Table Olive Wastewater Treatment Using the Clay Mineral Palygorskite as Adsorbent

Abstract

This study investigated the effectiveness of palygorskite (Pal) as an adsorbent for removing total phenolic content (TPC), dissolved chemical oxygen demand (d-COD), and color from treated olive wastewater (TOW). Experiments were conducted to evaluate the impact of varying Pal dosages (2.5–20 g/L), initial TPC concentrations (80–400 mg/L), and pH (2–9). The results showed that increasing the Pal dosage improved the removal efficiency of TPC and d-COD, though there were diminishing returns beyond 10 g L⁻¹, which indicates equilibrium adsorption behavior. The maximum TPC and d-COD removal reached 68% and 55%, respectively, while color removal exceeded 95% regardless of dosage. Adsorption was most efficient at lower TPC concentrations and an acidic pH (2–3), with up to 85% TPC removal. This suggests that pH-dependent phenolic ionization enhances Pal adsorption. Color removal remained consistently high across all conditions, highlighting palygorskite’s mesoporosity and affinity for chromophoric compounds. These findings affirm the potential of Pal as a cost-effective and versatile adsorbent for TOW treatment.

1. Introduction

The edible olive and olive oil production industry is a significant contributor to the Mediterranean and EU economy, accounting for 31% of the world’s total olive oil production, with Spain and Greece being the primary producers [1]. Specifically, Greece produces around 200,000 tons of table olives, from which two-thirds are exported worldwide [2]. The processing of olives for consumption includes several steps, such as debittering, fermentation, and rinsing, which demand high amounts of fresh water and chemicals such as lactic acid, NaOH, and NaCl [3]. All these stages generate substantial quantities of waste products and effluents, which are abundant in organic compounds like polysaccharides, tannins, and phenolic compounds, and are also characterized by low salinity and high concentrations of suspended solids [4]. The effluent consistency results in relatively high chemical oxygen demand (COD) ranging from 3000 mg L⁻¹ to 20,000 mg L⁻¹, acidic pH, and low biodegradability due to the antimicrobial activity of phenolic compounds, making table olive wastewater (TOW) a serious environmental threat [4,5]. In particular, phenols and phenolic derivatives, as water pollutants, pose crucial health and environmental issues since they can cause unpleasant odors and a brownish color in water and have a high carcinogenic potential [6].

In many producing regions, TOW is usually retained in evaporation ponds, causing bad odors and, in some cases, surface and groundwater pollution [7,8]. In response, researchers are exploring low-cost, robust, and sustainable technologies for TOW treatment. Among these, adsorption has emerged as a particularly promising solution. This technique, based on the accumulation of contaminants onto the surface of a solid material, offers simplicity, low energy requirements, and effectiveness for both organic and inorganic pollutants [9]. Moreover, the adsorption method facilitates regeneration, cleansing, and subsequent reutilization of the adsorbent on multiple occasions [10].

In recent years, the use of adsorbent materials such as zeolites, clay minerals, and other porous media has gained considerable attention in the treatment of phenol-rich effluents, including those from table olive processing [11]. Several studies have evaluated the removal capacity of zeolites or zeolite composites [6,12,13,14], activated carbon [9,15,16], and clay minerals such as montmorillonite [17] in the removal of phenolic compounds and polyphenols from olive processing wastewater. Among these, clay minerals have emerged as particularly promising adsorbents over the past decade. This is primarily due to their natural abundance, low cost, non-toxicity, and notably high specific surface area, as well as their capacity to remove large organic molecules, including ionic dyes and polyphenols [11,18]. This removal is facilitated by the formation of organo-mineral complexes, either on the external surfaces or within the mesoporous structure of the clays [19,20]. In contrast, although zeolites also offer a high surface area and favorable adsorption characteristics at a low cost, their microporous structure presents significant limitations in the adsorption of larger molecules due to restricted mass transfer [21].

Palygorskite, also known as attapulgite, is a fibrous clay mineral characterized by a ribbon-like structure formed by alternating tetrahedral and octahedral sheets, which results in a channel-like basal space of approximately 6.4 × 3.7 Å [22]. Its microporosity is related to its crystalline structure and fiber length, while, according to the usual mean pore diameter, it is characterized by mesoporosity [23]. Despite the micro-dimensions, the inner space of palygorskite contains coordinated water in the edges of channels, which can interact with hydrogen bonds between CO and OH_2, as Zhuang et al. [24] reported for indigo dyes, or develop electrostatic interactions with other molecules [22]. Also, the permanent active sites on the surface of palygorskite mostly interact with organic compounds and polyphenols. Depending on the molecular size of the compound and external conditions such as pH, penetration of the mineral’s inner space mainly occurs through electrostatic forces, particularly in the case of anionic dyes [20]. Palygorskite’s unique ribbon-like crystal structure, combined with mesoporosity, surface chemistry, and a selective affinity for large organic molecules, is advantageous.

