Comparative Study of the Structural and Adsorptive Performance of Biomass-Derived Graphene Materials

TL;DR

This paper introduces a new eco-friendly method to create graphene from agricultural waste, which shows strong potential for use in energy and water purification.

Contribution

A novel synthesis route for graphene from biomass with improved porosity, graphitic ordering, and chemical purity.

Findings

Graphene sheets with surface areas of 1300–1800 m²/g and pore diameters below 100 nm were produced.

Functionalized graphene achieved up to 80% adsorption efficiency for metal ions in aqueous solutions.

The method outperforms conventional biomass-derived graphene in terms of porosity and purity.

Abstract

This study presents the development of an environmentally benign and economically viable methodology for the synthesis of graphene-containing carbon materials derived from renewable agricultural residues, specifically walnut shells, rice husks, and apricot stones. The proposed synthesis route involves sequential stages of controlled pre-carbonization, desilicification, chemical activation with potassium hydroxide (KOH), and subsequent mild exfoliation, resulting in the formation of few-layer graphene with a high degree of structural ordering. Pre-carbonization carried out at 523–573 K, followed by activation at 1123 K, yields graphene sheets exhibiting a specific surface area of 1300–1800 m2/g, a carbon content of 60–90%, and an average pore diameter below 100 nm. The synthesized materials were subjected to comprehensive physicochemical characterization using BET surface area analysis, Raman spectroscopy, FTIR spectroscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM), and atomic absorption flame emission spectrophotometry. Raman spectroscopic analysis revealed an I_G/I_2D intensity ratio of approximately 1.5–2.0, indicating the presence of graphene structures consisting of approximately four to five layers. To enhance adsorption performance, the graphene-containing carbon materials were further functionalized with sulfuric acid, and the successful incorporation of surface functional groups was confirmed by FTIR spectroscopy. The adsorption properties of the functionalized graphene-containing carbon materials were evaluated in aqueous solutions containing sodium, potassium, calcium, and magnesium salts, demonstrating adsorption efficiencies of up to 80%. Compared to conventional biomass-derived graphene synthesis methods, the developed approach produces materials with enhanced porosity, higher graphitic ordering, and improved chemical purity. These characteristics highlight the strong potential of the synthesized graphene-containing carbon materials for applications in energy storage systems, adsorption-based water purification technologies, and environmentally sustainable nanotechnological applications.

Figures

Full text

1. Introduction

The production of carbon nanomaterials from renewable biomass feedstocks has attracted significant attention in recent years, driven by growing environmental concerns and the urgent need to replace fossil-based carbon precursors with sustainable alternatives [1,2,3,4,5,6]. Among various carbon nanostructures, graphene and graphene-containing carbon materials are particularly promising due to their high specific surface area, excellent electrical conductivity, and tunable pore structure, which collectively enhance their performance in adsorption and environmental remediation applications [7,8,9]. However, conventional synthesis techniques, such as chemical vapor deposition and mechanical exfoliation, typically rely on expensive precursors and involve energy-intensive or environmentally unfriendly processes, thereby limiting their large-scale applicability.

Consequently, agricultural biomass residues have emerged as attractive and sustainable carbon sources for the production of graphene-like porous carbons, owing to their low cost, abundance, and renewability [9]. Rice husks, apricot stones, and walnut shells are representative agro-industrial wastes generated in large quantities and commonly disposed of through open burning or landfilling, which contributes to environmental pollution. The valorization of these residues into high-value carbon nanomaterials is consistent with the principles of the circular bioeconomy and promotes sustainable waste management strategies [10]. Notably, these biomass precursors differ substantially in their chemical composition and structural characteristics: rice husk is rich in silica, whereas apricot stones and walnut shells are predominantly lignocellulosic materials with lower ash content and more condensed aromatic carbon structures. Such compositional diversity provides a valuable framework for systematically investigating the influence of precursor properties on the nanostructure and adsorption behavior of the resulting carbon materials.

