_Th and the adsorption capacity of the column q_o can be determined from a plot of C_t/C₀ against t at a given flow rate using non-linear regression from eqn (13):

3. Results and discussion

3.1 Characterization of IL, Mt, Na⁺-Mt and Mt-IL

¹H NMR, ¹³C NMR and FT-IR analyses were carried out and thermal stabilities were confirmed by TGA-DTA/DSC (see ESI†). ¹H NMR of IL (400 MHz, CDCl₃): δ (ppm) 0.81 (t, 3H), 1.2 (m, 26H), 1.8 (q, 2H), 4.0 (s, 3H), 4.2 (t, 2H), 7.4 (d, 1H), 7.5 (d, 1H) and 10.2 (s, 1H). ¹³C NMR of IL (400 MHz, CDCl₃) at δ (ppm): 14.1, 22.7, 26.3, 29.0, 29.3, 29.4, 29.5, 29.58, 29.6, 29.7, 30.3, 31.9, 36.7, 50.2, 121.7, 121.5, and 137.6). FT-IR of IL v = 3477 and 3430 (N–H stretch, 2° amines), 3145, 3083 and 3063 (C–H stretch, aromatics), 2950, 2915, 2850 (C–H stretch, alkanes), 2051, 1631, 1573 (C–C stretch (in-ring), aromatics), 1473, 1427 (C–H bend, alkanes), 1381, 1343, 1313 (C–H rocking, alkanes), 1175 (C–N stretch, aromatic amines), 1022, 862, 792, 739, 716, 662, 623, 512, 481, 434, 401 (C–H aromatics).

High-resolution mass spectrometry was used to confirm the mass of the IL. The result of the HRMS is as follows: HRMS of [C₂₀H₁₃N₂]⁺ (m/z): 307.3110; calculated mass is 307.3113 g mol⁻¹ (see ESI† for the HRMS spectra). Thermogravimetric analysis (TGA) showed that the ionic liquid is thermally stable up to 350 °C (additional spectral data can be found in the ESI†). Other properties are shown in Table 1.

Table 1 Properties of the surfactant and amaranth used in this study

Compound	Formula	Structure	Molecular weight (MW) g mol^−1	Appearance	Melting point (°C)
1-Methyl, 3-decahexyl imidazolium	C20H39N2^+		307.311	Golden/solid at room temperature	62–64
Amaranth	C20H11N2Na3O10S3		604.473	Powder/dark red	120

---

3.2 Chemical composition of montmorillonite

EDX analysis was used in determining the chemical composition of Na⁺-Mt. Table 2 indicates that silica and alumina are the main components of the Na⁺-Mt, with others such as sodium, potassium, iron, magnesium, and calcium in trace amounts. SiO₂ and Al₂O₃ will be the main adsorbing site.³⁹

Table 2 Showing the chemical composition of montmorillonite

Constituent	% weight
SiO2	63.51
Al2O3	21.9
Na2O	4.11
Fe2O3	3.1
CaO	3.22
K2O	2.1
MgO	2.06

---

The XRD of Na⁺-Mt was compared with IL intercalated montmorillonite (Mt-IL) as shown in Fig. 1a and b. IL intercalation of the Mt has decreased the intensity of the spectra compared to the unmodified and made the width of the peak at half-peak height broader, suggesting that IL intercalation caused extensive delamination.^40–42 A similar result has been reported by Lee and Kim.⁴³ Also the peaks at 001, 002 and 020 reflect at 9.53 Å, 4.87 Å and 4.18 Å, respectively which are an integral part of montmorillonite, and the sharp reflection at the 2θ value of 27.15 (3.28 Å) corresponds to quartz in Na⁺-Mt.¹⁸ When no water molecules are intercalated between the unit layers of montmorillonite the interlayer distance is at (001) and reflection around 9.53 Å.⁴⁴ When intercalation of IL in the interlayer of Na⁺-Mt occurs, 001, 002 and 020 reflections shift to 11.65, 6.24, and 4.79 Å respectively. A similar trend had been reported by Singla and co-workers and Takahashi and co-workers.^5,18 IL does not show any XRD peak due to its liquid nature. Furthermore, comparing Na⁺-Mt with Mt-IL, the (001) d-spacing value increases from 9.53 to 11.65 Å and signifies the intercalation of IL in Mt-IL (crystalline swelling). Fig. 1c shows that the intensity of all the peaks in the Mt-IL spectrum after adsorption of amaranth dye has increased significantly compared to Mt-IL before adsorption. This suggests that the amaranth dye has been intercalated along with IL already within the Mt.

