Introduction

Contamination of water resources is one of the greatest current and future environmental concerns. The sixth sustainable development goal (SDG) calls on nations to provide clean and safe water for ecosystem sustenance. Pharmaceutical compounds have been increasingly detected in the aqueous environment in recent decades (Bunting et al. 2021; Adeleye et al. 2022). Carbamazepine (CBZ) is the most prescribed psychotropic medication for the treatment of epilepsy being widely consumed worldwide. Due to its stable heterocyclic structure, CBZ is one of the most persistent drugs in the environment and has a removal rate of only 7%, approximately, in wastewater treatment plants (WWTPs). This low removal rate is attributed to its difficult biodegradation and photodegradation (Koroša et al. 2020; Wang et al. 2021; Vázquez-Tapia et al. 2022). Research carried out in different countries documents the presence of CBZ in various aquatic environments such as wastewater, surface waters, groundwater, and even drinking water (Cardoso-Vera et al. 2021; Almeida et al. 2021; Adeola et al. 2022).

Faced with such a problem, developing technologies aimed at removing this drug from water is paramount. Organoclays have been widely applied for drug sorption in aqueous matrices (Maia et al. 2019; de Andrade et al. 2020; Guégan and Le Forestier 2021; França et al. 2022). Due to its hydrophobic nature and new adsorption sites, the sorption capacity of organic contaminants by organically modified clays has increased substantially. Clay minerals from the smectite and vermiculite groups are the most used ones, as their expandable characteristics allow interaction with various organic compounds (Lagaly et al. 2013). Bentonite clay, for example, consisting predominantly of clay minerals from the smectite group (≥ 50%), mainly montmorillonite (Mt), has been used widely due to its properties and availability.

Organoclays can be used as alternative adsorbents in fixed-bed adsorption columns for wastewater treatment (de Andrade et al. 2020; Spaolonzi et al. 2022; Antonelli et al. 2024). Another potential application is as a reactive material for permeable reactive barriers (PRB) for in situ treatment of contaminated groundwater (Lee et al. 2012; Wang et al. 2020). Cationic surfactants are organic modifiers frequently used in the synthesis of organoclays (De Oliveira et al. 2017, 2020; Ltifi et al. 2018; Shattar et al. 2019; França et al. 2020; Mostafa et al. 2023). However, it is important to assess whether the intercalation of these chemicals in clay minerals could add toxicity to the environment. This is a gap in recent work. Therefore, this research focuses on the development and toxicity assessment of new alternative adsorbents for removing CBZ from contaminated water. Four organoclays were prepared from cationic surfactants of the same chemical nature, but with different alkyl chain lengths. They were extensively characterized before and after the adsorption process. The influence of pH, contact time, adsorbent dosage and initial CBZ concentration were investigated. The desorption of CBZ from the contaminated material and the interaction mechanism were also elucidated. Furthermore, the toxicity of the materials and the medium, after treatment, were evaluated. To the best of our knowledge, this is the first study to perform this.

Experimental

Chemicals and materials

Carbamazepine (CBZ, C15H12N2 O, 236.27 g mol−1) was obtained in powder (99% pure) from a compounding pharmacy. Surfactants hexadecyltrimethiamonium bromide (HDTMA, Chem-Impex Int'l Inc., 99.49%) and octadecyltrimethylammonium bromide (ODTMA, AmBeed., 97%) were used as modifiers in the preparation of organoclays. Bentonite clay sample (BCN1), CEC of 85.71 ± 3.96 cmol (+) kg−1, was provided by the company Bentonisa – Bentonita do Nordeste SA (Paraíba, Brazil).

Preparation of organoclays

Initially, 100 mL of surfactant solutions were prepared in concentrations corresponding to 50% and 100% of bentonite CEC and left under stirring in magnetic stirrers. A 4.0 g sample of bentonite was dispersed in each surfactant solution under stirring, remaining for 24 h at room temperature (25 ºC). The solids obtained (BCN1-HDTMA50, BCN1-HDTMA100, BCN1-ODTMA50, and BCN1-ODTMA100) were filtered and washed intensely with distilled water until the release of residual Br- ions, tested by AgNO3 solution. Afterward, the material was dried in an oven at 60 °C for 24 h and then disaggregated using a mortar and pestle.

Carbamazepine sorption

The batch equilibrium technique evaluated the sorption capacity of carbamazepine by the synthesized materials. The influence of the pH (2.0–12.0), contact time (0.5–360 min), adsorbent dosage (30–60 mg), and initial CBZ concentration (15–100 mg L−1) was evaluated.

