Development of New Efficient Adsorbent by Functionalization of Mg3Al-LDH with Methyl Trialkyl Ammonium Chloride Ionic Liquid

The present paper describes a new way of obtaining an efficient adsorbent material by functionalization of Mg3Al layered double hydroxides (LDH) with methyl trialkyl ammonium chloride-ionic liquid (IL) using two methods: ultrasound and cosynthesis. Layered double hydroxides are good solid support for the functionalization with ionic liquids due to their well-ordered structure. The immobilization of the ILs in suitable solid supports combine the advantages of the ILs with the properties of the solid supports bringing more benefits such as use of lower quantity of ILs and avoiding of ILs loss in the aqua phase which overall decrease the treatment costs. In case of ultrasound method of functionalization is assured a uniform distribution of IL on the solid surface, but through immobilization by cosynthesis due to the tunable properties of LDH, is assured an intercalation of the ILs between the LDH layers. This fact was highlighted by the X-ray diffraction (RXD), scanning electron microscopy (SEM) analyses and Fourier-transform infrared (FTIR) spectroscopy of the obtained adsorbent. The added value brought by the functionalization of Mg3Al with the studied IL was underlined by the adsorption studies conducted in the treatment process of water with diclofenac content. Kinetic, thermodynamic, and equilibrium studies were performed. DCF adsorption onto the studied materials correspond to a chemisorption, the pseudo-second-order kinetic model describing the most accurately the experimental data. DCF adsorption onto the studied materials occurs as a heterogeneous process, with the experimental data fitting best with the SIPS isotherm. The sample obtained through cosynthesis developed a maximum adsorption capacity of 648 mg/g.


Introduction
Water is an essential nutrient used in many fields, and for this reason keeping the water sources as fresh as possible, by pollution prevention or by development of efficient method of treatment, is the main concern of humanity [1]. Water treatment technologies were intensively studied, and after a primary treatment, it is more often necessary to use an advanced water treatment technology. This is especially the case when it is necessary to remove trace contaminants, which, even if they are found in low concentrations, are very harmful for living organisms such as: heavy metals and drugs. From the studied advanced water treatment technologies, adsorption is one of the most efficient due to its ease of operation and high-performance results [2][3][4]. The most important factor in the adsorption processes to obtain great efficiencies is represented by the used adsorbent materials. Therefore, the development of new efficient adsorbent materials is still a challenge. An efficient adsorbent material must meet the following requirements: high adsorption kinetics and capacity [5], excellent selectivity [3], great thermal and chemical stability [4,6], and ease of synthesis [5,7].

Adsorbent Materials Characterization
The RX diffractograms of the synthesized Mg 3 Al raw material and of the ultrasonication and cosynthesis functionalized LDH (Mg 3 Al IL-US and Mg 3 Al IL-COS, respectively) are presented in Figure 1, and the lattice parameters and the crystallites size are shown in Table 1.
All the three adsorbent materials present the only crystalline phase those of the layered double hydroxide. The crystallinity of Mg 3 Al and Mg 3 Al IL-US samples is evidenced by the characteristic peaks corresponding to (003) and (006) planes (at low values of 2θ angles) and corresponding to (110) and (113) planes (at higher values of 2θ angles), which are sharp and symmetrical [29]. The ultrasound method of functionalization with the studied IL does not affect the crystallinity of Mg 3 Al IL-US sample. In case of Mg 3 Al IL-COS, this well-ordered structure is disturbed because, in this case, the IL is not just attached on the surface of Mg 3 Al, it could be found in the interlayer space, leading to a decrease in the sample crystallinity, and therefore, a more amorphous sample is obtained. In this case, the corresponding peaks to (003) and (006) planes are shifted to lower 2θ angles, presenting a flattering of the maxima/a lower intensity. The data presented in Table 1 show that the interlayer space increases from 23.58 to 24.39 Å, leading to a decrease in the crystallite size from 4.86 to 3.38 nm. These results prove the intercalation of the IL between the LDH layers [14,22,30]. In contrast, in the case of the Mg 3 Al IL-US sample, the distance between the layers does not show a significant change, indicating the functionalization with IL by ultrasound at the surface of the adsorbent.
