FACTORIAL OPTIMIZATION OF HEMP STALK ACTIVATED CARBON AND MECHANISTIC INSIGHT INTO THE STRUCTURE-ACTIVITY RELATIONSHIP OF IRON-MODIFIED ADSORBENTS FOR METHYLENE BLUE REMOVAL

Abstract

This study utilized hemp stalks as raw material and employed a 3 × 2 × 3 full factorial design to optimize the preparation process for hemp stalk activated carbon. Through tetracycline adsorption experiments, the most effective activated carbon sample was identified and characterized by scanning electron microscopy (SEM). Building upon this, iron-loaded activated carbon was prepared via three different methods: coprecipitation (Fe-AC-1), conventional impregnation (Fe-AC-2), and water-bath impregnation (Fe-AC-3). The effects of activated carbon dosage, initial concentration, and adsorption time on methylene blue adsorption were systematically investigated. Adsorption mechanisms were analyzed through adsorption isotherms, kinetic models, and Brunauer Emmett Teller (BET) characterization. Results indicated that Fe-AC-3 exhibited the highest adsorption efficiency. Adsorption isotherm results indicated that Fe-AC-1 conformed to the Freundlich model, while Fe-AC-2 and Fe-AC-3 followed the Langmuir model. The adsorption performance ranking was Fe-AC-3 > Fe-AC-2 > Fe-AC-1. Adsorption kinetics revealed that all three conformed to pseudo-second-order kinetics, with performance ranking consistent with the isotherm results. The BET characterization report indicates that Fe-AC-3 exhibits a relatively more concentrated pore size distribution and slightly higher pore accessibility, demonstrating excellent adaptability for methylene blue adsorption. This finding elucidates the specific reasons for the differing adsorption performances of the three activated carbons towards methylene blue.

Keywords:
hemp stalk activated carbon; methylene blue; iron-loaded activated carbon; adsorption isotherm; adsorption kinetics; pore structure.

INTRODUCTION

With the rapid development of the textile industry, pollution from dyeing and printing wastewater has become increasingly severe.¹-³ According to current statistical data, the printing and dyeing industries generate enormous amounts of wastewater, which together account for 11% of the annual industrial wastewater emissions in China, or 2-2.3 billion tons.⁴ This is wastewater considered one of the major sources of pollution affecting water and soil, posing a significant threat to the ecological environment and living organisms.⁵ Textile dyes are a class of complex aromatic compounds which inhibit plant growth, enter the food chain, possess recalcitrance and bioaccumulation, and pose risks of toxicity, mutagenicity and carcinogenicity.⁶,⁷ The discharge of dye wastewater poses severe threats to both the ecological environment and human health.⁸ Dyes can be broadly categorized by their properties and applications into disperse dyes, direct dyes, cationic dyes, reactive dyes, acid dyes, basic dyes, vat dyes, etc.⁹ Methylene blue (MB) is one of the cationic dyes widely used in dyeing silk, wood, and cotton which is the most commonly used model pollutant for evaluating the adsorption performance of adsorbents.¹⁰

Compared to biodegradation and chemical treatment technologies, physical adsorption methods are widely adopted due to their low preparation costs, high adsorption efficiency, and environmental friendliness.¹¹,¹² Activated carbon represents a class of adsorbents characterized by high adsorption capacity, large specific surface area, substantial pore volume, and abundant functional groups.¹³,¹⁴ However, traditional commercial activated carbon on the market mostly uses coal or wood as raw materials, which have problems such as strong resource dependence, high production costs, and large carbon emissions. Therefore, finding a low-cost and simple-to-prepare biomass material that can reach excellent adsorption performance has become the main focus of researchers.¹⁵ Industrial hemp plants are 1-3 m in height, with tall stems which are the core structural support of the plant. Rich in cellulose and largely hollow, hemp stalks are an abundant lignocellulosic biomass resource. While the flowers, leaves, and seeds are widely utilized for various purposes, hemp stalks are often discarded or burned as agricultural waste, resulting in a low utilization rate.¹⁶ Therefore, using hemp stalks to prepare adsorbents for wastewater treatment can not only reduce preparation costs and improve the utilization rate of hemp stalk wastes, but also provide additional income for farmers, avoiding environmental pollution caused by their disposal or on-site incineration. Although hemp stalks possess a high cellulose content and are materials with significant adsorption potential, the adsorption capacity of raw unactivated hemp stalks for pollutants is extremely low.¹⁶ This is primarily because, compared to activated hemp stalk carbon, the surface of raw hemp stalks is relatively smooth and dense, resulting in a low specific surface area and a scarcity of active functional groups.¹⁷,¹⁸ Consequently, using raw hemp stalks alone as an adsorbent is inefficient for treating pollutants and has limited practical value; an activation process is required to produce a highly efficient activated carbon adsorbent with a large specific surface area, a rich pore structure and an abundance of active functional groups, thereby truly meeting the requirements of wastewater treatment. For example, Huang et al.¹⁷ prepared hemp stalk-based adsorbents via chemical modification and investigated their adsorption capacity for Cr^VI, finding that the maximum adsorption capacity of the adsorbent reached 396.88 mg g^-1. Meanwhile, Chew et al.¹⁹ successfully prepared nitrogen-doped porous activated carbon from hemp stalks. Compared with commercial activated carbon of a similar specific surface area, the hemp stalk-derived activated carbon exhibited approximately twice the adsorption capacity for tetracycline. Due to the high cellulose content in hemp stalks, the preparation of hemp stalk activated carbon using chemical activation methods can provide more adsorption sites and enhance adsorption capacity.²⁰ However, the reagents typically used in chemical activation methods, such as strong acids or strong bases, are corrosive to equipment, and the subsequent elution of residual phosphoric acid requires large amounts of fresh water, leading to increased equipment maintenance and water costs.²¹ Furthermore, the large volume of phosphorus-containing wastewater requires sophisticated treatment facilities, significantly increasing environmental management and overall costs, thereby limiting its application in sectors such as food and pharmaceuticals.

Conventional activated carbon exhibits limitations in adsorbing cationic dyes, including restricted adsorption capacity, suboptimal treatment efficiency, and weak selectivity.²²,²³ Relevant studies²⁴,²⁵ indicate that loading iron onto traditional activated carbon enables electrostatic adsorption with cationic dyes, enhancing both adsorption performance and selectivity. Currently, methods for preparing iron-loaded activated carbon primarily include coprecipitation, conventional impregnation, reduction, and water-bath impregnation.²⁶,²⁷ The principles of these different preparation methods vary, and the adsorption performance and mechanisms for cationic pollutants remain unclear.