Fig. 1 (a–c) showing the XRD of Na⁺-Mt, Mt-IL and Mt-IL after adsorption of amaranth.

To ascertain that IL in excess of CEC of Na⁺-Mt is arranged in the silicate layer of Na⁺-Mt, the cation exchangeable capacity of Mt-IL was determined to be 112 meq per 100 g. Comparing the CEC with that of Na⁺-Mt (89 meq per 100 g), it shows that all exchangeable cations present in the Na⁺-Mt had been exchanged in excess for IL, indicating 125% increase of the Na⁺-Mt CEC resulting in an aggregation on the surface of the clay and an overall positive surface.

Takahashi and co-workers have also reported the intercalation of ionic liquid in the interlayer of montmorillonite.¹⁸ To further ascertain that quaternary ammonium cations in 1-methyl, 3-decahexyl imidazolium were intercalated into the silicate lattice of Na⁺-Mt, FT-IR of the pure montmorillonite, Na⁺-Mt and Mt-IL spectra were recorded in the region 380–4000 cm⁻¹ as shown in Fig. 2.

Fig. 2 FT-IR spectra of Mt, Na⁺-Mt, Mt-IL and Mt-IL-Amaranth.

The presence of peaks in the region of 3392–3314 and 1639 cm⁻¹, represent the stretching and bending modes respectively of the –OH of water within the montmorillonite.⁷ It is observed that the 3314 cm⁻¹ band intensity on Mt-IL decreases when compared with Na⁺-Mt which implies that water of hydration is lost as the cation (Na⁺) is replaced by the cationic surfactant (alkyl chain). A similar result was also reported by Zaghouane-Boudiaf and Boutahala.¹⁰ The peaks at 2920 and 2851 cm⁻¹ are assigned to symmetric and asymmetric stretching vibrations of the methylene groups in the ionic liquid while peaks 1573 and 1467 cm⁻¹ are their bending vibrations respectively,⁴⁵ showing the intercalation of ionic liquid molecules between the silica layers. These peaks were not present in the unmodified montmorillonite. The peaks at 1029 and 459 cm⁻¹ are the bands for Si–O–Si stretching and bending respectively.^46–49 There is little difference between the spectra for Mt-IL before and after adsorption of amaranth dye (Mt-IL–amaranth). There was a slight shift in the peaks after adsorption of dye from 2921.70 and 2852.30 cm⁻¹ to 2919.60 and 2851.87 cm⁻¹ (symmetric and asymmetric stretching vibrations of the methylene groups), also from 1574.73 and 1467.42 cm⁻¹ to 1573.37 and 1468.12 cm⁻¹ (bending vibrations of the methylene groups) respectively. The presence of the OH peak is noticed after adsorption of dye signifying the presence of water in the interlayer of Mt (3341.31 cm⁻¹).

The surface morphologies of the Na⁺-Mt and Mt-IL samples are analysed by SEM, with Mt-IL in Fig. 3b showing more porosity and irregularity than Na⁺-Mt in Fig. 3a. The modified montmorillonite is slightly expanded, which indicates delamination of the layers of the montmorillonite due to the presence of cations of the salt.⁵⁰ Fig. 3a (Na⁺-Mt) also shows a clump like morphology with smooth surface as compared with Fig. 3b that shows agglomerated sheet like morphological characteristic with rough surface. Similar results have been reported by Anggraini and co-workers.⁵¹ Fig. 3c shows the morphology of the Mt-IL after adsorption and indicates that the adsorbed dye has blocked the pores of the material.

Fig. 3 SEM images of Na⁺-Mt (a) and Mt-IL (b) before adsorption and after adsorption (c).

Fig. 4a shows the N₂ adsorption desorption isotherm of Na⁺-Mt, where the material exhibited a type IV isotherm which is a characteristic of mesoporous material according to IUPAC classification (material having pores ranging from 1.5–100 nm).⁵² The adsorbent also showed a H₃ hysteresis loop over the range of relative pressure P/P₀ 0.40–1.00 which signifies that the material has pore blocking (forming slit shaped pores) with non-uniform pore size distribution and shape. The average pore diameter of the sample was within the mesoporous and macroporous range, with most of it being mesoporous. In Fig. 4b Na⁺-Mt shows narrow pore size distribution with most pores centred at about 7.62 nm and a very broad range of pores from 7.62–249 nm.