by the characteristic peaks corresponding to (003) and (006) planes (at low values o angles) and corresponding to (110) and (113) planes (at higher values of 2θ angles), w are sharp and symmetrical [29]. The ultrasound method of functionalization with studied IL does not affect the crystallinity of Mg3Al IL-US sample. In case of Mg3A COS, this well-ordered structure is disturbed because, in this case, the IL is not jus tached on the surface of Mg3Al, it could be found in the interlayer space, leading decrease in the sample crystallinity, and therefore, a more amorphous sample is obtai In this case, the corresponding peaks to (003) and (006) planes are shifted to lowe angles, presenting a flattering of the maxima/a lower intensity. The data presented in ble 1 show that the interlayer space increases from 23.58 to 24.39 Å, leading to a decr in the crystallite size from 4.86 to 3.38 nm. These results prove the intercalation of th between the LDH layers [14,22,30]. In contrast, in the case of the Mg3Al IL-US sample distance between the layers does not show a significant change, indicating the functio ization with IL by ultrasound at the surface of the adsorbent.  Scanning electron microscopy (SEM) images and Energy Dispersive X-ray Spec copy (EDX) spectra of studied adsorbent materials are shown in Figure 2.
The morphology of the Mg3Al sample is presented in the form of hexagons arran in order in overlapping layers, specific to layered double hydroxides. [8] The sample f tionalized by cosynthesis shows a network change: the layers are disordered due to intercalation from place to place of the ionic liquid between LDH layers. The sample face is presented like cotton flowers. For the sample functionalized by ultrasound met this network disorder does not occur, as only the attachment of the ionic liquid to surface of the solid support is observed, which leads to its homogenization. Mg3Al f tionalization with the studied IL was also highlighted by EDX analysis in which the p ence of nitrogen and chlorine could be observed.  Scanning electron microscopy (SEM) images and Energy Dispersive X-ray Spectroscopy (EDX) spectra of studied adsorbent materials are shown in Figure 2.
The morphology of the Mg 3 Al sample is presented in the form of hexagons arranged in order in overlapping layers, specific to layered double hydroxides [8]. The sample functionalized by cosynthesis shows a network change: the layers are disordered due to the intercalation from place to place of the ionic liquid between LDH layers. The sample surface is presented like cotton flowers. For the sample functionalized by ultrasound method, this network disorder does not occur, as only the attachment of the ionic liquid to the surface of the solid support is observed, which leads to its homogenization. Mg 3 Al functionalization with the studied IL was also highlighted by EDX analysis in which the presence of nitrogen and chlorine could be observed.

pH Influence upon the Adsorption Capacity Developed by the Studied Materials
In adsorption processes, the initial pH of the solutions plays an important role since it influences the surface of the adsorbent. The experimental results on the study of diclofenac adsorption (DCF) on the studied materials under conditions of pH variation are presented in Figure 3.

pH Influence upon the Adsorption Capacity Developed by the Studied Materials
In adsorption processes, the initial pH of the solutions plays an important role since it influences the surface of the adsorbent. The experimental results on the study of diclofenac adsorption (DCF) on the studied materials under conditions of pH variation are presented in Figure 3.  With the increase of the initial pH of DCF solutions from 4 to 7, the adsorption capacity of all studied materials increases. At pH values equal to 8, it decreases and at even higher values, there is a new increase of the studied parameters, but this is due to the precipitation of DCF and not due to an adsorption process. The aspects presented show that the optimal pH value with which to reach the maximum adsorption capacity under With the increase of the initial pH of DCF solutions from 4 to 7, the adsorption capacity of all studied materials increases. At pH values equal to 8, it decreases and at even higher values, there is a new increase of the studied parameters, but this is due to the precipitation of DCF and not due to an adsorption process. The aspects presented show that the optimal pH value with which to reach the maximum adsorption capacity under the given working conditions is pH = 7. According to literature, diclofenac is a weak organic acid, and at pH = 7.4, more than 90% of it exists in ionized anionic form in the solutions [31,32]. For these reasons, using Mg 3 Al-LDH as adsorbent material, the optimum pH for DCF adsorption proved to be 7 due to electrostatic interaction between the positive charge of the LDH surface and DCF, which is found in ionized anions. This agrees with other results presented in literature [3,25]. This pH value was used in subsequent adsorption studies.