To investigate the adsorption performance and mechanisms of activated carbon samples prepared by different methods toward the cationic dye MB, this study uses hemp stalks as raw materials to prepare activated carbon. Based on this, iron-loaded activated carbon was synthesized by coprecipitation, conventional impregnation, and water-bath impregnation at a fixed carbon-iron mass ratio of 18:1. The obtained samples were named Fe-AC-1, Fe-AC-2, and Fe-AC-3, respectively. MB was selected as the model adsorbate. The adsorption performance of iron-loaded activated carbons prepared by the three different methods was investigated under various experimental conditions, including adsorbent dosage, initial MB concentration, and contact time. Adsorption isotherm and kinetic models were fitted to the experimental data to elucidate the adsorption mechanisms, providing experimental evidence and theoretical guidance for the modification and design of other carbon-based composite materials, thereby promoting the development of highly efficient, environmentally friendly, and cost-effective water pollution control technologies.

EXPERIMENTAL

Materials

Materials used in this study included hemp stalks (provided by Kunming Institute of Botany, Chinese Academy of Sciences); phosphoric acid (analytical grade reagent (AR) grade); pure water; ethanol (AR grade); tetracycline (Shanghai McLean Biochemical Technology Co., Ltd.); sodium hydroxide (AR grade); ferrous sulfate (AR grade); ferric chloride (AR grade); methylene blue (AR grade); a 0.22 µm filter membrane. The experimental instruments are listed in Table 1.

Table 1
Experimental instruments

Name	Manufacturer	Model
Blower drying oven	Boxun	BGZ-30
Muffle furnace	Zhongxing Weiye Century	SX2.5-10
Constant-temperature water bath shaker	Bona Technology	THZ-82
Ultra-violet spectrophotometer	Shanghai Qinghua Scientific	-
Scanning electron microscope	Micromeritics	Regulus 8100
Specific surface area and pore size analyzer	Micromeritics	ASAP 2460 3.00
Grinder	-	-
Standard test sieves (60 and 100 mesh)	-	-

Table 2
Three-factor full factorial design (3 × 2 × 3)

Number	Mass ratio of H3PO4 to hemp stalks	Calcination temperature / °C	Calcination time / h
1AC-500-1	1:6	500	1
2AC-500-1	1:7.5	500	1
3AC-500-3	1:6	500	3
4AC-500-3	1:7.5	500	3
5AC-500-4	1:6	500	4
6AC-500-4	1:7.5	500	4
7AC-550-1	1:6	550	1
8AC-550-1	1:7.5	550	1
9AC-550-3	1:6	550	3
10AC-550-3	1:7.5	550	3
11AC-550-4	1:6	550	4
12AC-550-4	1:7.5	550	4
13AC-600-1	1:6	600	1
14AC-600-1	1:7.5	600	1
15AC-600-3	1:6	600	3
16AC-600-3	1:7.5	600	3
17AC-600-4	1:6	600	4
18AC-600-4	1:7.5	600	4

Material preparation

Preparation of hemp stalk activated carbon

Raw materials pretreatment process: fresh hemp stalks were washed three times with tap water to remove surface dirt and impurities, and then soaked in pure water for 30 min to eliminate soluble organic matter. The washed hemp stalks were placed in stainless-steel trays and dried in an oven at 60 ± 2 °C. After drying, the hemp stalks were ground using a plant grinder and sieved through a 60-mesh screen.

Preparation process and parameter optimization of activated carbon: during the preliminary experiments, different mass ratios of phosphoric acid to hemp stalk powder (2:1, 3:1, and 4:1) were evaluated. It was observed that when the mass ratio of phosphoric acid to hemp stalk powder was lower than 6:1, the amount of phosphoric acid solution was insufficient to completely wet the hemp stalk powder, leaving a considerable portion of the material uncoated by the liquid phase. At a mass ratio of 6:1, the phosphoric acid solution was able to wet most of the hemp stalk powder, whereas at a mass ratio of 7.5:1, the solution completely wetted the powder and fully penetrated its internal structure.

Consequently, using phosphoric acid (H₃PO₄) at a mass fraction of 40%, formulated in two mass ratios:

Experimental group 1: phosphoric acid:hemp stalk mass ratio of 6:1 (e.g., 12 g phosphoric acid impregnates 2 g raw material).

Experimental group 2: phosphoric acid:hemp stalk mass ratio of 7.5:1 (e.g., 15 g phosphoric acid impregnating 2 g raw material).

The phosphoric acid solution was added to the hemp stalk powder, and the mixture was thoroughly stirred in a porcelain crucible, then left to soak naturally for 24 h and dried in an oven at 105 °C for 12 h. A 3 × 2 × 3 full factorial design was employed, considering three factors: calcination temperature (500, 550, and 600 °C), H₃PO₄-to-hemp stalk mass ratio (6:1 and 7.5:1), and calcination time (1, 3, and 4 h).

The activated hemp stalk powder was placed in a muffle furnace for calcination. After calcination, the sample was allowed to cool naturally to below 200 °C before removal from the furnace to avoid structural damage caused by rapid cooling. The calcined product was washed with deionized water until the filtrate reached a neutral pH and then dried in an oven at 105 °C until a constant weight was achieved. Subsequently, the dried sample was ground and sieved through a 100-mesh sieve. The samples were named according to the three experimental factors: calcination temperature, calcination time, and mass ratio of hemp stalks to activator. The naming convention was AC-calcination temperature-calcination time; for example, 1AC 500-1 and 2AC-500-1.

Preparation of iron-loaded activated carbon by coprecipitation method

Ferrous sulfate (0.3 g) was dissolved in 100 mL of deionized water. Subsequently, 2 g of prepared activated carbon (14AC-600 1) was added to the solution, followed by the dropwise addition of 1 mol L^-1 sodium hydroxide solution to adjust the pH of the mixture to 7.5-8.0. The suspension was stirred at room temperature until complete mixing and then allowed to undergo coprecipitation for 24 h. The precipitate was collected by filtration using a Büchner funnel and repeatedly washed with deionized water until the pH of the washing solution reached 6.5-7.0. The solid was then washed once with 50 mL of ethanol to facilitate subsequent drying and reduce particle agglomeration. After filtration until the filter cake was free of visible moisture, the sample was dried in an oven at 120 °C for 2 h. The dried sample was placed in a crucible and heated in a muffle furnace to 400 °C, maintained at this temperature for 2 h, and then cooled to obtain the final product, designated as Fe-AC-1. The preparation process is illustrated in Figure 1.