The tests were conducted under agitation at 110 rpm in an incubator (Tecnal, model TE-420) at 25 °C, using samples of organoclays dispersed in 20 mL of carbamazepine solution. After each test, solids were recovered by centrifugation, and carbamazepine concentrations were monitored by UV–Vis molecular absorption spectrophotometry (Agilent Technologies Cary 60 UV–Vis spectrophotometer) at 285 nm. The amount of adsorbed drug (q, mg g−1) was calculated by the difference in concentration before and after the tests (Eq. 1).

$$q=\frac{\left({C}_{i}-{C}_{e}\right)V}{m}$$
(1)

, where \({C}_{i}\) and \({C}_{e}\) are the initial and final (equilibrium) concentrations of CBZ (mg L−1), respectively, \(V\) is the total volume of the drug solution (L) and \(m\) is the mass of the organoclay (g).

To investigate the influence of pH, 50 mg of solids were dispersed in 20 mL of CBZ solution (15 mg L−1) and left under stirring for 24 h. At that moment, pHs 2, 4, 7, 10, and 12 were adjusted with NaOH or HCl solutions. Then, tests were carried out to determine the equilibrium time and adsorption kinetics at different contact times (0.5–360 min) and at initial concentrations of 15 and 100 mg L−1 of CBZ without pH adjustment. The following tests evaluated the influence of the adsorbent dosage using 30, 40, 50, and 60 mg of organoclay and were carried out in the equilibrium time and without pH adjustment. Finally, to obtain the sorption isotherms, the initial concentrations of CBZ were varied in the range of 15–100 mg L−1 in optimized pH, solids dosage, and contact time. All experiments were performed in triplicate.

The following kinetic models were adjusted to the experimental data obtained from tests with contact time variation: pseudo-first-order (Lagergren 1898) and pseudo-second-order (Ho and McKay 1999). The experimental data obtained from the tests with a variation of the initial CBZ concentration were adjusted to the following isothermal equilibrium models: linear, Langmuir (1918), and Freundlich (1906) (Eqs. 1–5 in Supplementary Material).

Desorption tests

Desorption tests were conducted with organoclays contaminated with CBZ, obtained from equilibrium adsorption tests. After adsorption, contaminated clays were recovered by centrifugation and dried at room temperature for 72 h. The desorption assays were then carried out in 125 mL Erlenmeyer flasks, where 60 mg of contaminated clay was placed in contact with 20 mL of eluent and left under constant agitation at 110 rpm for 24 h at 25 °C. Four different eluents were evaluated: distilled water, basic NaOH solution (0.5 mol L−1), acid HCl solution (0.5 mol L−1) and methanol (30% vv−1). The percentage of desorbed CBZ was calculated by Eq. (2).

$${\text{Des}} \left(\%\right)=\frac{{C}_{d}{V}_{d}}{{q}_{e}{m}_{d}}*100$$
(2)

, where qe represents the adsorption capacity at equilibrium (mg g −1); Cd represents the concentration of CBZ in aqueous phase after desorption (mg L −1); Vd is the volume of distilled water (L) and md is the mass of the adsorbent (g).

Toxicity against Artemia salina

Artemia salina (microcrustaceans) was used as in vivo assays to evaluate the toxicity of materials and medium before and after adsorption (Rodrigues Sousa et al. 2023; Dias et al. 2023). As proposed by Meyer et al. (1982), initially, Artemia salina cysts were added to a solution with a salinity of 12 ppm and provided under protection under controlled lighting conditions. After 48 h, the nauplii hatched. An aqueous extract of the materials was prepared at a concentration of 10 mg of material per mL of saline solution under stirring for 24 h. Afterwards, the content was centrifuged and the concentrated extract was obtained. They were then diluted in saline solution at concentrations of 0.1, 0.5, 1.0 and 5.0 mg mL−1 to perform the test. Ten live nauplii were collected and placed in test tubes containing diluted aqueous extracts of the organoclays. The systems were maintained under controlled lighting conditions and after 24 and 48 h the survival rate of the microcrustaceans was determined. Saline solution was used as a negative control.

In a similar way, the effects of the treatment process were also evaluated, investigating the mortality of nauplii at a concentration of 100 mg L−1 of CBZ, initial condition before adsorption, as well as at concentrations of 0, 20, 40, 60 and 80 (v v−1) for a total volume of 10 mL, where 100% (v v−1) corresponds to the undiluted sample. Dilutions were carried out in saline solution. Toxicity was expressed as the mean lethal concentration required to kill 50% of a saline population of Artemia (LC50). Figure 1 illustrates the reported experimental procedure. Both tests were performed in triplicate.