Kinetic Studies
Studies aimed the influence of time upon the adsorption process present a particular importance, as they provide information for determining the time required to reach the adsorption equilibrium; furthermore, the results can be used to understand the variables that influence the adsorption of the solute. Experimental data regarding the adsorption capacity dependence of the studied materials as a function of the contact time for 3 initial concentrations of DCF at 3 different temperatures are presented in  However, this is insignificant, and it is not economically justified to conduct the adsorption process at higher temperatures. The time of 60 min was considered the optimal contact time, as this was respectively the time required to reach the adsorption equilibrium.
Adsorption kinetics is an important feature that defines the efficiency of the adsorption process. To clarify the mechanism of diclofenac adsorption onto the studied materials and to identify the model that validates the experimental data, three kinetic models were used: pseudo-first-order kinetic model, and pseudo-second-order kinetic model, and intraparticle diffusion model [28,33].
For the modeling of the kinetic data, the integrated form of the Lagergren model was used, as described by the relation: where: qt-adsorption capacity of the adsorbent at time t, mg/g; qe-adsorption capacity of the adsorbent at equilibrium, mg/g; k1-adsorption rate constant, min −1 ; t-stirring time, min.
From the linear dependencies ln(qe − qt) as a function of t represented in Figure S1      For all the studied materials and for all the initial DCF concentrations, the adsorption capacity of the materials increases with the contact time increasing up to 60 min, after which it remains constant. At stirring times of 60 min, the adsorption capacity remains practically constant, probably due to occupying of the active sites from the adsorbent surface [26]. It is important to know the time required to achieve the equilibrium between adsorbent and adsorbate to estimate the time required for water treatment to extend the treatment process from the laboratory scale. The equilibrium between adsorbent and adsorbate is achieved quickly in 60 min. The behavior of the adsorption capacities of the studied materials' function of time is similar, regardless of the initial concentration of DCF in the solution. Along with the increase of the initial concentration of DCF in aqueous solutions, the adsorption efficiency of the studied materials is intensified. The adsorption capacity showed a slight increase in tandem with an increasing working temperature. However, this is insignificant, and it is not economically justified to conduct the adsorption process at higher temperatures. The time of 60 min was considered the optimal contact time, as this was respectively the time required to reach the adsorption equilibrium.
Adsorption kinetics is an important feature that defines the efficiency of the adsorption process. To clarify the mechanism of diclofenac adsorption onto the studied materials and to identify the model that validates the experimental data, three kinetic models were used: pseudo-first-order kinetic model, and pseudo-second-order kinetic model, and intraparticle diffusion model [28,33].
For the modeling of the kinetic data, the integrated form of the Lagergren model was used, as described by the relation: where: q t -adsorption capacity of the adsorbent at time t, mg/g; q e -adsorption capacity of the adsorbent at equilibrium, mg/g; k 1 -adsorption rate constant, min −1 ; t-stirring time, min.
From the linear dependencies ln(q e − q t ) as a function of t represented in Figure S1 (Supplementary Information), the velocity constants k 1 were calculated, and the results are presented in Tables 2-4. The pseudo-second-order kinetic model, proposed by Ho and McKay, can be expressed by the equation: where: k 2 -the rate constant of the pseudo-second-order kinetic adsorption model (g/mg/min) q e -the amount of DCF adsorbed at equilibrium (mg/g); q t -the amount of DCF adsorbed at time t (mg/g).
The graphical representations t/q t as a function of time t allowed the evaluation of the rate constants k 2 (from the slope of the lines), respectively, of the quantity of diclofenac adsorbed at equilibrium q e (from the ordinate to the origin) ( Figure S2 Supplementary Information).
The kinetic model of intraparticle diffusion is described by the Weber-Morris equation, which was used to identify the rate determinant stage if the intraparticle diffusion describes the kinetic process: where: k int is the rate constant of the kinetic model of intraparticle diffusion (mg/g/min 1/2 ). The values of the k int rate constants were estimated from the slopes of the q t representations as a function of t 1/2 ( Figure S3 Supplementary Information).
According to Equation (3), the dependence of q t as a function of t 1/2 should lead to a straight line, the slope of which represents the rate constant k int in which case the intra-particle diffusion is the determining speed step. If the line does not pass through the origin, the intraparticle diffusion may be accompanied by the film diffusion.
