Screening Agricultural Residues as Sustainable Alternative Sorbents for the Active Removal of Methylene Blue

Abstract

This study investigates the potential of several sustainable agricultural by-products—including olive stones, cork, and almond shells, which are locally available in Alentejo, Portugal—as low-cost adsorbents for the removal of methylene blue (MB) from synthetic wastewater. The biomass residues were evaluated both in their raw form and after conversion into activated carbons (ACs) through chemical activation with KOH at 973 K. The produced ACs exhibited well-developed surface areas (760–1103.5 m² g⁻¹) and porous structures (0.31–0.51 cm³ g⁻¹). The adsorbents were characterised in terms of their chemical and textural properties. Raw biomass materials presented acidic surface groups, whereas the ACs presented neutral or basic groups. Batch adsorption experiments were conducted to assess the effects of adsorbent particle size, solution pH, initial MB concentration, stirring speed, contact time, and temperature on dye removal efficiency. Among all tested materials, the ACs achieved superior MB adsorption capacities, ranging from 244.2 to 317.6 mg g⁻¹, compared to the untreated biomass adsorbents, which showed capacities between 34.1 and 46.4 mg g⁻¹. The adsorption data were best described by the Langmuir isotherm model, while the kinetic data closely followed the pseudo-second-order (PSO) model. Thermodynamic analysis revealed that MB adsorption was spontaneous and endothermic; however, the relatively low enthalpy values indicated that physical interactions contributed significantly, particularly in the case of the raw biomass adsorbents. This suggests that the PSO model may also be applicable when physical adsorption is the dominant mechanism. This work demonstrates the novel use of cork, olive stone, and almond shell biomasses and their derived ACs as sustainable adsorbents, highlighting an integrated approach that simultaneously promotes efficient wastewater treatment, waste valorisation, and circular economy-driven socio-economic development.

1. Introduction

Sustainable water management has emerged as one of the most pressing challenges facing the scientific community today. Population growth and the indiscriminate use of water resources have led to the introduction of various contaminants, rendering water unsuitable for many daily uses. Among the most prevalent and problematic contaminants found in different effluents are dyes (e.g., methylene blue (MB), methyl orange, and methyl red), heavy metals (e.g., chromium, arsenic, zinc, and lead), chemical compounds (e.g., inks and phenolic compounds), pharmaceutical compounds (e.g., antibiotics, analgesics, and anti-inflammatories), and pesticides (e.g., fungicides, herbicides, and acaricides) [1].

The presence of contaminated water poses significant health hazards to humans, animals, and aquatic life. For this reason, the removal of pollutants from wastewater prior to its discharge into water bodies is imperative. To this end, a range of physical, chemical, biological, and combined treatment methods have been employed, including ozonation, advanced oxidation processes, activated sludge treatment and biofiltration, constructed wetlands, photocatalytic treatments, coagulation and flocculation, adsorption, and other emerging or hybrid approaches [1,2].

Due to its superior performance, operational simplicity, and ease of maintenance, adsorption using a broad range of adsorbents [2,3] is widely recognised among the various treatment methods reported in the literature [2]. Even in photodegradation processes, adsorption onto the photocatalyst surface has been shown to enhance methylene blue removal from aqueous solutions [4]. Nevertheless, some adsorbents, such as activated carbons (ACs), despite their excellent performance, are associated with high production costs. Therefore, researchers are striving to identify more economical and sustainable alternatives that can overcome this limitation and enable the treatment of the large volumes of wastewater generated daily [5].

The use of natural raw biomass waste for wastewater treatment, as data show in Table 1, offers a sustainable, circular economy-driven, and environmentally friendly alternative to conventional adsorbents such as biochar and activated carbons. Silva et al. reported that the cost of AC per gram of fluoxetine removed from water was 3.13 € g⁻¹ using granular AC and 0.41 € g⁻¹ using cork waste [6].

As shown in Table 2, the adsorption capacities of activated carbons (ACs) for MB typically range between 100 and 500 mg g⁻¹, while the market price of AC generally exceeds 5 € kg⁻¹ [7]. Based on these values, a simplified cost estimation suggests that the adsorbent cost required to remove 1 kg of MB from water may range from approximately 3 to 25 € kg⁻¹. In contrast, the adsorption capacities of raw biomass materials generally range from 50 to 100 mg g⁻¹ (Table 1), and these residues are often available at little or no cost, aside from harvesting and processing expenses. Considering harvesting costs in Portugal of approximately 21.4–27.4 € ton⁻¹ [8], the estimated adsorbent cost required to remove 1 kg of MB would range from approximately 0.22 to 2.74 € kg⁻¹. These estimates highlight raw biomass wastes as potentially more economical and sustainable alternatives, notwithstanding their lower adsorption capacities.

Although the use of raw biomass in applications such as biofuel production [9,10], biogas generation [11], and adsorption-based water treatment offers considerable environmental and sustainability benefits, certain drawbacks continue to limit investment in biomass valorisation. In particular, the dispersed nature of natural biowaste across the landscape increases the logistical and operational costs associated with its collection, transportation, and processing [12].

Furthermore, raw biomass can release soluble organic compounds into water, potentially increasing pollution levels and biochemical oxygen demand [13,14,15]. Dyes can be classified based on their origin (natural or synthetic), water solubility (soluble or insoluble), application method (reactive, disperse, or mordant), and particle charge in aqueous solution (anionic, cationic, or non-ionic) [1,16]. As synthetic dyes are readily available in a wide variety of shades and are easy to apply, they are extensively used in both industry and research [16,17,18,19]. According to Kamenická, more than 100,000 different types of dyes are currently available on the market, with one of their principal applications being in the textile industry [17]. Nevertheless, no precise databases exist to reliably estimate the total amount of dyes released into wastewater each year [18].

Table 1. Untreated biomass raw material used to remove methylene blue from water.