Recent investigations have highlighted palygorskite as a potential adsorbent for organic pollutants, particularly phenols and ionic dyes. For example, acid-treated palygorskite demonstrated high affinity toward cationic dyes, with adsorption capacities exceeding 180 mg g⁻¹ for crystal violet and 81.9 mg g⁻¹ for methylene blue, due to electrostatic interactions and surface charge properties [18]. Moreover, when sodium–humate–modified palygorskite was applied for methylene blue uptake, 227 mg g⁻¹, with ~99.7% removal efficiency at an initial dye concentration of 150 mg L⁻¹, was attained [25]. Instead, the non-chemically modified palygorskite achieved 49.27 mg g⁻¹ for methylene blue and 44.82 mg g⁻¹ for Congo red, corresponding to 98.54% and 89.63% removal efficiencies, respectively, under optimized conditions [26]. These findings confirm that both raw and modified Pal exhibit promising adsorption performance for phenols and ionic dyes, governed by mechanisms such as electrostatic attraction, hydrogen bonding, and ion exchange. However, its performance as an adsorbent for TOW treatment or the removal of phenolic compounds has not been well evaluated. While the effect of mineral chemical modification has been highlighted in most studies [27,28,29], its performance in real TOW in its raw form as a sustainable and low-cost material has not yet been evaluated.

The aim of this study was to evaluate the efficacy of palygorskite (Pal) clay in the removal of total phenolic compounds (TPC), brownish color, and general organic matter (as d-COD) from raw TOW through batch adsorption experiments. The innovation of this study stems from the first-time evaluation of raw palygorskite (Pal) clay as an adsorbent for complex TOW, targeting the simultaneous removal of total TPC, brownish color, and dissolved organic load. While previous studies have explored the adsorption of phenols and ionic dyes onto Pal using single-compound or synthetic solutions, this work uniquely investigates Pal’s performance within a real agro-industrial effluent, characterized by high COD, diverse polyphenols, and challenging organics. By employing untreated palygorskite, the study emphasizes a low-cost, sustainable alternative to conventional adsorbents such as activated carbon or zeolites, whose efficiency in such complex matrices is often limited. This approach not only assesses Pal’s natural affinity for phenols and dyes but also provides practical insights for its scalable application in the treatment of high-strength, phenol-rich agro-industrial wastewaters, where adsorption capacity and decolorization efficiency are critical.

2. Materials and Methods

2.1. Sample Preparation

The Pal sample was supplied from Geohellas S.A. (Ventzia basin, Grevena, Greece) as a commercial product. The raw sample was washed with distilled water and dried at 40 °C for 48 h, then sieved until powder diffraction was obtained (<50 μm). No other chemical purification or modification was applied to the clay sample in order to maintain the simplicity of the procedure for the prospect of larger-scale application.

2.2. TOW Characteristics

Real TOW was selected from a local table olive processing industry in Agrinio, Greece. The wastewaters were derived from the washing of equipment following table olive processing during the production of packed table olives. The TOW selected was fresh and contained a large amount of total suspended solids (TSS) at the range of 350–400 mg L⁻¹, dissolved COD (d-COD) 2600–3000 mg L⁻¹, total phenolic compounds (TPCs) 300–400, moderate brownish color, salinity ranging from 15 to 21 mg L⁻¹ and pH value of 5.5 ± 0.3. Before batch experiments, the TOW was filtered using filtration paper sheet 0.45 μm for TSS removal and kept at 4 °C for a maximum of four days.

2.3. Pal Sample Characterization

The washed and dried Pal sample was mineralogically characterized via X-Ray diffraction (XRD) and N₂ physisorption, while its fibrous morphology was evaluated with scanning electron microscopy (SEM). Specifically, the XRD pattern was obtained for the sample in a 2θ range of 2° to 60° and at a scanning rate of 2°/min, using an XRD Bruker D8 Advance diffractometer (Bruker Corporation, Billerica, MA, USA), with Ni-filtered CuKα radiation (λ = 1.5418 Å). The N₂ physisorption was performed in a Quantachrome iQ series apparatus. For the experiments, 0.15 g of the Pal sample was degassed at 180 °C under vacuum for 2.5 h and then re-weighed before the measurements of N₂ adsorption at 77 K. The Brunauner–Emmet–Teller (BET) equation was applied in the range of 0.05–0.3 P/P₀ to calculate the specific surface area (SSA). The total pore volume and pore distribution were determined as well using the DFT method. Finally, Pal’s typical morphology and elemental analysis were verified with scanning electron microscopy with Energy Dispersive X-ray Spectroscopy (SEM-EDS, Thermo Fisher Scientific, Waltham, MA, USA), using a Zeiss SUPRA 35 VP (Carl Zeiss AG, Oberkochen, Germany) system operating at 5 kV voltage. Fourier transform infrared spectroscopy (FT-IR, Shimadzu IR Tracer–100 spectrometer, Shimadzu, Kyoto, Japan) was performed in the range of 500–4000 cm⁻¹ to verify Pal’s typical bands, and detect newly formed ones after TOW adsorption.

