Hydrotalcite Supported on Polycaprolactone:Poly(methyl methacrylate) Fiber Membranes for Chlorogenic Acid Removal

Abstract

Polyphenols are organic molecules extracted from various fruits, such as coffee and citrus, that possess biological activity and antioxidant properties. However, the presence of polyphenols in the environment is hazardous to water quality and living health. Among a variety of water remediation methods, adsorption remains a staple in the field. Therefore, this work aims to develop porous polycaprolactone: poly(methyl methacrylate) (PCL:PMMA) membranes as a support for hydrotalcite immobilization for the removal of chlorogenic acid polyphenol (CGA) from aqueous solutions. Due to the hydrophilic nature of hydrotalcite, the adsorbent was functionalized with hexadecyltrimethylammonium bromide (CTAB) to increase its affinity for CGA, resulting in a removal efficiency of approximately 96%. Composite fiber membranes were prepared by solution-blowing spinning with specific amounts of hydrotalcite added (i.e., 1 to 60 wt%). A 3:1 PCL:PMMA blend resulted in superior mechanical traction (0.8 MPa) and stress deformation (70%) compared to pure PCL (0.7 MPa and 37%) and PMMA (0.1 MPa and 5%) fibers. PCL:PMMA membranes with 60% LDH-CTAB exhibited CGA removal rates equal to 55% in the first cycle while maintaining the capacity to remove 30% of the polyphenol after five consecutive reuses. Removal rates up to 90% could also be achieved with an appropriate adsorbent dose (2 g L⁻¹). Adsorption was found to follow pseudo-second-order kinetics and was adequately described by the Langmuir model, saturating LDH-CTAB active sites in four hours. PCL:PMMA:LDH-CTAB composites can be considered a potential alternative to support adsorbents for water remediation.

1. Introduction

Chlorogenic acid (CGA), also known as trans-5-O-caffeoyl-D-quinate or caffeoylquinic acid, is a nearly ubiquitous polyphenol that can be found in several other plants and fruits, including coffee, grapes, apples, carrots, eggplant, and tobacco leaves [1,2]. Polyphenol, specifically CGA, possesses important biological properties such as antioxidant, anti-inflammatory, and antimicrobial [2]. Hence, this compound may be found in wastewater from agribusiness, becoming a potential pollutant to the environment, interacting with living species, and affecting the potable water quality. It is reported the great affinity of CGA with proteins allows different interactions such as Van der Waals, electrostatic interactions, hydrogen bonds, and hydrophobic. As a consequence, CGA can affect their structure and functionality, solubility, and other properties, turning hard zones into the metabolism of plants and animals [3]. On the other hand, the recovery of this polyphenol supplies an economic value once it can be useful in the proper concentrations as a supplement in the human diet, helping in many health benefits, ranging from neuro- and cardioprotective effects to improved cardiovascular health and skin appearance [4,5].

In this context, the adsorption process allows the removal of contaminants from polluted environments (mainly water) and their subsequent recovery, reconversion, and reuse by the reverse desorption processes [6,7]. Adsorption is widely used in water and wastewater treatment to remove contaminants such as pesticides [8], heavy metals [9], dyes [10], and, last but not least, polyphenols [11]. The most prominent adsorbents, such as activated carbon [12], zeolites [13], silica [14], clays [15], and layered double hydroxides (LDHs), also known as anionic clays [16], are used in water remediation applications. LDHs have been widely used as high-efficiency adsorbents due to their lamellar structure, which provides a remarkable regenerative capacity [17]. LDHs have an organized structure formed by octahedral hydroxide layers (such as mineral brucite Mg(OH)₂) containing divalent cations in their sites and trivalent cations replacing the octahedral sites [18]. The general LDH formula is [M²⁺_1−xM³⁺_x(OH)₂]^x+(Aⁿ⁻)_x/n.mH₂O, where M²⁺ is a divalent metal cation, while M³⁺ is a trivalent cation; Aⁿ⁻ is an intercalated anion of charge n; x is the M³⁺/(M²⁺+M³⁺) ratio, and m is the number of water molecules [19]. Anionic species in the interlamellar space are necessary to balance the positive charge of the “brucite-like” layers. In addition, water molecules can occupy the free space in the interlamellar regions. Aside from structural properties, surface modifications are also conducive to the high affinities of LDH for different types of adsorbates [20].

