Comparison of Granular and Pellet Olive Stone-Based Activated Carbon in Adsorption-Based Post-Combustion CO₂ Capture

Abstract

In the present study, we evaluate the CO₂ uptake capacities of four activated carbons (ACs) obtained from olive stones. Two of the samples were generated using a chemical process utilizing phosphoric acid, thereafter undergoing carbonization in a nitrogen steam, yielding both granular and pellet forms, designated CH-ACG-410 and CH-ACP-410, respectively. The third sample, labeled CO-ACG-390, was produced by carbonization under a steam-nitrogen flow, while the fourth sample, designated PH-ACG-850, was prepared by a physical process involving water vapor at 850 °C. The carbon materials obtained in granular and pellet form were subjected to textural characterization using N₂ and CO₂ adsorption isotherms at 77 K and 273 K, respectively. Additionally, surface chemistry was analyzed using FTIR, Boehm titration, and TPD-MS. The materials were also assessed for CO₂ adsorption in a binary mixture consisting of 10% CO₂ and 90% N₂ at two temperatures, 25 and 50 °C. The results demonstrated that all prepared adsorbents exhibited competitive CO₂ capture performance, with the CH-ACP-410 sample (pellet form), showing the highest adsorption capacities, achieving approximately 4.6 wt. % at 25 °C and 2.2 wt. % at 50 °C. This superior behavior can be attributed to the conditioning methods applied to this material, which significantly influenced its textural properties and, consequently, its CO₂ adsorption capability.

1. Introduction

Finding an effective solution to climate change represents one of the primary challenges of our era, driven primarily by increasing concentrations of greenhouse gases in the atmosphere. Carbon dioxide (CO₂) is the most critical contributor to global warming. Since the start of the Industrial Revolution in the mid-19th century, 1850s, carbon dioxide levels have risen significantly due to the combustion of fossil fuels and biomass [1]. To mitigate “hard-to-abate” CO₂ emissions from industrial processes, one of the only available strategies is carbon capture and storage (CCS). Among the various CCS technologies, adsorption stands out as one of the most practical, effective, and cost-efficient methods for CO₂ capture [2]. Consequently, significant research has focused on a wide array of solid adsorbents, including zeolites [3], activated aluminas [4], carbon-based materials [5,6], silicas [7], and amine-based adsorbents [8,9,10]. More recently, advanced porous materials such as metal–organic frameworks (MOFs) [11,12,13], zeolitic imidazole frameworks (ZIFs) [14], and porous organic polymers (POPs) [15,16,17] have been explored for their tunable properties.

However, composite materials and waste-derived sorbents represent important avenues for cost-effective and sustainable capture [18]. Among these, activated carbons offer a series of significant advantages for large-scale CO₂ capture, including high adsorption capacity, fast adsorption and desorption kinetics, chemical stability, availability, and low regeneration energy [19].

Furthermore, activated carbons can be derived from waste biomass—a low-cost, renewable, and globally abundant carbon source that aligns with the principles of green synthesis and a circular economy. Examples include olive stones [20], almond shells [21], coconut shells [22], and sewage sludge [23].

Usually direct pyrolysis and activation are used to convert these precursors into porous carbon compounds. Activation, necessary for creating the pore structure, is typically performed using one of three methods: chemical, physical, or a combination of both [24]. Chemical activation involves subjecting the precursor material to a dehydrating agent, such as phosphoric acid, zinc chloride, or potassium hydroxide, at temperatures between 400 and 1000 °C, subsequently followed by the elimination of the dehydrating agent through multiple washes with deionized water [25].

Physical activation typically occurs in a two-step process: first, the carbonaceous raw materials are carbonized in an inert atmosphere, usually below 800 °C; subsequently, the resultant char undergoes activation through partial gasification at temperatures ranging from 700 to 1000 °C, utilizing carbon dioxide, steam, or a combination of both [26].

This study focuses on synthesizing activated carbons from local biomass residues, specifically olive stones, to serve as CO₂ adsorbents in post-combustion conditions. The primary objective is to optimize carbon’s porosity through various activation processes and conditions to maximize CO₂ adsorption capacity while also aiming to minimize setup costs and simplify synthesis. To achieve this, we developed a series of activated carbons in granular and pellet forms using three different methods: chemical activation, physical activation, and a combined process. This approach allows us to investigate their textural and structural properties while comparing their kinetics and CO₂ uptake capacities.

2. Materials and Methods

2.1. Preparation of Activated Carbon Samples

Olive stones (OSs), a by-product of Tunisian olive oil factories, were used in this study as a precursor to prepare activated carbon sorbents via different activation processes, as described in our lab [27,28].

