2.1 Materials

The mesoporous silica used in this study was derived from CFA [⁴⁵], MB dye (Merck), Sigma Aldrich NaOH (98 %) and HCl (32 %).

2.2 Characterization of Silica

Powder X-ray diffraction (XRD) measurements of the mesoporous silica were carried out using a PANalytical X'Pert Pro X-ray diffractometer equipped with a CuK_ X-ray source (wavelength 1.5405). The diffractograms were recorded in the 2-theta range of 10–80 degrees at 40 kV and 40 mA. Following the scan, prominent peaks were identified and matched with patterns in the software library, selecting the highest percentage agreement.

Sample morphologies were analyzed using a high-resolution scanning electron microscope (SEM) (Carl Zeiss Gemini SEM500). Chemical compositions were determined with the BRUKER S8 TIGER XRF spectrometer. Compound identification in the spectral range of 400–4000 cm⁻¹ was performed using a ThermoFisher Scientific Nicolet iS50 Fourier transform infrared (FTIR) spectrometer. Specific surface area and pore size were assessed using a Brunauer–Emmett–Teller (BET) analyzer (Micromeritics 3flex). Known sample amounts (40–60 mg) were loaded into the BET tube, degassed under vacuum (10–5 torr) at 100 °C for 1 h, and then further degassed overnight at 250 °C.

2.3 Point of Zero Charge of Adsorbent (Silica)

Potentiometric titrations (PMTs) were conducted under normal atmospheric conditions for a blank solution and suspensions containing different masses of immersed silica at a constant temperature of 25.0 °C ± 0.1 °C [^{46, 47}]. A small amount of 1 M KOH was added to each vessel with a given amount of silica to deprotonate part of the surface sites, making the surface negatively charged. The aqueous suspensions with the silica were equilibrated for 24 h before titration. The suspensions were then titrated with small increments of 0.5 M aqueous HNO₃ solution. The pH was measured and recorded after each addition of the acidic solution relative to its volume. The same titration process was also followed for the blank solution. A blank titration was performed in the absence of the silica adsorbent to establish a baseline for accurate interpretation of results. It accounts for pH changes resulting solely from the titrant, ensuring that subsequent shifts in pH during sample titration are attributed only to interactions with the adsorbent surface.

2.4 Adsorption Studies

Design-Expert software was used in this study to design experiments and model experimental data through response surface methodology (RSM) using the central composite design (CCD) technique, and the results were thoroughly analyzed using ANOVA (analysis of variance) [⁴⁸]. RSM is a statistical and mathematical method that evaluates the effects of independent variables and investigates the interaction of several parameters that affect the process by varying them all at once [⁴⁹]. Four independent variables were studied, as shown in Tab. 1: contact time, initial pH, adsorbent dosage, and initial dye concentration.

2.5 Batch Equilibrium Studies

2.6 Kinetics and Equilibrium Studies

2.7 Reusability Performance of Silica (Adsorbent)

2.8 Desorption Studies

To restore desorption, 100 mL of an MB aqueous solution at a concentration of 100 mg L⁻¹ was mixed with 1 g of adsorbent in a 250 mL Erlenmeyer flask. The flask was placed on a shaker and rotated at 600 rpm for 120 min. The concentration of MB in the centrifuged solution was then measured using a UV–Vis spectrophotometer. The used mesoporous silica was separated from the solution, rinsed with distilled water, and dried promptly at 110 °C for 3 h. Once fully dried, the mesoporous silica was transferred to an Erlenmeyer flask containing 100 mL of 95 % ethanol. The flask was placed on a shaker under the same conditions as before for 4 h. Samples of 5 mL were taken from the container at 30-min intervals over 4.5 h. The concentration of the desorbed substance was measured using a spectrophotometer to assess the effectiveness of ethanol regeneration.

3.1 Adsorbent Characterization

3.1.1 XRD Analysis

The noisy spectra from amorphous silica, showing a broad peak between 20 and 30 (Fig. 1), suggest the small particle size, nearing the nanoscale. This characteristic makes the mesoporous silica a promising adsorbent [⁵⁰]. SEM imaging reveals the absence of crystalline or regular particle arrangement, confirming the amorphous nature of mesoporous silica in line with other studies. This structural irregularity leads to high porosity and a large surface area per unit volume, providing more active sites for adsorption and easier access to accommodate adsorbates [⁵¹].