Figure 1
Flowchart for the preparation of iron-loaded activated carbon by the coprecipitation method

Preparation of iron-loaded activated carbon by conventional impregnation method

Prepared activated carbon (14AC-600-1, 2 g) was added to 20 mL of 0.1 mol L^-1 FeCl₃ solution. The mixture was placed in a constant-temperature shaking incubator and agitated at 120 rpm for 24 h to facilitate the penetration of Fe³⁺ ions into the surface and pore structure of the activated carbon. The impregnated sample was separated by vacuum filtration and washed 3-4 times with deionized water. Subsequently, the sample was dried in an oven at 105 °C and then calcined in a muffle furnace at 400 °C for 2 h, yielding the iron-loaded activated carbon designated as Fe-AC-2.²⁸

Preparation of nano-hydroxy iron oxide-loaded activated carbon via water-bath impregnation method

A portion of prepared activated carbon (14AC-600-1) was immersed in 1 mol L^-1 HCl solution for 24 h. The acid-treated activated carbon was then washed with deionized water until the washing solution reached a neutral pH and dried at 50 °C. Subsequently, 0.5 g of FeCl₃.6H₂O was dissolved in 20 mL of deionized water, and 2 g of the acid-treated activated carbon was added to the solution. The container was sealed and placed in a water bath at 90 °C for 2 h. The resulting product was washed three times with deionized water and dried at 50 °C for subsequent use.²⁷

Adsorption experiment design

Adsorption experiment of hemp stalk activated carbon on tetracycline

Tetracycline (TC) and MB are typical cationic organic pollutants widely found in aquatic environments. Under certain pH conditions, the dominant adsorption mechanisms of these two pollutants on activated carbon are consistent, being primarily governed by electrostatic attraction, hydrogen bonding and π-π stacking interactions. Due to similarities in their adsorption behavior and mechanisms, both MB and TC can be used as typical pollutants for evaluating the adsorption performance of activated carbon.²⁹ However, MB detection is prone to matrix interference, whereas the ultraviolet visible (UV-Vis) spectrophotometric determination of TC is well established, providing accurate quantification and high sensitivity. Therefore, TC was selected as the probe for adsorbent screening to rapidly identify the optimal adsorbent.

For each activated carbon sample, 4 mg of absorbent was weighed and added to a 20 mg L^-1 of TC solution. Samples were taken at different time points, filtered, and the residual TC concentration and adsorption capacity in the filtrate were determined using a UV-Vis spectrophotometer with a 1 cm path length quartz cuvette at a wavelength of 357 nm. The adsorption equilibrium time for a given activated carbon sample was determined when the residual TC concentration showed essentially no change at two consecutive time points and the adsorption capacity had stabilized.

To investigate the effect of initial TC concentration, 4 mg of each activated carbon sample was added to TC solutions with initial concentrations of 2, 4, 6, 8, 10, and 20 mg L^-1, respectively. The suspensions were placed in a constant-temperature shaker at 25 °C and 150 rpm. After reaching adsorption equilibrium, aliquots were collected for analysis.

For adsorption kinetic experiments, 4 mg of each activated carbon sample was added to 10 mL of TC solution (20 mg L^-1). The mixtures were shaken at 25 °C and 150 rpm, and aliquots were collected at 10, 30, 60, 90, 105, and 120 min to determine the residual TC concentration. All samples were analyzed in triplicate.

Determination of MB standard curve

A series of MB solutions with concentrations of 2, 3, 4, 5, and 6 mg L^-1 were accurately prepared. The absorbance of each solution was measured at 664 nm using a UV-Vis spectrophotometer equipped with a 1 cm path length quartz cuvette. Each sample was measured in triplicate, and a standard curve was plotted based on the data obtained, as shown in Figure 2, with coefficient of determination (R²) = 0.99979.

Figure 2
Standard curve of MB at a wavelength of 664 nm

MB adsorption removal experiments

Using the cationic dye MB as the adsorbate, this study investigated the differences in the adsorption performance of three types of iron-loaded activated carbon towards MB solutions under various conditions, including the dosage of iron-loaded activated carbon, the initial concentration of the MB solution, and the adsorption time. Furthermore, the experimental data were fitted using adsorption kinetic and isotherm models to further elucidate the adsorption mechanisms.

Experimental methods for determining adsorption capacity and adsorption efficiency

The absorbance of adsorbent solution at its maximum absorption wavelength was measured by UV-Vis spectrophotometry to determine the mass concentration of adsorbate in the solution. The equilibrium adsorption capacity (q_e) and adsorption efficiency (R) were calculated according to Equations 1 and 2, respectively.

(1) $${\mathrm{q}}_{\mathrm{e}}=\frac{({\mathrm{C}}_0-{\mathrm{C}}_{\mathrm{e}})\times \mathrm{V}}{\mathrm{m}}$$

(2) $$\mathrm{R}=\frac{{\mathrm{C}}_0-{\mathrm{C}}_{\mathrm{t}}}{{\mathrm{C}}_0}\times 100\mathrm{\%}$$

where C₀ is the initial mass concentration of the adsorbate (mg L^-1), C_e is the equilibrium mass concentration of the adsorbate (mg L^-1), C_t is the residual mass concentration of the adsorbate at a time t (mg L^-1), V is the volume of the solution (L), and m is the mass of the adsorbent (g).

RESULTS AND DISCUSSION

Determination of the optimal adsorption performance of activated carbon sample

Eighteen samples of activated carbon derived from hemp stalks were individually added to solutions of TC at different concentrations; once adsorption equilibrium was reached, the adsorption capacity and adsorption efficiency were determined. The results showed that, when only 4 mg of activated carbon was added to a 20 mg L^-1 TC solution, the adsorption efficiency of TC for the three samples (12AC-550-4, 13AC-600-1 and 14AC-600-1) all exceeded 99.3%. To identify the activated carbon sample with the best adsorption performance, these three samples were individually added to a TC solution, and samples were taken at different time points to determine the adsorption capacity. As shown in Figure 3, the adsorption capacity of the activated carbon for TC gradually increased over time. Compared with 12AC-550-4 and 13AC-600 1, the increase in adsorption capacity of 14AC-600-1 was more pronounced between 0 and 30 min. During the period from 30 to 90 min, the adsorption capacity growth of 14AC 600-1 remained significantly higher than that of 13AC-600-1, and this difference continued to increase.

Figure 3
Time-dependent adsorption capacity curves of TC onto activated carbon samples (note: error bars represent the standard deviation of triplicate measurements)

In summary, the activated carbon sample 14AC-600-1 exhibited the best adsorption performance; therefore, all subsequent experiments were conducted using this sample.