Fig. 4 (a and b) N₂ adsorption desorption isotherm and pore size distribution of Na⁺-Mt respectively.

The surface area of Na⁺-Mt was 183.2 m² g⁻¹ and with a pore volume of 0.52 cm³ g⁻¹. There is a decrease in the surface area (175 m² g⁻¹) and pore volume (0.32 cm³ g⁻¹) for Mt-IL, which is due to blocking of the pores by the ionic liquid cations. Also, Mt-IL in Fig. 5a and b also showed a type IV isotherm and H₃ hysteresis, but due to the modification by 1-methyl, 3-decahexyl imidazolium, the pore size has increased and is more macroporous than mesoporous.

Fig. 5 (a and b) N₂ adsorption desorption isotherm and pore size distribution of Mt-IL respectively.

The thermal study of the material shows loss of water observed in the Mt-IL TG-DTA spectra (see ESI†) which is due to the removal of the externally adsorbed and interlayer water at about 100 °C. The decomposition of the IL was observed in the range of 300 to 450 °C and an endothermic peak is observed around 700 °C which corresponds to the decomposition of the structural hydroxyl groups in the aluminosilicate.^18,53

3.3 Batch adsorption

The effect of amaranth on Na⁺-Mt and Mt-IL doses from 0.005–0.1 g was carried out and showed an increase in adsorption of amaranth on Na⁺-Mt and Mt-IL as their masses increased. This could be attributed to electrostatic attraction between the amaranth dye ion and the positively charged surface of the materials (Na⁺-Mt and Mt-IL). Interaction between the rings on the amaranth dye molecule and the Mt-IL may also occur (π–π interaction). Adsorption reached its maximum at about 0.01 g with 90% of the dye adsorbed (Fig. 6a). It was also observed that after adsorption reached equilibrium, the adsorbed dye showed a continuous decline as the amount of sorbent increased. Daneshvar and Ghosh,^54,55 reported the same trend⁵⁴ and suggested that it was as a result of increasing particle interaction and aggregation, leading to a reduction of total surface area. On the other hand, there is a gradual increase in the percentage removal of amaranth dye as the mass of Na⁺-Mt increases which is due to the increase in surface area with a high dosage of the adsorbent. Na⁺-Mt could adsorb about 30% at 0.1 g which could be attributed to fewer adsorption sites and lack of electrostatic interaction⁵⁶ between amaranth dye and Na⁺-Mt. Since the optimized dose is 0.01 g for Mt-IL, all other optimization experiments were performed with 0.01 g of sorbent.

Fig. 6 (a and b) showing the effect of dose of adsorbent and pH respectively.

The effect of pH showed that maximum adsorption was at pH 2, and decreases gradually as the pH tends towards basic medium (Fig. 6b). Other researchers reported similar trends for adsorption of anionic dye.^39,57 This can be explained by electrostatic attraction between the negatively charged adsorbate and positively charged adsorbent. At low pH levels a positively charged ion dominates the surface of the adsorbent because of the activities of the functional group on the adsorbent and the presence of IL in excess on the adsorbent. Amaranth is an anionic dye with predominantly negative charges at low pH. This negative charge on the dye at low pH results in a significantly high electrostatic attraction between the positive and negative charges on the adsorbent and the dye respectively.^57,58 Therefore, an increase in adsorption of the dye will occur at the Mt-IL surface. When pH is increased there will be more negative ions on the surface of the adsorbent, causing repulsion which will reduce adsorption. Also, at higher pH values excess OH⁻ is present in the solution, which can lead to competition of adsorption sites with amaranth ions. Nevertheless, there is still significant amount of dye adsorbed onto Mt-IL at higher pH values. This can be attributed to chemisorption,³⁹ i.e. chemical reaction between the adsorbent aromatic ring and the ring of the amaranth dye. All other experiments were carried out on Mt-IL only at pH of approximately 2. The adsorption capacity of Na⁺-Mt (6 mg g⁻¹) is much lower than Mt-IL.

3.4 Adsorption kinetics of amaranth

The effect of contact time on the extent of amaranth adsorption of different concentrations in the plot of removal efficiency against time as shown in Fig. 7 indicated that the adsorption of amaranth on the adsorbent increased with time. Equilibrium time for 20 mg L⁻¹, 50 mg L⁻¹, 100 mg L⁻¹ and 200 mg L⁻¹ was about 120 min. Ahmad, also reported equilibrium times of 120 min for amaranth onto alumina reinforced polystyrene.⁵⁷

Fig. 7 Plot of removal efficiency% against time (min) for four different concentrations at pH 2.

