Antimony- and Bismuth-Based Ionic Liquids as Efficient Adsorbents for the Removal of Dyes

Abstract

A series of ionic liquids consisting of anilinium cations with varying alkyl chains and metallic (Sb and Bi) halides as anions have been synthesized and thoroughly characterized by using multinuclear (¹H and ¹³C) NMR, FT-IR, Raman and XPS techniques. They have been exploited as adsorbents for the dye’s removal, such as malachite green, rhodamine B and Sudan II, from the aqueous solution. Various parameters like the effect of stirring rate, pH, reaction time, adsorbent amount and initial dye concentration have been optimized. Both antimony- and bismuth-based ionic liquids exhibit high adsorption efficiencies and have comparable performance for each dye. Kinetic data have been analyzed by applying kinetic models, and the best-fitted model was found to be pseudo-second order with an R² value greater than 0.98. Adsorption capacity has been determined by analyzing the sorption data using the Langmuir and Freundlich equations, and the Langmuir isotherm model has been found to be the best fitting. The maximum adsorption capacities (q_max) derived from the Langmuir isotherm for malachite green, Sudan II and rhodamine B by M-Sb ILs were 217.36, 162.10 and 62.94 mg·g⁻¹, whereas by M-Bi ILs, the adsorption capacities were slightly higher, at 230.18, 170.00 and 64.21 mg·g⁻¹, respectively. Kinetic studies indicated pseudo-second-order behavior (R² > 0.98), while thermodynamic analysis demonstrated an endothermic adsorption, and a spontaneous reaction was carried out by a physisorption process. These findings accentuate the potential of Sb- and Bi-based ionic liquids as efficient and reusable adsorbents for removing dyes from wastewater.

1. Introduction

Generally, organic dyes are essential chemical compounds that are significantly used as the coloring agents in various industrial sectors, including textiles, paper, printing, cosmetics, leather, plastics, and pharmaceuticals [1]. Thousands of different types of pigments and dyes are being used globally, and approximately 700 kilotons of dyes are produced in industries annually [2]. One kilogram of textile consumes nearly 200 L of water, while a standard textile mill requires approximately 1.6 million liters of water per day [3]. So, the wastewater contains a huge amount of dye pollutants that are a major threat not only to the biosphere but can also bring harmful impacts to both human [4] and aquatic life by reducing the transmission of sunlight [5,6].

Dye pollutants frequently contain hazardous aromatic substances and heavy metals [7] (e.g., lead, cadmium, chromium, etc.) that lead to the carcinogenicity, mutagenicity and malfunctioning of the brain, kidney, liver [8], central nervous system and reproductive system of human beings [9]. So, developing an economical and efficient approach to treating these contaminant organic dyes from the soil and water is a dire necessity. Numerous chemical, physical and biological methods have been used in the past [10], but due to the stable and resistant nature of dyes under several conditions, such as heat, light, and oxidizing agents, the adsorption technique has been adopted because it is highly proficient, sustainable and economic for wastewater treatment [11,12].

Numerous adsorbents like zeolites [13], nanomaterials [14], activated carbon [15], biosorbents [16] and miscellaneous materials [17] have been frequently used for removing organic dyes [18]; however, due to significant limitations and challenges, including hazardous by-products [19], agglomeration [20], lower efficiency [21], small surface area [21] and clearance of sludge [22], there is an utter requirement for an economical substitute with enhanced adsorbing and recyclability to remove organic dyes/pollutants from wastewater [23]. Ionic liquids (ILs) have emerged as promising materials for wastewater treatment, offering exceptional properties like non-volatility, less toxicity, a broad range of constancy, low liquefying temperature, a tendency to adsorb a variety of materials and eco-friendliness that have enabled them to be potential contenders as adsorbents for the removal of dyes from wastewater [24]. ILs are salts typically composed of different organic cations with inorganic or organic anions and can be altered with relative ease by a variety of suitable anion and cation combinations; thus, they can be designed for task-specific purposes [25,26].

Metal-substituted ionic liquids, also written as halometallate ionic liquids, have the structural building block [MX₄]²⁻ (M = Fe, Bi, Sb, Pd, Zn, Co, Cd, Ni, Cu; X = Cl, Br, I) and are a safer alternative to common volatile organic solvents [27,28] because they are thermally [29,30] and chemically stable [31,32], have limited (if any) volatility [33] and have strong catalytic activity [34]. Metallic salts, complexes and organometallic compounds can easily dissolve in ILs, providing a polar medium for a catalytic reaction [35,36]. Metal incorporation in ILs increases the hydrophobic character [37], which is estimated to be effective in regeneration [38], separation [39] and reuse after the catalysis [40].

In previous studies, phosphonium- and imidazolium-based ILs have been effectively used for dye adsorption from aqueous media. However, traditional ILs and adsorbents face significant limitations, including poor recyclability, decreased adsorption proficiency owing to aggregation and complications in separation after adsorption. Moreover, conventional adsorbents, such as zeolites and activated carbon, lack selectivity for specific dye pollutants and may possibly produce secondary waste during disposal processes [41,42]. So, considering these facts and in continuation of our prior work ([40,43,44] and [45]), four novel antimony- and bismuth-containing anilinium-based hydrophobic ILs have been synthesized and used as adsorbents in the extraction and sorption of three organic dyes (malachite green (MG), rhodamine B (RhB) and Sudan II). These synthesized ILs offer enhanced hydrophobicity, which improves their separation efficiency and recyclability after adsorption. Additionally, their strong interactions with various organic dyes significantly improve the performance of dye removal compared to conventional adsorbents and traditional ILs. They have been characterized comprehensively, including by UV–Visible spectroscopy, and their adsorption performance has been assessed and substantiated with detailed kinetic and thermodynamic analyses.