Before use, olive stones were washed several times with deionized water to remove impurities, and then dried at room temperature to obtain grains approximately 1–3 mm in size. Four activated carbons were prepared in this study, as shown in Figure 1: two samples were chemically activated with H₃PO₄, one in the form of granular material named CH-ACG-410 and the second in the form of pellet noted CH-ACP-410. In this process, the precursor mass was mixed with a 50% (w/w) phosphoric acid solution using a 1:3 weight ratio. The chemically impregnated olive stones suspension was heated to 110 °C and maintained at that temperature for nine hours. The resulting filtrate was subjected to a two-step carbonization process, first at 170 °C for 30 min, followed by a second stage at 410 °C for 2.5 h. The adsorbent was extensively washed with deionized water to remove all residual H₃PO₄, continuing until the wash solution reached a pH between 6 and 7 [29]. The CH-ACP-410 was prepared similarly by pressing the impregnated raw material into cylindrical pellets before calcination [28]. The combined activated carbon, noted CO-ACG-390, was synthesized by impregnating olive stones with H₃PO₄ under the same conditions as the first product. The dry solid was thermally activated in a horizontal tube furnace under a nitrogen and water vapor atmosphere, using a two-step protocol: 30 min at 170 °C, followed by 2.5 h at 390 °C. The fourth sample, named PH-ACG-850, was physically activated with steam and nitrogen in two steps: initial carbonization of the precursor in a fixed-bed reactor at 600 °C (1 h, N₂ flow), followed by steam activation of the produced char at 850 °C for 6 h (70 vol.% H₂O in N₂, total flow rate of 10 NL/h) [30].

2.2. Characterization of Samples

2.2.1. Pore Structure Characterization

The textural properties of the synthesized activated carbons were determined via gas physisorption analysis. Nitrogen adsorption isotherms were measured at −196 °C using an ASAP 2010 analyzer (Micromeritics Instrument Corp., Norcross, GA, USA), while carbon dioxide adsorption was assessed at 0 °C with a TriStar 3000 instrument (Micromeritics Instrument Corp., Norcross, GA, USA). Prior to analysis, each sample was subjected to overnight degassing at 100 °C under vacuum. The specific surface area (S_BET) was calculated by applying the Brunauer–Emmett–Teller (BET) model to the N₂ adsorption data in the relative pressure (P/P₀) interval of 0.1–0.3 [31]. The total pore volume was determined from the volume of N₂ adsorbed at a relative pressure (P/P₀) of 0.99. Narrow micropore volumes (W₀), corresponding to pores narrower than 0.7 nm, were derived from CO₂ adsorption isotherms by applying the Dubinin–Radushkevich (DR) equation [32]. The average micropore width (L₀) was calculated via the Stoeckli–Ballerini relation [33]. Furthermore, pore size distributions (PSD) were derived from the CO₂ isotherms using the non-local density functional theory (NLDFT).

2.2.2. Chemical Surface Characterization

The Point of Zero Charge (pH_PZC), which reflects a material’s acid-base surface properties, was determined for all samples via a mass titration method adapted from Savova et al. [34]. The surface chemistry of the activated carbon samples was characterized by determining the identity and concentration of functional groups via Fourier Transform Infrared Spectroscopy (FTIR), Boehm titration, and Temperature Programmed Desorption (TPD) analysis. FTIR spectra were performed on a Perkin Elmer 8400 instrument (PerkinElmer, Inc., Waltham, MA, USA), covering the spectral range of 4000–400 cm⁻¹ at a resolution of 4 cm⁻¹. For analysis, sample pellets were prepared by compressing a mixture of finely ground carbon (1 wt%) with KBr. However, the Boehm titration method was adapted by Heidari et al. (2014) [35], and for TPD analysis, approximately 20 mg of the activated carbon sample was loaded into a platinum crucible and heated from ambient temperature to 1000 °C under an argon flow using a Setaram TGA92 thermogravimetric analyzer coupled with a mass spectrometer. During heating, oxygen-containing surface groups decompose, evolving CO and CO₂ gases. Carboxyl groups, lactones, and anhydrides primarily release CO₂, whereas CO is generated from phenols, carbonyls, quinones, pyrones, and anhydrides. The mass spectrometer tracked these emissions by monitoring the mass-to-charge (m/z) ratios of 28 and 44 for CO and CO₂, respectively [36].

2.3. CO₂ Adsorption Measurements

The CO₂ adsorption behavior of the prepared materials under conditions representative of post-combustion capture was evaluated using a Mettler Toledo TGA/DSC1 thermogravimetric analyzer. The CO₂ uptake was calculated from the mass evolution profiles recorded during the samples’ exposure to a 10% CO₂ (balance N₂) gas stream. Before performing the binary-mixture adsorption tests, the samples were heated at 100 °C for 1 h under pure N₂ at a flow rate of 100 cm³·min⁻¹ to dry them and remove any residual gases. Subsequently, the samples were allowed to cool to the adsorption test temperature, either 25 or 50 °C. Once temperature stabilization was achieved, the gas flow was switched from pure N₂ to a 10% CO₂ mixture (balance N₂), maintaining the same flow rate. The sample mass increased during CO₂ adsorption until equilibrium was reached. The total mass gain, expressed as a weight percentage, is denoted as q_N2+CO2. Further details on the calculation of CO₂ capacity are available elsewhere [37]. Regeneration was then performed under a pure N₂ flow at the adsorption temperature.

2.4. Kinetic Studies

2.4.1. Adsorption Rate

The CO₂ adsorption kinetics of activated carbon (AC) samples are essential for evaluating dynamic performance and for understanding how textural and surface chemical properties influence uptake capacity. In this study, two kinetic models were employed to analyze the CO₂ adsorption kinetics of the AC samples: the pseudo-first order model and the Avrami fractional-order model. These models were selected for their established effectiveness in characterizing the adsorption kinetics of physical adsorbents [38].