Biomass Waste	Optimal Adsorption Conditions	MB Adsorbed mg g−1	Reference
Onion membrane	Dosage = 0.48 g L−1; T = 50 °C; pH = 11; Contact time = 1 h Contact time = 8 h	1055 1202	[20]
Citrus limetta peel waste	Dosage = 2.0 g L−1; T = 25 °C; pH = 4; Contact time = 3 h	227.3	[21]
Subble Tectona Grandis Adansonia digitata L. Bamboo flowers	Dosage = 1 g L−1; Co = 5–400 mg L−1; T = 25 °C; pH = 6; contact time = 24 h	63.7 27.9 156.8 42.8	[15]
Potato peels; date palm leaves; dragon fruit peels; corn husk; sugarcane bagasse; Chlorella vulgaris microalgae, plant leaves	Variable	0.25 to 72.0 4.5 to 149.0	[22] (review)
Barley (Hordeum vulgare) bran Enset (Ensete entricosum midrib leaf)		63.2 35.5	[23]
Potato (Solanum tuberosum) peel	Dosage = 2.0 g L−1, Co = 10–40 mg L−1; T = 25 °C; contact time = 5 h	15.77	[24]
Garlic peel	Co = 25–200 mg L−1; T = 50 °C; contact time = 5 h	142.86	[25]
Dragon fruit peels (Hylocereus polyrhizus) peels	Dosage = 0.2–2 g L−1; pH = 5; Co = 50–400 mg L−1; T = 30 °C; contact time = 2 h	192.31	[26]
Laminaria digitata; Horse chestnut husk Hazelnut husk Rapeseed residue	Dosage = 2.0 g L−1; Co = 200 mg L−1; contact time = 24 h; T = 30 °C	500 137 120 85	[27]

Table 1. Untreated biomass raw material used to remove methylene blue from water.

Biomass Waste	Optimal Adsorption Conditions	MB Adsorbed mg g−1	Reference
Onion membrane	Dosage = 0.48 g L−1; T = 50 °C; pH = 11; Contact time = 1 h Contact time = 8 h	1055 1202	[20]
Citrus limetta peel waste	Dosage = 2.0 g L−1; T = 25 °C; pH = 4; Contact time = 3 h	227.3	[21]
Subble Tectona Grandis Adansonia digitata L. Bamboo flowers	Dosage = 1 g L−1; Co = 5–400 mg L−1; T = 25 °C; pH = 6; contact time = 24 h	63.7 27.9 156.8 42.8	[15]
Potato peels; date palm leaves; dragon fruit peels; corn husk; sugarcane bagasse; Chlorella vulgaris microalgae, plant leaves	Variable	0.25 to 72.0 4.5 to 149.0	[22] (review)
Barley (Hordeum vulgare) bran Enset (Ensete entricosum midrib leaf)		63.2 35.5	[23]
Potato (Solanum tuberosum) peel	Dosage = 2.0 g L−1, Co = 10–40 mg L−1; T = 25 °C; contact time = 5 h	15.77	[24]
Garlic peel	Co = 25–200 mg L−1; T = 50 °C; contact time = 5 h	142.86	[25]
Dragon fruit peels (Hylocereus polyrhizus) peels	Dosage = 0.2–2 g L−1; pH = 5; Co = 50–400 mg L−1; T = 30 °C; contact time = 2 h	192.31	[26]
Laminaria digitata; Horse chestnut husk Hazelnut husk Rapeseed residue	Dosage = 2.0 g L−1; Co = 200 mg L−1; contact time = 24 h; T = 30 °C	500 137 120 85	[27]

Methylene blue, for instance, is widely used across chemical, biological, medical, and textile applications. However, it can cause several adverse health effects, including nausea, diarrhoea, anaemia, and irritation of the skin, mouth, throat, oesophagus, and stomach [16,19]. As a cationic dye, MB is particularly challenging to remove from aqueous solutions. Consequently, considerable research has focused on its sustainable removal, particularly through the use of untreated biomass waste and biomass-derived activated carbons, as part of broader biomass valorisation and environmentally friendly water treatment strategies, as summarised in Table 1 and Table 2.

The main objective of this work is to investigate, for the first time, the application of locally sourced untreated biomass by-products—namely cork sawdust, olive stones (OS), and almond shells (AS), which are abundant agricultural by-products in the Alentejo region of Portugal—as sustainable, low-cost alternative adsorbents for the removal of MB from aqueous solutions.

Table 2. Carbon materials, prepared from different biomass waste, used to remove MB from water (data with “*” are expressed as % removal).

Precursor	Experimental Conditions on MB Removal	MB Adsorbed (mg g−1) or (%) *	Reference
Acacia wood (Microwave—CO2: 150 mL min.)	Contact time = 180 min.; T = 30 °C; pH = 7	210.21	[28]
Date stones—mixed with 100 mL of H2 SO4 (2N); heated under reflux for 8 h.	Dosage = 1 g L−1; Co = 100–800 mg L−1; t = 60 min.; pH = 5.57; T = 50 °C	515.46	[23,29]
Areca leaf plate waste (biochar)	Dosage = 0.2 g L−1; contact time = 90 min.; pH = 8.0; T = 35 °C, Co = 10 to 200 mg L−1.	80.8 *	[30]
Wood Pinus caribaea hydrochar (hydrocarbonisation done at 200 or 240 °C, for 12 or 24 h, in acidic or basic medium)	Dosage = 1.0 g L−1; contact time = 360 min.; pH = 11.0; T = 25 °C; Co = 300 mg L−1	149.0 94.7 *	[31]
Leaves from enset; banana; Agave Samaniana; Aloe vera; carrot, grape, magnolia; Euonymus japonicus; coconut; Saccha-rum; Dipterocarpus; Artichoke; rubber; tealeaves; mango; potato; pineapple;	Variable	30 to 500.0	[22] (review)
Sol–gel—prepared from aminated lignin and sodium alginate	Dosage = 1.0 g L−1; Co = 400 mg L−1; pH = (5.0, 7.0 and 12.0); contact time = 5–700 min.; T = 25 °C.	388.81	[32]
Chestnut thorns shell—ACs	Co = 20–700 mg L−1; pH = (5.0, 7.0 and 12.0); contact time = 24 h.; T = 25 °C.	305.81	[33]
Bamboo flowers—ACs	Dosage = 1.0 g L−1; Co = 50 mg L−1; contact time = 12 h; T = 30–50 °C	374.75	[34]
Almond shell—ACs	Dosage = 1.0 g L−1; Co = 100–500 mg L−1; contact time = 24 h; T = 25 °C;	213.0 to 487	[35]
Coconut Shell—ACs	Dosage = 5.0 g L−1; Co = 25–500 mg L−1; contact time = 6 h; T = 25–45 °C;	200.0	[36]