2. Results and Discussion

2.1. ¹H NMR

The ¹HNMR spectrum for (M-IL) shows the singlet peak at 3.6 ppm with an intensity of 9, representing the protons from three methyl groups, and the ¹HNMR for (E-IL) shows the peak for –CH₂ at 1.18–1.21 ppm and for the methyl group –CH₃ at 3.36 ppm; thus, this confirms the alkylation of the nitrogen atom. The signal was initially at 3.06 ppm with an intensity of 6 in N,N-dimethylaniline. The downfield shifting of the methyl peak is the result of the quaternization of nitrogen, which withdraws the electron density toward itself due to positive charge formation. The aromatic H shifts occur at 6.96 and 7.27 ppm for the aromatic hydrogen in N,N-dimethylaniline. The formation of a fourth bond to the nitrogen atom gives a deficiency of electron density, for which nitrogen withdraws electron density toward itself from neighboring atoms, due to which the aromatic hydrogen shifts downfield to 8.01 and 7.60 ppm, which provides evidence for the formation of iodide-based ILs (Figures S2 and S3).

2.2. ¹³C NMR

The ¹³CNMR spectra for (M-IL) and (E-IL) have different carbon environments in DMSO. The methyl carbon in (M-IL) shows a chemical shift most upfield, around 56.89 ppm, which was originally at 41.3 ppm in N,N-dimethylaniline. Similarly, the signals at 9.34 and 55.0 ppm for the -CH₃ and -CH₂ groups, respectively, are observed for (E-IL). The shifting of the signal to downfield confirms the formation of the N-C bond. The quaternization of nitrogen causes the de-shielding of methyl carbon, thus shifting it from 40 to 56.89 ppm. While the aromatic carbon located on the benzene ring adjacent to the nitrogen shows the shift at 151 ppm, this could be explained based on the shift of electron density from methyl groups to the benzene ring through the nitrogen atom. The other aromatic carbons show the signal at 120.9 ppm and 130.5 ppm, respectively (Figures S4 and S5).

2.3. Raman Spectra

¹H and ¹³C NMR were used to confirm the intermediate structures, and they can indicate the proton and carbon environments. However, in an anion exchange step, the cations are not changing; it is the anions, which possess no carbon or hydrogen atom. Therefore, repeating the NMR step would not give any more useful information. Synthesized ionic liquids were characterized by Raman spectroscopy for the confirmation of metal chlorides. The Raman spectra of antimony- and bismuth-containing ionic liquids are shown in Figure 1. The Raman bands in the range of 0–500 cm⁻¹ are primarily designated to internal modes of metal chlorides. The Raman bands in the range of 220 cm⁻¹ are assigned to the angular distortion mode of Cl-Bi-Cl, whereas the Bi-Cl symmetric stretch was observed at 256 cm^−1, and the asymmetric stretch mode was observed at 296 cm⁻¹, respectively. Similarly, the Raman spectrum of Sb-ILs shows a band at 231 cm⁻¹ for Cl-Sb-Cl angular deformations. The band in the range of 270–300 cm⁻¹ is attributed to the symmetric stretching of Sb-Cl, while the band at 344–356 cm⁻¹ is attributed to the asymmetric vibration of [SbCl₄]^–. These values correspond very well with the literature values [46,47].

2.4. X-Ray Photoelectron Spectroscopy

The X-ray photoelectron spectroscopic (XPS) analysis was carried out to investigate the oxidation state, composition and chemical environment of synthesized ILs. The survey spectra in Figure 2a and Figure 3a authenticated the existence of N, C, Cl and corresponding metals (Sb and Bi) in them. The peak fitting of Sb 3d spectra (Figure 2b) clearly consists of doublet peaks at binding energy values of 530.5 and 539.8 eV, with a spin–orbit separation of 9.3 eV. However, the overlapped signal Sb 3d shows the existence of O 1s at 530 eV. There is no significant shift in binding energies, which signifies the occurrence of Sb⁺³, and the values align well with previous studies [48]. Meanwhile, high-resolution spectra of Bi 4f in Figure 3b also illustrate two peaks at the binding energies of 160.4 and 165.7 eV, with a spin–orbit separation of 5.3 eV. No significant fluctuations in peak values indicated that Bi is in the +3 oxidation state, and this is well in agreement with the literature [49]. The high-resolution spectra for chlorine demonstrated the spin–orbit doublets at 2p core levels as 2p_3/2 and 2p_1/2, as shown in Figure 2 and Figure 3c. Similarly, the carbon core level spectrum can be fitted by three peaks (lower to higher BE values) that are denoted as C_aromatic (remaining five carbons within the aromatic head group), sp² C-N (phenyl carbon with nitrogen atom) and sp³ C-N (sp³ hybridized carbons attached with nitrogen outside the ring), respectively, as displayed in Figure 2 and Figure 3d. Moreover, the peak at 401.5 eV can be assigned to the N 1s of aniline (R-N⁺), as represented in Figure 2 and Figure 3e; however, an additional signal was observed in the case of M-Sb at 402.5 of C=NH⁺, possibly due to a side reaction [50].

2.5. Adsorption Studies for MG, RhB and Sudan II Dyes

To check the adsorption efficiency of prepared ILs, three different dyes were used: MG, RhB and Sudan II, using a 10 mgL⁻¹ dye concentration, a time of 40 min, 0.01 g of adsorbent dose and a 100 rpm stirring rate. It is observed that the Bi-ILs show better efficiency than the corresponding Sb-ILs, and the ethyl chain-containing ILs have better performance than the methyl chain-ILs, as shown in Figure 4.

2.6. Optimization of Parameters for Adsorption Process

Parameters like stirring rate, pH, time, dose of adsorbents and dye concentration have a great effect on adsorption efficiencies. The effects of these parameters are studied in detail and optimized to obtain the maximum adsorption efficiencies. The adsorption behavior of ILs on these dyes and their optimization parameters are illustrated in Figure 5.