The experimental values and the calculated ones for the adsorption capacity at equilibrium, q e , and the values of the rate constants and of the regression coefficients for all the studied cases are presented in Tables 2-4.   According to the data presented in Tables 2-4, the application of the pseudo-first-order kinetic model led to large differences between the values of the equilibrium adsorption capacity determined experimentally and those obtained by modeling. This observation, correlated with the lower values of the regression coefficients, led to the conclusion that the pseudo-first-order kinetic model does not describe the adsorption process of diclofenac onto the studied materials.
The linear dependencies t/q t as a function of t ( Figure S2), obtained by applying the kinetic model of pseudo-second-order, indicate a good concordance of this model with the experimental data. Moreover, the correlation coefficients resulting from the linear representations have values higher than 0.98 for all the studied situations. Also, the values of the adsorption capacity obtained experimentally are close to the values obtained from the graphical representations for all the studied materials at all 3 initial concentrations of DCF from aqueous solutions and at all 3 temperatures. Therefore, the DCF adsorption process on the studied materials is faithfully described by the pseudo-second-order kinetic model. The adsorption efficiency is much higher for the functionalized Mg 3 Al samples.
From the q e dependencies as a function of t 1/2 ( Figure S3), the DCF adsorption on the studied materials has a complex mechanism which takes place in two stages corresponding to the two levels in the diagrams. The first stage corresponds to an instant adsorption process, characterized by the diffusion of the solute at the external surface of the adsorbent. The second level, with a lower slope, can be attributed to the adsorption process that takes place gradually at the pores of the adsorbent materials and corresponds to the rate determining the speed [34].
The best fit of DCF adsorption data with the pseudo-second-order kinetic model indicates that the adsorption process correspond to a chemisorption as the rate-limiting mechanism, which occurs by electron sharing or exchange between DCF and active sites/functional groups of the studied adsorbent materials [27].

Thermodynamic Studies
To assess the nature of the adsorption process of DCF onto the studied materials, the thermodynamic parameters, such as standard enthalpy (∆H 0 ), entropy (∆S 0 ), and Gibbs free energy (∆G 0 ), were evaluated using the following equations: where: ∆G is Gibbs free energy (KJ/mol), ∆S is the entropy and heat of adsorption (J/mol K), ∆H is enthalpy (kJ/mol), T is the absolute temperature (K), R is universal gas constant (8.314 J/(mol·K)), and K d is the distribution coefficient. The thermodynamic variables ∆H 0 and ∆S 0 were obtained from the slope and intercept of Von't Hoff plots of lnK d versus 1/T (Figure 7), and the resulting values, together with the Gibbs free energy ∆G 0 calculated with Equation (5) and the correlation coefficients, are presented in Table 5.
where: ΔG is Gibbs free energy (KJ/mol), ΔS is the entropy and heat of adsorption (J/mol K), ΔH is enthalpy (kJ/mol), T is the absolute temperature (K), R is universal gas constant (8.314 J/(mol•K)), and Kd is the distribution coefficient. The thermodynamic variables ΔH 0 and ΔS 0 were obtained from the slope and intercept of Von't Hoff plots of lnKd versus 1/T (Figure 7), and the resulting values, together with the Gibbs free energy ΔG 0 calculated with Equation (5) and the correlation coefficients, are presented in Table 5.