Recent studies have demonstrated the successful synthesis of biomass-derived graphene-like materials through multistep processing routes involving controlled pre-carbonization, removal of inorganic components, chemical activation, and subsequent exfoliation [7,11]. In the present work, this synthesis strategy is systematically applied to rice husks, apricot stones, and walnut shells under identical experimental conditions, enabling a direct and meaningful comparison of the resulting materials. After each processing stage, the samples were thoroughly washed and dried; although this leads to cumulative mass loss, it significantly improves structural purity and pore development [12].

Accordingly, the primary objective of this study is to conduct a comparative evaluation of the structural, morphological, and adsorption properties of graphene-like carbon materials derived from three distinct biomass precursors. The findings are expected to provide valuable insights into the rational design of sustainable, low-cost graphene-based sorbents for environmental remediation and energy-related applications.

2. Results

2.1. Selection and Optimization of Synthesis Methods for Raw Biomass Precursors of Graphene Materials

Rice husks, apricot stones, and walnut shells were selected as raw biomass precursors for the synthesis of graphene-containing carbon materials due to their wide availability, renewable origin, and favorable physicochemical properties (Figure 1). Apricot stones and walnut shells were collected from agricultural areas in the Karasai district of the Almaty region, whereas rice husks were sourced from the village of Bakanas in the Balkhash district of the same region. These locations represent major agricultural production zones where such residues are generated continuously and in substantial quantities, ensuring a stable and low-cost supply of raw materials suitable for scalable processing.

In previous studies, the optimal purification and thermal treatment conditions for converting rice husks into graphene-containing carbon materials were systematically investigated [7,12]. These studies demonstrated that the efficiency of graphene formation is highly dependent on the parameters of biomass pretreatment, carbonization, desilication, and chemical activation. Specifically, the combination of controlled pyrolysis under oxygen-free conditions followed by alkaline activation was shown to yield carbon materials with highly developed porous structures and enhanced graphene-layer formation.

Building upon these results, the present study applies the same optimized technological protocol to other agricultural residues, namely apricot stones and walnut shells. Employing identical purification and processing conditions enables a direct comparison of the structural evolution of different biomass precursors under equivalent synthesis parameters. This approach not only enhances the reliability of the comparative analysis but also allows evaluation of the universality and adaptability of the previously developed method for producing graphene-containing carbon materials from diverse lignocellulosic and silica-rich biomass sources. General characteristics of the synthesized samples are summarized in Table 1.

Figure 2 presents the mass yields obtained at each stage of the multistep synthesis process for the three biomass precursors. The pre-carbonization step resulted in a substantial mass reduction for all samples, primarily due to the removal of volatile organic components and physically adsorbed moisture. The yield after pre-carbonization was approximately 55% for rice husks and walnut shells, whereas apricot stones exhibited a slightly lower yield of 45%, likely attributable to their higher content of thermally labile organic fractions. It should be noted that, at each processing stage, the samples were repeatedly washed with distilled water and subsequently dried, further contributing to the observed mass loss.

Further mass loss was observed during the desilication and purification stage. This effect was most pronounced for rice husks, for which the yield decreased to 49%, reflecting the removal of the inherent silica fraction characteristic of this biomass type. In contrast, apricot stones and walnut shells retained 42% and 52% of their initial mass, respectively, consistent with their lower inorganic content.

Subsequent chemical activation with alkali led to an additional decrease in mass, associated with the burn-off of disordered carbon and the development of a porous structure. At this stage, the mass yield across all samples ranged from 16% to 18%. Following the final exfoliation step, the yield of graphene-containing carbon materials reached 11% for rice husks, 12% for apricot stones, and 13% for walnut shells. Consequently, walnut shells provided the highest final product yield, whereas rice husks yielded the lowest amount of graphene-containing carbon materials. This trend correlates with the initial ash and mineral content of the raw biomasses, where a higher inorganic fraction leads to greater mass loss during alkaline and oxidative treatments. Moreover, cumulative washing and drying procedures throughout the synthesis sequence likely contributed to the overall mass reduction compared with purely thermal conversion processes.

Overall, these results indicate that all three biomass sources are suitable for the synthesis of graphene-containing carbon materials, with walnut shells representing the most efficient precursor in terms of final yield.