3.1.2 BET Analysis

The N₂ adsorption–desorption isotherm was conducted to assess silica's specific surface area and pore size distribution. As depicted in Fig. 2, the isotherm follows a type IV pattern based on the BET method, signifying a variety of adsorption behaviors, including monolayer and multilayer adsorption, as well as capillary condensation [⁵¹]. These phenomena are typical in mesoporous materials such as silica, which are good adsorbents. The isotherm displays a hysteresis loop in the range of 0.4 < P/P_o < 0.8, indicating ongoing capillary evaporation and condensation of N₂ during the adsorption process [⁵²]. The pore-size distribution, derived from the adsorption branch of the N₂ isotherms using the Barrett–Joyner–Halenda (BJH) method, reveals that most pore diameters in silica range from 2 to 50 nm, with an average pore diameter of 3.93 nm. The template-free silica has a surface area of 793.51 m² g⁻¹ and a purity of 99.1 % [⁴⁵].

3.1.3 Point of Zero Charge (PZC)

The point of zero charge (PZC) is a crucial surface parameter that characterizes the acid–base behavior of solids, mainly mineral oxides, in electrolytic suspensions. It represents the solid surface's pH level with a net zero charge, indicating that the positive and negative surface sites are in balance [^{46, 48, 53, 54}]. Fig. 3 shows the experimental curves for the PMT technique recorded for SiO₂.

Fig. 3 demonstrates that silica's PZC is around 5, identified at the common intersection of the three mass lines. When the pH falls below the PZC, silica's surface gains a positive charge; above the PZC, the surface takes on a negative charge. This variation in surface charge significantly impacts the adsorption process. Research on the PZC of silica derived from CFA is limited; however, one study by de Oliveira reports a PZC value of 7 for silica sourced from CFA [⁵⁵].

3.2 Optimization of Methylene Blue (MB) Adsorption

Tab. 2 shows the 30 experimental results for the adsorption of MB using silica extracted from CFA. The removal efficiency of silica ranged from 24.6 % at pH 3, 0.1 g dosage, 90 min and 50 ppm MB concentration to 100 % at pH 7, 3.1 g dosage, 50 min, and 50 ppm MB concentration.

3.3 Analysis of the Adsorption Process and Modeling

The ANOVA results for the MB removal efficiency by silica are shown in Tab. 3. The Model F-value of 2727.00 implies that the model is significant. There is only a 0.01 % chance that an F-value this large could occur due to noise. P-values less than 0.0500 indicate model terms are important. A, B, C, AB, AC, AD, BC, BD, CD, A², B², ABC, ABD, A²D, and AB² are significant model terms. The F-value of the adsorbent dose (11 795.51) is significant more important than other parameters, which shows that it has the most effect on the MB removal efficiency.

3.4 Interaction Effects Plots of MB Removal Efficiencies

3.4.1 Adsorbent Dosage and pH

The interaction effects of mesoporous silica dosage and initial as a function of response in the MB adsorption process are depicted in Fig. 4. The initial pH of the solution can influence the adsorbent surface charge, promoting or inhibiting dye adsorption on the adsorbent surface [^{48, 57}]. As observed in Tab. 2, the value of MB removal is as low as 24.6 at pH 3 but increases to 100 % at pH 7. As depicted in Fig. 4, augmenting both the dosage and pH leads to a notable enhancement in MB removal. This effect arises from the increased dosage providing a more significant number of active sites for MB molecule attachment [⁵⁸].

Various studies have shown that negatively charged functional groups on the adsorbent surface are required for basic dye adsorption. According to research, the surface of silica is surrounded by H₃O⁺ ions at low pH when pH < pH_ZPC [⁵⁹]. As a result of the competition between H₃O⁺ ions and the electrostatic repulsion between the MB dye molecules and the positively charged active adsorption sites on the surface of silica, dye molecule adsorption is reduced [^{58, 60}]. At higher pH values, such as pH > pH_ZPC, the amount of hydroxyl ions (OH⁻) in the solution increases, increasing the number of negatively charged sites and increasing the interaction between positive dye molecules and the negative surface of the prepared silica [^{61, 62}].

3.4.2 Initial MB Concentration and Dosage

The relationship between MB concentration and adsorbent dosage, as depicted in Fig. 5, is essential in water treatment adsorption techniques. MB concentration denotes the dye quantity in the solution, whereas adsorbent dosage represents the material capable of adsorbing the dye. Higher MB concentrations typically increase adsorbent capacity due to a greater abundance of dye molecules [^63-65]. Increasing adsorbent dosage at lower concentrations enhances adsorption effectiveness by providing more active sites and vice versa, as shown in Fig. 5. The maximum removal efficiency for this surface plot was reached at 3.1 g rather than at 6.1 g. However, a saturation threshold exists where further increments in concentration or dosage yield diminishing returns, indicating full utilization of available adsorption sites [⁴⁸]. Understanding this interaction is pivotal for optimizing adsorption processes in water treatment.