2.7. Stirring Rate Effect

The effect of stirring rate was checked out by varying the stirring rate to 100, 150, 200, 250 and 300 rpm in a solution of 25 mgL⁻¹ dye concentration for a time of 40 min and with an adsorbent dose of 0.01 g. Figure 5a illustrates the effect of stirring rate on the adsorption efficiencies of malachite green, Sudan II and rhodamine B. The adsorption efficiencies were enhanced with a varying stirring rate. It was found that with increasing the stirring rate, the adsorption was quicker than with slow stirring, which could be explained by the fact that fast stirring keeps the dye particles suspended and allows more contact of solid–liquid, favoring the transfer of dye molecules to the adsorbent material [51]. However, the stirring speed was optimized at 200 rpm because a higher stirring rate causes the splashing of solution, which may result in loss of material, and the faster stirring rate did not increase adsorption significantly [52]. To run the experiment smoothly, the stirring rate was set at an intermediate rate of 200 rpm. The maximum adsorption efficiencies for MG were found to be about 73% for bismuth-based ionic liquids (M-Bi) and around 55% for antimony-based ionic liquids (M-Sb), which were higher than Sudan II, followed by rhodamine B, respectively.

2.8. Effect of Time

The time effect on the percentage removal of dyes was studied by taking an initial dye concentration of 25 mgL⁻¹, a 0.01 g adsorption dose and a 200 rpm stirring rate. The UV spectra were taken at constant intervals of time to study the constant effect. The removal of dye was quicker at earlier points of contact times and gradually declined until equilibrium was reached. A high rate of adsorption was observed in the initial contact time, corresponding to the presence of a greater number of adsorption sites on the synthesized materials. The optimum time required for the maximum adsorption of maximum dyes with a 25 mgL⁻¹ solution was found to be 40 min for both types of bismuth- and antimony-based ionic liquids, whereas the sorption efficiency did not increase significantly on further increasing the time [53]. The faster and higher adsorption capacities for bismuth-based ionic liquids are due to the larger size of bismuth, providing more adsorption sites. Figure 5b indicates the actual variation of adsorption with changing time. Malachite green was removed 10 times faster than rhodamine B with M-Bi, and M-Sb removed the malachite green only two times faster than rhodamine B. Sudan II showed significantly faster adsorption than rhodamine B. Hence, 40 min was optimal for removing the dyes.

2.9. Effect of Adsorbent Dose

The effect of adsorbent amount was assessed by taking 40 mL of the initial dye concentration of 25 mgL⁻¹, a 200 rpm stirring rate, and 40 min as the equilibrium time. The UV spectra of all the dyes at different dye adsorbances were taken at optimized conditions with changing doses. The adsorbent dosage varied from 10 to 50 mg for both types of ionic liquids, and the adsorption efficiency was enhanced with the amount of adsorbent. This may be attributed to the fact that the adsorption sites increase with an increase in adsorbents, leading to elevated adsorption efficiency [41]. The adsorbent dose of 30 mg was found to be the optimum for both types of ILs for removing the MG, RhB and Sudan II dyes (Figure 5c).

2.10. Effect of Dye Concentration

The effect of initial dye concentration on the adsorption properties of adsorbents is illustrated in Figure 5d. To assess the impact of dye, solutions of dyes from 1 ppm to 40 ppm were prepared. Each solution was agitated under optimized dose, pH and time. The experiments showed that adsorption efficiencies decreased with the increase in dye concentration for all dyes, whereas the adsorption capacities increased. This decrease in efficiency may be attributed to the fact that the number of adsorption sites is limited in adsorbents, so further sorbate molecules could not find a site to get attached. The increase in adsorption capacities is due to the maximum coverage of the adsorption sites [54,55].

2.11. Effect of Metal and Alkyl Chain Length

The UV spectra results indicate that Bi-ILs demonstrate better adsorption efficiency than Sb-ILs, which may be attributed to the larger atomic radius and higher polarizability of bismuth, leading to enhanced electrostatic interactions with organic dye molecules. Additionally, Bi-ILs may exhibit greater Lewis acidity, strengthening the adsorption process [56]. Furthermore, the ethyl-substituted ILs displayed slightly higher adsorption capacity than their methyl counterparts, which can be ascribed to the number of carbon atoms, probably providing more active sites for adsorption, increased hydrophobic interactions and improved molecular orientation [55]. However, the increase in adsorption capacity remains insignificant due to the small difference in alkyl chain length between methyl and ethyl groups [Figure 4]. The absorption spectrum of Sudan II dye undergoes not only a reduction in intensity but also a noticeable change in shape when Bi-based adsorbents (M-Bi and E-Bi) are introduced. The shift in peak positions or the emergence of new spectral features could indicate chemical interactions, such as complexation or changes in the electronic environment of the dye molecules [57].

2.12. Kinetic Studies

The mechanism of molecule adsorption onto adsorbent can be investigated by several kinetic models. To examine the sorption mechanism, characteristic constants of adsorption were determined by using the Lagergren equation and a pseudo-second-order equation based on solid phase adsorption, respectively.

2.13. Pseudo-First-Order Kinetics

The equation of Lagergren or pseudo-first-order kinetics is represented in Equation (1).

$$d{q}_t/dt=k_1(q_e-q_t)$$

(1)

This equation is used to check the rate and kinetics of adsorption. The integrated form of the equation can be written as

$$log{(q}_e-q_t)=log{q}_e-{\displaystyle \frac{k_1}{2.303}}t$$

(2)

where q_e and q_t are the amount adsorbed at equilibrium and time ‘t’, respectively; and k is the rate constant. k values and q_e values can be calculated from the slope and intercept, respectively. If the sorption system occurred via the pseudo-first-order kinetics, the experimental values should fit the linear plot at all concentrations. The important constants calculated from slope and intercept, respectively, are shown in

Table 1. Kinetic parameters.