To further enhance the sorption performance of the biomass-derived graphene-containing carbon materials, an additional surface functionalization step was performed using concentrated sulfuric acid. This treatment facilitates the formation of oxygen- and sulfur-containing functional groups (including –OH, –COOH, and –SO_3_H), which increase the surface polarity and hydrophilicity of the graphene-containing carbon materials, thereby improving their interactions with aqueous pollutants. The elemental composition (C, H, and N) of the synthesized samples is presented in Figure 3.

Elemental analysis confirmed systematic changes in the composition of the samples following acid treatment. For all materials, the carbon content increased relative to the initial state: from 75.5 to 80.6 wt.% for RHdGr, from 71.9 to 86.3 wt.% for WShdGr, and from 63.1 to 66.9 wt.% for ASdGr. This increase is attributed to the removal of residual volatile organic compounds and inorganic impurities during the washing and drying procedures, resulting in a more carbon-rich framework.

The hydrogen content also increased moderately, reaching 1.8 wt.% for RHdGr, 2.1 wt.% for WShdGr, and 3.7 wt.% for ASdGr, consistent with the formation of hydroxyl- and sulfonic-type surface functionalities. Nitrogen remained at low concentrations but exhibited a slight increase after treatment, suggesting the stabilization of nitrogen-containing groups on the graphene surface.

Overall, the simultaneous increase in heteroatom content and carbon enrichment indicates successful surface functionalization. This modification is expected to enhance adsorption performance by improving surface wettability and promoting stronger interactions with aqueous pollutants through hydrogen bonding and acid–base interactions.

2.2. Study of the Physicochemical Properties of Biomass-Derived Graphene

The surface morphology of graphene-containing carbon materials synthesized from apricot stones, rice husks, and walnut shells was investigated using scanning electron microscopy (SEM) (Figure 4). The analysis was conducted at the National Open Nanotechnology Laboratory using a Quanta 3D 200i Dual System (FEI). Powdered samples were mounted on an aluminum holder, coated with a thin layer of gold and palladium, and placed in the SEM chamber. Imaging was performed under high-vacuum conditions at an accelerating voltage of 3.00 kV, employing both Everhart–Thornley and Through-the-Lens detectors.

As shown in Figure 4, the morphology and particle size of the graphene-containing carbon materials are strongly influenced by the type of biomass precursor. The RHdGr sample obtained from rice husks consisted of cap-shaped, highly brittle particles with sizes ranging from approximately 80 to 500 nm, indicating relatively uniform graphene-layer formation and efficient carbonization. In contrast, ASdGr (apricot stones) and WShdGr (walnut shells) exhibited aggregated, brittle fragments with particle sizes ranging from 90 to 600 nm and 85 to 700 nm, respectively. These observations suggest that differences in the lignin and cellulose content of the raw biomass influence the carbonization and structural formation mechanisms.

Raman spectroscopy was employed to evaluate the presence of graphene-containing carbon materials and estimate the number of graphene layers. This technique allows for the differentiation between nanostructured graphene, amorphous carbon, and graphite. The Raman spectra of ASdGr, WShdGr, and RHdGr are presented in Figure 5. All samples display the characteristic D, G, and 2D bands associated with sp^2^-carbon frameworks. The G band, corresponding to the in-plane stretching vibration of graphitic sp^2^ carbon, appears at 1582 cm^−1^ for ASdGr, 1575 cm^−1^ for WShdGr, and 1569 cm^−1^ for RHdGr. The slight downshift observed in RHdGr may be attributed to differences in structural ordering and residual strain within the carbon lattice.

The D band, which arises from lattice disorder and finite crystallite size, is observed at 1354–1356 cm^−1^ for WShdGr and RHdGr, whereas in the ASdGr spectrum this band is significantly weaker, indicating a lower degree of structural defects. The presence of the second-order 2D band at approximately 2708–2712 cm^−1^ in all spectra confirms the formation of few-layer graphene-like structures. The relatively broad shape of the 2D band suggests turbostratic stacking rather than highly ordered graphite.