The positive value of ΔH 0 indicates that the adsorption process of DCF onto the studied materials is endothermic in nature. In case of the functionalized samples (Mg3Al IL-COS and Mg3Al IL-US), the feasibility and spontaneity of the adsorption process was confirmed by the obtained negative values of ΔG 0 , and these values are more negatively with the temperature increasing and the initial concentration of DCF increasing. This suggests that at higher concentrations of DCF in the solution, more active groups are available in the solution, assuring a better contact with the adsorbent surface and leading to a greater adsorption efficiency. The positive value of ΔS 0 suggested an increase in randomness at the solid/liquid interface during adsorption of DCF onto the functionalized samples. In case of raw Mg3Al, due to the absence of the functional group given by the studied IL from the adsorbent surface, it diminished the randomness at the solid/liquid interface, a fact that was also suggested by the obtained negative value ΔS 0 .    The positive value of ∆H 0 indicates that the adsorption process of DCF onto the studied materials is endothermic in nature. In case of the functionalized samples (Mg 3 Al IL-COS and Mg 3 Al IL-US), the feasibility and spontaneity of the adsorption process was confirmed by the obtained negative values of ∆G 0 , and these values are more negatively with the temperature increasing and the initial concentration of DCF increasing. This suggests that at higher concentrations of DCF in the solution, more active groups are available in the solution, assuring a better contact with the adsorbent surface and leading to a greater adsorption efficiency. The positive value of ∆S 0 suggested an increase in randomness at the solid/liquid interface during adsorption of DCF onto the functionalized samples. In case of raw Mg 3 Al, due to the absence of the functional group given by the studied IL from the adsorbent surface, it diminished the randomness at the solid/liquid interface, a fact that was also suggested by the obtained negative value ∆S 0 .

Equilibrium Studies
Equilibrium studies provide information on the maximum adsorption capacity and affinity of the materials used as adsorbents in relation to the target pollutants. Isotherm studies are essential to interpret the interaction between the adsorbent and adsorbate. The modeling of experimental data allows the identification of the process mechanism and the theoretical evaluation of the adsorption capacities at equilibrium. In this paper, the nonlinear equation of Langmuir, Freundlich, and Sips isotherms were used for modeling the experimental data.
Langmuir isotherm supposes that the adsorption of the target pollutant takes place as a complete monolayer coverage of the adsorbent surface, and its equation is given as: where: q e -amount of DCF adsorbed at equilibrium (mg/g) C e -equilibrium concentration of DCF in solution (mg/L) q m -maximum adsorption capacity in Langmuir model (mg/g) K L -Langmuir constant (L/mg).
In contrast the Freundlich isotherm supposes that the adsorption occurs in multilayers. The nonlinear form of Freundlich isotherm is given by Equation (9): where: K F -Freundlich isotherm constant (mg 1−1/n L 1/n /g) n-Freundlich exponent.
To predict the heterogenous adsorption systems and localized adsorption without adsorbate-adsorbent interaction, a combined form of Langmuir and Freundlich isotherm, the Sips model, is used. This model is given by the next equation: where: q mS -maximum adsorption capacity in Sips model (mg/g) K S -Sips isotherm constant (L m /mg m ) m-Sips exponent.
The graphical representation of the studied isotherm for the DCF adsorption onto the studied materials is presented in Figure 8, and the obtained parameters are summarized in Table 6.
The functionalization of adsorbent materials with the studied ionic liquid leads to an increase in their adsorbent efficiency, which means that the functional groups from the ionic liquid contribute to the process of DCF removal from aqueous solutions. Through ultrasonic functionalization of the LDH, the adsorption capacity of the obtained Mg 3 Al IL-US material increases by two-times greater than the raw Mg 3 Al material. The sample obtained by cosynthesis (Mg 3 Al IL-COS) develops a maximum adsorption capacity fivetimes higher than the capacity developed by the raw LDH. In this case, with IL being between the LDH layers, as was observed from the RX and SEM analysis of the samples, the DCF adsorption occurs due to a synergetic effect between the functional groups of IL and the porosity of the LDH surface. By functionalizing the LDH with the studied IL by cosynthesis, the cover of the LDH surface in the form of a film is avoided, which leads to a decrease in the porosity of the material, and implicitly, the decrease in adsorption in the LDH pores of DCF, as it happened in the case of samples obtained by ultrasound. Thus, in this case, both the functional groups of IL and the surface of LDH contribute to the removal of DCF from aqueous solutions, obtaining the highest value of the maximum adsorption capacity.