Fig. 1
figure 1

Experimental procedure for assessing the toxicity of organoclays and medium containing carbamazepine against Artemia salina

Full size image

Characterizations

The cation exchange capacity (CEC) of bentonite was obtained using the ammonium acetate method (da Rocha et al. 2019). The powder method received X-ray diffraction patterns using an X-ray diffractometer, model D8 Advance from Bruker with Cu-Kα radiation (λ = 1.5406 Å) with a monochromator and Bragg- Bethane configuration. The degree of organof-unctionalization was determined based on CHN elemental analysis obtained by a microelement analyzer Perkin-Elmer PE-2400. Spectroscopic studies were carried out in a Fourier transform infrared spectrometer by Perkin Elmer, model Frontier, equipped with an accessory for inserting the tablet, using a helium–neon gas laser (HeNe) (detector MIR TGS). Samples were prepared in KBr pellets with a concentration of 1% sample/ KBr (m m−1). Spectra were obtained in the scan range of 4000 to 400 cm−1 in 16-scan mode, with a resolution of 4.0 cm−1. Thermogravimetry data were obtained using a Shimadzu DTG-60 thermal analyzer with a heating rate of 10 °C min−1 under 100 mL min−1 of nitrogen (N2) flow in the 30–900 °C range. The morphology of the materials was analyzed using a TESCAN VEGA3 scanning electron microscope (SEM), operating in high and low vacuum, coupled with Oxford Instruments energy dispersion X-ray spectrometry (EDS). The textural properties were analyzed by N2 adsorption–desorption at 77K, using Quantachrome Autosorb-iQ Instruments. Brunauer–Emmett–Teller (BET) and Barrett- Joyner—Halenda (BJH) determined surface areas and pore size distributions.

Results and discussion

Characterizations

X-ray diffraction

X-ray diffractions of natural clay and organoclays are shown in Fig. 2. The XRD pattern of bentonite confirms the typical characteristics of a smectite mineral, showing the reflection of the clay mineral montmorillonite (Mt) as the predominant phase (ICDD: 00-003-0015), in addition to impurities of quartz (Qz, ICDD: 01-078-1254). Characteristic Mt reflections at 2θ = 5.89° suggested a basal spacing (d001) of 1.50 nm for the BCN1 bentonite.

Fig. 2
figure 2

XRD patterns for a pure bentonite and organobentonites, b BCN1-HDTMA50, c BCN1-HDTMA100, d BCN1-ODTMA50, and e BCN1-ODTMA100

Full size image

After the incorporation of surfactants, the characteristic peak of Mt was shifted to values lower than 2θ, reflecting in the increase of basal spacing values. The increase in the basal spacing suggests the intercalation of organic ammonium cations in the interlamellar space of the montmorillonite. The values observed were 1.54, 1.83, 2.49, and 2.93 nm for BCN1-HDTMA50, BCN1-ODTMA50, BCN1-HDTMA100 and BCN1-ODTMA100, respectively. Increasing the amount of organic modifier from 50 to 100% of the CEC altered the spacing significantly. However, it was observed that there was no significant change in spacing for the 50% HDTMA load, whereas for 50% ODTMA, due to the longer chain, the increase was more evident. Thus, the basal spacing of the organoclay increases with the length of the alkyl chain and the concentration of the surfactant.

The intercalated surfactant arrangement also interferes with basal spacing. They can be intercalated in the all-trans (lateral) conformation or in the paraffin conformation (He and Zhu 2017). The arrangements still differ into monolayer, bilayer or pseudo trilayer. The arrangements can be suggested based on the difference between the d(001) values and the thickness of the Mt layer [∼ 1.0 nm, (Bergaya et al. 2011)] and from the knowledge of the height of the head group of surfactants [~ 0.51 nm, (He and Zhu 2017)], which is greater than the alkyl chain height (~ 0.40 nm). Thus, BCN1-HDTMA50 has a lateral monolayer arrangement. As for the clay BCN1-ODTMA50, the intercalated height (0.83 nm) is greater than the head height group (~ 0.51 nm), but not enough to be considered a bilayer. The most reasonable hypothesis is that it is a monolayer, but not lateral, due to the longer chain and the associated steric factors. This occurs when the interlayer space cannot accommodate the surfactant parallel to the surface of the clay mineral, so the alkyl chain orientation can move away from the surface, resulting in increased spacing. For solids modified with 100% CEC, the differences were 1.49 and 1.93 nm for BCN1-HDTMA100 and BCN1-ODTMA100, which correspond to a pseudo-trilayer arrangement. Figure 3 illustrates the possible arrangements arising from the intercalation of surfactants in the interlamellar space of montmorillonite.

Fig. 3
figure 3

Basal spacing and intercalation of surfactants in the interlamellar space of montmorillonite

Full size image

CHN elemental analysis

The amount of surfactant incorporated into the bentonite was determined based on the elemental analysis of C and N (Table S1). As shown by the carbon percentages, the organic matter content increased with increasing amount of surfactant used in the synthesis and, more subtly, with increasing organic chain length of the modifying surfactant. Regarding the extent of the exchange process, almost complete exchange was obtained for samples BCN1-HDTMA50 and BCN1-ODTMA50, which used was 50% CEC, while BCN1-HDTMA100 and BCN1-ODTMA100 were saturated to 72.8 and 71.7% of their CEC values, respectively. The total exchange of inorganic by organic cations about the CEC of clay minerals is not always achieved, and similar results (França et al. 2020; Madejová et al. 2021).