Qualitative analysis of the D-to-G band intensity ratio (I_D/I_G) reveals differences in structural organization among the samples. ASdGr exhibits the lowest D-band contribution, indicating a higher proportion of ordered sp^2^ domains. In contrast, WShdGr and RHdGr display more pronounced D-band signals, consistent with the presence of defect sites introduced during chemical activation and exfoliation. These observations align with the heterogeneous nature of the biomass precursors and underscore the influence of inorganic components and lignocellulosic composition on the evolution of the graphene-like structure during synthesis.

The number of graphene layers was estimated using previously reported methodologies, and the results are summarized in Table 2.

Overall, Raman spectroscopy confirms that the applied multi-stage processing route successfully produces few-layer graphene-containing carbon materials, with the degree of structural ordering varying depending on the biomass precursor.

The microstructure of the graphene-containing carbon materials derived from apricot stones (ASdGr), walnut shells (WShdGr), and rice husks (RHdGr) was further investigated using high-resolution transmission electron microscopy (HRTEM) (FEI (now part of Thermo Fisher Scientific), Hillsboro, OR, USA). Measurements were performed at an accelerating voltage of 200–300 kV, with powdered samples deposited onto copper grids coated with an ultrathin carbon film. All samples exhibited thin, sheet-like carbon structures with wrinkled and folded regions, characteristic of few-layer graphene resulting from structural relaxation and defect formation during exfoliation (Figure 6).

For ASdGr, TEM images revealed relatively large, transparent nanosheets with smooth surfaces and limited stacking. The weak contrast of the sheets indicates the presence of few-layer graphene domains, with only minor regions of aggregation observed. These findings suggest that the exfoliation process was particularly effective for this precursor and are consistent with the Raman spectra, where the weaker D band indicates a lower degree of structural disorder.

The WShdGr sample displayed partially stacked graphene layers with pronounced folds and corrugations. The higher contrast of these regions suggests a slightly greater number of overlapping layers compared with ASdGr. Nevertheless, extended graphitic domains were still evident, confirming the preservation of a graphene-like lattice. This morphology indicates that the lignocellulosic structure of walnut shells supports graphene domain formation, although some residual stacking persists after exfoliation.

In the case of RHdGr, TEM images showed thinner graphene sheets coexisting with fragmented carbon domains. This morphology is likely associated with the inherent silica content of rice husks and the more intensive structural transformations occurring during desilication and chemical activation. The corrugated and partially fragmented appearance aligns with the higher defect density observed in the Raman spectra.

Collectively, TEM analysis confirms that all three biomass precursors yield graphene-containing carbon nanosheets, while the degree of stacking, fragmentation, and structural disorder depends on the type of biomass. Integration of TEM and Raman results demonstrates that ASdGr forms the most extended few-layer graphene domains, RHdGr exhibits a more defective and nanofragmented structure, and WShdGr occupies an intermediate position.

Fourier-transform infrared (FTIR) spectroscopy was employed to investigate the organic, polymeric, and inorganic constituents of the synthesized graphene-containing carbon materials. Analyses were conducted at IRC-CNR on solid dispersions prepared by homogenizing 0.5–0.8 wt.% of powdered samples with potassium bromide (KBr), followed by pelletizing under a pressure of 10 tons for 10 min. Spectra were recorded in transmission mode over the 3400–400 cm^−1^ range using a Nicolet 5700 spectrometer (Nicolet (Thermo Electron Corporation), Madison, WI, USA). The FTIR profiles of the biomass-derived graphene-containing carbon materials after washing and subsequent acid functionalization are presented in Figure 7.

The spectra of ASdGr and WShdGr exhibit the characteristic broad absorption envelope associated with extended aromatic carbon frameworks. Apart from the C=C stretching vibration and several overlapping skeletal vibration bands within 900–1500 cm^−1^, no significant additional features are observed, indicating the predominance of condensed sp^2^-hybridized domains.