Table 2. Carbon materials, prepared from different biomass waste, used to remove MB from water (data with “*” are expressed as % removal).

Precursor	Experimental Conditions on MB Removal	MB Adsorbed (mg g−1) or (%) *	Reference
Acacia wood (Microwave—CO2: 150 mL min.)	Contact time = 180 min.; T = 30 °C; pH = 7	210.21	[28]
Date stones—mixed with 100 mL of H2 SO4 (2N); heated under reflux for 8 h.	Dosage = 1 g L−1; Co = 100–800 mg L−1; t = 60 min.; pH = 5.57; T = 50 °C	515.46	[23,29]
Areca leaf plate waste (biochar)	Dosage = 0.2 g L−1; contact time = 90 min.; pH = 8.0; T = 35 °C, Co = 10 to 200 mg L−1.	80.8 *	[30]
Wood Pinus caribaea hydrochar (hydrocarbonisation done at 200 or 240 °C, for 12 or 24 h, in acidic or basic medium)	Dosage = 1.0 g L−1; contact time = 360 min.; pH = 11.0; T = 25 °C; Co = 300 mg L−1	149.0 94.7 *	[31]
Leaves from enset; banana; Agave Samaniana; Aloe vera; carrot, grape, magnolia; Euonymus japonicus; coconut; Saccha-rum; Dipterocarpus; Artichoke; rubber; tealeaves; mango; potato; pineapple;	Variable	30 to 500.0	[22] (review)
Sol–gel—prepared from aminated lignin and sodium alginate	Dosage = 1.0 g L−1; Co = 400 mg L−1; pH = (5.0, 7.0 and 12.0); contact time = 5–700 min.; T = 25 °C.	388.81	[32]
Chestnut thorns shell—ACs	Co = 20–700 mg L−1; pH = (5.0, 7.0 and 12.0); contact time = 24 h.; T = 25 °C.	305.81	[33]
Bamboo flowers—ACs	Dosage = 1.0 g L−1; Co = 50 mg L−1; contact time = 12 h; T = 30–50 °C	374.75	[34]
Almond shell—ACs	Dosage = 1.0 g L−1; Co = 100–500 mg L−1; contact time = 24 h; T = 25 °C;	213.0 to 487	[35]
Coconut Shell—ACs	Dosage = 5.0 g L−1; Co = 25–500 mg L−1; contact time = 6 h; T = 25–45 °C;	200.0	[36]

Following a circular economy and sustainability approach, locally available bioresources in their raw, untreated form are evaluated as substitutes for conventional adsorbents such as ACs in water treatment applications, thereby contributing to biomass waste valorisation and resource recovery. Furthermore, thermochemical conversion of these biomass residues into ACs significantly enhances their surface area, porosity, and MB adsorption capacity.

2. Materials and Methods

2.1. Use of Raw Materials as Adsorbents for Methylene Blue Removal from Water

Locally sourced agricultural waste-derived materials were used directly as sustainable adsorbents for the removal of MB from aqueous solutions. The raw biomass adsorbents were obtained from different sources: cork was collected during the bark-stripping process from trees, olive stones were supplied by a local olive oil processing facility, and almond shells were sourced from domestic almond consumption. Prior to use, the raw materials underwent minimal pre-treatment, consisting of air-drying at ambient temperature for 7 days, followed by oven-drying at 373 K until constant mass was achieved. The dried biomass was subsequently milled to reduce particle size, yielding fractions of either below 0.63 mm or between 0.63 and 1.25 mm. The particles were then washed with distilled water at 353 K for 24 h to remove surface impurities, dried, and stored for subsequent use, following a previously reported experimental procedure [15].

2.2. Activated Carbon Production from Agricultural and Processing Industries Waste

Activated carbons (ACs) were prepared from cork, almond shells, and olive stones following approaches previously established by our research group. The ACs were produced via chemical activation using potassium hydroxide (KOH) as the activating agent at a 1:2 mass ratio (biomass: KOH, w/w). KOH was supplied by VWR, Carnaxide, Oeiras, Portugal. As described in previous work [37], the precursors were mixed with KOH and subjected to controlled heating at a rate of 10 K min⁻¹ up to a final temperature of 973 K, at which they were held for 30 min. Carbonisation and activation were carried out under a nitrogen flow of 70 mL min⁻¹ in a semi-industrial horizontal tubular furnace (TR–334/2018, Thermolab, Águeda, Portugal). After activation, the ACs were cooled and subsequently washed with distilled water until the wash solution reached a pH close to 7. Finally, the resulting carbon materials were oven-dried at 373 K until constant mass was achieved, ensuring complete removal of residual moisture.

2.3. Characterisation of Natural Adsorbents and Activated Carbons

The raw adsorbents and their respective ACs were characterised using the following analytical techniques: determination of the pH at the point of zero charge (pHpzc), elemental analysis (EA), nitrogen adsorption at 77 K, and attenuated total reflectance Fourier-transform infrared spectroscopy (ATR-FTIR). The experimental procedures are described in detail in a previous study [15].