Infrared spectroscopy

Figure 4a shows the spectra obtained for natural and modified bentonite. The range of BCN1 offers typical clay mineral bands. At 3621 cm−1, the band attributed to the asymmetric elongation vibration of the structural hydroxyl M − OH (M = Al3+, Mg2+ or Fe3+) was observed. Also, a broad band at 3380 cm−1 was attributed to the OH elongation modes of adsorbed and intermediate water, as well as silanol (SiOH); a band at 1634 cm−1 associated with water deformation. At the same time, a broad band peak near 980 cm−1 is attributed to the elongation vibration of the Si–O groups (Zhou et al. 2019; França et al. 2020; Cunha et al. 2023).

Fig. 4
figure 4

a FTIR spectra for (a) BCN1 raw bentonite and organoclays, (b) BCN1-HDTMA50, (c) BCN1-HDTMA100, (d) BCN1-ODTMA50, and (e) BCN1-ODTMA100 b TG curves for samples of pure bentonite (BCN1) and organoclays

Full size image

The first evidence of the incorporation of HDTMA+ and ODTMA+ into the clayey material is given by the notorious decrease in the intensity and widening of the peak at 3380 cm−1, region of elongation of adsorbed water molecules, and in the reduction in the intensity of the peak at 1634 cm−1 associated with water deformation. The presence of organic cations causes changes in the surface affinity of montmorillonite, which changes from a hydrophilic character to a hydrophobic one.

The intercalation of surfactants in bentonite is also confirmed by the appearance of bands at 2925 cm−1 and 2851 cm−1 (Fig. 2a, region (i)) referring to asymmetric and symmetric vibrations, respectively, of C–H elongation in –CH2 groups (Brito et al. 2018; Zhou et al. 2019). These bands' frequency and width depend on the interlamellar region's surfactant packing density. At relatively low densities, the frequency changes to a higher wave number concerning the pure surfactant bands (2916 and 2848 cm−1) (He and Zhu 2017), as verified in the organophilic bentonites synthesized in this work. It is reflected in the increase of gauche conformers introduced in the alkyl chain and suggests a disorderly arrangement of intercalated organic cations.

In region (ii), bands were also identified at 1488 cm−1 and 1470 cm−1, attributed to –CH3 methyl scissoring vibrations, and at 735 cm−1 (region (iii)), associated with –CH3 vibrations. The increase in surfactant or alkyl chain load, led to the increase in intensity of –CH3 vibrations in scissor mode increases, and the split band appears (1473 and 1463 cm−1 for pure surfactant) (He and Zhu 2017). The division of peaks in this region was more straightforward and more intense for clays BCN1-ODTMA100, followed by BCN1-HDTMA100, and more subtle for modified bentonites at 50% CEC. Therefore, for all organoclays, the typical peaks of the presence of alkyl surfactants were sharp and without change in wavenumbers.

Thermogravimetric analysis (TG)

TG analyses of samples before and after surfactant incorporation are shown in Fig. 4b. In addition, Table S2 presents the main thermal degradation events of bentonite BCN1 and its organophilic derivatives. Two characteristic mass loss events in bentonite clays were identified. The first event occurred in the temperature range from 31 to 225 °C, with a maximum temperature of 63 °C observed by the DTG and is related to the dehydration of the physiosorbed water on the surface and the water present in the interlamellar region. The second event occurred at the maximum temperature of 467 °C, observed by DTG, and is associated with the dihydroxylation of structural OH groups.

The first event is associated with the loss of water in the materials, and it was observed that, for organoclays, dehydration occurred at lower temperatures and in smaller amounts when compared to that obtained for natural clay. It reflects the hydrophobicity of the materials after organo-functionalization with HDTMA and ODTMA surfactants, which reduce the surface energy of the bentonite and convert the hydrophilic surface of the silicate into a hydrophobic one (Ghemit et al. 2019; Cunha et al. 2023). For organoclays, except for the dihydroxylation event of structural OH units that occurred around 565 ºC for all clays, the other events were attributed to the thermal decomposition of intercalated surfactants.

Scanning electron microscopy (SEM)

The micrographs obtained for the unmodified bentonite samples (Fig. 5a) and for the organoclays (Fig. 5b–e) are shown in Fig. 5. As observed, the surface morphology of BCN1 bentonite appears smoother and less agglomerated when compared to the surface morphology of organoclays. Clays BCN1-HDTMA50 and BCN1-ODTMA50 did not show significant morphological alterations about bentonite BCN1; only greater agglomeration was identified due to intercalation of alkylammonium surfactants. However, the BCN1-HDTMA100 and BCN1-ODTMA100 clays showed substantial morphological changes concerning the raw clay, such as the formation of aggregates attributed to the higher concentration of surfactants used in organophilization.