3.4.3 Time and Adsorbent Dosage

The prolonged contact between MB and the adsorbent, as shown in Fig. 6, shows limited influence on removal efficiency, suggesting that MB adsorption reaches equilibrium quickly, rendering extended contact negligible for efficiency gains [^{66, 67}]. Conversely, increasing the adsorbent dosage markedly improves removal efficiency, highlighting a direct relationship between adsorbent quantity and adsorption capacity. This underscores the pivotal role of adsorbent dosage in optimizing removal efficacy, as evidenced by the apparent trend depicted in Fig. 6, with the highest MB removal at 100 % at 3.1 g dosage and 50 min.

3.5 Equilibrium Studies—Adsorption Isotherms

Adsorption isotherms describe and explain the relationship between adsorbent and adsorbate, which is important for optimizing adsorption processes [⁶⁸]. As the adsorption process approaches equilibrium, the equations illustrate the distribution of an adsorbate between the liquid and solid phases [⁶⁹]. Mapping the experimental data to various isotherm models is a critical step in determining a suitable model for process design [^{69, 70}]. In this study, the Langmuir and Freundlich isotherm models, widely used in the scientific community, were used to analyze the experimental adsorption equilibrium data.

3.5.1 The Langmuir Isotherm Model

The Langmuir isotherm describes monolayer adsorption on homogeneous surfaces with minimal interaction. When a dye molecule occupies one site in the Langmuir isotherm, no other adsorption occurs at that site [⁵⁶]. The following equation (Eq. 4) depicts the linear form of Langmuir isotherms:

$$()

where C_e (mg L⁻¹) is the equilibrium concentration of MB dye in the solution, q_e (mg g⁻¹) is the adsorption capacity at equilibrium, Q_m (mg g⁻¹) is the maximum adsorption capacity, and K (L mg⁻¹) is the effective dissociation constant relating to the affinity binding site. The intercept and slope of the linear plot of C_e/q_e against C_e yield the values of Q_m and K.

The linear and nonlinear Langmuir isotherm is depicted in Fig. 7:

From the two plots, the nonlinear Langmuir isotherm (b) has the highest correlation coefficient (R² = 0.9881) value than the linear form (R² = 0.9992). The model variables are summarized in a Tab. 4:

Table 4. Summary of the Langmuir model constants.

Langmuir
R2	Qm [mg g−1]	KL [L mg−1]	RL
0.9881	32.76	0.06	0.253

The maximum adsorption capacity of silica, Q_m = 32.76 mg g⁻¹, shows a strong interaction between the adsorbate and adsorbent (Q_m >> 1). The RL value of the Langmuir model (see equation below) demonstrates the type of adsorption. The separation coefficient, RL, is defined as a dimensionless transition variable. RL >1 indicates an undesirable adsorption type, RL = 1 indicates a linear adsorption type, RL = 0 indicates irreversible adsorption, and 0 < RL < 1 indicates a favorable adsorption type [⁷¹]. The calculated RL value is 0.253 which is well within the 0 < RL < 1, which means that the adsorption of the MB onto silica is favorable:

$$()

3.5.2 Freundlich Isotherm Model

The Freundlich isotherm expresses the interaction of dye adsorption on a heterogeneous surface. The amount of dye adsorption in the Freundlich isotherm varies with the exponential distribution of sites and adsorption energies [⁷²]. The linear Freundlich isotherm equation (Eq. 6) is expressed as

$$()

where K_F and n represent the Freundlich adsorption isotherm constants indicating adsorption capacity and intensity, respectively. The Freundlich isotherm constants K_F and 1/n can be reported using the plot of ln q_e versus ln C_e shown in Fig.8a and the nonlinear Freundlich shown in Fig.8b.

The magnitude of the exponent, 1/n, in the adsorption process indicates the adsorption's favorability [⁷³]. If the values are 1/n < 1, it means that as new adsorption sites appear, the type of isotherm required, the adsorption intensity, and the adsorption capacity increase. If 1/n > 1, the adsorption bond weakens, reducing the dye adsorption capacity [⁷⁴].

The Freundlich plots in Fig. 8 (linear and nonlinear) show that the nonlinear has a higher correlation than the linear. The calculated Freundlich adsorption constants n and K_F are 1.56 and 2.38, respectively. As 1/n < 1 (1/n = 0.64), the adsorption capacity increases and physical adsorption is advantageous (Note: physical adsorption process, n > 1, linear adsorption process, n = 1, and chemical adsorption, n < 1), indicating that the value of n is more than 1 [⁷⁵].