• Pseudo-first order model

The pseudo-first-order model, also known as the linear driving force (LDF) model [39], is widely used in adsorption kinetics. It operates on the premise that the adsorption rate is directly proportional to the concentration of available adsorption sites [40,41], as expressed by the following Equation (1):

$${\displaystyle \frac{dq}{dt}}=k_1\left(q_e-q_t\right)$$

(1)

Here, q_e and q_t represent the adsorption capacities at equilibrium and at a given time t, respectively, while k₁ is the pseudo-first-order rate constant. Given the boundary conditions q_t = 0 at t = 0 and q_t = q_e at t =∞, Equation (1) integrates to obtain Equation (2):

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

(2)

The pseudo-first-order model is suited for predicting CO₂ physisorption on solid sorbents, as it describes a reversible adsorbent–adsorbate interaction [42].

• Avrami’s fractional model

Originally developed to model phase transitions and crystal growth rates in materials [43], the Avrami equation has recently been adapted as an effective method for predicting CO₂ adsorption kinetics on biomass-derived activated carbons [38], modified activated carbon [44] and carbon hydrochar [45]. Its general form is given by Equation (3):

$${\displaystyle \frac{dq}{dt}}=k_A^{n_A} t^{n_A-1}\left(q_e-q_t\right)$$

(3)

In this model, k_A (min⁻¹) is the rate constant and n_A is the Avrami exponent, a parameter related to the adsorption mechanism that helps identify interactions during the process [32]. An exponent n_A = 1, it indicates homogeneous adsorption [46], while values of 2, 3 and 4 signify one-, two- and three-dimensional growth, respectively [38,47]. The integrated form of Equation (3), given by Equation (4), is as follows:

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

(4)

• Validation of the kinetic model

In this study, the standard deviation was used to assess the divergence between the measured data and model predictions. Furthermore, the sum of squared errors (SSE) and the coefficient of determination (R²) were calculated using Equations (5) and (6):

$$SSE (\%)= \sqrt{{\displaystyle \frac{{\displaystyle \sum {(q_{t,\exp } - q_{t,pred})}^2}}{N}}}\times 100$$

(5)

$$R^2= 1 -({\displaystyle \frac{{\displaystyle \sum _{i=1}^N{(q_{t,\exp } - q_{t,pred})}^2}}{{\displaystyle \sum _{i=1}^N{(q_{t,\exp } - \overline{q_{t,\exp }})}^2}}} ) ({\displaystyle \frac{N-1}{N-P}})$$

(6)

Here q_t,exp and q_t,pred are the experimentally measured and model-predicted adsorption capacities, respectively; N is the number of experimental data points, and p denotes the number of parameters in the model [48].

2.4.2. Adsorption Mechanism

Adsorption kinetics is crucial for assessing an adsorbent’s performance. However, many kinetic models do not explicitly differentiate the underlying adsorption mechanisms; therefore, it is essential, especially for practical applications, to consider models that account for diffusion mechanisms and identify the process’s rate-limiting step. The intra-particle diffusion model, proposed by Weber and Morris [49] and based on Fick’s second law, is given by Equation (7). It identifies the consecutive mass transfer stages during adsorption.

$$q_t = k_{id} t^{1/2}+ C$$

(7)

Here k_id is the intra-particle diffusion rate constant, and C is a constant related to the boundary-layer thickness. A higher C value indicates a more pronounced boundary-layer effect [50,51]. This model stipulates that diffusion is involved in adsorption when a linear plot of q_t versus t^1/2 is obtained. Furthermore, intra-particle diffusion is identified as the sole rate-determining step only when the corresponding linear plot intersects the origin. Nevertheless, multilinearity in the q_t − t^1/2 plot is commonly observed, indicating an adsorption process that proceeds in distinct stages.

Most frequently, the Weber–Morris model identifies three sequential steps. The first corresponds to external diffusion (boundary layer diffusion), the second to the gradual adsorption stage (i.e., intra-particle diffusion proper), and the third to the final equilibrium stage [42]. However, the third step is generally considered rapid and not rate-determining [52]. In each stage, the slope of the linear segment represents the rate parameter, k_id,i (where i denotes the stage number). The rate-limiting corresponds to the process stage with the lowest slope [42].

3. Results and Discussion

3.1. Textural Characterization

The four prepared carbon samples were analyzed using N₂ and CO₂ adsorption at −196 °C and 0 °C, respectively (see Figure 2). Given that CO₂ adsorption occurs at much higher temperatures than N₂ diffusion restrictions are reduced, and the CO₂ molecules are easily transported throughout the ultra-micropores (<0.7 nm), complementing the information provided by the N₂ isotherms [48].

The textural properties derived from the N₂ and CO₂ adsorption–desorption isotherms for all ACs are presented in

Table 1. Textural properties and CO₂ adsorption capacity of activated carbon samples.