2.4. Batch Kinetic Experiments

A series of batch kinetic experiments was conducted using Pal sample under various adsorbent dosages (2.5, 5, 10, and 20 g L⁻¹) in 200 mL filtered TOW solution using 600 mL beakers. To evaluate the effect of initial TPC concentration, the raw TOW was diluted with distilled water till the final TPC concentrations of (80, 150, 200, 300, and 400 mg L⁻¹) were obtained. The impact of pH on d-COD, TPC, and color removal was evaluated for pH values (2, 3, 4, 5, 6, 7, 8, and 9). Solution pH was adjusted using either H₂SO₄ or NaOH for acidic and basic values, respectively. The adsorption process for all the experiments was carried out using the jar tester VELP Scientifica JLT6 (VELP Scientifica, Usmate Velate, Italy) at 200 rpm. Samples were collected at different time intervals (1, 2.5, 5, 10, 15, 20, 30, and 40 min) and were centrifuged at 5500 rpm for 3 min. The supernatant was filtered through Whatman filters (0.45 μm) to remove the finest suspended particles. All the experiments were conducted in duplicate, and the final removal is expressed as the mean value of the % removal efficiency from the duplicate experiments with standard deviation, calculated from Equation (1):

$$R\%={\displaystyle \frac{C_0-C_e}{C_0}}\times 100$$

(1)

where C₀ is the initial pollutant concentration (mg L⁻¹), and C_e is the concentration after adsorption in equilibrium (mg L⁻¹). The same equation was applied for decolorization as well, where initial and equilibrium color absorbance was used instead of concentration. The adsorption capacity of Pal for TPC was calculated as q_e (mg g⁻¹) based on Equation (2).

$$q_e={\displaystyle \frac{{(C}_0-C_e)V}{m}}$$

(2)

where C₀ and C_e are the initial and equilibrium TPC concentrations in TOW solution (mg L⁻¹), respectively; V is the solution volume (L); and m is the adsorbent weight (g).

2.5. Analytical Methods

Determination of TPC as syringic acid equivalents was performed using Folin–Ciocalteau’s phenol reagent according to Singleton et al. [30]. The final concentrations were determined spectrophotometrically at 760 nm using a Hach Lange DR 5000 UV-Vis spectrophotometer (Hach Company, Loveland, CO, USA). The d-COD determination was followed using the closed reflux dichromate procedure, as described in Standard Methods [31], using a Wastewater Treatment Photometer (HANNA HI 83,214, HANNA Instruments, Limena, Italy). The color of the TOW was determined by measuring the absorbance of the sample at the 200–800 nm wavelength range using a UV-Vis spectrophotometer (Hach Lange DR 5000, Hach Company, Loveland, CO, USA). Since 400 nm was found to be the optimum wavelength for decolorization detection, measurements on that wavelength were selected. The pH values were determined with a HANNA HI 5521 multiparameter instrument (HANNA Instruments, Limena, Italy), using a pH/ORP sensor (HI7698194–1, HANNA Instruments, Limena, Italy).

2.6. Adsorption Isotherm Models

The Langmuir and Freundlich adsorption isotherms were used to determine the adsorbent surface sites–adsorbate ions relationship, specifically, Pal interactions with TPC that TOW contained. The Langmuir model [32] indicates monolayer adsorption according to Equation (3) as:

$$q_e={\displaystyle \frac{q_m{K}_L{C}_e}{1+K_L{C}_e}}$$

(3)

Its linearized form is illustrated in Equation (4).

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

(4)

where q_e is the number of exchanged ions (mg g⁻¹), C_e is the equilibrium concentration of TPC in solution (mg L⁻¹), K_L is the Langmuir adsorption constant (L mg⁻¹) related to the affinity between the adsorbate and the bindings sites of the adsorbent, and q_max (mg g⁻¹) the maximum adsorption capacity of Pal for TPC removal. From the C_e/q_e ratio vs. C_e plot, the values of q_max and K_L were calculated from the slope and intercept of the plot, respectively.

The Freundlich isotherm expresses heterogeneous adsorption surfaces with unequal active sites and energies of adsorption [33] and can be expressed as in Equation (5):

$$q_e=K_F{C}^{{\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$n$}\right.}}$$

(5)

Its linearized form is illustrated in Equation (6).

$${lnq}_e={lnK}_F+{\displaystyle \frac{1}{n}}{lnC}_e$$

(6)

where q_e is the number of exchanged ions (mg g⁻¹), C_e is the equilibrium concentration of TPC in solution (mg L⁻¹), K_F is the adsorbent capacity, and n is the Freundlich constant. When 1/n value is 0 < 1/n < 1, adsorption is considered favorable and a physical process [34]; when 1/n = 1, adsorption is linear and irreversible; and when 1/n > 1, adsorption is a chemical process and unfavorable.