But the correlation constants of Freundlich isotherms are lower than the Langmuir models. As a result, it is possible to argue that MB adsorption on silica is a monolayer adsorption type. According to the Langmuir isotherm, Q_m = 32.76 mg g⁻¹), Q_m >> 1, and value of RL for MB adsorption was 0 < RL < 1, indicating that MB adsorption was favorable [⁵⁶].

The MB adsorption capacity results from this study demonstrate a moderate performance of our mesoporous silica when compared with literature data (Tab. 5). However, this efficiency is noteworthy because the material was synthesized without the use of structure-directing agents, seeding with pure silica sources, or any functionalization to improve MB adsorption. This simplicity in synthesis and processing not only reduces production costs but also highlights the inherent efficacy and versatility of our mesoporous silica, making it a promising, sustainable option for future applications in water treatment and dye removal.

Table 5. Comparison of adsorption capacity of template free synthesized silica with other adsorbents.

Silica source	Silica source	Surface modification/Template	Surface area [m2 g−1]	Pore size [nm]	MB removal efficiency [%]	Maximum adsorption capacity, Qmax [mg g−1]	References
Mesoporous silica	CFA	N/A	793.51	3.93	100 %	32.76	This work
Poly-HIPE@nano-silica	Sigma Aldrich	Polystyrene coated	30.95	17.78	–	23.1	[76]
Cysteine-functionalized mesoporous silica	Sigma Aldrich chemicals	CTAB	652	5	–	140	[77]
Mesoporous silica	Rice husk	CTAB	150.28	3.62	95	4.94	[44]
SBA-15 Al-SBA-15	CFA	P123	1014 871	7.06 8.34	90.5 ∼100	160.15 190.1	[78]
Silica gel desiccant	Commercial desiccant	–	633	2.3	85.7	34.30	[79]
Micro-mesoporous silica	ZSM-5 Zeolite	CTAB	840	2.5	73–91	303	[42]

3.6 Kinetics

The kinetic study is an essential part of the adsorption process in water and wastewater treatment as it can provide valuable data about reaction pathways and dye adsorption reactions [⁸⁰]. Furthermore, adsorption kinetics can predict the rate of solute adsorption, which is dependent on the retention time or adsorption reaction at the solid solution interface [⁸¹]. As a result, the adsorption rate or kinetic of adsorption is an important factor to consider when selecting adsorbent materials, designing processes, controlling operations, and predicting adsorption amount [⁷¹].

In this study, pseudo first-order (PFO) and pseudo second-order (PSO) adsorption mechanism of kinetic models were evaluated using Eqs. 7 and 8 [⁸²].

Adsorption rates were calculated using both aforementioned models, and the results are shown in Fig. 9a PFO and Fig. 9b PSO plots.

The obtained kinetic parameters for both models are also shown in Tab. 6. The results showed that the pseudo-second-order model had a better fit than the pseudo-first-order model, with a regression coefficient (R²) of 0.9988, a maximum adsorbent amount of 7.35 mg g⁻¹, and a rate constant of 0.076 g mg⁻¹ min⁻¹.

The pseudo-second-order model is typically applied in situations where the adsorption involves a multicomponent system, the process is relatively slow, or the adsorption follows an irreversible reaction. This kinetic model characterizes the adsorption of adsorbates onto adsorbents [^{83, 84}], where chemical interactions (chemisorption) between adsorbates and functional groups on the adsorbent's surface determine its adsorption capacity [⁸⁵].

3.7 Adsorption Thermodynamics

The thermodynamic analysis of MB adsorption onto silica provides key insights into the process's feasibility and underlying interactions and is comparable with other studies [^{86, 87}]. The negative enthalpy change ($$) confirms the exothermic nature of the adsorption, indicating that heat is released, and minimal energy is required for the process to occur. This behavior is driven by strong molecular interactions, including electrostatic attraction between the positively charged MB molecules and the negatively charged silica surface, hydrogen bonding with silica's silanol groups, van der Waals forces, and π–π interactions involving MB's aromatic rings [^88-91]. The negative entropy change (ΔS = −64.52811 J (mol K)⁻¹ = −64.53 J (mol K)⁻¹) suggests a decrease in randomness as MB molecules transition from the bulk solution to a more ordered state on the silica surface [⁹²].

Additionally, as seen in Tab. 7, the negative Gibbs free energy values (ΔG ranging from − 3.29 kJ mol⁻¹ at 303 K to − 2.65 kJ mol⁻¹ at 313 K) confirm that adsorption is thermodynamically spontaneous and favorable at lower temperature ranges [^{93, 94}]. However, the decreasing spontaneity at higher temperatures (e.g., ΔG = −1.51 kJ mol⁻¹) at 328 K) is consistent with the exothermic nature of the process, as increased thermal energy reduces the driving force for adsorption [⁹⁵]. These findings highlight silica's efficiency as an adsorbent, requiring minimal energy input for dye removal. Its strong adsorption capacity and spontaneous nature make it a cost-effective and environmentally viable option for wastewater treatment applications.