Comparing the parameters of the used isotherm for the experimental data modeling, the DCF adsorption onto the studied materials is best described by the Sips model. In this case, we obtained the correlation coefficients close to unity and maximum adsorption capacity calculated from the model close with those experimentally obtained. This indicates that the DCF adsorption onto the studied materials occurs as a heterogeneous process, not as a monolayer covering. Figure 9 presents the FTIR spectra of adsorbents before and after DCF adsorption. The FTIR spectrum of Mg 3 Al (Figure 9a) exhibits characteristic peaks of layered double hydroxide with carbonate as interlayer anion: the broad and strong absorption band in the range of 3600-3200 cm −1 is due to the O-H stretching vibration of the surface and interlayer water molecules. The broad shoulder at 3050 cm −1 could been attributed to hydrogen bounding between water and carbonate anions in the interlayer galleries [35]. The adsorption band with maximum at 1640 cm −1 is due to the water bending vibration of the interlayer water molecules. The absorption band located around 1500 cm −1 is due to the stretching vibrations of carbonate anions. The presence of a multiple band between 1200-1600 cm −1 , and the lack of even a single intense band at 1383 cm −1 indicates the lowering of the symmetry of carbonate anions from the planar D 3h to the C 2ν symmetry. Weakly resolved shoulder present at 850 cm −1 could be attributed to bending vibrations of carbonates. In addition, the spectrum contains absorption bands at 650 cm −1 (bending vibrations of carbonate bonds) and 420 cm −1 (vibrations of Mg-OH, Al-OH and Zn-OH bonds from octahedral networks of layered double hydroxide) [36].
The adsorbents prepared by cosynthesis (Mg 3 Al IL-COS, Figure 9b) and by ultrasonication (Mg 3 Al IL-US, Figure 9c) present, besides the characteristic absorption bans of layered double hydroxide, a doublet of shoulders (at 2860 and 2930 cm −1 ) that overlap to the shoulder assigned to hydrogen bonding in the LDH, and they could be assigned to the ν 3 vibrations of -CH 3 group in the IL [22]. This doublet is more evident in the FTIR spectrum of Mg 3 Al IL-COS sample, and it proves the presence of IL in the layered double hydroxide structure. obtained by cosynthesis (Mg3Al IL-COS) develops a maximum adsorption capacity five-times higher than the capacity developed by the raw LDH. In this case, with IL being between the LDH layers, as was observed from the RX and SEM analysis of the samples, the DCF adsorption occurs due to a synergetic effect between the functional groups of IL and the porosity of the LDH surface. By functionalizing the LDH with the studied IL by cosynthesis, the cover of the LDH surface in the form of a film is avoided, which leads to a decrease in the porosity of the material, and implicitly, the decrease in adsorption in the LDH pores of DCF, as it happened in the case of samples obtained by ultrasound. Thus, in this case, both the functional groups of IL and the surface of LDH contribute to the removal of DCF from aqueous solutions, obtaining the highest value of the maximum adsorption capacity.  Comparing the parameters of the used isotherm for the experimental data modeling, the DCF adsorption onto the studied materials is best described by the Sips model. In this case, we obtained the correlation coefficients close to unity and maximum adsorption capacity calculated from the model close with those experimentally obtained. This indicates that the DCF adsorption onto the studied materials occurs as a heterogeneous process, not as a monolayer covering. Figure 9 presents the FTIR spectra of adsorbents before and after DCF adsorption. The FTIR spectrum of Mg3Al (Figure 9a) exhibits characteristic peaks of layered double   The adsorbents prepared by cosynthesis (Mg3Al IL-COS, Figure 9b) and by ultrasonication (Mg3Al IL-US, Figure 9c) present, besides the characteristic absorption bans of layered double hydroxide, a doublet of shoulders (at 2860 and 2930 cm −1 ) that overlap to the shoulder assigned to hydrogen bonding in the LDH, and they could be assigned to the ν3 vibrations of -CH3 group in the IL [22]. This doublet is more evident in the FTIR spectrum of Mg3Al IL-COS sample, and it proves the presence of IL in the layered double hydroxide structure.