2.7. Adsorption Kinetic Models

The adsorption rate of TPC on Pal can be estimated via the application of kinetic models, while the optimal adsorption mechanism and the surficial characteristics of Pal adsorbent were approached [35]. Specifically, the pseudo-first-order kinetic model or Lagergren equation describes the adsorption process of a solid–liquid system, whose adsorption rate depends on the adsorbent’s sorption capacity [36]. The pseudo-second order kinetic model expresses the proportional adsorption rate to the available active sites on the adsorbent’s surface [37]. The nonlinearized forms of the kinetic models are expressed by Equations (7) and (8), respectively.

$$q_t=q_e(1-e^{-k_{1}t})$$

(7)

$$q_t=\left({\displaystyle \frac{k_2{q}_e^2}{1+k_2{q}_e}}\right)t$$

(8)

and the linearized equations, respectively, by Equations (9) and (10):

$$ln\left(q_e-q_t\right)=ln{q}_e-k_1$$

(9)

$${\displaystyle \frac{t}{q_t}}={\displaystyle \frac{1}{k_2}}+{\displaystyle \frac{1}{q_e}}t$$

(10)

where q_t is the amount of adsorbed pollutant at time t (mg g⁻¹), and k₁ (1 min⁻¹) and k₂ (g mg⁻¹ min⁻¹) are the rate constants of TPC sorption for the pseudo-first and second-order kinetic models, respectively. The k₁ value can be determined from the slope of the linear plot of ln(q_e – q_t) vs. t, and k₂ from the intercept of the linear plot of t/q_t vs. t.

The coefficient of determination (R²), as well as the residual sum of squares of errors (RSS), were used as statistical factors to quantify the deviation to measure the uncertainty in the error distribution in order to identify the most suitable kinetic/isotherm model [38].

3. Results and Discussion

3.1. Pal Characterization

The XRD pattern of the Pal sample was characterized by palygorskite’s mineral reflections at 2θ° values 8.3°, 16.1°, 19.7°, 26.7°, 34.5°, and 42°, in accordance with the standard card (Figure S1), indicating that palygorskite is the main mineralogical phase (97%) in the Pal sample. The 2θ° reflection at 6^o indicated the predicted occurrence of saponite (Sap) as an impurity (3%), since this mineral coexists with Pal in the Ventzia basin deposit (Figure 1). The presence of saponite impurities in palygorskite samples could enhance phenol adsorption by providing additional high-energy adsorption sites, interlayer accommodation, and coordination bonding in the interlayer space [19,20]. The palygorskite dominance in the sample was also morphologically verified via SEM observation, where the characteristic fibrous morphology of palygorskite mineral solely characterizes the Pal sample (Figure 2a,b). Figure 2a clearly shows that the length of the fiber ranges from 200 nm to 2 μm, as well as the typical agglomeration that the fibers form, known as bundles. Figure 2b shows the dominance of the fibrous morphology in the examined area of the Pal sample. The EDS analysis verified that Al, Si, O, Mg, and Fe constitute the major composition of Pal, with atomic concentrations of 0.64%, 19.6%, 66.1%, 9.24%, and 3.96%, respectively (Table S1, Figure S2). The Mg- and Fe- substitutions in the composition of Pal are typical of the Ventzia basin [39].

Following the N₂ adsorption–desorption isotherm of Pal can be presented in Figure 3a, which exhibits a characteristic Type IV isotherm with a H₃-type hysteresis loop, according to the IUPAC classification. This isotherm shows a gradual increase in N₂ uptake across a relative pressure range of 0.1 to 0.9, followed by steep adsorption at high P/P₀ (>0.9). This suggests capillary condensation within the mesopores, while the limited loop at 0.7 < P/P₀ < 0.9 is typical of low degree mesoporosity [40,41]. The overlap at P/P₀ < 0.4 and the pronounced adsorption at low relative pressures (P/P₀ < 0.1), indicated the micropores co-existence [24]. The pore size distribution (Figure 3b) verifies the low-degree mesoporosity, which ranges from 4 nm to 15 nm, whereas the N₂ adsorption in pores below 2.5 nm confirms the micropores derived from the crystalline channels of palygorskite. The specific surface area (SSA) value, average pore diameter or widths, and the total pore volumes of the Pal sample were determined as shown in

Table 1. Physicochemical properties of Pal sample.