It was also observed that the adsorption of amaranth is very rapid in the first 10 min at all concentrations but later decreased with time until equilibrium was reached. A large number of free adsorbent sites available for sorption are likely to be responsible for the high removal of the amaranth dye at the beginning of the experiment. The decrease in adsorption rate was as a result of a reduction in available sites and also because of repulsive forces among adsorbed dye molecules and unmodified adsorption sites present in the solution.⁵⁹

To have a better understanding of the adsorption kinetics, mechanism and the potential rate-controlling step involved in the process of adsorption, several kinetic models were used. These include Lagergran-first-order model, pseudo-second-order kinetic model and intra-particle diffusion. Table 3 showed that pseudo-second-order fitted best with R² ≥ 0.99 for all the concentrations at different temperatures. From the result in the above models, amaranth adsorption on to Mt-IL was most probably chemisorption and took place through exchange of ions on the surface of the active sites which is in agreement with other studies.^57,60–62 Also, the adsorption capacity (q_e) calculated for all the concentrations were very close to q_e experimental as shown in Table 3. Similar results were reported by Ahmad and Zargar.^57,63,64 According to Ho and co-workers and Liao and co-workers,^36,61 three steps were involved in the pseudo second order kinetic model, which are: (i) the dye molecules diffuse from liquid phase to liquid–solid interface; (ii) the dye molecules move from liquid–solid interface to solid surface; and (iii) the dye molecules diffuse into the particle pores.

Table 3 Experimental values for q_e (mg g⁻¹) and parameters for pseudo-first-order, pseudo-second-order and intra-particle diffusion

Concentrations (mg L^−1)	Temp	qe,exp (mg g^−1)	Pseudo-first-order			Pseudo-second-order			Weber–Morris model
			qe,cal (mg g^−1)	ke (h^−1)	R^2	qe,cal (mg g^−1)	k (g mg^−1 h^−1)	R^2	qe (mg g^−1)	kid (mg g^−1 h^−1/2)	R^2
200	313 K	239.360	6.869	0.002	0.697	222.222	0.000	0.987	142.790	4.525	0.519
100		199.787	0.714	0.002	0.817	181.818	0.004	0.998	174.480	0.652	0.170
50		168.853	2.115	0.004	0.157	178.571	0.003	0.999	159.710	0.566	0.278
20		73.493	3.492	0.004	0.671	74.074	0.000	0.999	47.640	1.611	0.747
200	303 K	224.853	4.119	−0.008	0.455	196.078	0.001	0.995	255.000	−2.720	0.217
100		192.853	3.650	0.098	0.817	227.273	0.000	0.996	165.590	2.026	0.468
50		165.440	4.268	0.004	0.768	166.667	0.001	0.999	126.030	2.454	0.721
20		60.267	3.160	0.004	0.824	60.606	0.002	0.998	41.449	1.164	0.739
200	293 K	246.613	5.933	−0.008	0.190	238.095	0.000	0.923	194.990	3.843	0.276
100		204.587	100.324	0.098	0.545	227.273	0.000	0.996	156.100	3.791	0.699
50		173.973	2.883	0.005	0.912	178.571	0.003	0.999	164.810	0.757	0.651
20		60.480	4.147	0.001	0.84	66.667	0.002	0.998	50.656	0.865	0.886

---

The data were further fitted into the intra-particle diffusion model to examine the relative contribution of surface and intra-particle diffusion to the kinetics process. The plot not passing through the origin indicated some degree of boundary layer control which further showed that the intra-particle diffusion was not the only rate controlling step, but other processes also controlled the rate of adsorption.^57,61,65 Surface diffusion mechanism is likely attributed to electrostatic attraction, since amaranth will be negatively charged at pH 2 and the surface of the material is positivity charged.