Adsorbent	Dyes	Pseudo-1st-Order Kinetics			Pseudo-2nd-Order Kinetics			Elovich Plot
		k1 (min−1)	qe (mg·g−1)	R2	k2 (min−1)	qe (mg·g−1)	R2	α (mg·g−1 .min−1)	β (g.mg−1)	R2
M-Sb	MG	0.225	8.1	0.972	0.00077	146.41	0.991	1.007	0.029	0.989
	RhB	0.114	7.64	0.946	0.00139	56.49	0.998	1.03	0.084	0.995
	Sudan II	0.0306	42.52	0.75	0.00012	165.56	0.98	18.77	0.038	0.965
M-Bi	MG	0.204	36.59	0.929	0.00015	56.08	0.989	2.26	0.033	0.984
	RhB	0.05	4.55	0.96	6.1E-05	85.47	0.985	5.48	0.193	0.983
	Sudan II	0.0202	14.53	0.91	0.0009	134.95	0.99	39.16	0.035	0.97

. These values correspond to the regression of 0.92 to 0.97.

2.14. Pseudo-Second-Order Kinetics

The pseudo-second-order reaction can be expressed in non-linear form as Equation (3):

$$d{q}_t/dt=k_2{(q_e-q_t)}^2$$

(3)

where q_t is the adsorbed amount (mg·g⁻¹) at a given time, q_e is the adsorbed amount at equilibrium and k is the rate constant. To explain the adsorption process, the equation is used in its linear form as given (Equation (4)):

$${\displaystyle \frac{t}{q_t}}={\displaystyle \frac{1}{k_2{q_e}^2}}+{\displaystyle \frac{1}{{\mathrm{q}}_{\mathrm{e}}}}t$$

(4)

In a normal sorption experiment, initially, due to the large number of vacant sorption sites being available, the adsorbed amount increases rapidly and gradually decreases with time until equilibrium is reached.

The correlation coefficient values were higher than those of pseudo-first-order kinetics, indicating that adsorption on adsorbent occurred through second-order kinetics, based on the correlation with pseudo-second-order kinetics. Although the adsorption followed pseudo-second-order kinetics, which often implies chemisorption, the low enthalpy values and good reusability suggest that the rate-limiting step is more likely governed by physisorption involving weak electrostatic interactions or van der Waals forces [58]. The rate constants, regression correlation and adsorption capacity at equilibrium calculated from the plots are tabulated in Table 1.

2.15. Elovich Plot

The Elovich equation also provides evidence for the nature of adsorption systems, describes the kinetics on heterogeneous surfaces and gives insights about the mechanism of adsorption. The integrated form of the equation is written as

$$q_t={\displaystyle \frac{1}{\beta }}ln\alpha \beta +{\displaystyle \frac{1}{\beta }}\mathrm{l}\mathrm{n}(t)$$

(5)

The constant $\beta$ (mg·g⁻¹) is an important parameter of this equation, which is called an adsorption constant, and is associated with the activation energy of chemisorption and the extent of surface coverage. α is associated with the initial adsorption rate in the Elovich equation. Their values depend on temperature; with an increase in temperature, the constants α and β increase, showing an enhanced rate of adsorption and desorption. The Elovich model is typically linked with chemisorption; it can also represent certain physisorption processes, especially those that involve heterogeneous surfaces. In this study, malachite green and rhodamine B have shown a good fit to the Elovich model, and their α and β values are listed in Table 1. However, the adsorption process, due to low values of enthalpy and the higher recyclability of ILs, is described predominantly as physisorption that possibly involves weak interactions.

From the experimental values and plot regression values, it is seen that the adsorption process followed pseudo-second-order kinetics for M-Bi and M-Sb for malachite green, rhodamine B and Sudan II, supporting the involvement of surface interactions as the rate-limiting step.

2.16. Adsorption Isotherm Models

Adsorption isotherms were used to assess the interaction between adsorbate particles and adsorbent surfaces that play a vital role in the optimization of adsorbents. The Langmuir and Freundlich models were preferred to examine the adsorption of dye on ILs.

2.17. Langmuir Isotherm

The Langmuir model proposes that intermolecular forces are indirectly related to the distance between sorbate and sorbent molecules and consequently assumes the formation of a monolayer on the surface of the adsorbent. The Langmuir isotherm further predicts that each active site on the adsorbent molecule is occupied by only one molecule. Once the site is occupied, no further molecules can bind to that site. Moreover, the adsorption capacity of an adsorbent is limited for sorbate, beyond which it cannot adsorb additional molecules; this point is referred to as saturation, and the corresponding value is known as the maximum adsorption capacity of the adsorbent. The linear form of the Langmuir isotherm is given by Equation (6).

$${\displaystyle \frac{{\mathit{C}}_{\mathit{e}}}{{\mathit{q}}_{\mathit{e}}}}={\displaystyle \frac{1}{{\mathit{K}}_{\mathit{L}}{\mathit{q}}_{\mathit{m}}}}+{\displaystyle \frac{{\mathit{C}}_{\mathit{e}}}{{\mathit{q}}_{\mathit{m}}}}$$

(6)

Here, C_e is the equilibrium concentration or final concentration, q_e is the adsorbed amount at equilibrium, q_m is the maximum adsorption capacity of the adsorbent and K_L is the Langmuir constant (dm³/mg), related to the affinity of the adsorbent for an adsorbate.

The plot of C_e vs. C_e/q_e gives a straight line, and K_L and q_m are derived from the slope and intercept, respectively. The Langmuir isotherm can also be analyzed using the non-linear form of the following equation:

$$q_e={\displaystyle \frac{q_{max}. K_L.C_e}{{1+K}_L{. C}_e}}$$

(7)

Therefore, non-linear regression was performed to directly estimate q_max and K_L, ensuring accurate parameter estimation and minimizing error distortion. The goodness-of-fit metrics (R²) confirmed the suitability of the Langmuir model. The Langmuir model has been studied at room temperature, and both adsorbents fit the Langmuir isotherm model. R², q_m and K_L values are given in

Table 2. Adsorption parameters.