Scanning electron microscopy (SEM) characterization of 14AC-600-1

SEM was employed to characterize the 14AC-600-1 activated carbon sample. Figure 4 shows that, at a magnification of 2000×, the surface of some particles is locally rough, and rich pore structures are formed between the particles. Activated carbon particles have irregular angular shapes with a wide particle size distribution, with most particles having a diameter less than 10 µm. At a magnification of 4000×, rich interparticle macropores are formed between irregular carbon particles. These large pores between particles can provide preliminary diffusion channels for adsorbate molecules. At a magnification of 10000×, the surface of activated carbon particles is relatively flat, with some small carbon chips attached locally, and multiple microcracks visible on the surface. The macroscopic appearances of the as-prepared activated carbon samples are presented in Figure 5.

Figure 4
SEM image of 14AC-600-1

Figure 5
(a) 14AC-600-1; (b) Fe-AC-1; (c) Fe-AC-2 and (d) Fe-AC-3

Effect of adsorbent dosage on adsorption performance

A 150 mg L^-1 MB solution was prepared, and 50 mL of this solution was transferred into conical flasks for each experimental group. At 25 °C and an agitation speed of 120 rpm, 10, 20, 30, 40, and 50 mg of activated carbon were added sequentially to the flasks. After shaking for 120 min, the suspensions were allowed to stand for 24 h. Figure 6 shows a comparison of the adsorption performance of Fe-AC-3 activated carbon before and after treatment of the MB solution. After filtration through a membrane filter, the absorbance was measured using a UV-Vis spectrophotometer. The adsorption capacity and adsorption efficiency of the MB solution were determined from the absorbance data. A curve-fitting analysis was conducted to investigate the variations, revealing the relationship between the amount of activated carbon added and the adsorption capacity and efficiency, as shown in Figures 7 and 8.

Figure 6
Visual comparison of MB solutions before and after adsorption by activated carbon (Fe-AC-3) under different adsorbent dosages

Figure 7
Effect of adsorbent dosage on the adsorption efficiency of MB using different iron-loaded activated carbons (Fe-AC-1, Fe-AC-2, Fe-AC-3)

Figure 8
Effect of adsorbent dosage on the adsorption capacity of MB using different iron-loaded activated carbons (Fe-AC-1, Fe-AC-2, Fe-AC-3)

The activated carbon dosage significantly affects the adsorption performance of MB. As the dosage of activated carbon increases, the adsorption efficiency rises accordingly. The nano-hydroxy iron oxide-loaded activated carbon (Fe-AC-3) prepared by water-bath impregnation method exhibits the optimal adsorption performance for MB solution, with a maximum adsorption efficiency of 98.79% and a maximum adsorption capacity of 629 mg g^-1, showing an overall steady upward trend. For Fe-AC-2, when the activated carbon dosage reached 40 mg, the adsorption efficiency increased sharply from 77 to 95%, followed by an immediate decrease. This behavior can be attributed to the conventional impregnation process used for preparing the activated carbon, in which impregnation under atmospheric pressure results in incomplete microporous filling, leading to the enrichment of Fe³⁺ primarily on the outer surface. During the drying process, solvent evaporation drives the migration of dissolved species toward the particle surface.²⁹,³⁰ Subsequent calcination at 400 °C promotes atomic diffusion and grain boundary migration, leading to particle sintering and grain growth,³¹ which reduce the dispersion of the iron oxide particles. When an excessive amount of activated carbon is added, the low dispersion of the iron oxide readily leads to agglomeration of the activated carbon particles,³² thereby reducing the overall specific surface area of the activated carbon, diminishing active sites, and causing a decrease in adsorption efficiency. In contrast, when iron-loaded activated carbon is prepared using the coprecipitation method, the iron oxide formed in situ on the activated carbon surface from the iron precursor is uniformly distributed across the surface,³³ thereby increasing the dispersion of the iron oxide. Even when an excess of activated carbon is added, no significant agglomeration occurs; active sites gradually become saturated as the dosage increases, and the adsorption efficiency levels off gradually. Compared to the conventional impregnation method, the water-bath impregnation method under low-temperature conditions is conducive to the formation of small-sized, uniformly dispersed nanoparticles.²⁷ Furthermore, the proportion of oxygen-containing functional groups on the surface of activated carbon pre-treated with hydrochloric acid is significantly increased, and the number of hydrophilic polar groups is enhanced,³⁴ which facilitates the adsorption and dispersion of iron oxides and reduces the likelihood of agglomeration.

Effect of initial solution concentration on adsorption performance

MB solutions with initial concentrations of 50, 100, 150, 200, and 250 mg L^-1. For each concentration, 50 mL of the solution was accurately measured and placed in a 250 mL conical flask. Keeping all other experimental conditions constant, 10 mg of each of the three activated carbon samples was added to each concentration group. After shaking at a constant temperature of 25 °C for 120 min, the solutions were left to stand for 24 h. A comparison of the adsorption performance of Fe-AC-3 before and after MB adsorption is presented in Figure 9. After filtration through a membrane filter, the adsorption efficiency and adsorption capacity were determined using a UV-Vis spectrophotometer. The relationship between the initial concentration of MB and the adsorption efficiency as well as adsorption capacity of each material is shown in Figures 10 and 11.

Figure 9
Visual comparison of MB solutions before and after adsorption by activated carbon (Fe-AC-3) at different initial concentrations

Figure 10
Effect of initial MB concentration on the adsorption efficiency of different iron-loaded activated carbons (Fe-AC-1, Fe-AC-2, Fe-AC-3)

Figure 11
Effect of initial MB concentration on the adsorption capacity of different iron-loaded activated carbons (Fe-AC-1, Fe-AC-2, Fe-AC-3)

As can be seen from the Figures 10 and 11, the adsorption efficiency of all three activated carbons decrease as the initial concentration of the MB solution increases. At an initial concentration of 50 mg L^-1, all three activated carbons exhibited relatively high adsorption efficiencies. However, when the initial concentration exceeded 50 mg L^-1, the adsorption efficiency decreased significantly. Nevertheless, the adsorption efficiency of activated carbons Fe-AC-2 and Fe-AC-3 remained consistently higher than that of Fe-AC-1. Regarding the variation in adsorption capacity, at an initial concentration of 50 mg L^-1, there was little difference in adsorption capacity among the three activated carbons. As the concentration gradually increased, the adsorption capacity of activated carbon Fe-AC-3 remained consistently higher than that of activated carbons Fe-AC-1 and Fe-AC-2. This indicates that activated carbon Fe-AC-3 exhibits superior adsorption performance in MB solutions with high initial concentrations.