Physicochemical Properties	Pal Sample
Specific surface area (m2 g−1)	185
Total pore volume (cm3 g−1)	0.307
Micropore volume (cm3 g−1)	0.022

.

The characterization of Pal via FTIR revealed well-defined, typical structural bands (Figure 4). Specifically, a sharp hydroxyl stretching at approximately 3618 cm⁻¹, assigned to structural Al–OH–Al groups, alongside broader absorptions at 3540 cm⁻¹ due to coordinated and adsorbed water molecules [42]. The bending mode near 1651 cm⁻¹ corresponds to water in the clay channels, while significant Si–O–Si tetrahedral inversion bands appear near 1197 cm⁻¹, with additional Si–O stretching at 976 cm⁻¹ and Al–Al–OH deformation at 912 cm⁻¹ [43]. In TOW spectra, characteristic phenolic features appear: a broad –OH band near 3400 cm⁻¹, as well as expected aromatic C = C stretching around 1600–1580 cm⁻¹, and conjugated C–O or glycosidic vibrations in the 1200–1000 cm⁻¹ region. The band at approximately 1450 cm⁻¹ may be faint or absent, depending on the concentration of hydroxytyrosol or verbascoside present in the raw wastewater [44,45]. After adsorption (Pal/TOW spectra), two newly formed bands at ~1445 cm⁻¹ and 1480 cm⁻¹ occurred in Pal. The band at 1445 cm⁻¹ is specifically attributed to aromatic ring vibrations often associated with phenolic OH or CH deformation modes, as phenols engage in π-π interactions, hydrogen bonding, or coordination with surface hydroxyls of Pal, while similar patterns are noted in kaolinite-rich systems [46,47,48]. Additionally, a band around 1479 cm⁻¹ is consistently attributed to aromatic C = C stretching of phenols adsorbed on clay surfaces [46]. Moreover, the broad –OH region (3400–3650 cm⁻¹) intensity increased after phenol loading, interpreted as the incorporation of phenolic OH groups into hydrogen-bond networks with Pal hydration or surface sites [46,47]. Thus, the FTIR evidence confirmed that phenolic compounds derived from TOW were successfully adsorbed onto Pal via hydrogen bonding or coordination with surface hydroxyls of Pal.

3.2. Adsorption Batch Kinetic Experiments

3.2.1. Effect of Adsorbent Dosage

The impact of varying dosages of TPC, d-COD, and color removal from TOW using different amounts of Pal adsorbent (from 2.5 to 20 g L⁻¹) was assessed, as illustrated in Figure 5a–c. Increasing the adsorbent dosage was found to enhance the removal efficiency of all the TOW substances analyzed, though not substantially. TPC removal was found to be 65 ± 2.7% and 68 ± 3.2% when 10 g/L and 20 g/L of Pal were used, respectively. This compares to 60 ± 1.8% and 62 ± 2.2% when 2.5 and 5 g/L were used, respectively (see Figure 5a). The lowest examined masses (2.5 and 5 g/L) reached maximum removal efficiency within the first 2.5 min, whereas 10 and 20 g/L of Pal demanded 5 min. After this maximum point, a slight decrease in removal efficiency was observed until a plateau was reached after 30 min. This pattern could be attributed to the increasing availability of active sites with adsorbent increase at the same volume of TOW, enabling higher removal [38]. However, because no major rises were observed, this shows that the equilibrium between TPC molecules anchored to the adsorbent and free TPC molecules approaches [49]. Similar results were obtained in the study of Guo et al. [28] in which raw and chemically modified palygorskite were used to adsorb phenol solutions. This indicates that, for a given mass of adsorbent, further increases in adsorbent mass do not improve equilibrium adsorption capacity. Subsequently, the d-COD exhibited a similar pattern to TPC removal, with 10 and 20 g L⁻¹ achieving the highest removal efficiency of 55 ± 2.5% at 5 and 20 min, respectively, while the other masses achieved 50 ± 3% within the first 2.5 min (Figure 5b). This d-COD decrease is significant for the capacity of clay minerals. Lastly, Pal was found to be an excellent option for TOW decolorization, since more than 95% decolorization was achieved within 2.5 min, independently of the adsorbent dosage (Figure 5c). The palygorskite mineral has been shown in many research studies to be an efficient adsorbent for a variety of dyes [26,50] and colored wastewaters [51,52], verifying its capacity to remove large and potentially complex organic molecules. In the present study, the TOC color could mainly be attributed to polyphenols, which Palygorskite was capable of removing. However, these decolorization rates exceeded TOC removal, indicating that Palygorskite is capable of adsorbing other color-responsible compounds. This highlights the dominance of mesoporosity, as indicated by the N₂ physiosorption analysis of the present study.