Adsorbent	Adsorbate	Langmuir				Freundlich
		qe (mg·g−1)	qm (mg·g−1)	KL (L.mg−1)	R2	n	KF (mg·g−1)	R2
M-Sb	MG	217.1	246.2	0.278	0.995	1.99	4.09	0.98
	RhB	62.94	66.45	0.176	0.99	1.49	1.28	0.97
	Sudan II	162.1	176.67	0.761	0.92	2.46	63.43	0.982
M-Bi	MG	230.1	261	2.63	0.997	2.48	4.71	0.98
	RhB	64.21	78.57	0.078	0.99	1.84	3.68	0.97
	Sudan II	170.4	187.96	1.19	0.866	2.08	51.41	0.942

. Adsorption values of bismuth-incorporated ionic liquids are found to be higher than those of antimony-containing ionic liquids.

2.18. Freundlich Isotherm

The adsorption mechanism of heterogeneous systems is explained by the Freundlich isotherm. The Freundlich empirical equation can be expressed as

q_e = K_FC^1/n.

(8)

where q_e is the equilibrium adsorption capacity, and C is the concentration at equilibrium. A linear form of the Freundlich equation can be written by taking the natural log of the following equation:

$$ln{q}_e=ln{K}_F+{\displaystyle \frac{1}{n}}ln{C}_e$$

(9)

where K_F is the Freundlich constant, which gives the measure of adsorption capacity; and 1/n is the adsorption intensity of the adsorbent. The value of the Freundlich constant and the heterogeneity factor can be measured from the intercept and slope of the isotherm plot ln of C_e vs. ln q_e.

Based on R² values, the adsorption mechanism can be better described by the Langmuir isotherm, which represents the homogeneity of the adsorbent surface. The 1/n values represent the favorability of the adsorbent–adsorbate system. The values of the Freundlich plot are given in Table 2 for both the adsorbent and its adsorption properties for dyes. The intercept K_F obtained from the Freundlich isotherm gives roughly the idea of adsorption capacity, and the slope, which is the inverse of n, gives the adsorption intensity. The value of n = 1 suggests homogeneous adsorption and predicts that interactions are absent between the adsorbed species. The value of 1/n less than 1 verifies the favorability of adsorption, and if the value of 1/n is higher than 1, the adsorption is not favorable, in which case the adsorption capacity decreases because of a weaker bond. In the current study, it was found that the 1/n value is less than 1 in the Langmuir plot, which means that adsorption is favorable for the Langmuir isotherm model.

2.19. Thermodynamic Studies

To calculate the thermodynamic parameters, the Van’t Hoff equation is used: [59]

$$ln{\displaystyle \frac{q_e/q_o}{C_e/C_0}}=-\Delta G^o=-{\displaystyle \frac{\Delta H^o}{RT}}+{\displaystyle \frac{\Delta S^o}{R}}$$

(10)

$\Delta G°$, $\Delta H°$ and $\Delta S^o$ of adsorption can be calculated from linear and non-linear forms of the Van’t Hoff equation, as described in the literature, and apparently, no remarkable difference could be observed either in sign or in magnitude [60].

The standard enthalpy of reaction shows that adsorption was endothermic, and further values below 50 kJ/mol represent that the adsorption was the physisorption type in both cases. However, the values of entropy and standard Gibbs free energy show that adsorption was spontaneous. The calculated thermodynamic parameters are shown in

Table 3. Thermodynamic parameters.

Adsorbent	Adsorbate	ΔS° (J·mol−1·K−1)	ΔG° (J·mol−1)	ΔH° (J·mol−1)
M-Sb	MG	566.5	−1920.0	−46,660.0
	RhB	651.9	−3350.0	−45.34
	Sudan II	478.7	−3580.0	−16,840.0
M-Bi	MG	−1.619	−3420.0	−20,340.0
	RhB	471.4	−4110.0	−3.13
	Sudan II	475.1	−3420.0	−16,480.0

.

Comparing the Sb- and Bi-based adsorbents, both have almost similar characteristics of adsorption. Bi-IL- and Sb-IL-based adsorbents both followed the Langmuir isotherm and pseudo-second-order kinetics. The rates are faster for bismuth-based ionic liquids, which could be ascribed to the size and electronegativity difference. The rate constant values are listed in Table 1. The adsorption capacities of both the adsorbents are comparable and have higher capacities for malachite green than Sudan II, followed by rhodamine B. The enthalpy of reaction suggested that both followed endothermic adsorption, and their adsorption was spontaneous, with Gibbs free energy, entropy and enthalpy values listed in Table 3.

2.20. Comparison with the Literature

The adsorption properties and capacities are compared with the previously available adsorbents for malachite green and rhodamine B in Table 4. Most of the adsorbents have followed pseudo-second-order adsorption kinetics and Langmuir isotherms, with few exceptions. The rate constants of prepared adsorbents have shown comparable rate constants and have great adsorption capacities. Similar materials, i.e., imidazolium, Chitosan, and phosphonium ILs, have shown only 84, 9.20 and 150 mg·g⁻¹, respectively, for malachite green, whereas the metal-based ionic liquids in this work showed adsorption capacities of 217 mg·g⁻¹ and 230 mg·g⁻¹ by M-Sb and M-Bi, respectively.

The other materials used for dye removal by adsorption are mainly carbonaceous materials and usually ashes. Rice husk has exhibited an adsorption capacity of 17.98 mg·g⁻¹ for malachite green following pseudo-second-order kinetics and the Freundlich isotherm, i.e., multilayer adsorption. Waste apricot has given the high adsorption capacities for malachite green of about 116.3 mg·g⁻¹, and activated slag adsorbed about 74 mg·g⁻¹ of malachite green following the Langmuir isotherm.