Table 4 showed that Langmuir was best fitted for the data with R² > 0.99 for all the temperatures. This further confirmed the previous statement that the adsorption process of the dye onto Mt-IL was chemisorbed onto its surface. The Langmuir isotherm fitted the experimental data for all the temperatures very well, which implied that the active sites on Mt-IL were homogeneously distributed, since the Langmuir equation assumed that the surface was homogeneous. Furthermore, the maximum sorption capacities (q_m) were 243.902, 238.095 and 263.158 mg g⁻¹ at temperatures 313 K, 303 K and 313 K respectively, and it was concluded that Langmuir is the best fit isotherm. To the best of the authors' knowledge the adsorption capacity is high compared with other previous research carried out on amaranth which the authors attribute to the ionic liquid used in the modification of the montmorillonite. Also, from Dubinin–Raduskevich the mean sorption energy (E) derived from the D–R isotherm model can be used to distinguish chemical and physical adsorption. Table 4 showed E values of 10.00, 7.91 and 8.34 kJ mol⁻¹, which were in the range of 8–16 kJ mol⁻¹ which suggested a dominating chemisorption⁶⁶ effect which is in agreement with other kinetics models. This is also evident in the XRD studies (Fig. 1c) where the intensity of the peaks increased which suggested a chemical reaction between IL on the Mt and the amaranth dye after adsorption.

Table 4 The q_m and b values in Langmuir equation, the K_f and 1/n values in Freundlich equation, q_s and β values in Dubinin–Radushkevich, bT and AT values in Temkin models and their respective correlation coefficient values

Temp	Langmuir model			Freundlich model			Dubinin–Radushkevich model				Temkin model
	qm (mg g^−1)	b (L mg^−1)	R^2	Kf (L mg^−1)	n (−) (g L^−1)	R^2	qm (mol g^−1)	β (mol^2 kJ^−2)	E (kJ mol^−1)	R^2	bT (g kJ mg^−1 mol^−1)	AT (L mg^−1)	R^2
313 K	243.902	0.0229	0.996	77.657	4.08	0.862	205.799	5 × 10^−9	10.000	0.939	75.53	8.044	0.935
303 K	238.095	0.099	0.995	54.435	3.25	0.654	250.711	9 × 10^−9	8.452	0.797	41.165	0.836	0.851
293 K	263.158	0.091	0.992	59.973	3.305	0.598	250.711	8 × 10^−9	7.906	0.831	56.875	2.302	0.861

---

Thermodynamic studies showed negative values of ΔG at temperatures 313, 303 and 293 K. The values of ΔH, ΔS and the R² are listed in the ESI.† The positive value of ΔH indicated that the adsorption process was endothermic in nature and the positive value of ΔS may be due to the increase in the dye concentration at the solid phase.^57,67–70 The negative values of ΔG showed that the process was thermodynamically feasible and spontaneous. This is similar to reports by other researchers.^55,57,68

Table 5 indicates that the adsorption capacities obtained from this study were far higher having a q_e value (263.20 mg g⁻¹) compared to that reported in previous literature for the adsorption of amaranth dye from other adsorbents.

Table 5 Comparison of adsorption capacities of MT-IL with other materials

pH	Adsorption capacity (mg g^−1)	Material	Ref.
2	10.00	Alumina reinforced polystyrene	57
2	14.90	Peanut hull	71
3–9	1.05	Iron oxide nanoparticles coated with cetyltrimethylammonium bromide	63
2–11	18.80	Activated carbon	72
2	100.17	Activated de-oiled mustard	72
3	3.44	Pomegranate peel	73
6–12	65.04	Tamarind pod shells	74
9	95.12	Activated carbon	75
2	263.20	Mt-IL	Present work
2	6.00	Na-Mt	Present work

---

3.5 Column studies

The effect of different concentrations from 50 to 200 mg L⁻¹ of amaranth dye influent with a constant mass of Mt-IL and constant solution flow rate of 3.2 mL min⁻¹ is shown in the breakthrough curve in Fig. 8. Fig. 8 showed increased breakthrough times with decreased influent concentrations of amaranth dye and steeper breakthrough curves at increased influent concentrations. The reason for this is because at higher influent concentrations, transportation of amaranth dye to the adsorbent is faster because of an increased diffusion coefficient or mass transfer coefficient which is the driving force for the adsorption process.⁷⁶ As a result this leads to higher gradients of the breakthrough curves at increased concentrations. The larger the influent concentration, the steeper the slope of the breakthrough curve and the shorter the breakthrough time.⁷⁷ These results also showed that a change in concentration of influent affects the saturation rate, and implies that the diffusion process is concentration dependent. Similar trends were obtained from other work.^78–80

Fig. 8 Breakthrough curves showing the effect of influent concentration on amaranth dye.