Fig. 5
figure 5

SEM micrographs of samples of a crude bentonite BCN1; b BCN1-HDTMA50; c BCN1-HDTMA100; d BCN1-ODTMA50 and e BCN1-ODTMA100

Full size image

Regarding the type of surfactant used, observing the images of clays with a concentration of 100% CEC (Fig. 4c, e), an increase in agglomeration and surface roughness in organophilized clay with HDTMA. This result is like to the one observed by Bianchi et al. (2013), who concluded that the longest alkylammonium chain (ODTMA), despite the slight chain difference (addition of two carbons), presented an ordered intercalation and flatter plates.

Surface area

Figure S1 shows the nitrogen adsorption and desorption isotherms for crude and organophilic bentonite samples. The profile of the isotherms resembles a type IV isotherm with a hysteresis loop compatible with the H3 type, according to the classification of the International Union of Pure and Applied Chemistry (IUPAC) (Sing et al. 1985), suggesting the presence of mesoporous adsorbents. The H3 hysteresis loop is observed in the presence of non-rigid aggregates of plate-like particles that originate from slit-shaped pores, characteristics consistent with the structure of lamellar materials such as clays (Acevedo et al. 2021).

The adsorption–desorption isotherms were used to verify the specific surface area (SBET), the volume, and the pore diameter of the clays (both calculated by the BJH method), as shown in Table 1. There was a significant change in the surface area values from raw clay to organophilic clays. The insertion of long-chain surfactants contributes to the reduction of the surface area of the material. The higher the surfactant load used in the modification, the smaller the surface area. The difference was even more pronounced for clays BCN1-ODTMA50 and BCN1-ODTMA100 (SBET = 9.9 and 3.7) due to the slightly longer alkyl chain.

Table 1 Surface area, pore volume and average pore diameter for bentonite and organoclays
Full size table

The average pore volume and diameter were also calculated for each sample (Table 1). The pore volume of BCN1 raw clay, of approximately 0.04 cm3 g−1, decreased around 75% with organophilization, reaching an average of 0.01 cm3 g−1, since there were no significant changes between the organoclays. Finally, the insertion of surfactants in the material's crystalline structure caused an increase in pore diameter, which is more significant for clay BCN1-ODTMA100 (12.74 nm).

Table 2 Kinetic parameters obtained from nonlinear regression of data to pseudo-first order and pseudo-second order models for the sorption of carbamazepine in organophilic bentonites under the experimental conditions of: 50 mg of clay, 25 °C, and natural pH
Full size table

Carbamazepine sorption

Influence of pH

The effect of pH on CBZ sorption is shown in Fig. 6a. The results showed that there was no significant difference in CBZ sorption with the pH change for the four analyzed clays. The low dependence of pH on sorption can be explained by the nature of the drug. Figure 6b shows the CBZ speciation diagram. The diagram shows that this molecule is neutral between pH 3 and 11. The pKa values for this molecule are pKa1 ~ 1.0 and pKa2 = 13.9, which correspond respectively to protonation (RCONH3+) and deprotonation (RCONH−) of the amide group. Thus, CBZ is a neutral nonpolar organic molecule, with a stable heterocyclic structure over a wide pH range, reason why it is also one of the most persistent drugs in the environment.

Fig. 6
figure 6

a Effect of pH on CBZ sorption by BCN1-HDTMA50, BCN1-HDTMA100, BCN1-ODTMA 50 and BCN1-ODTMA 100 at 25 °C at Ci = 15 mg L−1. b Speciation of carbamazepine

Full size image

The pH of the solution also affects the surface chemistry, causing a change in the surface charge of the adsorbents. However, behaviors similar to those obtained in this study were observed for the sorption of CBZ in different materials, such as magnetic carbon nanotubes (Deng et al. 2019), phosphorus-doped carbonaceous material (Turk Sekulic et al. 2019), modified diatomaceous earth with iron oxide (Jemutai-Kimosop et al. 2020)activated carbon fiber (Zhao et al. 2020), zirconia/porous carbon nanocomposites (Chen et al. 2020) and organically modified montmorillonite (Kryuchkova et al. 2021). Thus, especially in hydrophobic materials, the nature of CBZ stands out in relation to the change in the surface charge of the materials. The synthesized organoclays, therefore, manage to maintain a good CBZ adsorption capacity in a wide range of pH, allowing its application in different media without the need to modify this variable.

Kinetic studies

Adsorption kinetics was evaluated by contacting 50 mg of organoclays with 20 mL of CBZ solution at 15 mg L−1 and 100 mg L−1 (Fig. S2), lower and upper limits of the concentration range used for obtaining equilibrium isotherms. Even in the first minutes of interaction (30 min), all adsorbents reached their maximum CBZ sorption efficiency. After 120 min, all systems under study are guaranteed to be at equilibrium at the two initial drug concentrations evaluated. Brito et al. (2018) obtained a similar result for anionic dye sorption by organophilized bentonites with HDTMA and ODTMA at 100% CEC.