The adsorption capacities by mango leaf powder and fly ash were used by Sharma et al. and Chang et al. [61,62] respectively, and their adsorption capacities were a maximum of 3.31 mg·g⁻¹ and 10.0 mg·g⁻¹, respectively; both followed a single-layer adsorption. Whereas ionic liquids have not been used as adsorbents for rhodamine B removal, other materials, like fly ash and mango leaf powder, have shown adsorption capacities of only 10 mg·g⁻¹ and 3.3 mg·g⁻¹, respectively. Most adsorbents do not adsorb rhodamine B, for it is a hard dye. The sago waste activated carbon has an adsorption capacity of 28 mg·g⁻¹ for the removal of rhodamine B. Duo Lite Resin C-20 has shown a 28 mg·g⁻¹ adsorption capacity for rhodamine B, while antimony- and bismuth-based ionic liquids have shown adsorption capacities of 62 and 64 mg·g⁻¹, respectively.

Table 4. Comparison with the previously reported studies.

Adsorbent	Adsorbate	Isotherm Model	Reaction Kinetics	Adsorption Capacity (mg·g−1)	References
Activated slag	MG	Langmuir	-	74.2	[63]
Waste apricot	MG	Langmuir Model	-	116.3	[64]
Self-assembled ionic liquid-based organosilica (SAIBO)	MG	Langmuir Model	Pseudo-2nd Order k2 = 0.01	19.23	[65]
[BMIM][PF6]	MG	-	Pseudo-2nd Order k2 = 0.056	84	[41]
Chitosan-based ionic liquids	MG	Langmuir Model	Pseudo-2nd Order k2 = 0.055	9.20	[66]
[PC6C6C6C14] [Tf2N]	MG	-	-	150	[67]
Rice Husk	MG	Freundlich Model	Pseudo-2nd Order k2 = 0.032	17.98	[68]
M-Sb	MG	Langmuir Model	Pseudo-2nd Order k2 = 0.00014	217.36	This work
M-Bi	MG	Langmuir Model	Pseudo-2nd Order k2 = 0.00077	230.18	This work
Sago waste activated carbon	Rhodamine B	Langmuir Model	Pseudo-2nd Order	16.12	[69]
Hyper cross-linked polymeric adsorbent	Rhodamine B	Freundlich Model	Pseudo-2nd Order k2 = 0.0002	2.1	[70]
Fly ash	Rhodamine B	Langmuir Model	-	10.0	[61]
Mango leaf powder	Rhodamine B	Langmuir Model	Pseudo-1st Order k1 = 0.053	3.31	[62]
Duo-lite C-20 resin	Rhodamine B	Langmuir Model	Pseudo-1st Order k1 = 0.069	28.57	[71]
M-Sb	Rhodamine B	Langmuir Model	Pseudo-2nd Order k2 = 0.00006	62.21	This work
M-Bi	Rhodamine B	Langmuir Model	Pseudo-2nd Order k2 = 0.00013	64.94	This work
Organoclays MCNTS	Sudan IV	Freundlich Model	Pseudo-2nd Order	22.7	[72]
Coastal soil	Sudan IV	-	-	0.43–4.6	[73]
M-Sb	Sudan II	Langmuir Model	Pseudo-2nd Order	162.3	This Work
M-Bi	Sudan II	Langmuir Model	Pseudo-2nd Order	170.1	This Work

Table 4. Comparison with the previously reported studies.

Adsorbent	Adsorbate	Isotherm Model	Reaction Kinetics	Adsorption Capacity (mg·g−1)	References
Activated slag	MG	Langmuir	-	74.2	[63]
Waste apricot	MG	Langmuir Model	-	116.3	[64]
Self-assembled ionic liquid-based organosilica (SAIBO)	MG	Langmuir Model	Pseudo-2nd Order k2 = 0.01	19.23	[65]
[BMIM][PF6]	MG	-	Pseudo-2nd Order k2 = 0.056	84	[41]
Chitosan-based ionic liquids	MG	Langmuir Model	Pseudo-2nd Order k2 = 0.055	9.20	[66]
[PC6C6C6C14] [Tf2N]	MG	-	-	150	[67]
Rice Husk	MG	Freundlich Model	Pseudo-2nd Order k2 = 0.032	17.98	[68]
M-Sb	MG	Langmuir Model	Pseudo-2nd Order k2 = 0.00014	217.36	This work
M-Bi	MG	Langmuir Model	Pseudo-2nd Order k2 = 0.00077	230.18	This work
Sago waste activated carbon	Rhodamine B	Langmuir Model	Pseudo-2nd Order	16.12	[69]
Hyper cross-linked polymeric adsorbent	Rhodamine B	Freundlich Model	Pseudo-2nd Order k2 = 0.0002	2.1	[70]
Fly ash	Rhodamine B	Langmuir Model	-	10.0	[61]
Mango leaf powder	Rhodamine B	Langmuir Model	Pseudo-1st Order k1 = 0.053	3.31	[62]
Duo-lite C-20 resin	Rhodamine B	Langmuir Model	Pseudo-1st Order k1 = 0.069	28.57	[71]
M-Sb	Rhodamine B	Langmuir Model	Pseudo-2nd Order k2 = 0.00006	62.21	This work
M-Bi	Rhodamine B	Langmuir Model	Pseudo-2nd Order k2 = 0.00013	64.94	This work
Organoclays MCNTS	Sudan IV	Freundlich Model	Pseudo-2nd Order	22.7	[72]
Coastal soil	Sudan IV	-	-	0.43–4.6	[73]
M-Sb	Sudan II	Langmuir Model	Pseudo-2nd Order	162.3	This Work
M-Bi	Sudan II	Langmuir Model	Pseudo-2nd Order	170.1	This Work

2.21. Regeneration and Recyclability of Adsorbents/ILs

The regeneration and reusability of adsorbents are crucial constraints to evaluate their overall activity and practical applications [67]. After the complete separation of dyes with the hydrophobic antimony- and bismuth-based ILs (M-Sb & M-Bi), ethanol was used to wash the ILs. In this way, ILs can be recovered and reused in a new cycle after filtration. Interestingly, Sudan II has a 2.953 mg/kg saturation limit in water [53], suggesting that the solubility of these hydrophobic dyes increases in the presence of ILs in aqueous solution, and the same aqueous phase of ILs could be used several times without reaching saturation, as shown in Figure 6.