	N2 Adsorption at −196 °C		CO2 Adsorption at 0 °C
Samples	SBET (m2·g−1)	VT (cm3·g−1)	Smic (m2·g−1)	L (nm)	E0 (Kj·mol−1)	W0 (cm3·g−1)	CO2 Uptake (mmol·g−1)
							0.15 bar	1 bar
CH-ACP-410	1288	0.49	936	0.69	27.14	0.32	1.53	4.60
CH-ACG-410	978	0.34	685	0.67	27.54	0.23	1.15	3.61
CO-ACG-390	1219.88	0.51	678	0.74	26.09	0.25	1.05	3.90
PH-ACG-850	865.52	0.44	768	0.60	29.44	0.22	1.43	3.93

.

Figure 2a shows that the three carbonaceous materials (CH-ACP-410, CH-ACG-410, and CO-ACG-390) exhibit completely reversible isotherms corresponding to type I under the IUPAC classification. These samples exhibit strong monolayer adsorption at low relative pressure (P/P₀ < 0.3), followed by a horizontal plateau in the 0.3–0.9 range due to saturation of active sites. This type of isotherm is characteristic of essentially microporous materials. The common factor among the first three samples is that they are all activated with phosphoric acid and carbonized at approximately 400 °C. Yorgun and Yildiz reported an 82% microporous volume for activated carbons derived from Paulownia wood through chemical activation with H₃PO₄; they explained this by noting that 400 °C is the optimal carbonization temperature for chemical activation. These micropores are also formed during the activation step due to the release of heteroatoms, such as H and O, from phosphoric acid [53].

For the physically activated carbon PH-ACG-850, the nitrogen adsorption–desorption isotherm is of type IV, characterized by strong monolayer adsorption (P/P₀: 0–0.3), followed by a plateau corresponding to the progressive filling of the pores (P/P₀: 0.3–1). The sharp rise in the adsorption branch and the presence of a hysteresis loop at high relative pressures indicate the simultaneous presence of micro and mesopores. It is because the PH-ACG-850 sample is physically activated by water vapor. Thus, the nature of the activating agent influences the material’s pore distribution. For example, the use of carbon dioxide as an oxidizing agent promotes the development of microporosity, while water vapor favors larger pore dimensions. It contributes to the formation of mesopores in our case [54].

Figure 2b displays the experimental CO₂ adsorption isotherms obtained at 0 °C. As observed, all the carbon materials present almost the same shape. Their adsorption curves confirm type I behavior, indicating a strong adsorbate-adsorbent interaction [36].

Figure 3a,b display the pore size distribution (PSD) and cumulative pore volume as determined by CO₂ adsorption at 0 °C.

The evolution of the pore-size distribution of the prepared adsorbents indicates that all samples exhibit a broader distribution spanning 0.4–0.9 nm. Sample CH-ACP-410 exhibits the largest micropore volume in the narrowest range between 0.4 and 0.7 nm. A large volume in the ultra-micropore range (smaller than 0.7 nm) is highly desirable for CO₂ adsorption under post-combustion conditions [28]. These results suggest that the pelletizing process reduces interparticle space and decreases mesopore volume, generating narrower micropores.

The CO₂ uptakes at atmospheric pressure (1 bar) and sub-atmospheric pressure (0.15 bar), determined from the CO₂ adsorption isotherms at 0 °C for the prepared adsorbents, are also reported in Table 1. These pressures were chosen from the literature because they mimic typical conditions for post-combustion CO₂ removal from flue gas.

As shown in Table 1, the pelletized AC (CH-ACP-410) exhibits the highest CO₂ adsorption capacity, followed by the physically activated sample (PH-ACG-850), at both low (0.15 bar CO₂) and atmospheric pressure. Specifically, the CO₂ uptake for these two samples is approximately 1.53 and 1.43 mmol·g⁻¹ at 0.15 bar, and 4.60 and 3.93 mmol·g⁻¹ at 1 bar, respectively. These values are highly competitive compared to those reported in the literature [37]. Notably, the pelletized sample outperforms its granular counterpart despite being activated under similar conditions. As previously discussed, this enhanced performance can be attributed to the pelletizing process, which reduces interparticle space and decreases mesopore volume, ultimately generating a greater volume of narrower micropores. A correlation between the porosity data and CO₂ uptake reveals that, at both atmospheric and sub-atmospheric pressures, adsorption capacity is more strongly influenced by micropore surface area than by total surface area. This finding aligns with the observation that the top-performing samples, CH-ACP-410 and PH-ACG-850, possess narrower micropore volumes than the other two adsorbents. The importance of these narrow micropores is further underscored by the study of Xin Liu, which demonstrated that CO₂ uptake at 1 bar is directly proportional to the volume of micropores smaller than 1 nm [55].

3.2. Determination of Surface Oxygen Functional Groups

Given that CO₂ adsorption on carbon materials depends on textural properties, surface chemistry, and specific adsorbate-adsorbent interactions, this section examines the correlation between the chemical characteristics of activated carbon and its CO₂ adsorption capacity.

Table 2. Point of Zero Charge (pH_PZC) and Surface Functional Group Concentrations as determined by Boehm titrations.