Elemental analysis was performed in triplicate using a Eurovector EuroEA analyser (EuroVector, Pavia, Italy). ATR-FTIR spectroscopy was employed to identify the surface functional groups present in the raw adsorbents and ACs. Spectra were collected over the range of 400–4000 cm⁻¹ using a PerkinElmer Spectrum Two IR spectrophotometer (ILC, Lisboa, Portugal), equipped with an attenuated total reflectance (ATR) accessory. Textural characterisation was carried out by nitrogen adsorption at 77 K, with the corresponding isotherms measured using a Quadrasorb gas adsorption manometric instrument (Quantachrome Instruments, Bobadela, Portugal), employing nitrogen of 99.999% purity supplied by Air Liquide, Sines, Portugal. Prior to measurement, the samples were outgassed in a MasterPrep unit (Quantachrome Instruments, Bobadela, Portugal) at 423 K for 5 h, at a heating rate of 2 K min⁻¹. The pHpzc values were determined by the mass titration method, using suspensions containing 7% (m/v) of adsorbent in a 0.1 mol L⁻¹ sodium nitrate solution. The suspensions were maintained in a thermostatic bath under continuous stirring at 20 rpm. After 48 h, the pH of the supernatant was recorded and taken as the pHpzc of each adsorbent.

2.4. Dye Adsorption

The physicochemical properties of MB are presented in

Table 3. Methylene blue properties.

Methylene Blue		Molecule Structure	Reference
Molecular formula	C16H18ClN3S	Protonated form Deprotonated form
Molar mass (g mol−1)	319.85
pKa	3.8		[4]
Solubility in water (g L−1), 298 K	43.6		[4]
Molar absorption coefficient (L mol−1 cm−1, at 664 nm)	~8.4104
Molecular volume (cm3 mol−1) Molecular diameter (Å)	241.9 8.9		[38]
Molecular area σ2 (Å2)	130–135		[39]
Length (Å) Width (Å)	13.5 or 14.47 9.5		[4]

. MB, with a purity greater than 85%, was purchased from Sigma-Aldrich. To investigate the effect of solution pH on MB removal, the pH was adjusted to values ranging from 3 to 11 using 0.1 mol L⁻¹ HCl and NaOH solutions.

To obtain MB adsorption isotherms, approximately 25 mg of raw adsorbents and 10 mg of ACs were accurately weighed into individual Erlenmeyer flasks, to which an exact volume of 25.0 mL of MB aqueous solution—with concentrations ranging from 0 to 400 mg L⁻¹—was added. The suspensions were maintained in a Linear–Grant thermostatic bath at 298 K, under a stirring speed of 20 rpm, for 24 h. Afterwards, the supernatant was carefully collected and diluted where necessary, and the MB concentration was determined by UV-Vis spectrophotometry using a Nicolet Evolution 300 spectrophotometer (Thermo Electron Corporation, Madison, WI, USA).

To quantify the MB concentration in solution, absorbance was measured at the characteristic wavelength of 668 nm. The amount of MB adsorbed per unit mass of adsorbent, Q_ads (mg g⁻¹), was calculated from the difference between the initial (C_i) and equilibrium (C_e) MB concentrations, the volume of solution (V, L), and the mass of adsorbent (m, g), according to Equation (1): Q_ads = (C_i − C_e)/m × V.

The influence of contact time on MB adsorption was evaluated through experiments conducted at 298 K and pH ≈ 7, with the amount of MB removed monitored over 48 h. The influence of temperature (T) was evaluated following the procedure described for isotherm determination, with T carefully controlled between 20 and 60 °C.

The MB adsorption isotherms were analysed by fitting the experimental data to the Langmuir and Freundlich models. The Langmuir isotherm assumes monolayer adsorption onto a homogeneous surface with a finite number of identical sites, and its linearised form is expressed by Equation (2): ${\displaystyle \frac{Ce}{Qads}}={\displaystyle \frac{Ce}{qm}}+{\displaystyle \frac{1}{KL qm}}$. The Freundlich model, which is related to the adsorption on heterogeneous surface adsorbents, can be expressed by Equation (3): log Q_ads = log K_F + 1/n log C_e. In the equations, K_L is the Langmuir constant, Q_ads, C_e and q_m are amount of adsorbate adsorbed, concentration on the equilibrium and the adsorbent maximum adsorption capacity, respectively, estimated by the Langmuir model, and K_F and n are the Freundlich constants.

3. Results and Discussion

3.1. Adsorbents Characterisation

A screening study was conducted to identify potential adsorbents for dye removal from the aqueous solutions. In previous works, the authors investigated several agricultural wastes, such as stubble, fibres from Adansonia digitata, bamboo flowers [15], and sawdust from Tectona grandis [37], for the removal of MB from the aqueous phase. In the present work, almond shells, olive stone, and cork waste were screened as adsorbents for MB removal from the liquid phase.

In the Alentejo region, the extraction of cork and the cultivation of olive trees for olive oil production are centuries-old practices. Although intensive almond cultivation is relatively recent, the by-products generated from these crops already represent a challenge for producers. The selected adsorbents were used only after a washing step to remove impurities, followed by crushing and grinding to reduce particle size. In a second phase, the different types of biomass were used as precursors for the production of activated carbons, as described in Section 2.2.

First, all adsorbents were chemically characterised through pHpzc determination, elemental analysis, and FTIR-ATR analysis. The three natural precursors presented pHpzc values lower than five, as shown in

Table 4. Elemental analysis and pHpzc values of the natural adsorbents and ACs (expressed as % w/w). (* %O₂ was calculated by difference, subtracting the sum of %N, %C, %H, and %S from 100%.)

Adsorbents	N	C	H	O2 *	pHpzc
OS	0.15	47.9	5.7	46.25	4.36
AC_OS_KOH_1_2	0.45	71.3	2.3	25.95	8.8
AS	1.01	46.25	6.37	46.37	4.73
AC_AS_KOH_1_2	3.23	55.4	0.72	40.65	8.6
Cork	0.59	61.7	6.57	31.14	3.93
AC_Cork_KOH_1_2	0.64	63.2	0.03	36.13	7.96

. The ACs presented pHpzc values ranging between 7.96 and 8.80, suggesting that chemical activation with KOH at 973 K renders the resulting carbon materials more basic in character.