3.2.2. Effect of Initial Concentration

The effect of the initial total phenolic content (TPC) concentration (ranging from 80 to 400 mg L⁻¹) on d-COD and color removal efficiency was evaluated over time, as shown in Figure 6a–c. Based on the kinetic experiments conducted, the Pal adsorption capacity for TPC increased with lower initial TPC concentrations, achieving a removal efficiency of up to 85 ± 3.4% compared to 64 ± 4.7% at TPC concentrations of 80 and 400 mg L⁻¹, respectively. Similarly to TPC, the d-COD decrease presented a similar pattern, since its removal efficiency reached up to 55 ± 3.1% instead of 42 ± 2.8% for 80 and 400 mg L⁻¹, respectively, corresponding to 600 and 3000 mg L⁻¹ d-COD feed concentration, respectively. This behavior was expected because, as the concentration of adsorbate was increased, the insufficient number of active sites on the Pal surface would limit the mass transfer process [53]. On the other hand, the decolorization remained almost stable and independent of TPC initial concentration, with removal rates ranging from 93 ± 3.5% to 98 ± 2.3%. This efficacy highlights that the decolorization process is mostly affected by the Pal adsorption capacity and not by adsorbate concentration, as Cifuentes-Cabezas [37] determined for phenolic compounds.

3.2.3. Effect of pH

The effectiveness of Pal at absorbing TPC and d-COD in TOW is strongly influenced by the pH level, while the ability to remove color remains consistent across the pH range (see Figure 7). The data indicate that color removal exceeds 90% at pH 2–9, peaking near pH 4–5. This suggests that chromophores predominantly adsorb via pore entrapment or strong surface interactions that are largely insensitive to pH-mediated surface charge variations [20]. Conversely, d-COD and TPC removal display distinct pH dependency: optimal efficiencies (around 65% d-COD and 74% TPC) are attained under acidic conditions (pH 2–3), decline steeply around pH 5–6, and partially rebound, though never attain initial levels at neutral to alkaline pH. This trend may be attributed to pH-induced changes in phenolic ionization, whereby protonated phenolic species exhibit stronger adsorption onto the palygorskite surface at low pH [54]. In contrast, at higher pH, deprotonation impairs adsorption due to electrostatic repulsion and competition with OH⁻ ions. This is due to the surface protonation of palygorskite. Similar behavior was observed in the study of Zhao et al. [29] for Bisphenol-A adsorption at basic pH values on palygorskite, due to the electrostatic repulsion between the anionic Bisphenol-A and the negatively charged sites of the Pal surface.

3.3. Adsorption Isotherm Models

The experimental data of TPC adsorption on Pal were fitted to Langmuir and Freundlich isotherm models in linear and nonlinear forms (

Table 2. Parameters of linear and nonlinear forms of Langmuir and Freundlich isotherms, describing TPC adsorption on Pal.

Parameters	Langmuir	Parameters	Freundlich
	Linear Fitting
R2	0.982	R2	0.992
qmax (mg g−1)	0.068	1/n	1.931
KL (L mg−1)	0.634	KF	1.244
RSS	1.054	RSS	0.035
	Nonlinear Fitting
R2	0.883	R2	0.892
qmax (mg g−1)	0.051	1/n	0.757
KL (L mg−1)	0.053	KF	1.848
RSS	5.737	RSS	5.293

). The linear Freundlich isotherm yielded the best fit (R² = 0.992; RSS = 0.035), indicating a heterogeneous adsorption surface with varying site energies, which is consistent with multilayer adsorption behavior. In contrast, the Langmuir model, which assumes a monolayer adsorption on a homogeneous surface, provided a lower fit (R² = 0.982; RSS = 1.054). Nonlinear fittings for both models resulted in significantly lower R² values (0.883 for Langmuir and 0.892 for Freundlich), further confirming that the linear Freundlich model most accurately describes TPC adsorption onto Pal under the studied conditions. Furthermore, the lowest RSS value was obtained by the linear Freundlich model, which confirms that the discrepancy between the experimental data and this model is the smallest compared to the other fittings. These findings were differentiated with other studies where clay minerals or zeolites were used for phenols removal, which presented a better fit in the Langmuir isotherm [17,28,55]. However, none of these studies evaluated the adsorption of phenolic compounds in wastewater, which could differentiate the interaction between TPC and the heterogeneous surface and active sites of Pal. Despite the Pal adsorption performance fitting the Freundlich model, the operating conditions tested were not optimal for adsorption, as the 1/n values were greater than 1, highlighting that TPC removal is probably a chemical process [34].

3.4. Adsorption Kinetic Models

Adsorption kinetics were evaluated using both pseudo-first order (PFO) and pseudo-second order (PSO) models in linear and nonlinear forms (

Table 3. Parameters of linear and nonlinear forms of pseudo-first order (PFO) and pseudo-second order kinetic models, describing TPC adsorption on Pal.