Nanomaterials are a special class of materials with improved intrinsic properties by the lower diameter dimension, higher surface, and reactivity [21,22,23,24,25]. It is special to combine different materials forming nanocomposites in a variable system, such as magnetic materials [8], ceramic filters [26], and fiber membranes [27], which are some platforms that can support active particles for pollutant remediation and facilitate recovery from aqueous medium. In this regard, organic-inorganic hybrid materials as fiber membranes can take advantage of synergistic effects by immobilizing inorganic nanoparticles in polymer fibers [28]. The great advantage is attributed to the porosity of the fibers and the diameter control, which can facilitate the diffusion of the particles and be favorable for contact with the active sites. Aside from that, a variety of biopolymers can be used to promote affinity for analytes and stability in water by controlling their hydrophilicity and hydrophobic properties. Solution blow spinning (SBS) is a processing technique that allows the production of fine fibers using polymer solutions or dispersions [29,30]. SBS fibers exhibit characteristics such as high specific surface area, porosity, rapid production, and low cost [31,32]. The innumerable combinations of solvents and polymers reported in the literature [33,34,35,36] demonstrate the versatility of SBS for membrane fabrication.

PMMA is a thermoplastic polymer with excellent resistance to UV radiation and weathering, ensuring outstanding durability in external environments. In addition, PMMA has good chemical resistance, making it suitable for use in chemically aggressive environments [37]. On the other hand, polycaprolactone (PCL), a semi-crystalline polyester, is a biodegradable polymer that has been widely used in various applications since biomedical, bag, and remediation applications. Aside from that, it is flexible, with moderate mechanical strength and good fatigue resistance [38,39], which are interesting properties for filter membranes. PCL also shows good compatibility with other polymers, allowing the possibility of formulating specific polymer blends to improve the properties of each component [40,41]. Thus, the PCL:PMMA blend demonstrated a potential use as support material for removal systems. It has good stability in water (hydrophobic properties), mechanical resistance, and posterior degradability.

For these reasons, the development of composite and/or hybrid systems aimed at exploiting the antimicrobial and antioxidant properties of polyphenols [42,43,44,45,46] has been pursued in recent years. The novelty involves the removal of polyphenols from aqueous media using fiber membranes specifically obtained by solution blow spinning. To the best of our knowledge, the present work is the first paper to investigate the adsorption of CGA on LDH particles supported on fiber systems. In this sense, CGA adsorption on hybrid PCL:PMMA:LDH-CTAB membranes was evaluated in the present work, showing novelty in the hybrid adsorption system investigated. Such an approach is a proof of concept that uses unexplored advanced materials for polyphenol remediation in an aqueous medium.

2. Materials and Methods

2.1. Materials

Commercial hydrotalcite (LDH), hexadecyltrimethylammonium bromide (CTAB, ≥98%), and poly (methyl methacrylate) (PMMA, MM: 350,000 g mol⁻¹) (Sigma-Aldrich Corporation, St. Louis, MO, USA). Polycaprolactone (PCL, MM: 50,000 g mol⁻¹) was purchased from Perstop (Malmö, Sweden). Chloroform (99%) and ethanol (99%) were purchased from LabSynth (Diadema (SP), Brazil). CGA was acquired from Toronto Research Chemicals (Toronto, Canada).

2.2. Hydrotalcite Modified with CTAB (LDH-CTAB)

Commercial hydrotalcite (Mg₆Al₂(CO₃)(OH)₁₆.4H₂O) was annealed at 600 °C for 6 h to remove the intercalated carbonate ion and increase LDH wettability, improving the CTAB functionalization. Heat-treated hydrotalcite will be referred to as LDH in the following discussion to avoid burdening the used nomenclature. CTAB treatment was carried out according to the procedure reported in the literature [11]. Briefly, 0.5 g of (heat-treated) LDH was dispersed in 100 mL of deionized water by sonication. Subsequently, 0.25 g of CTAB was added to the LDH dispersion and refluxed for 24 h at 90 °C. The white precipitate obtained was washed using ethanol and distilled water, followed by centrifugation (8000 rpm, 10 min, room temperature). The final product (LDH-CTAB) was dried at 40 °C in a circulation oven overnight.