At solution pH values higher than the pHpzc, the adsorbent surface tends to become negatively charged, which enhances electrostatic attraction and consequently increases the adsorption of cationic dyes. Conversely, at pH values below the pHpzc, the surface becomes more positively charged, leading to adsorbate–adsorbent repulsion, which reduces MB adsorption. Since MB is a cationic dye, adsorption would be expected to increase at pH values above the pHpzc. However, in this first stage, the influence of solution pH on MB adsorption was also evaluated.

As show in Table 4, raw biomass materials present higher hydrogen contents than their corresponding ACs. For AS and cork, the carbon content increased only slightly following chemical activation process. However, the chemical activation of OS promoted a pronounced increase in carbon content (from 47.9 to 71.3%). The opposite trend was observed for the oxygen content, with AS and OS, indicating a loss of oxygen-containing functional groups. These findings are consistent with previously reported results. The activation of different types of biomass with KOH at 973 K increases the carbon and nitrogen contents while reducing the hydrogen and oxygen contents of the produced ACs [40,41,42]. In contrast, cork showed an increase in oxygen content after activation, from 31.14% to 36.13%. This behaviour may be attributed to the formation of new oxygenated surface groups during KOH activation. Furthermore, the relatively small increase in carbon content (1.5%) also suggests that cork-derived activated carbon retained a significant proportion of oxygen-containing functionalities.

The yield of biomass chemical activation varied from 17.2% for AS to 24.8% for OS, as presented in Table 5. The differences observed in the activation yield may be related to the distinct lignocellulosic compositions of the biomasses used. The obtained yields seem to contradict the lignin content of the precursors. Cork contains 42.7%, 24.9%, and 16.8% suberin, lignin, and polysaccharides, respectively [43]. Olive stone contains 32.1%, 34.8%, and 26.9% lignin, hemicellulose, and cellulose, respectively [44], while almond shells contain 20.1–32.7%, 19.7–35.2%, and 22.8–40.5% lignin, hemicellulose, and cellulose, respectively [45]. Almond shells possess a high lignin content but also a less compact and more fractured structure, which facilitates contact between KOH and the internal surface.

The adsorbents were chemically characterised by ATR-FTIR analysis, and the resulting spectra of raw biomass materials and ACs are shown in Figure 1a and Figure 1b, respectively. The FTIR-ATR spectra of almond shells, cork, and olive stone display similar bands throughout the wavenumber range. The bands between 2923 and 2985 cm⁻¹ are characteristic of the asymmetric stretching of aliphatic C–H groups. This band is more intense in cork biomass due to its high suberin content.

The broad and intense band observed around 3300–3400 cm⁻¹ is attributed to O–H stretching vibrations in phenolic groups characteristic of hemicellulose and cellulose [46]. In the almond shell spectrum, the more evident band at 900 cm⁻¹ can be attributed to N–H deformation in amine groups. The scissoring movement of NH₂ was evidenced by a deformation band at 1640 cm⁻¹. These bands are consistent with the higher nitrogen content identified by elemental analysis in AS compared with OS and cork.

The bands around 1610 and 1635 cm⁻¹ are ascribed to C=C stretching vibrations [41]. The band at 1020 cm⁻¹, observed predominantly in OS and AS, is attributed to C=C bonds in aromatic groups. The bands at 1719 and 1748 cm⁻¹, which are more intense in cork biomass, are attributed to C=O stretching, characteristic of suberin, hemicellulose, and cellulose [46,47]. All these bands corroborate the presence of cellulose, hemicellulose, and lignin, which are the major constituents of these biomass materials.

The FTIR-ATR spectra of the ACs are shown in Figure 1b. Spectra of the three ACs are quite similar. The activation promotes the loosening of near-acidic functional groups, causing ACs to lose their fingerprints. The band around 3400 cm⁻¹ is considerably less intense compared with those of the original biomass materials. The band at 1620 cm⁻¹ is more evident on AC_AS_KOH_1_2 and AC_OS_KOH_1_2 and can be assigned to C=O aromatic structure stretching vibrations. The broad band between 1100 and 1325 cm⁻¹ predominantly in olive stone and almond shell ACs is attributed to the C-O stretching in a phenolic group. At low wavenumbers, around 540 and 595 cm ⁻¹, the presence of small bands confirms the existence of aromatic structures. Additionally, FTIR-ATR spectra were also collected for MB-saturated adsorbents. The solidity of functional groups and molecular vibrations bands are confirmed on all spectra, suggesting that the adsorption mechanism is primarily surface-based and does not appreciably alter the chemical composition of the adsorbents.

Table 5. Structural parameters of the ACs prepared from cork, OS and AS, through chemical activation, with KOH at 973 K. In Table 5, η is the activation yield, A_BET is the apparent surface area, V_s is the total volume, A_s is the external surface area, (V_s and A_s were obtained using the alfa-s method proposed by Carrott et al., 1987) [48], V_o is the microporous volume, E_o is the energy of adsorption and L_o is the mean pore size (V_o, E_o and L_o were obtained through the application of Dubinin–Radushkevich equation) [49].

Samples	η	ABET	αs		DR
			As	Vs	Vo	Eo	Lo
	%	m2g−1	m2 g−1	cm3 g−1	cm3g−1	kJ mol−1	nm
AC-AS_KOH_1_2	17.2	1103.5	130.9	0.51	0.27	18.7	1.48
AC-Cork-KOH_1_2	22.9	986	86	0.38	0.22	25.40	0.82
AC_OS_KOH_1_2	24.8	760	78	0.31	0.27	--	1.39

Table 5. Structural parameters of the ACs prepared from cork, OS and AS, through chemical activation, with KOH at 973 K. In Table 5, η is the activation yield, A_BET is the apparent surface area, V_s is the total volume, A_s is the external surface area, (V_s and A_s were obtained using the alfa-s method proposed by Carrott et al., 1987) [48], V_o is the microporous volume, E_o is the energy of adsorption and L_o is the mean pore size (V_o, E_o and L_o were obtained through the application of Dubinin–Radushkevich equation) [49].