3. Experimental

3.1. Materials

N,N-Dimethylaniline, methyl iodide, ethyl iodide, antimony(III) chloride, dimethyl sulfoxide (DMSO), bismuth(III) chloride, methanol, acetonitrile (ACN), tetrahydrofuran (THF), hydrochloric acid (HCl), malachite green (MG), rhodamine B (RhB) and Sudan II were purchased from Sigma-Aldrich Chemicals (Co., Gillingham, UK) and used without any purification.

3.2. Instrumentations

The melting points of synthesized ILs were measured by using an electrothermal melting point apparatus (model MP-D Mitamura Riken, Kogyo, Tokyo, Japan). The Bio-Rad Excalibur (model FTS 3000 MX, Bio-Rad, Hercules, CA, USA) and Bruker Advance Digital FT-NMR spectrometer (Fällanden, Switzerland) with 300 MHz were employed to record the IR spectra and multinuclear (¹H and ¹³C) NMR spectra, respectively. DMSO and deuterated ethanol were used as solvents for ¹H and ¹³C NMR. A Renishaw in Via Raman microscope (London, UK) was used to obtain the Raman spectra, having a 10 s acquisition time, around a 1 µm² laser spot and an excitation wavelength of 785 nm (130 mW, 10%). Optical studies were carried out using the UV-1700 Shimadzu (Kyoto, Japan).

3.3. Synthesis of Adsorbent

Antimony- and bismuth-based ionic liquids were prepared in a two-step reaction using a reported procedure [25], described in Scheme 1. In the first step, the alkylation of N,N-dimethylaniline was performed, which was followed by an anion exchange reaction. An equimolar ratio of alkyl (methyl and ethyl) iodide (0.025 moles) and N,N-dimethylaniline (0.025 moles) was prepared in 25 mL of THF separately and was stirred in a 100 mL round-bottom flask. The temperature of the reaction mixture was maintained at 70–80 °C and allowed to react for 24 h. Pale-green solid products (M & E-ILs) were obtained in the flask, which was filtered, washed with ethyl acetate and allowed to dry. The filtrate was stirred for an additional hour to ensure the complete precipitation of the product. An above 90% yield was obtained after the complete precipitation of the product. In a second step, metal was incorporated by a metathesis reaction. The synthesized N,N,N-trialkylanilinium iodide (M & E-ILs) and metal (antimony and bismuth) chlorides in the equimolar ratio of 1:1 were separately dissolved in acetonitrile and agitated for 6 h, although the reaction was rapid, and solid products were precipitated immediately at room temperature when the reactants were mixed. The light-yellow antimony and orange-colored bismuth-based solid products (Sb & Bi-ILs) were separated, washed with n-hexane and diethyl ether and allowed to air dry. The structures of these synthesized ILs are shown in Scheme 2.

3.4. Ionic Liquids Having Iodide as Anion (M- and E-ILs)

N,N,N-Trimethylanilinium iodide (M-IL): ¹H NMR (400 MHz, CD₃Cl), d: 7.69 (d, ²J_HH = 8.0 Hz, 2H), 7.54 (m, 3H), 3.50 (s, 9H). ¹³C NMR (150 MHz, CD₃Cl), d: 147.70, 130.51, 120.92, 56.88. FT-IR (cm⁻¹); nN–C (1006), sp³ C–H (2852), sp² C–H (2983), nC=C(1589).

N,N-Dimethyl-N-ethylanilinium iodide (E-IL): ¹H NMR (400 MHz, CD₃Cl), d: 8.01 (d, ²J_HH = 10.8 Hz 2H), 7.61 (m, 5H), 3.64 (s, 2H), 3.37 (s, 6H). ¹³C NMR (75 MHz, CD₃Cl), d: 143.53, 130.78, 120.13, 65.14, 55.06, 9.34. FT-IR (cm⁻¹); nN–C (1006), sp³ C–H (2852), sp² C–H (2983), nC=C (1589).

Antimony-based ionic liquids (Sb-ILs)

N,N,N-Trimethylanilinium antimony(IV) tetrachloride (M-Sb): Quantities used: N,N,N-Trimethylanilinium iodide (2 mmol), Antimony(III) tetrachloride (2 × 10⁻³ mol); m.p.: 217 °C; Pale yellow powder; anal. calcd%: C, 27.04; H, 3.53; Cl, 35.47; N, 3.50; Sb, 30.46; found %: C, 26.99; H, 3.51; N, 3.48; Cl, 35.46; Sb; 30.44, FT-IR (cm⁻¹); nN–C (1001), sp³ C–H (2858), sp² C–H (2987), nC=C (1589).

N,N-Dimethy-N-ethylanilinium antimony(IV)tetrachloride (E-Sb): Quantities used: N,N-Dimethyl-N-ethylanilinium iodide (2 mmol), Antimony(III) tetrachloride (2 × 10⁻³ mol); m.p.: 211 °C; Pale yellow powder; anal. calcd%: C, 35.48; H, 5.95; Cl, 29.92; N, 2.96; Sb, 25.69; found %: C, 35.50; H, 5.94; N, 2.97; Cl, 29.93; Sb; 25.7, FT-IR (cm⁻¹) nN–C (1000), sp³ C–H (2855), sp² C–H (2972), nC=C (1589).