2.3. PCL:PMMA Fibers

PCL:PMMA fibers were obtained using the solution blow spinning adapted method [47]. A polymeric solution (12% w v⁻¹) was prepared, solubilizing the polymer pellets in chloroform for 1 h under magnetic stirring and keeping them in a sealed flask. Aside from the pure polymers, the polymeric solution had variable PCL:PMMA ratios at 1:3, 1:1, and 3:1 (w w⁻¹). After solubilized, the polymeric solution was inserted into a 20 mL glass syringe (ArtGlass) and connected to a solution blow spinning (SBS) system comprising an ejection pump (NE-300, New Era, Buffalo, NY, USA). A typical SBS consists of a syringe containing the polymeric solution ejected through the nozzle connected with an external high-pressure gas. Generally, a shear of the drop solution exposed to the gas interface reduces the pressure in the inner zone of the cone-shaped solution flow. The polymer solution elongates with solvent evaporation, favoring the formation of elongated polymer chains into fibers deposited in a metal collector [29,31,48]. The fibers were spun at a pressure of 1.5 bar, at a working distance of 16 cm, a collector rotation speed of 150 rpm, and an injection rate of 7 mL h⁻¹.

2.4. PCL:PMMA:LDH-CTAB (PPLDH) Fibers

The composite (i.e., hybrid) membranes were prepared by the same methodology illustrated in Section 2.3, adding the LDH-CTAB (1, 5, 10, 20, 40, and 60% w w⁻¹) to the PCL:PMMA polymer solution (3:1). After solubilize, the fibers were obtained with an applied pressure of 2 bar, 150 rpm rotation, 16 cm working distance, and 7 mL h⁻¹. PCL:PPMA:LDH-CTAB fiber membranes were labeled PPLDH X, with X as the LDH amount.

2.5. Material Characterizations

X-ray diffractograms were acquired in the 3–80° 2θ range on a Shimadzu model LabX XRD-6000, Cu-Kα radiation of λ = 1.5406 Å, operating at 30 kV voltage and 30 mA current. ATR (attenuated total reflectance) IR (infrared) spectra were collected in the 4000 to 400 cm⁻¹ region with a Bruker Vertex 70 spectrophotometer, applying 32 scans and 4 cm⁻¹ resolution. Scanning electron microscopy (SEM) images and Energy Dispersive Spectroscopy (EDS) mapping were acquired with JEOL^® 6701F microscope, with gold and carbon coatings, respectively. Image J 1.53k software was used to extrapolate the average diameter of the fiber. Mechanical tests were carried out using the Dynamic mechanical, thermal analysis (DMTA) technique using a TA Instruments model Q800 equipment, operating in force control mode (clamp tension film) using the stress-strain method at 25 °C with a force ramp of 0.5 N min⁻¹ up to 18 N. Rectangular samples (10 specimens each) with height (12.8 mm) and width (6.3 mm) were used. Thickness measurements were performed on each sample. The contact angle was performed with a Theta Lite optical tensiometer equipped with a USB3 digital camera (2680 photos per second (fps)), a maximum resolution of 1280 × 1024 pixels, capturing images for 1 min at 1 fps.

2.6. Chlorogenic Acid (CGA) Removal Assays

2.6.1. LDH Loose Particles

Pristine and CTAB-functionalized LDH particles were tested for CGA polyphenol adsorption. Initially, 0.1 g of adsorbent was added to a Falcon^® tube containing 10 mL of a chlorogenic acid (CGA) aqueous solution (20 mg L⁻¹). This concentration was selected to maximize CGA UV-vis absorbance peak at λ_max = 323 nm while keeping absorbance within the Lambert-Beer law linear range. The adsorbing materials were stirred in the CGA solution for 24 h at room temperature and 150 rpm. The precipitate was then separated by centrifugation, and a supernatant aliquot was removed for UV-vis measurement. All analyses were carried out in triplicates. UV-vis measurements were performed with a Shimadzu UV-PC 1601 model.