Samples	η	ABET	αs		DR
			As	Vs	Vo	Eo	Lo
	%	m2g−1	m2 g−1	cm3 g−1	cm3g−1	kJ mol−1	nm
AC-AS_KOH_1_2	17.2	1103.5	130.9	0.51	0.27	18.7	1.48
AC-Cork-KOH_1_2	22.9	986	86	0.38	0.22	25.40	0.82
AC_OS_KOH_1_2	24.8	760	78	0.31	0.27	--	1.39

Activated carbons were structurally characterised through nitrogen adsorption isotherms, obtained at 77 K. The isotherms are presented in Figure 2 and data obtained from their analyses are show in Table 5. The BET surface area and the total porous volume of the ACs varied from 760 to 1103.5 m² g⁻¹ and from 0.31 to 0.51 cm³ g⁻¹, respectively. The shape of the isotherms are characteristics of microporous materials, with a well-developed external surface area. Pore size dimensions obtained through the application of the Dubinin Radushkevich equation corroborate the presence of microporous materials. As expect from the isotherms analysis, AC_AS_KOH_1_2 presented the highest surface area, porous volume and also external surface area. These characteristics are good indicators of the possible performance for removing MB from the aqueous phase.

The isotherm of AC_AS_KOH_1_2 displays a more open knee, consistent with a high external surface area, larger pores size and porous volume. Mean pore size corroborates the presence of micropores. The difference between total volume and microporous volume is more pronounced in the AC_AS_KOH_1_2 and makes clear the existence of a wider porosity when compared to the other ACs. The lower activation yield obtained for AC_AS_KOH_1_2 is consistent with a more extensive activation process, which promotes the widening and merging of micropores into larger pore structures.

3.2. Methylene Blue Adsorption

3.2.1. MB Kinetic Studies

This research has practical focus on the use of locally sourced raw biomass materials, as collected from the field or after conversion into ACs, as adsorbents to remove pollutants, mainly dyes, from wastewater. After an initial characterisation, the effects of key experimental parameters on MB adsorption were investigated, and optimal adsorption conditions were established. The influence of size of the adsorbent particles, the pH of the solution, the stirring of suspensions, and temperature on MB removal was evaluated.

As show in Figure 3, MB removal decreased with increasing cork sawdust particle size, as larger particles present a smaller specific surface area available for adsorption. Removal of MB from solution increased when increasing the pH of solution, up to approximately 11. The same trend was observed for almond shells and olive stones and is consistent with findings reported in previous work using other raw biomass materials [15].

The adsorption kinetics of MB were evaluated as a function of time using raw adsorbent (particle < 0.63 mm) at pH 7 and room temperature over 1620 min, as shown in Figure 4a. Kinetic experiments were conducted at different initial MB concentrations and under two stirring speeds of the suspensions. For low MB concentrations, adsorption equilibrium was reached very quickly, confirming that the active sites present on the adsorbent surface were sufficient for all MB molecules to accommodate. With natural adsorbents, using MB concentrations of 8 and 4 mg L⁻¹, the percentage removal reached 72 and 79%, respectively, on AS, after 45 min. For an initial MB concentration of 25 mg L⁻¹, a removal efficiency of 71% was only achieved after 500 min.

It is also evident that for the same initial MB concentration (8 ppm), the equilibrium was reached more rapidly on almond shells than on olive stone. On OS, the percentage removal reached only 45.5%, after a contact time of 45 min. Nevertheless, once equilibrium was reached, the maximum amounts of MB removed were 7.2 and 13.2 mg g⁻¹, for AS and OS, respectively. To evaluate the influence of the stirring speed, the MB kinetic studies (4, 8 and 25 mg L⁻¹) were done under a stirring of 20 rpm (l—low speed) and 80 rpm (h—high speed), on AS. At a short contact time, high stirring appears to accelerate MB adsorption. After a contact time of 520 min, the effect of the stirring speed was irrelevant, and the equilibrium was reached under all conditions tested.

Adsorption kinetic studies were also conducted using AC_AS_KOH_1_2 and AC_cork_KOH_1_2. For both ACs, an initial concentration of MB (52 ppm) was higher than that employed for the raw adsorbents (maximum 25 ppm). After 45 min the removal efficiencies reached 65.6% and 79% for AC_AS_KOH_1_2 and AC_cork_KOH_1_2, respectively.

The kinetic profiles (Figure 4b) reveal a sharp initial uptake of MB across all adsorbents, reflecting the abundant availability of active surface sites at the onset of contact. As time progresses, the adsorption rate progressively declines until equilibrium is established—a pattern consistent with a mechanism governed first by external diffusion and subsequently by intraparticle diffusion. To confirm that equilibrium was fully achieved at elevated MB concentrations, and for practical experimental consistency, all subsequent experiments were carried out with a fixed contact time of 24 h.

The MB kinetic data were analysed using the Lagergren pseudo-first-order (PFO) model (Equation (4)) and the Ho–McKay pseudo-second-order (PSO) model (Equation (5)), in order to gain deeper insight into the underlying adsorption mechanisms. The PFO model is expressed as Equation (4): ln (q_e − q_t) = ln q_e − K1 × t, and the PSO model as Equation (5): (1/q_t = 1/(K₂ × q_e ²) + 1/q_e × t).

Table 6. Parameters of the application of PFO and PSO models to MB adsorption on natural adsorbents and ACs.