Sorption kinetics were analyzed by nonlinear regression of experimental data to pseudo-first-order and pseudo-second-order models, following the parameters shown in Table 2. The coefficient of determination (R2) value and the slightest difference between the capacities of experimental sorption and models (qe,exp, and qe,mod) show that the experimental data, in all systems evaluated, were well adjusted to the pseudo-second-order model. The best fit to this model suggests that the rate-determining step is chemisorption. The sorption rate is regulated by the concentration of CBZ in the solution and the availability of active sites on the adsorbent. However, a conclusion about the mechanism cannot be made based on kinetics alone. A detailed mechanistic analysis is necessary to determine the specific sorption mechanism.

Kinetic experiments with natural bentonite clay were carried out under the same conditions and revealed the low sorption potential of CBZ by the clay, reaching an average sorption capacity of 1.92 ± 0.20 mg g−1, that is, around 5% removal after 120 min of contact (Fig. S3). This result reveals the versatility and importance of organophilization, which results in improved sorption of organic molecules.

Influence of adsorbent mass and type of surfactant

The adsorbent dosage is vital to establishing the best drug sorption efficiency with the synthesized materials. As shown in Fig. S4, lower adsorbent dosages resulted in higher values of adsorption capacity (q), reaching values of 34.34 ± 1.41 mg g −1, 32.16 ± 0.33 mg g −1, 29.45 ± 0.53 mg g −1 and 27.58 ± 0.96 mg g −1 for clays BCN1-HDTMA100, BCN1-ODTMA50, BCN1-HDTMA50 and BCN1-ODTMA100, respectively, at 30 mg of adsorbent and initial CBZ concentration of 100 mg L−1. Comparison of the efficiency of the synthesized organoclays with pure bentonite (BCN1), which presented an adsorption capacity of 3.10 ± 0.71 mg g−1 on average, shows that organophilization increased the adsorption capacity by approximately 10 times.

Despite the higher carbon content in the BCN1-ODTMA100 clay, it did not present the best adsorption performance. The low available surface area (3.7 m2 g−1), the smallest among all organoclays, and the arrangement of these molecules in the interlayer space, possibly offered a significant steric hindrance to CBZ sorption. On the other hand, the clays BCN1-HDTMA100 and BCN1-ODTMA50 showed very similar performance. Thus, it was observed that for the surfactant with a longer chain, and lower charge (50% CEC), a material was obtained with behavior very similar to the surfactant with a shorter chain, but with a higher charge (100% CEC).

However, the highest percentages of CBZ removal were obtained for the dosage of 60 mg of clay. At this dosage, the equilibrium sorption capacity was 22.91 ± 0.36 mg g−1, 22.58 ± 0.68 mg g−1, 21.61 ± 0.50 mg g−1 and 19.96 ± 0.11 mg g−1 for the clays BCN1-HDTMA100, BCN1-ODTMA50, BCN1-ODTMA100 and BCN1-HDTMA50, respectively. The CBZ removal percentages were maximum and equal to 69.7, 68.9, 65.1 and 61.4%, respectively, under the evaluated conditions.

Equilibrium isotherms

The isotherms were evaluated in the initial concentration range of 15–100 mg L−1 of CBZ, 60 mg of clay and 120 min of contact. According to the statistical analysis performed by Awad et al. (2019), the linear, Langmuir, and Freundlich models are the most used to describe the adsorption processes of organic pollutants in wastewater considering different clays. Thus, the models mentioned above were adjusted to the equilibrium data (Fig. 7), and the parameters are shown in Table 3. As observed, the isotherms obtained do not indicate a saturation trend, as more significant amounts of qe were observed with increasing Ce, indicating that the sorption is probably not restricted to a single layer. Considering the correlation coefficients (R2), the linear model was the one that best fitted all systems investigated. However, in general, the Freundlich model also presented a good correlation, and the n parameter values were close to one, further validating the linear fit. On the other hand, the Langmuir parameter qmax presents a very significant error and does not represent the systems studied.

Fig. 7
figure 7

Isotherms for carbamazepine adsorption by a BCN1-HDTMA50; b BCN1-HDTMA100; c BCN1-ODTMA50 and d BCN1-ODTMA100, under the experimental conditions of: 60 mg of clay, 25 °C, 120 min of contact, and natural pH

Full size image
Table 3 Parameters of isothermal CBZ sorption models obtained from non-linear regression of data to linear, Langmuir and Freundlich models
Full size table

The linear relationship observed between the CBZ concentration, and the sorption capacity is also characteristic of the hydrophobic partition process combined with the low drug concentrations evaluated in this study. Hydrophobic partitioning refers to the distribution of a substance between two immiscible phases, in this case, with the drug being sorbed onto the interspersed surfactants of the organoclay. The linear model suggests a consistent and proportional distribution of the drug between phases.

Desorption of carbamazepine

Adsorption is known to be associated with the generation of adsorbate-contaminated waste loads. Thus, it is essential to assess whether the adsorbent material has good desorption and regeneration efficiency to obtain a better process cost-effectiveness. In this study, distilled water, a basic solution (NaOH 0.5 mol L−1), an acid solution (HCl 0.5 mol L−1), and a solution with a polar organic compound (methanol 30% vv−1) were evaluated. As eluents for CBZ desorption and their efficiencies are shown in Fig. 8.