2.6.2. Filter Membranes

The adsorption test with the PCL:PMMA:LDH-CTAB fiber membranes was performed as described d in the previous section, using 0.1 g per membrane. The assays were conducted on the PCL:PMMA fibers (3:1) with different LDH-CTAB concentrations. The fibers were removed from the solution using metal tweezers, and the aliquot was analyzed by UV-vis spectroscopy. Also, UV-vis diffuse spectra were run on LDH fiber membrane adsorbents using a Shimadzu UV-vis equipment, 2600 model. All analyses were performed in triplicates.

The experimental adsorption capacity (q_exp), that is, the amount of CGA adsorbed per gram of adsorbent, was determined by the following Equation (1):

$$q_{exp}\left(mg CGA g^{-1} adsorbent\right)={\displaystyle \frac{({CGA}_{initial}-{CGA}_{final})}{C_{adsorb}}}$$

(1)

where, C_adsorb is the concentration of adsorbent (g L⁻¹), CGA_initial (time zero), and CGA_final (after 24 h) are the polyphenol concentrations in the supernatant. PPLDH 60 dosage effect was tested in the 0.5–2 g L⁻¹ adsorbent concentration range.

The kinetic study of CGA adsorption using the pseudo-first-order (PFO), pseudo-second-order (PSO), and Elovich models was carried out using non-linear equations. More detailed information on each model can be found in the literature [49].

$$\begin{array}{cc}{q}_t=q_e\left[1-{exp}^{\left(-K_1·t\right)}\right]\hfill & \hfill Modelo de PFO\end{array}$$

(2)

$$\begin{array}{cc}{q}_t={\displaystyle \frac{q_e^2·K_2·t}{[K_2·(q_e)·t+1]}}\hfill & \hfill Modelo de PSO\end{array}$$

(3)

$$\begin{array}{cc}{q}_t={\displaystyle \frac{1}{\beta }}\left[\ln \left(1+\alpha \beta t\right)\right]\hfill & \hfill Modelo de Elovich\end{array}$$

(4)

The Langmuir model (i.e., formation of a homogenous monolayer of adsorbate) was used to fit experimental data and infer the adsorption mechanism [13]. Specifically, Langmuir isotherm Equation (5) in its linear form was applied as follows:

$${\displaystyle \frac{C_e}{q_e}}={\displaystyle \frac{1}{q_{max}{K}_L}}+{\displaystyle \frac{C_e}{q_{max}}}$$

(5)

Ce (mg L⁻¹) is the CGA equilibrium concentration; qe is the equilibrium adsorption capacity (mg g⁻¹), and q_max is the maximum adsorption capacity (mg g⁻¹). The K_L Langmuir binding constant (L mg⁻¹) describes the adsorbent-adsorbate interaction. Linear Freundlich empirical Equation (6) was also used as follows:

$$\mathit{log}{q}_e={\displaystyle \frac{1}{n}}log(C_e)+\mathit{log}k$$

(6)

where Ce (mg L⁻¹) and qe (mg g⁻¹) have the usual meaning. k and n stand for Freundlich constant (mg^1−1/n L^1/n g⁻¹) and exponent, respectively. The former is related to the adsorption capacity, whereas the latter indicates the heterogeneity of the adsorbing surface. If n > 1 (1/n less than 1) the adsorption is favorable, while if n < 1 (1/n greater than 1) the adsorption is unfavorable.

Finally, PPLDH 60 reutilization capabilities were evaluated along five consecutive cycles for 24 h, using 0.1 g of adsorbent and a concentration equal to 20 mg L⁻¹ CGA. A fresh CGA solution was used in each cycle.

3. Results and Discussion

3.1. LDH Characterization

The morphological and structural characterization of LDH is shown in Figure 1. Hydrotalcite particles with regular shapes and sizes between 200 and 300 nm can be identified (Figure 1a). The stacking of LDH lamellae was also observed. The structural analysis by XDR (Figure 1b) shows hydrotalcite characteristic peaks at 2θ equal to 11°, 23°, 34°, 39°, 46°, 60° and 62° (ICDD n° 70-2151, space group R3), which belong to the (003), (006), (012), (015), (018), (110) and (113) basal planes, respectively [17,50]. As expected, the peaks are sharp and well-defined due to the basal constant spacing resulting from the intercalation of carbonate ions (CO₃²). Likewise, LDH-CTAB shows a highly crystalline diffraction pattern with some peak broadening, probably due to the CTAB functionalization. Such results are similar to those reported in the literature [51,52].