The column data were fitted into the Thomas model to determine the Thomas rate constant k_Th, and equilibrium uptake of amaranth dye per gram of the adsorbent (mg g⁻¹) q_o. Table 6 showed the parameters of the Thomas model, with R² values ranging from 0.941 to 0.998. This showed that the Thomas model is suitable for the adsorption of amaranth on Mt-IL. Also, from Table 6, as the influent concentration increased the value of column adsorption capacity q_o increased (393.63, 580.89 and 603.60 mg g⁻¹ at 50, 100 and 200 mg L⁻¹ respectively). The increased concentrations of solute in solution resulted in increased rates of transportation of solutes to the adsorbent thus leading to increased adsorption capacities.^77,79,81 Thus column studies also confirmed the high adsorption capacity of IL modified montmorillonite for the adsorption of amaranth dye.

Table 6 Thomas model parameters at different concentrations using non-linear regression analysis

C0 (mg L^−1)	v (mL min^−1)	kTh (mL min^−1 mg^−1)	qo (mg g^−1)	R^2
50	3.2	1.4 × 10^−4	393.64	0.941
100	3.2	0.86 × 10^−4	580.89	0.998
200	3.2	0.95 × 10^−4	608.60	0.996

---

4. Conclusion

In this study, a novel, cationic ionic liquid was synthesized for the modification of sodium-montmorillonite. The modified material is hydrophobic with a large pore volume to trap pollutants and is a mixture of mesoporous and macroporous. Adsorption of Mt-IL studies showed that 1-methyl, 3-decahexyl imidazolium intercalated in the interlayer of montmorillonite is a possible good alternative for the removal of anionic amaranth dye from wastewater. The maximum removal percentage was >90% at an optimum pH of 2. The adsorption behaviour was best described by the Langmuir isotherm with a high adsorption capacity of 263.2 mg g⁻¹ and the kinetics by pseudo-second-order. Positive values of enthalpy change suggested the endothermic nature of the process, while the negative free energy proved its thermodynamic feasibility and spontaneity. Column experiments using the Thomas model also showed high adsorption capacities of 393.64, 580.89 and 603.60 mg g⁻¹ at different concentrations. From the result in this research, ionic liquid modified montmorillonite is a good alternative for adsorption of dye from wastewater.

Acknowledgements

The authors wish to acknowledge the University of KwaZulu-Natal for the use of the instrumentation in this study.