3.7.1 Adsorption Mechanism

The adsorption of MB onto silica, as depicted in Fig. 11, occurs through inner-sphere and outer-sphere complexation, depending on the nature of interactions between MB molecules and the silica surface. Inner-sphere complexation involves strong, direct bonding, including electrostatic attraction between negatively charged silanol groups (SiO⁻) and cationic MB⁺, hydrogen bonding between silica's hydroxyl (–SiOH) groups and MB's functional groups, and π–π stacking/interaction between MB's aromatic rings and surface-bound organic compounds. These interactions enhance adsorption stability and affinity.

In contrast, outer-sphere complexation relies on weaker, indirect interactions, such as van der Waals forces, which facilitate temporary adhesion, and solvent-mediated adsorption, where water molecules create a barrier that inhibits direct bonding. The combination of these mechanisms ensures efficient MB adsorption, reinforcing silica's potential as an effective and sustainable adsorbent for dye removal in environmental applications [^88-91].

3.8 Reusability of the Adsorbent

In addition to its high adsorption capacity, the reusability of mesoporous silica adsorbent is crucial for ensuring economic viability in adsorption studies. In this study, 12 cycles of adsorption and regeneration were carried out using silica sorbent for MB adsorption (Fig. 12).

The cyclic adsorption experiments used 1 g of sorbent in a 100 mg L⁻¹ concentration of MB dye, at a pH of 7 (without adjustment) and a temperature of 25 °C. The consistency in MB removal capacity across all 12 cycles, with an average removal rate of approximately 97% ±  0.17 %, demonstrates the economic potential of template-free, high-purity silica synthesized from CFA. Despite these encouraging results, there is a lack of research on the reusability of silica derived from CFA in MB adsorption.

Comparative studies using commercial silica gel or modified silica have shown a decline in adsorption capacity within the first five cycles of reusability. For instance, Subhan's [⁴²] study on cetyltrimethylammonium bromide (CTAB)-synthesized micro-mesoporous silica (MMZ, BET 840 m² g⁻¹ and pore diameter 3.91 nm) from ZSM-5 zeolite reported a slight decrease in maximum MB adsorption over successive cycles. This reduction may result from a decline in surface area and weight loss during MMZ collection in adsorption and regeneration cycles. Similarly, Koyuncu's [⁴³] study examined the reusability of silica (BET 652 m² g⁻¹, pore diameter 17.7 nm) and carbon-silica hybrid aerogels (BET 519 m² g⁻¹, pore diameter 32.1 nm) made from paddy waste ash. The study found that although there was minimal change in dye adsorption capacity from the first cycle (83.5 %) to the fourth cycle (79.7 %), a significant drop occurred in the fifth cycle, with dye removal decreasing to 59.7 % [⁴³].

Usgodaarachchi's [⁴⁴] research on mesoporous silica nanoparticles derived from rice husk and surface-controlled amine functionalization (MSN-A, BET 150 m² g⁻¹ and pore size 3.62 nm) for MB adsorption observed a decrease in MB adsorption rate over five cycles. The equilibrium adsorption capacity decreased from 13.72 mg g⁻¹ in the first cycle to 11.05 mg g⁻¹ in the fifth cycle [⁴⁴].

These findings highlight the significance of further investigating the reusability of silica from various sources, mainly CFA, which is both an inexpensive and abundant waste material. Utilizing silica derived from CFA enhances its performance and cost-effectiveness in adsorption applications and contributes to environmental remediation by repurposing an industrial byproduct.

Fig. 13 illustrates the ability of silica to desorb adsorbed MB, a process that holds significant economic importance for the reutilization of the adsorbent.

This study explores the potential of ethanol as a desorbing agent for MB dye from spent silica. Ethanol's suitability for silica regeneration stems from its easy removal through water rinsing, low-cost, low-toxicity, and convenient recycling [⁹⁶]. Fig. 11 shows that silica exhibits a time-dependent increase in desorption potential at pH 7, reaching a desorption capacity of 5.5 mg L⁻¹ within 5 h. This suggests the release of physical-adsorbed MB, consistent with findings from previous research [⁷⁵]. The observed desorption aligns with the Freundlich isotherm model (n = 1.56, n > 1), indicating the effectiveness of the desorption process and applicability in describing the interaction between MB and silica.