The LDH interlamellar space was calculated before and after CTAB treatment to allow for any possible CTAB intercalation.

Table 1. Calculated unit cell parameters for LDH.

Sample	d (Å)	c (Å)	a (Å)
LDH	7.6	22.8	1.5
LDH-CTAB	7.5	22.5	1.5

shows the values of the estimated unit cell parameters, considering the R3 group with a rhombohedral configuration. Since the interlamellar space and lattice parameters are very close to those of pristine LDH, CTAB intercalation can be safely excluded. Consequently, the LDH-surfactant interaction mainly took place on the surface of the adsorbent. CO₃²⁻ ions prevented CTAB from penetrating the LDH structure, as they were firmly intercalated into the lamellar structure due to the high charge density and hydrogen bonding that can be established between CO₃²⁻ and OHs in the hydroxide layers.

The effectiveness of LDH modification with CTAB was evaluated in terms of CGA adsorption. For this purpose, 1 g L⁻¹ of LDH-CTAB was exposed to the CGA solution (10 mL of 20 mg L⁻¹ as described in Section 2.6.1) for 24 h. The adsorption results were compared with the original LDH. Figure 2 shows the UV-vis absorbance spectra of the initial CGA solution after contact with LDH and LDH-CTAB. Due to its extremely hydrophilic nature, bare LDH failed to adsorb relevant amounts of CGA. On the other hand, LDH-CTAB was able to remove CGA almost completely from the solution (96.9% removal rate). Such a significant removal rate can be attributed to the ability of CTAB (cationic surfactant) to solvate CGA by several mechanisms. CTAB nitrogen groups can form a chemical bond with the carboxyl groups on the CGA (i.e., amide bond) [3,53]. Concerning intermolecular interactions, electrostatic attractions between CGA ionized from carboxylate groups and the quaternary ammonium head of CTAB are thought to favor polyphenol adsorption [54]. Moreover, the possible hydrophobic interaction between CGA and CTAB long alkyl chains may be significant. Other intermolecular interactions such as van der Waals-like forces, π-π interactions (mediated by the aromatic rings of CGA), ion-dipole, and hydrogen bonding can provide an additional driving force for adsorption [55]. Hence, CGA removal can be considered a mixture of physisorption and chemisorption processes [56,57,58].

3.2. PCL:PPMA Fibers

SEM was used to evaluate fiber morphology in PCL:PMMA polymer membranes. Furthermore, frequency size distribution analysis was exploited to extrapolate the average fiber diameters. Figure 3a shows that the PCL fibers are in the submicrometric range with good homogeneity and a medium average size of (0.234 ± 0.089) µm. Conversely, PMMA fibers under the same spinning conditions were found to be in the micrometric range, with a medium average diameter (1.4 ± 0.5) µm. Not surprisingly, increasing the PMMA ratio in the PCL:PMMA blends resulted in larger fiber diameters: (0.8 ± 0.4) µm (3:1), (1.2 ± 0.6) µm (1:1) to (1.4 ± 0.6) µm (1:3). Therefore, the PCL:PMMA (3:1) blend was found to be the most homogeneous while maintaining a diameter in the submicron range. Such results are similar to those reported in the literature for PCL:PMMA fibers obtained by the electrospinning method [59,60].

FTIR spectrum analysis collected for the three blends showed both PCL and PPMMA features. In Figure 4a, PCL vibrations are visible at 2942 cm⁻¹ (asymmetric -CH₂ stretching), 2861 cm⁻¹ (symmetric -CH₂ stretching), 1729 cm⁻¹ (carbonyl stretching), 1294 cm⁻¹ (C-O and C-C stretching), 1247 cm⁻¹ (asymmetric C-O-C stretching), and 1165 cm⁻¹ (symmetric C-O-C stretching). PMMA characteristic bands were also observed at 2942 cm⁻¹ and 2861 cm⁻¹, but with lower relative intensity than those of PCL, at 1729 cm⁻¹ (carbonyl stretching), 1428 cm⁻¹ (CH₃ stretching), 1063 cm⁻¹ (C-O stretching) [61]. PCL bands at 2942 cm⁻¹, 2861 cm⁻¹, and 1170 cm⁻¹ can still be identified in the blends, being more prominent for the higher PCL percentages. Nevertheless, PMMA characteristic vibrations [62] belonging to the C=C were observed at about 1600 cm⁻¹ for all blends. Regions with significant absorption changes are highlighted in the spectra (orange lines).