Sample	pHPZC	Total Acidity (meq/g)	Total Basicity (meq/g)
CH-ACP-410	2.4	2.9	0.05
CH-ACG-410	2.68	3.13	0
CO-ACG-390	5.05	1.28	0.25
PH-ACG-850	7.52	0.43	1.86

summarizes the surface chemical properties of the prepared adsorbents, including total acidity and basicity, obtained from Boehm titration and the corresponding Point of Zero Charge for each material. From this table, the CH-ACP-410 and CH-ACG-410 samples, activated with phosphoric acid, show pH_Pzc values of 2.4 and 2.7, respectively, indicating a strongly acidic character. This is due to H₃PO₄ being a strong oxidant of carbon atom surfaces [56]. For the CO-ACG-390 sample, the pH_Pzc is higher at 5.06, but it still exhibits acidic character. This value can be attributed to the utilization of both phosphoric acid and water vapor as activating agents. However, the physically activated sample PH-ACG-850 has a pH_Pzc very close to neutral. It shows that the sample behaves like an amphoteric compound. As expected, the observed pH_pzc values correlate with the distribution of acidic and basic functional groups across the various carbon materials, with the total quantity of acidic functional groups decreasing as pH_pzc values increase. Thus, the results show relatively firm acidity up to 3 meq/g, for the two carbon materials CH-ACP-410 and CH-ACG-410, with almost negligible basicity. More acidic sites indicate a material rich in oxygenated groups.

In contrast, for sample PH-ACG-850, which was activated by water vapor, the total quantity of basic groups (1.85 meq/g) exceeds that of acidic groups. The pH_Pzc value around 7 confirms this result. This basic character can be partially attributed to the presence of free oxygen in the form of Lewis bases, a feature often associated with the π- electron system of the graphite layers [57].

The surface functional groups of the adsorbents were further characterized by FTIR spectroscopy; the resulting spectra for the prepared activated carbons are presented in Figure 4.

Examining the figure above, we observe that CH-ACP-410, CH-ACG-410 and CO-ACG-390 display a broad, intense absorption band between 3600 and 3200 cm⁻¹, centered at approximately 3427 cm⁻¹. This band corresponds to the O-H stretching vibration typically found in hydroxyl-containing compounds, and its broadening indicates extensive hydrogen bonding, suggesting a high degree of molecular association [58]. This indicates the presence of hydroxyl groups, such as carboxyl, phenolic, or alcoholic, in the carbon samples. This band is almost nonexistent for physically activated carbon. The presence of these hydroxyl (OH) groups, as evidenced by the very intense bands in the two samples CH-ACG-410 and CH-ACP-410, is responsible for their strongly acidic character. Furthermore, the spectra of the two chemically activated samples, CH-ACP-410 and CH-ACG-410, exhibit aliphatic C–H stretching vibrations. The asymmetric and symmetric stretches of methylene (>CH₂) groups are observed at 2924 cm⁻¹ and 2851 cm⁻¹, respectively [59], indicating the presence of aliphatic structures. The two samples CH-ACP-410 and CH-ACG-410 also showed a broad band in the primary fingerprint region (1622–1543 cm⁻¹), corresponding to aromatic C=C ring-stretching vibrations. The presence of primary alcohols is corroborated by C-O stretching bands in the 1051–1011 cm⁻¹ range. According to the results, the two ACs have very similar spectra. Furthermore, the presence of acidic oxides, as mentioned before, helps explain the acidic character of these samples and their relatively high oxygen content. As is well known, the oxygen-containing surface functionalities of carbonaceous materials play a key role in CO₂ adsorption, especially at low relative pressure. For all samples, the bands observed in the 1000–1350 cm⁻¹ region correspond to C-O stretching vibrations. The aromatic character of the carbon materials is supported by the presence of aromatic C-H stretching modes between 1000 and 400 cm⁻¹.

Temperature-programmed desorption (TPD) tests were utilized to monitor the desorption of surface oxygen species, as evidenced by the CO and CO₂ profiles obtained over the applied temperature range. The nature of the surface oxygen functionalities can be inferred from the type of CO_x groups evolved and their corresponding desorption temperatures. Specifically, CO₂ evolution is typically associated with the decomposition of carboxyl or anhydride groups, whereas CO release originates from carbonyl, ether, and quinone groups [60]. Figure 5 displays the CO₂ and CO evolution profiles obtained from the prepared carbon samples during heating at 15 °C min⁻¹ under an argon flow of 50 cm³ min⁻¹. As shown in this figure, the AC pellet (CH-ACP-410) exhibits two maxima in the CO₂ profile at approximately 400 and 350 °C. For the CH-ACG-410, three small peaks are observed at around 300 °C, 630 °C, and 835 °C. For the CO-ACG-390 sample, CO₂ desorption shows two peaks: an intense peak at 322 °C and a small peak at 860 °C. However, the PH-ACG-850 sample exhibits a broad shoulder from 60 to 470 °C, followed by a continuous increase in peak starting at 568 °C, with a small shoulder at 690 °C. CO₂ released at lower temperatures generally corresponds to the decomposition of carboxyl groups, whereas CO₂ evolved at higher temperatures is linked to the decomposition of lactone or anhydride functionalities [61,62,63]. However, the CO desorption curve is characterized by two peaks at 661 and 826 °C for the CH-ACP-410 sample, a continuous increase starting from 400 °C, with a shoulder at 635 °C and an intense peak at 836 °C for CH-ACG-410; it starts from 60 °C for CO-ACG-390, with a shoulder at 347 °C and an intense peak at 895 °C, which begins from 600 °C and continues to increase up to 1000 °C for PH-ACG-850. The high-temperature CO desorption observed for all treated carbonaceous materials indicates the presence of stable oxygen-containing functional groups with basic or neutral character, such as ethers, phenols, and quinones [60,62,64].