Parameters	PFO	Parameters	PSO
	Linear Fitting
R2	0.690	R2	0.999
qecal (mg g−1)	0.867	qecal (mg g−1)	0.085
qeexp (mg g−1)	12.48	qeexp (mg g−1)	12.48
k1 (min−1)	0.048	k2 (g mg−1 min−1)	0.008
RSS	0.680	RSS	0.004
	Nonlinear Fitting
R2	0.846	R2	0.983
qe (mg g−1)	12.27	qecal (mg g−1)	12.36
qeexp (mg g−1)	12.48	qeexpl (mg g−1)	12.48
k1 (min−1)	3.226	k2 (g mg−1 min−1)	1.705
RSS	0.035	RSS	0.004

). The linear PSO model demonstrated an almost perfect fit (R² = 0.999; RSS = 0.004), indicating that the adsorption rate is likely controlled by chemisorption processes involving valency forces through sharing or exchange of electrons. In contrast, the PFO model showed a poor fit (R² = 0.690; RSS = 0.680), suggesting that physical adsorption is not the primary mechanism. Notably, the nonlinear forms yielded higher calculated q_e values (q_ecal = 12.3 mg g⁻¹) for both models, which are close to experimental q_e (q_eexp = 12.48 mg g⁻¹), but only the PSO nonlinear model maintained a strong fit (R² = 0.983). This finding highlights the PSO’s suitability to describe the process by supporting the close agreement between experimental and model-predicted equilibrium adsorption capacity. These results are consistent with previous reports that identified PSO kinetics as the most representative for phenol adsorption on complex adsorbents, while the adsorption process of TPC can be described as chemisorption, as the fit of the Freundlich isotherm with 1/n > 1 indicated as well, and the rate-determining step is proposed to be the surface adsorption [55].

3.5. Suggested Mechanism for TOW Treatment by Pal

The treatment of real table olive wastewater (TOW) using natural palygorskite (Pal) was investigated to assess its adsorption efficiency and potential for future scale-up applications. The primary focus was the removal of total phenolic content (TPC), decolorization, and a decrease in dissolved chemical oxygen demand (d-COD), all of which contribute significantly to the organic load of TOW. These removals are governed by a spectrum of physicochemical interactions between clay functional groups and organic solutes, including hydrogen bonding, electrostatic interactions, interlayer intercalation, and coordination bonding mechanisms.

Batch adsorption experiments combined with characterization techniques (FTIR, XRD, N₂ physisorption, and EDS) revealed that multiple concurrent interaction mechanisms are likely. Given the complexity of real wastewater, the competition among diverse organic compounds for adsorption sites may affect selectivity and capacity. Notably, despite achieving 95% decolorization, only 65% TPC removal was observed. Although polyphenols, particularly hydroxytyrosol, are recognized as major contributors to TOW color, the partial mismatch suggests the co-presence of additional chromophoric species such as humic substances, tannins, flavonoids, and iron salts (e.g., ferrous gluconate or lactate), which are commonly added during olive stabilization.

FTIR spectra before and after adsorption further support the hypothesis of complex adsorption pathways. A broad band between 3000 and 3500 cm⁻¹, centered at 3397 cm⁻¹, indicates –OH stretching vibrations consistent with hydrogen bonding and potential presence of ferrous gluconate [56]. This is reinforced by EDX data showing high iron content in Pal, supporting the possible formation of Fe–phenol complexes via coordination bonds [57]. Hydroxytyrosol, the dominant phenolic compound in TOW, exhibits characteristic FTIR bands near 3200–3400 cm⁻¹ for phenolic –OH stretching, ~1600 cm⁻¹ for aromatic C=C stretching, and ~1275 cm⁻¹ for C–O–H bending—all of which were altered upon adsorption, indicating interactions with surface hydroxyl and silanol groups of Pal [44,46].

Thermodynamic parameters further illuminate the nature of adsorption [58]. Data from phenol adsorption on Moroccan clays show increased capacity with rising temperature (e.g., from 5.84 mg/g at 30 °C to 6.80 mg/g at 50 °C for acid-activated clays), indicating an endothermic process [46]. The negative Gibbs free energy (ΔG°) across the temperature range confirms spontaneity, while positive enthalpy (ΔH°) and entropy (ΔS°) suggest that adsorption is favored by increased molecular disorder at the clay–solution interface. Comparable studies reinforce these findings. Huang et al. [59] reported enthalpy values of −49.8 kJ/mol (pH 4) and −38.6 kJ/mol (pH 8) for tannin adsorption on organo-attapulgite, indicating that hydrogen bonding and hydrophobic interactions dominate, with more favorable bonding occurring under acidic conditions. Similarly, Khachay et al. [60] demonstrated that phenol adsorbs strongly on sepiolite—a clay mineral closely related to Pal—through hydrogen bonding via its –OH group with surface silanol (–SiOH) groups and π–π interactions with siloxane surfaces. The highly negative adsorption energy supports a thermodynamically stable adsorption process.