Prior to incorporating the LDH-CTAB into the polymer fibers, the best polymer blend in terms of mechanical properties was determined. This intrinsic property is essential to evaluate the maximum capacity of the membrane to recover and maintain an intact underflow. Mechanical resistance evaluation of the PCL:PMMA membranes was based on the tensile strength of the fibers (Figure 4b). The PMMA fibers showed the lowest tensile strength and deformation values, (0.09 ± 0.01) MPa and (5 ± 2) %, respectively. On the other hand, the PCL fibers showed better mechanical response, with average values of (0.6 ± 0.3) MPa and (35 ± 2) %, respectively. PCL:PMMA blends were found to behave similarly to PCL. In fact, the mechanical property values were PCL:PMMA 3:1 [(0.81 ± 0.06) MPa, (71 ± 3) %] and PCL:PMMA 1:3 [(0.81 ± 0.06) MPa and (63 ± 7)%]. Lo and coworkers [63], PCL:PMMA 1:1 blend showed tensile strength (0.7 ± 0.1) MPa and deformation of (5 ± 1)% values between PCL and PMMA. The poor tensile strength and modest elongation of PMMA can be justified by its glassy nature, whereas PCL was shown to be beneficial in improving mechanical resistance. Due to good homogeneity, submicrometric fiber diameters, and good mechanical response, PCL:PMMA 3:1 was chosen as the support for the LDH-CTAB adsorbents.

3.3. PPLDH Adsorbing Membranes

PCL:PMMA (3:1):LDH-CTAB (PPLDH) composite fibers with different LDH concentrations ranging from 1 to 60% (w w⁻¹) were prepared by SBS. Figure 5 shows the SEM images of the obtained composite fibers. The addition of LDH to the PCL:PMMA 3:1 solution promoted an increase in fiber diameter from (0.8 ± 0.4) µm to (1.5 ± 0.8) µm (1%), (1.4 ± 0.6) µm (5%), (1.3 ± 0.6) µm (10%), (1.0 ± 0.5) µm (20%), (1.2 ± 0.5) µm (40%), and (1.3 ± 0.5) µm (60%). Superior fiber heterogeneity was observed at the lower LDH concentrations. It is possible that the increase in mean fiber diameters, commonly associated with the particle intercalation between the polymer chains, increases the solution viscosity and facilitates fiber entanglement [64,65].

SEM-EDS analyses confirm the presence of aluminum (Al) and magnesium (Mg) in LDH-CTAB immobilized on the PCL:PMMA fibers. The LDH-CTAB distribution showed agglomerated particles located in punctual regions of the fibers for the lowest concentrations of (1–10%, Figure 6a–c). On the other hand, better LDH homogenization into the polymer matrix was achieved for 20 wt% (Figure 6d). The composites with the highest LDH-CTAB concentrations (Figure 6d,e), PPLDH 40 (40%), and PPLDH 60 (60%) showed efficient particle coverages over the entire length of the fibers. Large LDH concentrations may also be beneficial for the removal of pollutants from aqueous solutions.

CGA polyphenol adsorption on PPLDH (1 g L⁻¹) can be appreciated in Figure 7a by CGA UV-vis absorbance decrease. Furthermore, diffuse reflectance spectra run on PPLDH after polyphenol adsorption clearly manifest CGA UV-vis absorption peaks, verifying its presence on the composite membrane. CGA removal was also effective for the lowest LDH-CTAB addition (1%). PPLDH 1 to PPLDH 40 showed adsorption removal rates from 3 to 20%, while by contrast, PPLDH 60 measured up to 55% of CGA removed. As expected, pristine PCL:PMMA 3:1 fiber membrane showed no response to CGA. One possible hypothesis is the heterogeneous agglomeration of the LDH particles in the polymeric fibers long, reducing the effective contact area. In addition, the particles may be packed more internally in the membranes once the SBS technique allows their intercalation into the polymeric chains, hampering the diffusion and interaction with the desired target. Thus, the excess of particles inserted can be more partially present on the surface, guaranteeing the adsorption process. The high LDH-CTAB concentration brought plenty of hydrophobic adsorption sites about to interact with the CGA through hydrogen bonding, amide bond, electrostatic interaction, etc. [66,67]. Therefore, the PPLDH surface was found to be highly hydrophobic (Figure 7b). Not surprisingly, as the LDH-CTAB wt% in the composite membrane was raised, the contact angle increased from around 98° to 135°. In Figure 7c, it is possible to verify that the UV-vis spectrum of the adsorbent membrane, after the CGA adsorption, exhibits a large band between 250–450 nm with no previous existence. This result indicates that the CGA removed from the aqueous medium is found adsorbed in the fibers.