Table 3. Quantities of CO and CO₂ evolved during TPD experiments.

Samples	CO (µmol/g)	CO2 (µmol/g)	O (µmol/g)	CO/CO2
CH-ACP-410	2600	1600	3000	1.6
CH-ACG-410	3800	700	3000	5.1
CO-ACG-390	500	400	1000	1.3
PH-ACG-850	5200	2600	5000	2

presents the quantities of CO and CO₂ evolved, determined by integrating the areas under the TPD profiles, and the resulting oxygen content.

Table 3 shows that the amount of oxygen functionalities that release CO is significantly greater than those that release CO₂. However, this result is inconsistent with the Boehm titration and FTIR spectra results. This occurs because, during heating, oxygen-containing functional groups—such as adjacent carboxyl groups, lactols, and cyclic lactones—decompose to evolve CO rather than CO₂. In addition, within micro-porous carbon, a CO₂ molecule can collide with the pore walls and be transformed into two CO molecules. On the other hand, the data in this table indicate that chemical activation of AC in pellet form yields higher surface oxygen content than activation of the granular form. The presence of this oxygen could create additional active sites on the AC pellet, thereby enhancing its CO₂ adsorption capacity.

3.3. CO₂ Capture from a CO₂/N₂ Binary Mixture

The adsorption–desorption kinetics were evaluated over a 60 min period to quantify the amount of CO₂ adsorbed from the binary mixture. Equilibrium time and adsorption capacity are the most economically significant factors for CO₂ adsorption in post-combustion systems.

Figure 6 displays the full adsorption experiments for all samples, plotting the total mass uptake as a function of time. Measurements were performed at atmospheric pressure with a feed consisting of 10% CO₂ (balance N₂) at temperatures of 25 and 50 °C.

As shown in the figures, the samples CO-ACG-390 and PH-ACG-850 exhibit rapid adsorption and desorption kinetics at both 25 °C and 50 °C. The majority of uptake occurs within the first few minutes of exposure to either pure N₂ or the mixed gas (10% CO₂/90% N₂). This behavior is attributed to the greater availability of active sites on the adsorbent surface during the initial stage, which diminishes over time as adsorption progresses.

In contrast, the curves for the two chemically activated samples, CH-ACP-410 and CH-ACG-410, show a slow approach to equilibrium at 25 °C, likely due to diffusion resistance within the pores at lower temperatures. This kinetic process becomes faster for both samples at 50 °C. This is because higher temperatures enhance the diffusion rate of adsorbate molecules through the external boundary layer and within the pores.

Fast adsorption and desorption kinetics are essential for the viability of adsorbents in CO₂ capture, particularly under post-combustion conditions, where rapid oscillation cycles are necessary to maximize the processed flow rate in industrial applications [28,46].

Figure 7 compares, on a mass basis, the equilibrium adsorption capacities of the samples in both 100% N₂ and a mixture consisting of 10% CO₂ (balance N₂).

The adsorption capacity of the samples decreased with temperature, declining from 25 to 50 °C, a trend consistent with an exothermic process. This enhanced mass transfer rate consequently reduces the residence time of adsorbate molecules at the active sites on the activated carbon (AC) [65]. For pure N₂, activated carbon pellets exhibit the highest capacity at both 25 and 50 °C. It is attributed to its large specific surface area and significant pore volume. However, the greater mass uptake of CO₂ from the binary mixture (10% CO₂ + 90% N₂) is related to its better, narrower microporosity development. As is well known, narrow micropores are crucial for CO₂ adsorption at the low CO₂ partial pressures typical of post-combustion applications [37].

On the other hand, a correlation analysis between chemical characteristics and CO₂ uptake reveals that surface chemistry has no significant influence on adsorption performance. This is due to the dominant physical adsorption mechanism identified in our experiments, in which CO₂ capture is governed primarily by weak van der Waals forces and pore-filling effects rather than by specific chemical interactions.

Although numerous carbon-based materials have been explored in recent years, biomass-derived carbons stand out as particularly attractive due to their abundance, low cost, and sustainability. To benchmark the performance of our olive stone-derived activated carbons, we compared their CO₂ adsorption capacity with data from the literature under similar post-combustion conditions.

Table 4. A comparison of the low- pressure CO₂ adsorption performances of porous carbon adsorbents derived from biomass.