Adsorbents	Qmax,exp Qmax (Exp) mg g−1	Pseudo-First-Order Model			Pseudo-Second-Order Model
		Qmax1,cal1 mg g−1	K1 h−1	R2	Qmax,cal2 mg g−1	Vo mg g−1 h−1	K2 h−1	R2
AS_8ppm	7.25	2.42	0.067	0.88	7.0	20.1	0.38	0.85
AS_25ppml	19.85	3.44	0.092	0.97	19.12	60.0	0.17	0.91
AS_25ppm	19.48	3.79	0.062	0.91	18.87	99.0	0.26	0.90
OS_8ppm	13.15	7.36	0.061	0.97	12.89	12.5	0.073	0.97
AC_AS_KOH_1_2	120.6	29.6	0.065	0.991	121.95	90.1	0.0063	0.96
AC_Cork_ KOH_1_2	130.68	62.7	0.28	0.991	129.87	386.6	0.0023	0.91

and Figure 5a,b present the fitting of the PFO model to the experimental data, while Figure 6a,b shows the corresponding fitting of the PSO model, for both the raw biomass and the activated carbons (ACs).

Both models showed a strong correlation with the experimental data. Although the R² values obtained from the PFO model were marginally higher than those from the PSO model, the latter yielded calculated adsorption capacities in closer agreement with the experimentally determined values. This suggests that the PSO model more accurately captures the MB adsorption behaviour on both raw adsorbents and activated carbons, implying that chemisorption is likely the rate-limiting step governing the process. Based on the data presented in Table 6, it can be concluded that MB adsorption onto both raw adsorbents and ACs is best described by a PSO model.

The increase in the stirring speed, from 20 rpm to 80 rpm in the suspension containing MB (25 mg L⁻¹) and AS, promoted an increase in the initial adsorption velocity from 60 to 90 mg g⁻¹ h⁻¹. As reported before, the initial adsorption velocity of MB also increases with MB initial concentration, as show in Table 6. The MB adsorption is faster in ACs when compared to their respective precursors (raw adsorbents), which highlights the high surface area of the ACs available to MB. Adsorption occurs very rapidly, which is a behaviour typically associated with physisorption. The application of PSO model as previously discussed, seems contradictory, as frequently chemisorption is the dominant mechanism linked to PSO. A thorough review of the literature indicates that the PSO model is a general kinetic model that can adequately describe the physisorption processes as well and is not limited exclusively to chemisorption [50]. In fact, this information is further supported by the thermodynamic studies.

3.2.2. MB Thermodynamic Studies

The effect of temperature on the MB sorption capacity of both raw adsorbents and ACs was investigated. The results presented in Figure 7 and

Table 7. Thermodynamics parameters of MB adsorption on natural raw materials and ACs. (Initial concentration of MB was 20 mg L⁻¹ for AS and OS, 46.8 mg L⁻¹ for cork, and 73.12 mg L⁻¹ for the Acs.)

Samples	Temperature (°C)	ΔG (KJ mol−1)	ΔH (KJ mol−1)	ΔS (J mol−1)
Cork	292	−2.29	19.88	74.99
	305	−2.85
	315	−3.78
	325	−4.67
	335	−5.15
AS	292	−6.39	1.57	27.5
	305	−6.91
	315	−7.14
	325	−7.12
	335	−7.37
OS	292	−5.88	1.23	25.03
	305	−6.58
	315	−6.76
	325	−6.75
	335	−7.05
AC_cork_KOH_1_2	298	−10.33	61.66	244.08
	303	−12.83
	313	−15.22
	323	−17.49
	333	−19.76
AC_AS_KOH_1_2	298	−11.97	36.10	160.4
	303	−13.24
	313	−14.78
	323	−16.78
	333	−17.41
AC_OS_KOH_1_2	298	−10.89	14.17	92.11
	303	−15.05
	313	−16.26
	323	−16.93
	333	−14.70

demonstrate that MB adsorption increased with rising temperature, indicative of an endothermic process. The standard thermodynamic parameters—Gibbs free energy change (ΔG°), enthalpy change (ΔH°), and entropy change (ΔS°)—were determined and are summarised in Table 7.

The results presented in Figure 7 confirm that MB adsorption capacity increased with rising temperature for both raw materials and activated carbons, consistent with an endothermic process. The effect of temperature was, however, minimal for ACs, AS, and OS. In contrast, a more pronounced temperature dependence was observed for cork, where the MB removal percentage increased from 76.4 to 97.2% as the temperature rose from 22 to 62 °C. This temperature-driven enhancement may be attributed to increased mobility of MB molecules, which facilitates their diffusion from the bulk solution to the active sites within the porous structure of the adsorbents.

The negative ΔG° values obtained for all adsorbents across the temperature range of 293–333 K confirm the spontaneous nature of the adsorption process and validate its thermodynamic feasibility. Furthermore, ΔG° became increasingly negative with rising temperature, corroborating the enhancement of adsorption spontaneity at higher temperatures. Both ΔG° and ΔH° were more negative for the ACs than for their respective raw material precursors, reflecting a stronger temperature dependence in the activated carbons. This behaviour suggests that intraparticle diffusion and surface interactions play a significant role in governing the adsorption mechanism.

The data presented in Table 7 confirm that MB adsorption is an endothermic process, consistent with findings previously reported for other adsorbent materials [15]. The relatively low adsorption energy values—particularly for the raw materials (AS, OS, and cork)—suggest that the process is predominantly governed by physisorption, characterised by weak van der Waals interactions between MB molecules and the adsorbent surface. The positive ΔS° values obtained for all adsorbent–adsorbate combinations reflect an increase in disorder at the solid–liquid interface upon MB adsorption, likely associated with the release of hydration water molecules from both the dye and the adsorbent surface. The thermodynamic parameters determined in this study are of the same order of magnitude as those reported by Tran et al. [51].