Being the most efficient configuration, PPLDH 60 fibers were used to establish the effect of the adsorbent dosage on CGA (20 mg L⁻¹) removal. Figure 8a shows that as the adsorbent concentration is increased from 0.25 to 1 g L⁻¹, the CGA removal rate rose from 15 to 50%. Similarly, both 1.5 and 2 g L⁻¹ doses resulted in a removal rate of 90%, with no significant difference among the two concentrations. This result indicates that LDH incorporation into membranes caused only a negligible removal efficiency loss compared to the powder adsorbent (Section 3.1). The experimental adsorption capacity (q_exp) of the PPLDH 60 system amounted to an average value of (11 ± 2) mg g⁻¹.

The Pseudo-First Order (PFO), Pseudo-Second Order (PSO), and Elovich kinetic models were applied to the adsorption data over time, and results are shown in Figure 8b and

Table 2. CGA adsorption kinetic parameters.

PFO Model		PSO Model		Elovich Model
qe1 (mg g−1)	8.67	qe2 (mg g−1)	9.50	α (mg g−1 min−1)	0.804
k1 (min−1)	0.022	K2 (g mg−1 min−1)	0.003	β (mg g−1)	0.653
R2	0.892	R2	0.967	R2	0.979
X2	0.234	X2	0.014	X2	1.17
Δq (%)	14.6	Δq(%)	3.74	Δq(%)	30.7

. The best fits were confirmed for the PSO models, which apply to chemisorption processes, where the formation of chemical bonds and the transfer of charges between adsorbent and adsorbed can occur in monolayers [68] , and for the Elovich model, which describes that adsorption occurs on a heterogeneous surface that can have a decrease in its adsorption capacity as the adsorbent is saturated and fewer adsorption sites are available [69]. The fit observed for the PFO model leads to the conclusion that the adsorption mechanism by physisorption is irrelevant. These results corroborate the heterogeneous nature of the membrane surface, composed of fibers of non-uniform diameter, especially due to the presence of hydrotalcite. The ester groups present in the polymers that make up the membrane, associated with the numerous hydroxyl groups of the CGA, also contribute to the chemical bonds between adsorbent and adsorbate, highlighting the chemisorption mechanism as preferential [70]. Statistical calculations, especially X2, show that the PSO model found the lowest values for this parameter, reinforcing its influence on adsorption processes.

Contact time experiments were best fitted with a pseudo-second-order model (Figure 8b) (Table 2). Pseudo-second order kinetics is commonly used to describe adsorption phenomena (normally chemisorptive ones) in which the adsorption capacity is connected to the number of available adsorption sites [71,72,73,74]. Furthermore, The k₂ (g mg min⁻¹) values can be regarded as quite high, especially when compared to the adsorption of organic acids on LDH, signifying rapid adsorption on the membrane [75].

The kinetic adsorption behavior of the PPLDH 60 fibers showed a burst effect in the first 4 h (240 min), as shown in Figure 8b, which is in agreement with the results [76,77] of the literature. Such a trend is consistent with the gradual saturation of the adsorption site, as suggested by the Langmuir model. CGA adsorption slightly increased from 4 to 24 h, indicating closeness to the plateau. In other words, most adsorption sites are saturated in a few hours. Nevertheless, PPLDH 60 reveals available sites for adsorption for 24 h until the equilibrium. Possibly, this difference can be related to the two different kinds of interaction taking place between the membrane and the adsorbate, that is, adsorption by electrostatic attraction among COO- groups and CTAB charged heads, and hydrophobic interaction between PCL, PMMA, and CGA aromatic moieties [11,78]. An adsorption schematization of CGA from LDH-CTAB supported in the PPLDH fiber membranes is shown in Figure 9 to illustrate the interaction between the adsorbent and the polyphenol.