Sodium diclofenac spectrum exhibits some characteristic IR-absorption bands: one peak at about 1600 cm −1 , attributed to the C=C bonds in benzene ring, two adsorption bands located at 1576 and 1507 cm −1 due to C=O stretching and C=C stretching of carboxyl group, and two peaks located at 1306 and 1283 cm −1 , attributed to C-N stretching vibrations [37]. After DCF adsorption, in the adsorbents' spectra, the vibration bands due to the carboxyl group from the DCF structure appear. The weak shoulders at 1500 and 1575 cm −1 , as observed in the Mg3Al spectrum after DCF adsorption (Figure 9a), at 1506 and 1579 cm −1 , as observed in the Mg3Al IL-COS spectrum (Figure 9b), and at 1504 and 1581 cm −1 , as observed in the Mg3Al IL-US spectrum (Figure 9c), could be attributed to the C=O stretching and C=C stretching of carboxyl group from diclofenac. This absorption band overlaps with the band attributed to stretching vibrations of carbonate anion; it is presumed that the carboxyl group replaced some carbonate anions placed onto the positive Sodium diclofenac spectrum exhibits some characteristic IR-absorption bands: one peak at about 1600 cm −1 , attributed to the C=C bonds in benzene ring, two adsorption bands located at 1576 and 1507 cm −1 due to C=O stretching and C=C stretching of carboxyl group, and two peaks located at 1306 and 1283 cm −1 , attributed to C-N stretching vibrations [37]. After DCF adsorption, in the adsorbents' spectra, the vibration bands due to the carboxyl group from the DCF structure appear. The weak shoulders at 1500 and 1575 cm −1 , as observed in the Mg 3 Al spectrum after DCF adsorption (Figure 9a), at 1506 and 1579 cm −1 , as observed in the Mg 3 Al IL-COS spectrum (Figure 9b), and at 1504 and 1581 cm −1 , as observed in the Mg 3 Al IL-US spectrum (Figure 9c), could be attributed to the C=O stretching and C=C stretching of carboxyl group from diclofenac. This absorption band overlaps with the band attributed to stretching vibrations of carbonate anion; it is presumed that the carboxyl group replaced some carbonate anions placed onto the positive surface of layered double hydroxides. The higher the amount of DCF adsorbed, as the equilibrium studies reveal, the more visible and intense this IR-absorption bands are.
Comparing the maximum adsorption capacity developed by the studied materials with the adsorption capacity, as reported in literature and developed by other materials in the removal process of DCF from aqueous solutions (Table 7), we obtained through functionalization of Mg 3 Al with methyl trialkyl ammonium chloride, especially through cosynthesis method, an efficient and high-performance adsorbent that could be used in the treatment process of waters with pharmaceutical content.

Adsorbent Obtaining and Characterization
Mg 3 Al-LDH was synthetized by coprecipitation method at low supra-saturation [39]. As metal precursor were used nitrate salts (Mg(NO 3 ) 2 ·6H 2 O and Al(NO 3 ) 3 ·9H 2 O). A mixed solution of 1 M concentration, containing the metal ions in the ratio, Mg:Al = 3:1, was added drop by drop into 100 mL, 1 M Na 2 CO 3 solution. The suspension was kept continuous under vigorous stirring, maintaining the pH around 10.5 value, with the help of 2M NaOH solution. After 24 h of maturation at 70 • C, the mass reaction was filtered and washed until pH = 7, using distilled water. The obtained product was dried for 24 h at 70 • C, and then crushed and sieved to a dimension lower than 90 µm. Further in the manuscript for this sample, Mg 3 Al symbol will be used.
The functionalization of Mg 3 Al with methyl trialkyl ammonium chloride was made using two methods. The first one assures the distribution of the studied IL onto the surface of the synthesized LDH because is used the ultrasound method [12,19,20]. For this purpose, a well-defined quantity of Mg 3 Al was treated with a well-established quantity of IL (in finality, to have 10% of IL in the adsorbent mass), which previously was dissolved in acetone. The obtained suspension was subject to 10 min of ultrasonication, using an ultrasound bath. After functionalization, the solvent was evaporated at 50 • C for 24 h. Further in the manuscript for this sample, Mg 3 Al IL-US symbol will be used. The second method of functionalization implies the intercalation of the ILs between the LDH interlayers [12]. In this case, the cosynthesis method was used, as was described above for the obtaining of Mg 3 Al, with the exception that, in this case, instead of Na 2 CO 3 solution, the IL dissolved in acetone was used. Further in the manuscript for this sample, Mg 3 Al IL-COS symbol will be used. To prove that Mg 3 Al was synthesized and its functionalization with methyl trialkyl ammonium chloride occurred, the obtained adsorbent materials were characterized through XRD and SEM analysis. A Rigaku Ultima IV X-ray diffractometer (Rigaku Analytical Devices Inc., Wilmington, MA, USA) (40 kV, 40 mA) with Cu Kα radiation was used for the recording of the X-ray diffractograms. SEM images were recorded using a Quanta FEG 250 microscope (FEI Company, Hillsboro, OR, USA), equipped with an EDAX/ZAF quantifier. The FTIR spectra of the samples prepared as KBr pellets were recorded by using a Shimadzu IRPrestige-21 FTIR spectrophotometer (Shimadzu Corporation, Kyoto, Japan) in the range 400-4000 cm −1 and a nominal resolution of 4 cm −1 .