Carbon Precursor	Sample	Activation Method	Post-Combustion Condition	CO2 Uptake	References
Pine sawdust	IH3	Carbonization + CO2	(10% CO2 + 90% N2) (P = 1 atm; T = 50 °C)	2.1 wt%	[37]
Olive stones	GKOSA40	Carbonization + CO2	(15% CO2 + 85% N2) (P = 1 atm; T = 40 °C)	2.4 wt%	[66]
Almond shell	AA740	Carbonization + CO2	(15% CO2 + 85% N2) (P = 1 atm; T = 25 °C)	5.1 wt%	[21]
Spent coffee grounds	CNHA29	KOH	(10% CO2 + 90% N2) (P = 1 atm; T = 50 °C)	2.8 wt%	[67]
Commercial AC	Commercial AC	Carbonization	(20% CO2 + 80% N2) (P = 1 atm; T = 25 °C)	3.1 wt%	[68]
			(20% CO2 + 80% N2) (P = 1 atm; T = 50 °C)	1.9 wt%
Coconut shell	Optimized AC	Carbonization	(20% CO2 + 80% N2) (P = 1 atm; T = 25 °C)	2.7 wt%
			(20% CO2 + 80% N2) (P = 1 atm; T = 50 °C)	1.5 wt%
Yellow mombin stones	YMKOH	Carbonization + KOH	(10% CO2 + 90%N2) (P = 1 atm; T = 40 °C)	4.2 wt%	[69]
			(50%CO2 + 50%N2) (P = 1 atm; T = 40 °C)	6.3 wt%
Olive stones	AC_KOH	Carbonization + KOH	(10%CO2 + 90%N2) (P = 1 atm; T = 25 °C)	4.21 wt%	[70]
	AC_K2CO3	Carbonization + K2CO3	(10%CO2 + 90%N2) (P = 1 atm; T = 25 °C)	4.13 wt%
Olive stones	CH-ACP-410	H3PO4	(10%CO2 + 90%N2) (P = 1 atm; T = 25 °C)	4.6 wt%	This study
			(10%CO2 + 90%N2) (P = 1 atm; T = 50 °C)	2.2 wt%

summarizes the CO₂ adsorption performance of various porous carbon materials derived from different precursors—such as olive stones, coconut shell, and coffee grounds—under low relative pressures.

As pointed out before and shown in Table 4, the nature of the precursor and the activation process are determining parameters for the CO₂ adsorption capacities of carbon materials. By analyzing this table, one can note that the CO₂ adsorption capacities reported in this study may not be the highest in the literature, but they are among the greatest for biomass-derived carbon. The pelletized carbon CH-ACP-410, identified as the optimal adsorbent in this study, shows at 25 °C, a CO₂ uptake of 4.6 wt.%, which is comparable to the values of 4.21 and 4.13 wt.% reported by Moussa et al. [70] for olive stone-based activated carbons prepared with KOH and K₂CO₃, respectively. At 50 °C, the adsorption capacity decreased to approximately 2.2 wt.%, closely aligning with the ~2.1 wt.% uptake reported by Plaza et al. [37] for CO₂-activated pine sawdust pellets under identical conditions.

The evaluation of adsorption kinetics was conducted by monitoring the dynamic mass uptake of the samples exposed to the gas mixture, using the mass at the end of the conditioning step in N₂ at 50 °C as a reference (see Figure 8). This incremental mass uptake corresponds to the adsorption of CO₂ from the mixture.

Figure 8a reports the predicted CO₂ adsorption by fitting the experimental data to the two kinetic models: pseudo-first order and Avrami’s fractional models.

As shown in Figure 8a, the experimental curves for all samples closely followed both the pseudo-first-order and Avrami fractional kinetic models, accurately in the first seconds (below 100 s) and at equilibrium (above 10 min). At the average time (between 0.6 and 10 min), the experimental data points significantly deviate from the fitted straight line. These findings imply that physisorption governs the initial adsorption of CO₂ from the binary mixture onto the adsorbent surface. However, the Avrami equation provides a more accurate description of CO₂ adsorption kinetics across all stages on the studied carbon adsorbents, as it accounts for complex reaction pathways [71,72,73]. This model has been successfully applied to a wide range of CO₂ adsorbents in previous studies [73,74,75]. Therefore, the Avrami model parameters were selected for a more detailed analysis of the CO₂ adsorption mechanism on the prepared adsorbents. These parameters are summarized in

Table 5. Values of the calculated kinetic Model parameters for the adsorption experiments of CO₂ for all samples.

		Pseudo-First Order			Avrami’s Fractional Order
Samples	qe,exp	k1	qe,calc	R2	kA	nA	qe,calc	R2
CH-ACP-410	0.832	2.133	0.823	0.967	2.220	1.062	0.823	0.967
CH-ACG-410	0.733	1.527	0.720	0.948	1.503	0.955	0.720	0.948
CO-ACG-390	0.701	1.113	0.706	0.944	1.161	2.463	0.704	0.997
PH-ACG-850	1.248	0.978	1.239	0.953	0.921	2.137	1.235	0.989

.

The examination of Avrami’s fractional order kinetic model parameters indicates that the apparent rate constant k_A of CH-ACG-410, CO-ACG-390 and PH-ACG-850 is lower than that of the CH-ACP-410 sample. This is likely due to greater diffusional resistance resulting from their less-developed porosity. The Avrami exponent n_A, determined from the experimental data, was 1.062 for CH-ACP-410 and 0.955 for CH-ACG-410. These values suggest a homogeneous CO₂ adsorption process on their surfaces, with the adsorption probability uniform across the surface at any given time [38]. In contrast, the Avrami exponents (n_A) for CO-ACG-390 and PH-ACG-850 were 2.46 and 2.14, respectively, indicating the coexistence of multiple adsorption mechanisms. This parameter is known to reflect changes in the adsorption mechanism over time [38]. Accordingly, the variation in n_A values between the CH-ACP-410, CH-ACG-410, CO-ACG-390, and PH-ACG-850 samples suggests that different adsorption pathways are involved.