Additionally, the Langmuir model (Figure 10a) proved to describe adequately the adsorption isotherm data compared to the Freundlich model (Figure 10b), as shown by the linear correlation coefficients (

Table 3. Langmuir and Freundlich isotherms parameters for CGA adsorption.

Langmuir Model			Freundlich Model
qmax/mg g−1	kL/L mg−1	R2	n	k/mg1−1/n L1/n g−1	R2
9.7	1.3	0.9067	10	15.6	0.0934

), especially concerning the adsorption capacity. Such findings have already been reported for phenolic acid adsorption onto surfaces functionalized with quaternary ammonium groups [79]. Therefore, it is highly plausible that the nitrogen atom from the CTAB head group can electrostatically interact with CGA carboxylate groups. Specifically, the adsorption isotherm shape resembles an “L-type” curve, which is usually indicative of progressive saturation of adsorption sites [80]. Particularly, CGA adsorption on PPLDH 60 “L-2” shape perfectly reflects organic molecule adsorption on a flat surface. Indeed, aside from minor surface roughness associated with the membrane texture, LDH incorporation into polymeric fibers resulted in an approximately flat surface whose geometrical features are dominated by PMMA and PCL network rather than by layered double hydroxide porosity. Since most of the composite adsorbent surface accounts for the fibers, CGA adsorption can be regarded as occurring on equivalent adsorption sites. In addition, L-type curves generally imply a quite strong kind of interaction between the substrate and the adsorbate. Needless to say, adsorption by electrostatic attraction between ions is way more intense than Van der Waals interactions despite being immensely weaker than chemisorption relying on chemical (e.g., covalent) bond formation [81]. For this reason, the K_L value reported in Table 2 is relatively high. Nonetheless, the intrinsically hydrophobic nature of both polymers may favor interaction with CGA.

An investigation concerning the adsorbent activity over prolonged periods was performed on the PPLDH 60 by cycling the material five times (CGA, 20 mg L⁻¹, 24 h). Figure 11 shows around (55 ± 2) % CGA removal rate after the first reuse cycle for adsorbent dosage of 1 g L⁻¹. As for the second cycle, (17 ± 1) % of CGA was adsorbed from a freshly prepared solution. Thus, 30% of the adsorbent activity was retained. The adsorbent activity was preserved in the sequential cycles from the 2nd to the 5th. This result indicates an excellent adsorption capacity of PPLDH 60 at a high initial CGA concentration (20 mg L⁻¹).

4. Conclusions

A layered double hydroxide (LDH) modified with a hexadecyltrimethylammonium bromide (CTAB) was successfully employed as adsorbing material for the removal of chlorogenic acid (CGA) in aqueous solutions. In particular, PCL:PMMA fiber membranes attained by solution-blow spinning possess remarkable mechanical properties and were functionalized with CTAB-modified LDH particles to facilitate adsorbent recovery. A composite comprised of PCL:PMMA 3:1 fibers functionalized with a 60 wt% LDH-CTAB (w w⁻¹) was able to reach a CGA removal rate of 90% for an adsorbent dosage of 2 g L⁻¹ with most polyphenol adsorption occurring in the first four hours of operation. Adsorption was mainly predicated on CTAB-CGA electrostatic interaction following the Langmuir isotherm model and pseudo-second-order kinetics. After removing 55% of the initial CGA (20 mg L⁻¹) in the first cycle, the system stabilized to an 18% removal rate, corresponding to the complete polyphenol elimination from a 3 mg L⁻¹ CGA solution for four consecutive cycles. Overall, CTAB-modified LDHs immobilized on PCL:PMMA membrane may allow for a “low-end” adsorption platform to be applied to polyphenol removal in aqueous solutions, especially in traces. Needless to say, further tests involving, for instance, adsorbent toxicity evaluation are required. In addition, future work will aim to produce LDH-based composite membranes by exploiting industrial waste such as mining tailings and post-consumer food trays. Nonetheless, the present adsorption system can be regarded as a promising environmental remediation platform to be evaluated for the treatment of CGA-contaminated wastewater originating from the fruit and coffee processing industries.