References

1. D. Avisar, O. Primor, I. Gozlan and H. Mamane, Water, Air, Soil Pollut., 2010, 209, 439–450.
2. A. Meleshyn and D. Tunega, Geoderma, 2011, 169, 41–46.
3. Q. Wu, Z. Li and H. Hong, Appl. Clay Sci., 2013, 74, 66–73.
4. D. L. Ho, R. M. Briber and C. J. Glinka, Chem. Mater., 2001, 13, 1923–1931.
5. P. Singla, R. Mehta and S. N. Upadhyay, Green Sustainable Chem., 2012, 2, 21–25.
6. B. Du, Z. Guo, P. A. Song, H. Liu, Z. Fang and Y. Wu, Appl. Clay Sci., 2009, 45, 178–184.
7. M. Boufatit, H. Ait-Amar and W. Mc Whinnie, Desalination, 2008, 223, 366–374.
8. M. Pospıšil, P. Čapková, Z. Weiss, Z. Maláč and J. Šimonık, J. Colloid Interface Sci., 2002, 245, 126–132.
9. M. Pospıšil, P. Čapková, D. Měřínská, Z. Maláč and J. Šimoník, J. Colloid Interface Sci., 2001, 236, 127–131.
10. H. Zaghouane-Boudiaf and M. Boutahala, Adv. Powder Technol., 2011, 22, 735–740.
11. F.-C. Huang, J.-F. Lee, C.-K. Lee and H.-P. Chao, Colloids Surf., A, 2004, 239, 41–47.
12. A. C. Newman, Chemistry of Clays and Clay Minerals, Longman Scientific and Technical, 1987.
13. S. Y. Lee and S. J. Kim, Clays Clay Miner., 2002, 50, 435–445.
14. S. H. Lee, D. I. Song and Y. W. Jeon, Environ. Technol., 2001, 22, 247–254.
15. M. Vakili, M. Rafatullah, B. Salamatinia, A. Z. Abdullah, M. H. Ibrahim, K. B. Tan, Z. Gholami and P. Amouzgar, Carbohydr. Polym., 2014, 113, 115–130.
16. R. Upson and S. Burns, J. Colloid Interface Sci., 2006, 297, 70–76.
17. Y. Hsu, M. Wang, C. Pai and Y. Wang, Appl. Clay Sci., 2000, 16, 147–159.
18. C. Takahashi, T. Shirai, Y. Hayashi and M. Fuji, Solid State Ionics, 2013, 241, 53–61.
19. H. Sajjadi, A. Modaressi, P. Magri, U. Domańska, M. Sindt, J.-L. Mieloszynski, F. Mutelet and M. Rogalski, J. Mol. Liq., 2013, 186, 1–6.
20. G. Absalan, M. Asadi, S. Kamran, L. Sheikhian and D. M. Goltz, J. Hazard. Mater., 2011, 192, 476–484.
21. S. Kamran, M. Asadi and G. Absalan, Anal. Methods, 2014, 6, 798–806.
22. T. Welton, Chem. Rev., 1999, 99, 2071–2084.
23. E. A. Turner, C. C. Pye and R. D. Singer, J. Phys. Chem. A, 2003, 107, 2277–2288.
24. J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D. Willauer, G. A. Broker and R. D. Rogers, Green Chem., 2001, 3, 156–164.
25. A. Jess, A. Große Böwing and P. Wasserscheid, Chem. Ing. Tech., 2005, 77, 1430–1439.
26. 4,517,094, 1985.
27. S. V. Dzyuba and R. A. Bartsch, J. Heterocycl. Chem., 2001, 38, 265–268.
28. G. Gillman and E. Sumpter, Soil Res., 1986, 24, 61–66.
29. S. Xu and S. A. Boyd, Environ. Sci. Technol., 1995, 29, 312–320.
30. H. Zhao, W. F. Jaynes and G. F. Vance, Chemosphere, 1996, 33, 2089–2100.
31. N. M. Soule and S. E. Burns, J. Geotech. Geoenviron. Eng., 2001, 127, 363–370.
32. H. Freundlich, Z. Phys. Chem., 1906, 57, 385–470.
33. I. Langmuir, J. Am. Chem. Soc., 1916, 38, 2221–2295.
34. M. Dubinin and L. Radushkevich, Chem. Zentralbl., 1947, 1, 875.
35. S. Lagergren, K. Sven. Vetenskapsakad. Handl., 1898, 24, 1–39.
36. Y.-S. Ho and G. McKay, Process Biochem., 1999, 34, 451–465.
37. W. J. Weber and J. C. Morris, J. Sanit. Eng. Div., Am. Soc. Civ. Eng., 1963, 89, 31.
38. H. C. Thomas, J. Am. Chem. Soc., 1944, 66, 1664–1666.
39. A. S. Özcan, B. Erdem and A. Özcan, J. Colloid Interface Sci., 2004, 280, 44–54.
40. Z. Li, P.-H. Chang, J.-S. Jean, W.-T. Jiang and C.-J. Wang, J. Colloid Interface Sci., 2010, 341, 311–319.
41. P. Kulshrestha, R. F. Giese and D. S. Aga, Environ. Sci. Technol., 2004, 38, 4097–4105.
42. L. S. Porubcan, C. J. Serna, J. L. White and S. L. Hem, J. Pharm. Sci., 1978, 67, 1081–1087.
43. S. Y. Lee and S. J. Kim, J. Colloid Interface Sci., 2002, 248, 231–238.