Fig. 8
figure 8

Efficiency of CBZ desorption of organoclays synthesized by different eluents: ultrapure water, sodium hydroxide, hydrochloric acid and methanol

Full size image

Methanol was the eluent with the highest desorption capacity among all synthesized clays, showing an average of 91.2, 83.1, 78.4 and 70.9% of desorption for the hybrids BCN1-HD50-CBZ, BCN1-OD100-CBZ, BCN1-OD50-CBZ, and BCN1-HD100-CBZ. Stronger interactions between BCN1-HDTMA100 clay and CBZ may be responsible for the lower percentage of desorption via methanol (70.9%) compared to the other clays. Basic and acidic eluents could have shown better desorption efficiency, with average percentages lower than 40%. These results corroborate those found in the adsorption tests. They are also because CBZ has more excellent solubility in organic solvents, such as methanol than in water, in addition to having high stability against electrostatic interactions, as discussed. Aghababaei et al. (2023) also observed this behavior in the desorption of CBZ from biochars.

Toxicity

In addition to low cost, good adsorption, and regeneration efficiency, some applications require the adsorbent material to be environmentally friendly. The result of the acute toxicity test of the materials, performed with Artemia salina, is shown in Fig. 9. For clays BCN1-HDTMA50, BCN1-ODTMA 50, and BCN1-ODTMA100, it was possible to observe that, in any extract concentration, the survival rate was greater than 50% (LD50), even after 48h. Thus, it is possible to state that these clays do not present toxicity, similar to the control test. However, for BCN1-HDTMA100 clay, a decrease in the survival rate of nauplii was observed when exposed to higher concentrations of the extract, reaching a survival rate equal to 0% in 48h. Therefore, it is not recommended to use the material in the environment in this proportion of material/volume of water to be treated.

Fig. 9
figure 9

Toxicity test against Artemia salina, during a period of 24 and 48 h, using materials a BCN1-HDTMA50, b BCN1-HDTMA100, c BCN1-ODTMA50 and d BCN1-ODTMA100

Full size image

Figure 10 presents the result for the bioassays performed in CBZ solutions. As noted, in the initial condition of 100 mg mL−1 CBZ (before treatment), all Artemia died within 24h. It can be attributed to the presence of the substance in the medium and the absence of saline conditions, validated by the result of the control in distilled water (Fig. 10, W Control). At the 80% vv−1 dilution, considerable toxicity was also detected, reaching LD50 after 48 h of observation. From the dilution of 60% vv−1, the medium showed no toxicity to the organisms tested. In this study, the residual concentration of CBZ was a maximum of 35 mg L−1 (in conditions of 60 mg of adsorbent) for the clay with the lowest adsorption performance. Thus, after treatment, the medium has no toxicity, and the survival rate is 100%.

Fig. 10
figure 10

Toxicity test against Artemia salina, during a period of 24 and 48 h, of the medium in the presence of carbamazepine at different concentrations (% v v−1). S control and W control are, respectively, the saline and distilled water control tests

Full size image

Characterization of organoclay-drug hybrids

FTIR analysis before and after adsorption

The infrared spectra of the organoclays before and after CBZ adsorption were analyzed, and the results are shown in Fig. 11. The FTIR spectrum for the drug is shown in Fig. S5. CBZ has –NH2, –C=O, and aromatic rings in its structure. Characteristic IR bands of CBZ were observed in the contaminated organoclays (Fig. 11), with the most relevant changes in the spectrum of the clays that obtained the best adsorption performance: BCN1-HDTMA100 and BCN1-ODTMA50. There was a broadening of the band around 1640 cm−1 and the formation of a shoulder around 1598 cm−1 due to the stretching C=O that occurs in this region and the aromatic vibration –C=C– that appears in this region; the appearance of a band at 1430 cm−1 referring to deformation –NH was also observed (Deng et al. 2019; Aghababaei et al. 2023). In the spectrum of the hybrid BCN1-OD50-CBZ, the contribution referring to deformation –NH was even more evident in the expressive increase in the band's intensity at 1470 cm−1 (Fig. 11c). Other IR bands characteristic of CBZ were not observed due to the adsorbed drug's low content with the FTIR instrument's detection limit.

Fig. 11
figure 11

FTIR analyses after adsorption of carbamazepine on organoclays a BCN1-HDTMA50, b BCN1-HDTMA100, c BCN1-ODTMA50, and d BCN1-ODTMA100

Full size image
TG analysis before and after adsorption

Analyzing the mass loss curve (TGA) for clays contaminated with CBZ in comparison with non-contaminated organoclays (Fig. 12), one can notice that all of them present a similar pattern to each other, except for the post-adsorption curve of CBZ for the BCN1-ODTMA50 clay. The region of most significant mass loss in all clays was from 250 to 450 °C, where surfactants and adsorbed CBZ thermal degradation occurs. Table S3 summarizes the mass loss data obtained. The slope of the mass loss curve for the BCN1-HD 50-CBZ hybrid has a subtle shape and is very similar to the uncontaminated clay, in agreement with the adsorption results, since it had the lowest performance. For the clays with the highest adsorption performance, BCN1-HDTMA100 and BCN1-ODTMA50, introducing the drug also resulted in a more significant weight loss of the respective hybrids. However, the behavior of the hybrid BCN1-OD50 -CBZ stands out.