3.2.3. MB Adsorption Isotherms

MB adsorption isotherms were obtained for AS, cork, and OS, as well as for their respective ACs prepared at 973 K, and are presented in Figure 8a,b. All three raw adsorbents demonstrated effective MB removal from aqueous solution, achieving removal percentages exceeding 85%, particularly at low MB concentrations. The maximum MB adsorption capacities of AS, cork, and OS were 46.42, 35.38, and 34.13 mg g⁻¹, respectively—values comparable to those reported in the literature for other raw biomass materials (Table 1) and for selected ACs (Table 2).

The conversion of biomass residues into ACs with high MB removal capacity represents a promising and innovative strategy for valorising by-products from local agricultural activities. The ACs derived from OS, AS, and cork exhibited maximum MB adsorption capacities of 244.2, 287.7, and 317.6 mg g⁻¹, respectively, as summarised in

Table 8. Results obtained by fitting the Langmuir and Freundlich equations to experimental isotherms data obtained on natural adsorbents and respective ACs, chemically produced with KOH, at 973 K (n_max—corresponds to the experimentally determined amount of MB adsorbed; n_mL—the amount of MB adsorbed, calculated using the Langmuir equation, both expressed in mg g⁻¹).

Adsorbents	nmax mg g−1	nmL mg g−1	KL L mmol−1	R2	KF mmolg−1 [L mmol−1]1/nF	nF	R2
AS	46.42	4.37	6.12	0.99	8.67	24.87	0.98
Cork	35.58	7.17	3.64	0.95	6.70	3.0	0.94
OS	34.13	3.98	0.63	0.98	6.33	2.82	0.95
AC_AS_KOH_1_2	287.69	434.78	0.66	0.99	197.42	11.10	0.95
AC_Cork_KOH_1_2	317.58	166.67	1.94	0.99	180.47	12.58	0.86
AC_OS_KOH_1_2	244.20	102.05	2.51	0.98	88.90	2.37	0.99

and illustrated in Figure 8b. Among these, AC_OS_KOH_1_2 displayed the lowest BET surface area and micropore volume, which can be attributed to the greater hardness and structural compactness of olive stone relative to almond shell and cork—physical characteristics that hinder chemical activation under identical experimental conditions. The reduced micropore volume available for MB uptake is consistent with the lower maximum adsorption capacity observed for this material.

MB adsorption isotherms were analysed by fitting the experimental data to the Langmuir and Freundlich models, with the corresponding graphical representations shown in Figure 9a–d for both raw adsorbents and ACs. The fitting parameters derived from both models are summarised in Table 8. The Langmuir model provided a superior fit to the experimental data, implying that adsorption occurs on a finite number of homogeneous surface sites of equal energy, leading to the formation of a single adsorbate monolayer on the external surface of the adsorbent, with no further adsorption taking place once all sites are fully occupied. These assumptions are characteristic of a monolayer adsorption mechanism and are conventionally associated with chemisorption. However, as discussed in the context of the kinetic modelling results, physisorption cannot be excluded and may in fact be the dominant mechanism, particularly given the low adsorption energy values and the rapid adsorption kinetics observed.

However, the relatively low enthalpy values recorded across all adsorbents (1.2–61.7 kJ mol⁻¹) are indicative of physical adsorption, suggesting that physisorption contributes significantly to the overall MB uptake [33]. Taken together, these findings support the interpretation that the PSO model can adequately describe adsorption systems in which physical and chemical interactions occur concurrently. Overall, the results demonstrate that both raw biomasses and their corresponding ACs are effective adsorbents for MB removal from aqueous solution. Chemical activation markedly enhances adsorption performance through the development of well-defined surface area and microporosity, while the underlying adsorption mechanism appears to involve a combination of physisorption and chemisorption, the relative contribution of which may vary depending on the adsorbent and the experimental conditions.

4. Conclusions

This study demonstrated that agricultural residues from the Alentejo region—namely cork waste, olive stones, and almond shells—can be successfully valorised as sustainable and low-cost adsorbents for methylene blue (MB) removal from aqueous solution. These locally available materials show considerable potential as alternatives to conventional adsorbents such as activated carbons, which remain comparatively expensive. Nevertheless, chemical activation with KOH significantly enhanced the textural properties of all three precursors, yielding activated carbons with well-developed surface areas and markedly improved adsorption capacities.

Batch adsorption performance was optimised under alkaline pH conditions, elevated temperatures, extended contact times, higher agitation speeds, and reduced particle sizes. The activated carbons exhibited substantially higher maximum adsorption capacities (244.2–317.6 mg g⁻¹) than their untreated biomass precursors (34.1–46.4 mg g⁻¹), reflecting the critical role of microporosity development in enhancing MB uptake. Equilibrium data were best described by the Langmuir model, while adsorption kinetics conformed to the pseudo-second-order (PSO) model—a framework applicable to systems where physisorption and chemisorption occur concurrently. Thermodynamic analysis confirmed that the adsorption process was spontaneous and endothermic across the temperature range studied, with relatively low enthalpy values indicating a predominantly physical adsorption mechanism. At low MB concentrations in particular, the untreated biomass materials may serve as effective, inexpensive, and locally sourced alternatives to conventional activated carbons, offering a promising route for the valorisation of regional agricultural by-products.

The adsorption results suggest that MB uptake is governed by a combination of external mass transfer, intraparticle diffusion, and predominantly physical interactions between MB molecules and the adsorbent surface. The high adsorption capacities and rapid kinetics observed for the activated carbons further underscore their potential as effective and low-cost adsorbents for wastewater treatment applications. Nevertheless, several aspects warrant further investigation before practical implementation can be realised, including performance evaluation in real wastewater matrices, behaviour in multicomponent systems, adsorbent regeneration and reusability, and operation under continuous-flow conditions. Collectively, the findings of this study demonstrate that the valorisation of cork waste, olive stones, and almond shells as low-cost adsorbents for water treatment is fully aligned with circular economy and sustainability principles, contributing meaningfully to the reduction in environmental impacts while supporting regional socio-economic development through the productive use of locally available agricultural by-products.