To investigate the role of mass-transfer resistance during the CO₂ adsorption process, the intra-particle diffusion model was applied. Figure 8b presents the Weber–Morris plot for CO₂ adsorption at 50 °C for all prepared adsorbents. In this model, a linear plot of q_t versus t^1/2 indicates that intra-particle diffusion is occurring; if the line also passes through the origin, it suggests that intra-particle diffusion is the sole rate-limiting step [38]. As shown in Figure 8b, the plots for all samples are not linear across the entire time range. Instead, they exhibit tri-linearity, indicating three distinct stages in the CO₂ uptake mechanism. The first linear segment corresponds to boundary layer (film) diffusion, the second to intra-particle diffusion within micropores, and the third to the saturation of active sites, marking the equilibrium stage. In this case, the linear fits for the second and third stages do not pass through the origin. This deviation suggests a difference in mass transfer rates between the initial and final adsorption steps, consistent with observations reported in the literature [42]. Therefore, intra-particle diffusion is not the sole rate-limiting mechanism; film diffusion also contributes significantly to the adsorption kinetics on the activated carbon samples. Between these two processes, only one governs the overall adsorption rate. This rate-controlling step can be identified by comparing the slopes of the linear portions preceding the equilibrium stage (region III), where the segment with the smallest slope corresponds to the rate-limiting step. The parameters of the intra-particle diffusion model and their corresponding correlation coefficients (R²) are presented in

Table 6. Intraparticle diffusion model parameters from the fittings of CO₂ at 50 °C for all samples.

Intraparticle Diffusion Model Parameters
	First Region			Second Region			Third Region
Samples	k1	C1	R2	k2	C2	R2	k3	C3	R2
CH-ACP-410	1.090	−0.226	0.997	0.194	0.537	0.895	0.006	0.709	0.935
CH-ACG-410	0.800	−0.184	0.996	0.156	0.449	0.904	0.008	0.676	0.953
CO-ACG-390	1.183	−0.697	0.982	0.170	0.440	0.854	0.0006	0.708	0.327
PH-ACG-850	1.898	−1.128	0.985	0.315	0.690	0.844	0.007	1.199	0.956

.

An analysis of the parameters obtained from the first and second linear regions reveals that, for all AC samples, the intra-particle diffusion rate constant (k₂) is consistently lower than k₁. This indicates that the diffusion of CO₂ molecules from the adsorbent surface into the micropores is the primary step governing the overall adsorption kinetics. Extrapolation of the linear segments back to the y-axis provides the intercept I, a parameter associated with the thickness of the boundary layer. A larger intercept value suggests a more pronounced boundary layer effect in hindering intra-particle diffusion. Among the samples, PH-ACG-850 exhibited the highest rate constants for both external mass transfer (k₁) and pore diffusion (k₂). This enhanced diffusion behavior is likely attributable to its greater degree of mesoporous development, as previously discussed.

4. Conclusions

Activated carbon materials for CO₂ capture under post-combustion conditions were synthesized from abundant low-cost biomass residues using practical, cost-effective methods. Two types of chemically activated carbons were compared: one granular and one pelletized. To further investigate the influence of the activation method on CO₂ adsorption performance, two additional samples were included for comparison—a combined activated carbon and a physically activated carbon.

Adsorption experiments were conducted using pure CO₂ at atmospheric pressure and a binary mixture of 10% CO₂ balanced with N₂. Among the samples, the chemically activated carbon pellet exhibited the highest CO₂ adsorption capacities, outperforming its granular counterpart prepared under identical conditions, as well as all samples produced via physical or combined activation. This superior performance underscores that chemical activation is more effective than the other methodologies for generating adsorbents with high CO₂ capture capability. The enhanced uptake is attributed to the pellet’s improved textural properties, resulting from a pelletizing process that minimizes void formation and promotes the development of narrow micropores over larger mesopores. As is well established, the volume of narrow micropores plays a decisive role in CO₂ uptake under post-combustion conditions, where the CO₂ partial pressure is inherently low. The kinetics of CO₂ adsorption on the activated carbon samples were evaluated using the pseudo-first-order and Avrami models. Among these, the Avrami model provided the best fit to the experimental data for all samples, suggesting the involvement of multiple adsorption pathways. To gain further insight into the underlying adsorption mechanism, the intra-particle diffusion model was applied. The fitting results revealed that the rate-limiting process comprises three distinct consecutive stages. The adsorption rate is subsequently governed by intra-particle diffusion within the abundant micropores of the activated carbon until equilibrium is attained. In summary, the chemically activated carbon pellets demonstrated excellent CO₂ adsorption performance under dry conditions. Their combination of low-cost precursors and favorable textural properties makes them promising candidates for large-scale applications. However, under realistic post-combustion conditions, the omnipresence of water vapor in flue gas presents a significant challenge for selective CO₂ capture using physical adsorbents. To fully assess the practical viability of these materials, future work should focus on adsorption studies using complex, humid gas mixtures (e.g., CO₂/H₂O/N₂) under dynamic and cyclic operating conditions. Such investigations are essential to evaluate competitive adsorption behavior and long-term material stability.