Considering the overall trend in the plots of adsorption efficiency versus adsorption capacity, it can be observed that whilst the adsorption capacity increases continuously with rising initial concentration, the adsorption efficiency decreases as the initial concentration increases. This phenomenon occurs because, as the initial concentration of the MB solution increases, the concentration gradient across the solid-liquid interface increases, thereby enhancing the mass transfer driving force. This drives more MB molecules to migrate from the bulk solution to the activated carbon surface, resulting in an increase in the amount adsorbed as the initial concentration rises. However, the number of effective free adsorption sites and the intrinsic adsorption capacity of the activated carbon are limited. This leads to increasingly intense competition for adsorption among MB molecules at the limited surface sites; consequently, the adsorption efficiency actually decreases as the initial concentration increases.³⁵,³⁶ In summary, activated carbon Fe-AC-3 demonstrated the best performance in terms of adsorption efficiency and adsorption capacity, and exhibited superior adsorption performance for high concentrations of MB.

Adsorption isotherm analysis

The activated carbon adsorption isotherm was modelled using the Langmuir and Freundlich equations. The Langmuir model is a theoretical model for monolayer adsorption, assuming a uniform adsorbent surface with fixed adsorption sites. Its expression is as follows:

(3) $${\mathrm{q}}_{\mathrm{e}}=\frac{{\mathrm{q}}_{\mathrm{m}}{\mathrm{K}}_{\mathrm{L}}{\mathrm{C}}_{\mathrm{e}}}{1+{\mathrm{K}}_{\mathrm{L}}{\mathrm{C}}_{\mathrm{e}}}$$

where q_e is the equilibrium adsorption capacity (mg g^-1), q_m is the monolayer saturation adsorption capacity (mg g^-1), K_L is the Langmuir adsorption equilibrium constant, and C_e is the equilibrium concentration of the adsorbate (mg L^-1).

The Freundlich model is an adsorption model for non-uniform surfaces. Its central assumption is that the adsorbent surface is heterogeneous; as the degree of coverage increases, the adsorption energy of the adsorption sites decreases logarithmically with increasing surface coverage. The expression is as follows:

(4) $${\mathrm{q}}_{\mathrm{e}}={\mathrm{K}}_{\mathrm{F}}{\mathrm{C}}_{\mathrm{e}}^{1/\mathrm{n}}$$

where q_e is the equilibrium adsorption capacity (mg g^-1), K_F is the Freundlich adsorption constant, 1/n is the adsorption intensity factor, and C_e is the equilibrium concentration of the adsorbate (mg L^-1).

Using non-linear regression methods, the adsorption isotherm data for the three types of activated carbon were fitted to the Langmuir and Freundlich models, as shown in Figure 12. The corresponding fitting parameters for each model are presented in Table 3. According to Table 3, the Langmuir model coefficient of determination (R²) for activated carbons Fe-AC-2 and Fe-AC-3 are 0.94 and 0.97, respectively, higher than those for the Freundlich model. Thus, the adsorption processes of Fe-AC-2 and Fe-AC-3 better conform to the Langmuir model, indicating that adsorption of MB on the surface of iron-loaded activated carbon prepared by conventional impregnation and water-bath impregnation methods is monolayer adsorption, with a relatively uniform distribution of adsorption sites and no significant intermolecular interactions between the adsorbate molecules. In contrast, the isotherm for activated carbon Fe-AC-1 showed a better fit to the Freundlich model, indicating that the adsorption of MB by the iron-loaded activated carbon prepared via the coprecipitation method is controlled synergistically by multiple active sites on a heterogeneous surface. Based on the Langmuir constant K_L, the highest K_L value obtained for activated carbon Fe-AC-3 indicates the strongest binding capacity toward MB and the optimal adsorption affinity. Considering both the maximum adsorption capacity (q_m) and the adsorption intensity factor (1/n), the adsorption performance of the three adsorbents can be ranked as follows: Fe AC 3 > Fe AC 2 > Fe-AC-1.

Table 3
Important parameters for adsorption isotherm fitting of three types of iron-loaded activated carbons (Fe-AC-1, Fe-AC-2, Fe-AC-3)

Category	Langmuir adsorption isotherm fitting parameter				Freundlich adsorption isotherm fitting parameter
	KL / (L mg-1)	qm / (mg g-1)	R2		KF / ( mg g-1)	1/n	R2
Fe-AC-1	0.01438	441.83	0.90249		84.25541	0.31013	0.96392
Fe-AC-2	0.00673	747.52	0.94425		70.53917	0.42986	0.83934
Fe-AC-3	0.02597	719.62	0.97681		69.97782	0.05955	0.94839

Figure 12
Adsorption isotherm fitting curves for MB on different iron-loaded activated carbons (Fe-AC-1, Fe-AC-2, Fe-AC-3): (a) Langmuir model; (b) Freundlich model

According to previous studies,³⁶-⁴² the adsorption isotherm fitting results for Fe-AC-3 indicate a theoretical maximum adsorption capacity of 719.62 mg g^-1, which is significantly higher than those reported for biomass-derived activated carbons prepared from palm kernel shells, fruit shells, castor seed cake and hemp seeds. Although its maximum adsorption capacity is lower than that of mangosteen shell-derived activated carbon, it is the highest among the hemp-based activated carbons listed in this series. The comparison is presented in Table 4.

Table 4
Comparison of maximum adsorption capacities of different activated carbons

Adsorbent	Adsorbate	qm / (mg g-1)	Reference
Palm kernel shell-derived activated carbon	MB	14.70	37
Fruit shell-activated carbon		204.08	36
Mangosteen shell activated carbon		1360.73	38
Castor seed cake activated carbon		485.03	39
Carbon-based hemp seed adsorbents		52.61	40
Activated carbon from hemp waste		400.25	41
Hemp hurds powder-based adsorbents		239.81	42
Iron-loaded hemp stalks activated carbona		719.62	present study

• ^a
  Activated carbon Fe-AC-3; q_m: maximum adsorption capacity; MB: methylene blue.

Adsorption kinetics study

Adsorption kinetics were modeled using pseudo-first-order and pseudo-second-order kinetic models. The pseudo-first-order kinetic model is an empirical model derived from the Lagergren equation. It assumes that the adsorption rate is proportional to the number of unoccupied sites on the adsorbent surface and is typically used to describe physical adsorption processes governed by the diffusion step of the adsorbate. It is expressed as follows:

(5) $$\ln ({\mathrm{q}}_{\mathrm{e}}-{\mathrm{q}}_{\mathrm{t}})=\ln {\mathrm{q}}_{\mathrm{e}}-{\mathrm{k}}_1\mathrm{t}$$

where q_t is the adsorption capacity at time t (mg g^-1), q_e is the equilibrium adsorption capacity (mg g^-1), k₁ is the pseudo-first-order adsorption rate constant (min^-1), and t is the adsorption time (min).