Adsorption Studies
To evaluate the adsorption performance of the obtained materials, the synthesized and characterized samples were used as adsorbent materials in the removal process of DCF from aqueous solutions. Their efficiency in the DCF adsorption process was studied according to a series of physico-chemical parameters: the initial pH of the aqueous solution, the stirring time of the reaction mass, the temperature, and the initial concentration of DCF in the solution, respectively.
In the first step, the influence of pH upon the adsorption efficiency developed by the studied materials was studied. In this sense, the adsorption studies were conducted at initial pH values of the mass reaction, ranged between 4-10. Outside this range, the layered double hydroxides solubilize. The other physico-chemical parameters studied were kept constant (initial concentration of DCF, C 0 = 10 mg/L; stirring time, t = 60 min; temperature T = 25 • C).
Studies regarding the influence of agitation time on the adsorbent capacity developed by the studied materials in the removal process of DCF were performed in the range of 15-180 min, using 3 different initial concentrations of DCF solutions (10, 20, and 30 mg DCF/L) at 3 temperatures (25, 40 and 55 • C), the initial pH of the solutions being around pH = 7.
To determine the maximum adsorption capacity of the studied materials developed in the removal process of DCF from aqueous solutions, studies were performed by varying the initial concentration of DCF solutions (5-400 mg/L), maintaining the other parameters at the optimal values previously determined: pH = 7, t = 60 min, T = 25 • C. In all adsorption studies, the solid-liquid ratio was (adsorbent material:solution containing DCF) S:L = 1:1 (0.025 g adsorbent material in 25 mL solution containing DCF).
The adsorption capacity of the studied materials was determined using the following equation: where: q e -equilibrium capacity developed by adsorbent materials, mg DCF/g adsorbent material. C 0 -initial concentration of DCF in aqueous solutions, mg/L. C e -equilibrium concentration of DCF, mg/L. V-volume of the solution containing DCF used in the adsorption process, L. m-mass of adsorbent material used in the adsorption process, g.

Conclusions
The present paper highlights the benefits brought by ionic liquids in the process of obtaining efficient adsorbent materials via its functionalization on layered double hydroxide solid support. As a solid support, we used Mg 3 Al LDH, which was functionalized with methyl trialkyl ammonium chloride by ultrasound and cosynthesis methods. The RX and SEM analysis performed on the obtained adsorbent materials proved that the functionalization occurred. The adsorption efficiency of the obtained materials was tested in the removal process of diclofenac from aqueous solutions. In all cases, the equilibrium between the adsorbent and adsorbate was achieved in 60 min, and the maximum removal efficiency was obtained at an initial DCF solution pH of around 7. The DCF adsorption onto the studied materials is best described by the pseudo-second-order kinetic model. The samples functionalized with the studied IL showed a better performance than that of raw Mg 3 Al. Correlating the results from the adsorbent characterization section with the results from the equilibrium and thermodynamic studies, the beneficial role of the IL from the adsorbent structure in the adsorption process of DCF from aqueous solutions is highlighted. Its presence contributes to the (1) disturbance of the well-ordered structure of Mg 3 Al, which leads to an increasing in the active sites from the adsorbent surface; (2) bringing more functional groups into the adsorbent structure, and providing a synergic effect in the adsorption process; (3) due to its functional group, assuring some electrostatic interaction with the target pollutant, and determining this random and heterogeneous adsorption of DCF from aqueous solutions.
Supplementary Materials: The following are available online. Figure S1: Linear representation of the pseudo-first-order kinetic model for DCF adsorption onto the studied materials, Figure S2: Linear representation of the pseudo-second-order kinetic model for DCF adsorption onto the studied materials, Figure S3: Representation of the intraparticle diffusion model for DCF adsorption onto the studied materials.