Fig. 12
figure 12

TG analysis after carbamazepine adsorption on organoclays a BCN1-HDTMA50, b BCN1-HDTMA100, c BCN1-ODTMA50, and d BCN1-ODTMA100

Full size image

The presence of CBZ in the BCN1-ODTMA50 clay conferred a significant decrease in the material's thermal resistance, which previously presented a total mass loss of around 19% and, after drug adsorption, approximately 60%. This behavior demonstrates weaker interactions between the active sites of the material and the CBZ molecules despite the good adsorption performance obtained. Meanwhile, the BCN1-HDTMA100 clay showed an increase of only 5% in total mass loss after drug introduction, demonstrating greater intensity in the interactions between adsorbate and adsorbent and more excellent thermal stability.

XRD analysis before and after adsorption

Figure 13 shows the XRD result of organoclays before and after CBZ adsorption. CBZ did not significantly change the basal spacing of the hybrids BCN1-HD50-CBZ, BCN1-HD100-CBZ, and BCN1-OD100-CBZ. França et al. (2020) also observed this behavior after the adsorption of diclofenac on organoclays. For the hybrid BCN1-OD 50-CBZ (d001 = 1.83nm). However, the packing density of surfactants in clay BCN1-ODTMA50 was lower than that found in clay BCN1-ODTMA100 (d001 = 2.93 nm), modified with the same type of surfactant. Consequently, drug entry is easier in BCN1-ODTMA50, which has access to more internal adsorption sites and can intercalate in different positions. Thus, the disappearance of reflection 001 indicates that such ease of drug incorporation led to the exfoliation of BCN1-ODTMA50, corroborating the adsorption performance of this clay, which was greater than that of BCN1-ODTMA100.

Fig. 13
figure 13

XRD patterns for samples of bentonite and organoclays a before and b after CBZ adsorption

Full size image

Carbamazepine—organoclay interaction mechanism

Previous adsorption studies proved that the hydrophobic interaction played an essential role in the adsorption of carbamazepine and other drugs (França et al. 2020, 2022; Kryuchkova et al. 2021; Yu et al. 2022; Aghababaei et al. 2023). Thus, among the possible adsorption mechanisms and because of the results discussed, it is suggested that the hydrophobic interaction is the dominant one between CBZ and the organoclays obtained in this study.

In the CBZ molecule, there are hydrophobic aromatic rings. The octanol–water distribution coefficient (Kow) of CBZ is 2.45, indicating an intermediate level of hydrophobicity compared to other pharmaceutical products. Thus, the hydrophobicity of the drug is the driving force that leads to greater sorption in the organic phase of the surfactants intercalated in the organoclays.

The CBZ also has –NH2 and –C=O groups in its structure. Thus, groups containing N and O can also form hydrogen bonds with the electronegative atoms on the bentonite clay's surface. However, given the low adsorption performance of non-organophilic bentonite clay (BCN1), this is believed to be a minimal contribution. Thus, the mineral fractions played a secondary role in the retention of CBZ. Figure 14 represents a general scheme of CBZ adsorption on organoclays based on the results obtained and the CBZ molecule dimensions.

Fig. 14
figure 14

Proposed interactions involved in the CBZ adsorption process on organoclays

Full size image

Conclusion

A complete investigation of the adsorption of carbamazepine on organoclays was carried out. The organophilization of bentonite clay with the surfactants HDTMA and ODTMA resulted in four organoclays with distinct characteristics, which differed by the amount of surfactant and the size of the organic chain. Characterization analyzes confirmed the intercalation of organic cations in the interlayer region of montmorillonite. The increase in the hydrophobicity of the bentonite clay allowed a better performance in the adsorption of CBZ. The kinetic and isothermal models of pseudo-second order and linear, respectively, best represented the experimental sorption data at 25 ºC. The CBZ desorption from contaminated organoclays was improved using a methanol solution. However, this area of research needs to be further investigated. The organic phase intercalated in the clays works as a means of partitioning the drug. Therefore, the hydrophobic interaction was the dominant mechanism for the adsorption of CBZ. Organoclays had no toxicity against Artemia salina, except for BCN1-HDTMA100 in a proportion of 5 mg mL−1. The adsorptive treatment reduced the toxicity of the medium.

From the point of view of environmental significance, low cost, and adsorption performance, organoclays show promising characteristics for application as drug adsorbents in the treatment of contaminated water.