The pseudo-second-order kinetic model assumes that the square of the number of unoccupied adsorption sites on the adsorbent surface is proportional to the adsorption rate; it is commonly used to describe adsorption processes governed by chemical adsorption mechanisms. The equation is as follows:

(6) $$\frac{t}{q_t}=\frac{1}{k_2{q}_e^2}+\frac{t}{q_e}$$

where q_t is the adsorption capacity at time t (mg g^-1), q_e is the equilibrium adsorption capacity (mg g^-1), k₂ is the pseudo-second-order adsorption rate constant (min^-1), and t is the adsorption time (min).

In the adsorption kinetics experiment, the initial MB concentration was 150 mg L^-1 and the adsorbent dosage was 10 mg. Samples were collected at time intervals of 20, 40, 60, 90, 120, 150, and 210 min, and the absorbance was measured at each corresponding time point. The processed data were fitted using pseudo-first-order and pseudo second-order kinetic models. The fitting results are presented in Figure 13, while the corresponding equations and key parameters are listed in Table 5. The data in the table indicate that the adsorption processes for all three activated carbons are better described by the pseudo-second-order kinetic model, with R² values of 0.991, 0.986, and 0.990, respectively. This suggests that the adsorption of MB by all three activated carbons is primarily a chemical adsorption process. Based on the pseudo-second-order adsorption rate constants, activated carbon Fe-AC-2 exhibits the fastest adsorption rate. However, considering the theoretical adsorption capacity (q_e), activated carbon Fe-AC-3 exhibits the highest adsorption capacity and the best adsorption performance, followed by Fe-AC-2, with Fe-AC-1 is slightly inferior to the other two adsorbents. These results are consistent with those obtained from the adsorption isotherm analysis.

Table 5
Fitting equations and kinetic parameters of adsorption models for three types of iron-loaded activated carbons (Fe-AC-1, Fe-AC-2, Fe-AC-3)

Category	Pseudo-first-order					Pseudo-second-order
	Equation	k1 / min-1	qe cal / (mg g-1)	R2		Equation	k2 / min-1	qe cal / (mg g-1)	R2
Fe-AC-1	ln(qe - qt) = 5.56596 - 0.00231t	0.00231	261.38	0.945		t/qt = 0.05789 + 0.00321t	0.000178	311.53	0.991
Fe-AC-2	ln(qe - qt) = 5.50158 - 0.00138t	0.00138	245.08	0.883		t/qt = 0.03773 + 0.00302t	0.000242	331.12	0.986
Fe-AC-3	ln(qe - qt) = 5.4838 - 0.00267t	0.00267	240.76	0.966		t/qt = 0.02375 + 0.00191t	0.000154	523.56	0.990

Figure 13
Kinetic model fitting curves for MB adsorption on different iron-loaded activated carbons (Fe-AC-1, Fe-AC-2, Fe-AC-3): (a) pseudo-first-order kinetic model; (b) pseudo-second-order kinetic model

BET characterization results of activated carbon

The pore size distribution parameters of the three activated carbons determined at 77 K are presented in Table 6. Figure 14 show the N₂ adsorption-desorption isotherms and the corresponding pore size distribution curves for activated carbon samples Fe-AC-1, Fe AC-2, and Fe-AC-3, respectively. The data in Table 6 indicate that activated carbon Fe-AC-2 possesses a BET specific surface area of 285.655 m² g^-1, exceeding that of Fe-AC-1 (248.388 m² g^-1) and Fe-AC-3 (240.838 m² g^-1). According to the International Union of Pure and Applied Chemistry (IUPAC) classification, the analysis of the N₂ adsorption-desorption isotherms of the three activated carbons showed that they all exhibit typical type IV characteristics and have obvious H4 hysteresis loops. This indicates that all three samples possess a hierarchical pore structure consisting of both micropores and mesopores, which are defined by IUPAC as pores with widths ≤ 2 nm and 2-50 nm. Combining the corresponding pore size distribution curves of the three activated carbons reveals that Fe-AC-1 possesses the largest pore volume. However, its pore size distribution is excessively broad, containing minor macropores alongside the predominant mesopores and micropores. This results in uneven surface energy, diminishing the effectiveness of adsorption sites, reducing the affinity for MB adsorption and decreasing pore accessibility. Notably, according to the IUPAC definition, pore accessibility is a key parameter that refers to the fraction of pore space in a porous medium that is actually accessible to fluid molecules of a specific size and can participate in fluid transport. In this work, this parameter is introduced to elucidate the intrinsic adsorption differences of MB on activated carbon arising from variations in pore structure. In contrast, Fe-AC-2 and Fe-AC-3 exhibit extremely similar pore structures, with smaller total pore volumes and more concentrated pore size distributions. Their most probable pore sizes are also very close, better matching the molecular dimensions of MB. However, compared to Fe-AC-2, Fe-AC-3 exhibits a slightly steeper peak shape; this subtle difference reflects a slightly more concentrated pore size distribution in Fe-AC-3, resulting in a slight improvement in pore accessibility. Based on a comprehensive assessment of the adsorption experimental results, adsorption kinetics and the fit of the isotherm model fitting, Fe-AC-3 demonstrates the optimal adsorption performance for MB.