In summary, adsorption of phenolic and chromophoric compounds from TOW onto Pal appears to proceed through a combination of mechanisms: (i) hydrogen bonding between –OH groups of hydroxytyrosol and surface silanol groups; (ii) electrostatic interactions facilitated by the clay’s variable surface charge; and (iii) coordination complex formation with Fe³⁺ species inherent to Pal. These interactions are further enhanced by increased temperature, which promotes entropic gain and molecular mobility. The presence of multiple organic species in real wastewater likely results in competitive and selective adsorption, explaining the observed disparities between TPC and color removal efficiencies.

3.6. Practical Implementations for Potential Pilot Application

The present study highlighted Pal’s adsorption capacity for TPC, d-COD, and decolorization of real, untreated TOW. These preliminary experiments shed light on the application of the fibrous clay mineral palygorskite on this kind of wastewater, which was not evaluated before. The determination of its removal efficiency was critical, since adsorption has been found to be the most attractive technology for olive-processing wastewaters, aiming to further recover phenols [37]. So far, the extraction and identification of phenols from a variety of wastewater treatments has been evaluated in chemical adsorbents such as resins or membranes [37,61,62,63], but there is a lack of a targeted approach concerning the TOW treatment with natural materials, which could further lead to phenol recovery and the investigation of material regeneration.

According to the batch kinetic experiments, the organic load of TOW was reduced by up to 70% from TPC and approximately 40% from d-COD, decreasing the final TOW concentrations at 150 mg L⁻¹ and 1450 mg L⁻¹ for TPC and d-COD, respectively, for the highest examined initial TOW concentrations. This could be considered a relatively heavy TOW with an organic load remaining after adsorption. However, this load is significantly reduced after Pal adsorption as a pretreatment step. The effluent could then undergo biological treatment, which is the usual method for managing industrial waste. At the same time, phenols could be desorbed and isolated from the Pal, as was achieved in the study by Tri et al. [64], where more than 50% desorption of phenols was achieved using a 20% ethanol mixture with synthesized Fe-nano zeolite as a cost-effective and sustainable method. Specifically, based on the data provided by Geohellas S.A. and published by Lazaratou et al. [65], the cost of Pal is 0.22 EUR kg⁻¹, meaning the cost of the specific experiments is 3.2 EUR kg⁻¹ d-COD and 18 EUR kg⁻¹ TPC when 20 g L⁻¹ were applied. This makes Pal a cost-effective solution for TOW treatment. The overall cost would be reduced if initial concentrations were lower, as found in other case studies, making it worthwhile to study the process in scale-up. This is especially true given the good mechanical performance of thermally treated Pal in continuous flow experiments [66]. The results of the present research showed that Pal is worth further investigation in this kind of research, via continuous flow experiments, conduction, desorption, or generation studies as an integrated approach for TOW treatment, or a pretreatment step.

4. Conclusions

The present study evaluated the adsorption behavior of palygorskite (Pal) as a natural, low-cost adsorbent for the treatment of table olive wastewater (TOW), with particular emphasis on total phenolic compounds (TPCs), color, and dissolved chemical oxygen demand (d-COD). Batch adsorption experiments demonstrated Pal’s efficacy in removing these pollutants, indicating that increased Pal dosage improved TPC and d-COD removal, highlighting that equilibrium conditions were reached beyond certain dosages. Notably, color removal remained consistently high (>95%) regardless of adsorbent amount, underscoring Pal’s strong affinity for chromophoric compounds, possibly due to surface interactions and complex adsorption mechanisms, including hydrogen bonding, electrostatic interactions, and coordination bonding, especially between phenolic compounds and iron inherent in Pal, as EDS and FTIR analysis proved. The discrepancies between color and TPC removal by Pal (95% compared to 65%, respectively) were most likely due to the presence of other chromophoric substances such as ferrous salts. Similarly, lower initial TPC concentrations favored higher removal efficiencies for both TPC and d-COD, suggesting site saturation limitations at elevated adsorbate levels. Across a wide pH range, Pal maintained its color removal efficiency, while TPC and d-COD adsorption were maximized at acidic pH (2–3), reaching 75% and 65%, respectively, reflecting enhanced protonated interactions and reduced electrostatic repulsion. Kinetic and equilibrium data further validated Pal’s adsorption performance, with the Freundlich isotherm and PSO kinetic order models presenting the best fit, rendering the TPC removal a chemisorption on a heterogeneous surface. However, the research findings support Pal’s suitability as a cost-effective and favorable treatment option. Practical application is bolstered by Pal’s low cost, high mechanical stability, and potential for phenol recovery and reuse, making it an attractive candidate for future scale-up, particularly as a pretreatment step before biological processing.