44. R. E. Grim, Clay Mineralogy, McGraw-Hill, New York, 1953.
45. Q. Zhou, R. L. Frost, H. He, Y. Xi and H. Liu, J. Colloid Interface Sci., 2007, 307, 357–363.
46. F. Adam, J. N. Appaturi, Z. Khanam, R. Thankappan and M. A. M. Nawi, Appl. Surf. Sci., 2013, 264, 718–726.
47. S. Endud and K.-L. Wong, Microporous Mesoporous Mater., 2007, 101, 256–263.
48. C. E. A. Kirschhock, R. Ravishankar, F. Verspeurt, P. J. Grobet, P. A. Jacobs and J. A. Martens, J. Phys. Chem. B, 1999, 103, 4965–4971.
49. E. M. Flanigen, H. Khatami and H. A. Szymanski, in Molecular Sieve Zeolites-I, ed. F. Gould Robert, American Chemical Society, Washington, DC, 1974, vol. 101, pp. i-iv.
50. O. A. A. dos Santos, C. Z. Castellib, M. F. Oliveirab, A. F. de Almeida Netob and M. G. da Silvab, Chem. Eng., 2013, 32, 307–312.
51. M. Anggraini, A. Kurniawan, L. K. Ong, M. A. Martin, J.-C. Liu, F. E. Soetaredjo, N. Indraswati and S. Ismadji, RSC Adv., 2014, 4, 16298–16311.
52. T. Allen, Particle Size Measurement: Volume 2: Surface Area and Pore Size Determination, Springer, London, 5th edn, 1997.
53. T. Bastow, S. Hardin and T. Turney, J. Mater. Sci., 1991, 26, 1443–1453.
54. E. Daneshvar, M. Kousha, M. S. Sohrabi, A. Khataee and A. Converti, Chem. Eng. J., 2012, 195–196, 297–306.
55. D. Ghosh and K. G. Bhattacharyya, Appl. Clay Sci., 2002, 20, 295–300.
56. H. Yoshida, A. Okamoto and T. Kataoka, Chem. Eng. Sci., 1993, 48, 2267–2272.
57. R. Ahmad and R. Kumar, Clean: Soil, Air, Water, 2011, 39, 74–82.
58. P. Baskaralingam, M. Pulikesi, D. Elango, V. Ramamurthi and S. Sivanesan, J. Hazard. Mater., 2006, 128, 138–144.
59. C. Smaranda, D. Bulgariu and M. Gavrilescu, Environ. Eng. Manage. J., 2009, 8, 1391–1402.
60. Y.-S. Ho and G. McKay, Chem. Eng. J., 1998, 70, 115–124.
61. P. Liao, Z. Zhan, J. Dai, X. Wu, W. Zhang, K. Wang and S. Yuan, Chem. Eng. J., 2013, 228, 496–505.
62. X.-E. Shen, X.-Q. Shan, D.-M. Dong, X.-Y. Hua and G. Owens, J. Colloid Interface Sci., 2009, 330, 1–8.
63. B. Zargar, H. Parham and A. Hatamie, Chemosphere, 2009, 76, 554–557.
64. A. Mittal and V. K. Gupta, J. Hazard. Mater., 2005, 117, 171–178.
65. G. Crini, H. N. Peindy, F. Gimbert and C. Robert, Sep. Purif. Technol., 2007, 53, 97–110.
66. F. G. Helfferich, Ion exchange, Courier Dover Publications, 1962.
67. V. Gupta, A. Mittal, V. Gajbe and J. Mittal, Ind. Eng. Chem. Res., 2006, 45, 1446–1453.
68. P. Liao, Z. Malik Ismael, W. Zhang, S. Yuan, M. Tong, K. Wang and J. Bao, Chem. Eng. J., 2012, 195, 339–346.
69. A. Bhatnagar and A. Jain, J. Colloid Interface Sci., 2005, 281, 49–55.
70. S.-H. Lin and R.-S. Juang, J. Environ. Manage., 2009, 90, 1336–1349.
71. R. Gong, Y. Ding, M. Li, C. Yang, H. Liu and Y. Sun, Dyes Pigm., 2005, 64, 187–192.
72. R. Jain and S. Sikarwar, Desalin. Water Treat., 2011, 28, 120–129.
73. H. A. A. Ali, H. A. K. Egzar, N. M. Kamal, N. Abdulsaheb and M. S. Mashkour, Int. J. Basic Appl. Sci., 2013, 13, 57–64.
74. A. Naidu, J. Biochem. Technol., 2014, 3, S189–S192.
75. J.-J. Lee, Clean Technol., 2011, 17, 97–102.
76. I. A. W. Tan, A. L. Ahmad and B. H. Hameed, Desalination, 2008, 225, 13–28.
77. Z. Aksu and F. Gönen, Process Biochem., 2004, 39, 599–613.
78. R. Han, Y. Wang, W. Yu, W. Zou, J. Shi and H. Liu, J. Hazard. Mater., 2007, 141, 713–718.
79. A. Ahmad and B. Hameed, J. Hazard. Mater., 2010, 175, 298–303.
80. S.-H. Lin, R.-S. Juang and Y.-H. Wang, J. Hazard. Mater., 2004, 113, 195–200.
81. K. Vijayaraghavan, J. Jegan, K. Palanivelu and M. Velan, J. Hazard. Mater., 2004, 113, 223–230.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra09483f

---