Table 6
Pore structure parameters of three types of iron-loaded activated carbon samples (Fe-AC-1, Fe-AC-2, Fe-AC-3)

Adsorbent	Specific surface area / (m2 g-1)	Total pore volume / (cm3 g-1)	Average pore size / nm	Median pore diameter / nm
Fe-AC-1	248.3885	0.2904	7.2143	1.0157
Fe-AC-2	285.6550	0.1868	5.9531	1.0156
Fe-AC-3	240.8377	0.1779	7.2054	1.0215

Figure 14
Nitrogen adsorption-desorption isotherm and pore volume distribution of: (a, b) Fe-AC-1, (c, d) Fe-AC-2, and (e, f) Fe-AC-3. (a), (c), and (e) Show the nitrogen adsorption-desorption isotherms, while (b), (d), and (f) show the BJH (Barrett-Joyner-Halenda) adsorption pore volume distribution (dV/dw)

CONCLUSIONS

This study optimized the preparation process for hemp stalk activated carbon and subsequently employed three different methods to produce iron-loaded activated carbon. Through a series of adsorption experiments and model fitting, the adsorption mechanisms and structure-activity relationships of three different iron-loaded activated carbons towards MB were determined. The study confirmed that the activated carbon (14AC-600-1), prepared under conditions of a calcination temperature of 600 °C, a calcination time of 1 h, and a phosphoric acid-to-hemp stalk mass ratio of 7.5:1, exhibited the best adsorption performance. The preparation of iron-loaded activated carbon based on this activated carbon sample can further optimize the surface properties and adsorption performance of the material, thereby broadening its potential applications in pollutant control. In experiments investigating the effects of activated carbon dosage and initial MB concentration on adsorption performance, the results indicated that nano-hydroxy iron oxide-loaded activated carbon Fe-AC-3, prepared using the water-bath impregnation method, exhibited the most outstanding adsorption performance, particularly for the adsorption of high-concentration MB. The results of adsorption isotherm and adsorption kinetics fitting indicate that activated carbon Fe-AC-1 primarily exhibits heterogeneous surface adsorption characteristics, whilst activated carbons Fe-AC-2 and Fe-AC-3 primarily exhibit monolayer adsorption characteristics. The primary adsorption mechanism for MB in all three activated carbons is chemical adsorption. Based on an analysis of specific parameters, the adsorption performance of the three activated carbons is ranked as follows: Fe-AC-3 > Fe-AC-2 > Fe-AC-1. The BET characterization results indicate that Fe-AC-1 has the largest pore volume, but its pore size distribution is too broad, reducing pore accessibility. In comparison, Fe-AC-2 and Fe-AC-3 exhibit relatively steeper and narrower peak shapes, with higher pore accessibility, and consequently demonstrate superior adsorption performance for MB.

In summary, this study utilized hemp stalks - a type of agricultural waste - as raw material to establish a 3 × 2 × 3 full factorial design optimization system for the preparation of biomass-derived activated carbon. It systematically compared three different methods for preparing iron-loaded activated carbon and established a correlation mechanism between macroscopic adsorption performance and microscopic pore structure. This work has enriched the theoretical framework for the resource utilization of agricultural and forestry wastes and the adsorption of organic cationic dyes by iron-loaded activated carbon, providing a systematic research approach for the performance screening and mechanism investigation of similar adsorbent materials. It has pioneered a new pathway for the high-value utilization of hemp stalks, an agricultural waste material, transforming them from waste into a valuable resource. This provides a novel, practical, low-cost, and highly efficient material for the treatment of organic dye wastewater in the printing and dyeing industry, offering both theoretical significance and practical engineering value.

However, this study currently examines only the adsorption performance of a single cationic dye and has not conducted in-depth investigations of mixed dye systems or actual printing and dyeing wastewater, making it difficult to comprehensively assess the applicability of the material in real aquatic environments. Furthermore, all experiments in this study were conducted under static batch conditions in the laboratory; no dynamic column adsorption experiments or continuous-flow treatment simulations were carried out, making it difficult to directly guide industrial production and application. To address these limitations, future research could expand to investigate adsorption systems involving mixed dyes and actual textile dyeing wastewater environments, conduct practical wastewater treatment experiments, further evaluate the cycle stability and environmental safety of the material, and carry out dynamic adsorption experiments to establish a more solid foundation for the practical engineering applications of the material.

DATA AVAILABILITY STATEMENT

All data supporting the findings of this study are included within the article. No additional data are available.

ACKNOWLEDGMENTS

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. We sincerely thank our supervisor for valuable suggestions on experiment design and manuscript revision. Gratitude is also extended to all members of our research group for their assistance during the experimental process, as well as to our laboratory for providing the necessary experimental resources.