Bismuth-based ionic liquids (Bi-ILs)

N,N-N-Trimethylanilinium Bismuth(IV) tetrachloride (M-Bi): Quantities used: N,N-Dimethyl-N-ethylanilinium iodide (2 mmol), Bismuth(III) tetrachloride (2 mmol); m.p.: 215 °C; Pale yellow powder; anal. calcd (%) C, 22.20; H, 2.90; Bi, 42.91; Cl, 29.12; N, 2.88; found %: C, 22.18; H, 2.94; N, 2.93; Cl, 29.09; Bi; 42.90, FT-IR (cm⁻¹) nN–C (1001), sp³ C–H (2875), sp² C–H (2970), nC=C (1478).

N,N-Dimethy-N-ethylanilinium Bismuth(IV) tetrachloride (E-Bi): Quantities used: N,N-Dimethyl-N-ethylanilinium iodide (2 mmol), Bismuth(III) tetrachloride (2 mmol); m.p.: 201 °C; Pale yellow powder; anal. calcd(%): C, 23.97; H, 3.22; Bi, 41.71; Cl, 28.30; N, 2.80; found %: C, 23.96; H, 3.20; Bi, 41.71; Cl, 28.31; N, 2.81, FT-IR (cm⁻¹) nN–C (1003), sp³ C–H (2851), sp² C–H (2969), nC=C (1478).

3.5. Dye Adsorption

Three dyes—malachite green, Sudan II and rhodamine B—were adsorbed on a prepared adsorbent according to the following procedure: A stock solution of 1000 ppm was prepared by dissolving a weighed amount of 100 mg of dye in 100 mL of deionized water (50% ethanol was used for Sudan II). The solutions were diluted to an appropriate concentration for conducting the experiments. A weighed amount of adsorbent was added to 40 cm³ of 25 ppm dye solution in water and stirred for different intervals of time (10% ethanolic solution was used for Sudan II). The following four parameters were varied to find the effect on the rate of adsorption and are discussed in detail below. The adsorbate and synthesized adsorbent (antimony- and bismuth-based ILs) were separated by centrifugation, followed by filtration. The lingering concentration of dye was determined by using a calibration curve attained from a UV/Vis spectrophotometer at its particular maximum wavelength.

$$\mathit{A}\mathit{d}\mathit{s}\mathit{o}\mathit{r}\mathit{p}\mathit{t}\mathit{i}\mathit{o}\mathit{n} \mathit{E}\mathit{f}\mathit{f}\mathit{e}\mathit{c}\mathit{i}\mathit{e}\mathit{n}\mathit{c}\mathit{y}(\mathit{S}\mathit{\%})={\displaystyle \frac{{(\mathit{C}}_{\mathit{i}}-{\mathit{C}}_{\mathit{e}})}{{\mathit{C}}_{\mathit{i}}}}\mathit{x}100$$

(11)

$$\mathit{A}\mathit{d}\mathit{s}\mathit{o}\mathit{r}\mathit{p}\mathit{t}\mathit{i}\mathit{o}\mathit{n} \mathit{c}\mathit{a}\mathit{p}\mathit{a}\mathit{c}\mathit{i}\mathit{t}\mathit{y}({\mathit{q}}_{\mathit{e}})={\displaystyle \frac{{(\mathit{C}}_{\mathit{i}}-{\mathit{C}}_{\mathit{e}})}{\mathit{m}}}\mathit{V}$$

(12)

where Ci = initial concentration of the dyes (mg L⁻¹);

Ce = the equilibrium concentration of the dye (mg L⁻¹);

m = the mass of adsorbent material (g);

V= the volume (dm³).

3.6. Adsorption Kinetics

The kinetic studies were carried out at various temperatures (293, 303, 313, 323 and 333 K). Adsorbates at initial concentrations of 25 ppm were agitated with optimized amounts of adsorbents for the optimized intervals of time, and the final concentrations were recorded spectrophotometrically.

3.7. Adsorption Isotherm

The equilibrium points of RhB, MG and Sudan II were determined by carrying out the agitation experiment of the optimized dose of adsorbents (Sb- and Bi-ILs) for each adsorbate with varying concentrations (5, 10, 15, 25 and 50 ppm) of all three adsorbates at varying pH values. Approximately 5 mL of each solution was taken out after intervals of time and centrifuged, and then final concentrations were determined spectrophotometrically at 555 nm for RhB, 617 nm for MG and 495 nm for Sudan II.

4. Conclusions

The Sb- and Bi-based ILs were successfully synthesized by a metathesis reaction. All of them were characterized by ¹HNMR, ¹³CNMR, FTIR, Raman and X-ray photoelectron spectroscopy. They showed great adsorption efficiencies for the removal of rhodamine B and malachite green. The parameters affecting adsorption efficiencies (stirring rate, pH, time and adsorbent dosage) were optimized. The adsorption efficiencies increased with an increase in stirring rate, pH, time and adsorbent dosage. With the optimization of factors, the adsorption efficiencies increased above 90%. Bi- and Sb-based ILs showed comparable adsorption capacities. The maximum adsorption capacity for malachite green, Sudan II and rhodamine B by M-Bi was 230.18 mg·g⁻¹, 170 mg·g−1 and 64.21 mg·g−1, respectively, whereas by M-Sb it was 217.36 mg·g−1, 162.1 mg·g⁻¹ and 62.94 mg·g⁻¹, respectively. The adsorption was homogeneous in nature, occurred through monolayer adsorption and followed pseudo-second-order kinetics. The thermodynamic parameters showed that the adsorption was endothermic and occurred spontaneously by a physisorption process.