REFERENCES

• ¹ Liu, Y.; Chen, J.; Duan, D.; Zhang, Z.; Liu, C.; Cai, W.; Zhao, Z.; Sustainability 2024, 16, 10867. [Crossref]
  » Crossref
• ² Adane, T.; Adugna, A. T.; Alemayehu, E.; J. Chem. 2021, 1, 5314404. [Crossref]
  » Crossref
• ³ Wang, X.; Jiang, J.; Gao, W.; Water Sci. Technol. 2022, 85, 2076. [Crossref]
  » Crossref
• ⁴ Song, Y.; Wang, L.; Qiang, X.; Gu, W.; Ma, Z.; Wang, G.; Journal of Water Process Engineering 2023, 55, 104242. [Crossref]
  » Crossref
• ⁵ Kishor, R.; Purchase, D.; Saratale, G. D.; Saratale, R. G.; Ferreira, L. F. R.; Bilal, M.; Chandra, R.; Bharagava, R. N.; J. Environ. Chem. Eng. 2021, 9, 105012. [Crossref]
  » Crossref
• ⁶ Lellis, B.; Fávaro-Polonio, C. Z.; Pamphile, J. A.; Polonio, J. C.; Biotechnology Research and Innovation 2019, 3, 275. [Crossref]
  » Crossref
• ⁷ Sudarshan, S.; Harikrishnan, S.; RathiBhuvaneswari, G.; Alamelu, V.; Aanand, S.; Rajasekar, A.; Govarthanan, M.; J. Appl. Microbiol. 2023, 134, lxac064. [Crossref]
  » Crossref
• ⁸ Dutta, S.; Adhikary, S.; Bhattacharya, S.; Roy, D.; Chatterjee, S.; Chakraborty, A.; Banerjee, D.; Ganguly, A.; Nanda, S.; Rajak, P.; J. Environ. Manage. 2024, 353, 120103. [Crossref]
  » Crossref
• ⁹ Benkhaya, S.; M’ Rabet, S.; El Harfi, A.; Inorg. Chem. Commun. 2020, 115, 107891. [Crossref]
  » Crossref
• ¹⁰ Kusuma, H. S.; Jaya, D. E. C.; Illiyanasafa, N.; Ikawati, K. L.; Kurniasari, E.; Darmokoesoemo, H.; Amenaghawon, A. N.; Chemosphere 2024, 356, 141867. [Crossref]
  » Crossref
• ¹¹ Mutegoa, E.; Discover Water 2024, 4, 69. [Crossref]
  » Crossref
• ¹² Cui, Q.; Huang, Y.; Ma, X.; Li, S.; Bai, R.; Li, H.; Liu, W.; Wei, H.; Molecules 2024, 29, 5841. [Crossref]
  » Crossref
• ¹³ Kyi, P. P.; Quansah, J. O.; Lee, C. G.; Moon, J. K.; Park, S. J.; Appl. Sci. 2020, 10, 2251. [Crossref]
  » Crossref
• ¹⁴ Faizal, A. N. M.; Putra, N. R.; Abdul Aziz, A. H.; Agi, A.; Zaini, M. A. A.; J. Cleaner Prod. 2024, 447, 141637. [Crossref]
  » Crossref
• ¹⁵ Tang, C.; Ni, Z.; Xu, C.; Luo, Y.; Cai, X.; Gao, Q.; Fang, Y.; Zhong, G.; Qiu, R.; Zhang, S.; Sep. Purif. Technol. 2024, 333, 125878. [Crossref]
  » Crossref
• ¹⁶ Suručić, L.; Andrić, D.; Jevtić, I.; Momčilović, M.; Suručić, R.; Penjišević, J.; Coatings 2025, 15, 519. [Crossref]
  » Crossref
• ¹⁷ Huang, W.; Xu, Y.; Chen, N.; Cheng, G.; Ke, H.; J. Environ. Chem. Eng. 2023, 11, 111441. [Crossref]
  » Crossref
• ¹⁸ Myeni, Z. R.; Dziike, F.; Motaung, T. E.; Deenadayalu, N.; Sustainability 2026, 18, 2197. [Crossref]
  » Crossref
• ¹⁹ Chew, T. W.; H’Ng, P. S.; Abdullah, B. C. T. G. L. C.; Chin, K. L.; Lee, C. L.; Hafizuddin, B. M. S. M. N.; TaungMai, L.; Materials 2023, 16, 7365. [Crossref]
  » Crossref
• ²⁰ Diao, Y.; Walawender, W. P.; Fan, L. T.; Bioresour. Technol. 2002, 81, 45. [Crossref]
  » Crossref
• ²¹ Puziy, A. M.; Poddubnaya, O. I.; Gawdzik, B.; Sobiesiak, M.; Sprynskyy, M.; C Journal of Carbon Research 2022, 8, 47. [Crossref]
  » Crossref
• ²² Jeyaraj, S. G.; Hemavarshini, S.; Shree Gg, K.; Aravind, J.; Physical Sciences Reviews 2024, 9, 3409. [Crossref]
  » Crossref
• ²³ Han, M.; Liu, Z.; Huang, S.; Zhang, H.; Yang, H.; Liu, Y.; Zhang, K.; Zeng, Y.; Nanomaterials 2024, 14, 1933. [Crossref]
  » Crossref
• ²⁴ Mohit, A.; Remya, N.; Priyadarshini, U.; Environmental Quality Management 2024, 33, 945. [Crossref]
  » Crossref
• ²⁵ Beljin, J.; Ðukanović, N.; Anojčić, J.; Simetić, T.; Apostolović, T.; Mutić, S.; Maletić, S.; Nanomaterials 2024, 15, 26. [Crossref]
  » Crossref
• ²⁶ Xu, J.; Liu, H.; Luan, W.; Wang, C.; Liu, C.; Yang, Y.; Environ. Eng. 2024, 42, 67. [Crossref]
  » Crossref
• ²⁷ Wang, S.; Liu, Y.; Li, S.; Liu, H.; Gu, J.; Yang, G.; Wang, J.; Wang, H.; Journal of Dalian Ocean University 2022, 37, 276. [Crossref]
  » Crossref
• ²⁸ Wang, H.; Zhao, L.; Chen, L.; Applied Chemical Industry 2025, 54, 2636. [Crossref]
  » Crossref
• ²⁹ Liu, Y.; Wang, J.; Xu, L.; Qian, D.; Mu, Y.; J. Environ. Chem. Eng. 2025, 13, 118317. [Crossref]
  » Crossref
• ³⁰ Liang, D.; Su, X.; Xie, Q.; Zhang, L.; Deng, Q.; Zhou, H.; Liu, J.; Journal of Mining Science and Technology 2024, 9, 988. [Crossref]
  » Crossref
• ³¹ Marica, I.; Stefan, M.; Boca, S.; Falamaş, A.; Farcău, C.; J. Colloid Interface Sci. 2023, 635, 117. [Crossref]
  » Crossref
• ³² Cheng, Q.; Wang, Y.; Zhang, J.; Conejo, A. N.; Liu, Z.; Chem. Eng. Sci. 2022, 262, 118038. [Crossref]
  » Crossref
• ³³ Hu, M.; Wu, H.; Xie, J.; Chen, P.; Chin. J. Environ. Eng. 2023, 17, 368. [Crossref]
  » Crossref
• ³⁴ Kozyatnyk, I.; Yakupova, I.; ACS Sustainable Chem. Eng. 2025, 13, 2500. [Crossref]
  » Crossref
• ³⁵ Kocaman, S.; Carbon Lett. 2023, 33, 883. [Crossref]
  » Crossref
• ³⁶ Wang, J.; Wang, R.; Ma, J.; Sun, Y.; Water 2022, 14, 3752. [Crossref]
  » Crossref
• ³⁷ Said, K. A. M.; Amin, M. A. M.; Yakub, I.; Rahman, M. R.; Kueh, A. B. H.; Hamdan, S.; Rahman, M. M.; BioResources 2023, 18, 5120. [Crossref]
  » Crossref
• ³⁸ Fan, Z.; Zhang, Z.; Zhang, G.; Qin, L.; Fang, J.; Tao, P.; Int. J. Environ. Sci. Technol. 2022, 19, 12315. [Crossref]
  » Crossref
• ³⁹ Ferreira, L. M.; de Melo, R. R.; Pimenta, A. S.; de Azevedo, T. K. B.; de Souza, C. B.; Biomass Convers. Biorefin. 2020, 12, 1181. [Crossref]
  » Crossref
• ⁴⁰ Mphuthi, B. R.; Thabede, P. M.; Modise, J. S.; Xaba, T.; Shooto, N. D.; Appl. Sci. 2023, 13, 9659. [Crossref]
  » Crossref
• ⁴¹ Kaya, H.; Cellul. Chem. Technol. 2025, 59, 441. [Crossref]
  » Crossref
• ⁴² Xu, P.; Li, B.; Yu, J.; Liu, L.; Xu, J.; Wang, Z.; Fan, Y.; Sci. Adv. Mater. 2019, 11, 661. [Crossref]
  » Crossref

Edited by

• Executive Editor handled this article:
  Gustavo F. S. Andrade

Publication Dates

• Publication in this collection
  31 Aug 2026
• Date of issue
  2026

History

• Received
  01 Apr 2026
• Accepted
  23 July 2026
• Published
  12 Aug 2026