In contrast, RHdGr displays a more complex spectral pattern due to the presence of multiple surface functional groups. In the washed state, the spectrum shows a prominent C=C stretching band near 1600 cm^−1^, along with several overlapping absorptions in the 900–1500 cm^−1^ region corresponding to C–C, C–H, and C–O stretching and bending vibrations. After 24 h of treatment with concentrated sulfuric acid, a distinct absorption band appears at approximately 1090 cm^−1^, which can be attributed to sulfonic (–SO_3_H) functionalities. These observations confirm that prolonged exposure to H_2_SO_4_ effectively introduces sulfonic groups onto the graphene-containing carbon surface, thereby modifying its chemical structure and surface reactivity.

The textural properties of the biomass-derived graphene-containing carbon materials were further assessed by BET surface area analysis, and the results are summarized in Table 3. The analysis revealed significant differences in the specific surface area and pore structure among the samples. The highest specific surface area was observed for RHdGr (1809.8 m^2^/g), indicating a highly developed porous network. Lower values were obtained for ASdGr (1458.7 m^2^/g) and WShdGr (1378.5 m^2^/g), reflecting differences in the structural evolution of the graphene-containing carbon materials derived from distinct biomass precursors.

The specific pore volume of the biomass-derived graphene materials also varies considerably among the samples. ASdGr and WShdGr exhibit relatively low pore volumes of 0.06 and 0.09 cm^3^/g, respectively, whereas RHdGr shows a markedly higher pore volume of 1.69 cm^3^/g. This observation aligns with its high specific surface area and indicates the formation of a highly developed meso-/macroporous network. The average pore size is comparable across all three materials, ranging from 25 to 26 nm, confirming the predominance of mesopores in the graphene structure regardless of the biomass source. The largest pore size was observed for RHdGr (26.1 nm). These results highlight that graphene derived from rice husks (RHdGr) possesses the most developed porous architecture and the highest specific surface area, making it particularly promising for adsorption processes and electrochemical applications.

Adsorption and salt rejection experiments were performed using a standard laboratory filtration system operated at room temperature. Membranes prepared from ASdGr, WShdGr, and RHdGr were mounted in a stirred filtration cell with an effective membrane area of 38 mm^2^, connected to a 300 mL feed reservoir. The total working volume of the feed solution was 200 mL, encompassing both the stirred cell and the storage tank. Pressure was supplied and regulated by a pump to maintain stable operating conditions. To ensure reproducibility of the membrane fabrication procedure, the steady-state deionized (DI) water flux was first measured for each membrane.

The concentrations of Na^+^, K^+^, Mg^2+^, and Ca^2+^ ions in both feed and permeate solutions were determined using atomic absorption flame emission spectrophotometry (Shimadzu AA-6200, Kyoto, Japan) at the Center for Physical and Chemical Research and Analysis (Almaty, Kazakhstan). This technique relies on the absorption of monochromatic radiation by atoms in a flame to quantify elemental concentrations. Measurements were conducted using a dual-beam optical configuration with a Cherny–Turner monochromator over a wavelength range of 190–900 nm, with selectable spectral slit widths of 0.2 and 0.7 nm and deuterium background correction. A titanium flame burner equipped with a Pt/Ir capillary and a corrosion-resistant spray chamber was used, operating with a C_2_H_2_–N_2_O flame (Table 4).

The results demonstrate that all three biomass-derived graphene membranes exhibit measurable salt rejection and adsorption capabilities. Among them, the RHdGr-based membrane achieved the highest ion removal efficiency, which correlates with its significantly larger specific surface area and pore volume. Membranes prepared from ASdGr and WShdGr also displayed stable filtration performance; however, their rejection efficiencies were comparatively lower, consistent with their less developed porous structures. These findings indicate that the textural characteristics of the graphene framework—such as surface area, pore volume, and pore architecture—play a critical role in governing ion transport and adsorption behavior.

3. Discussion

The choice of lignocellulosic precursor plays a critical role in the carbonization process and the subsequent formation of graphene-like structures. Although rice husk, apricot pit shells, and walnut shells share a common biogenic origin, they exhibit significant differences in structure and chemical composition, which ultimately govern their behavior during thermal conversion into carbon materials.

Rice husk differs markedly from nut shells due to its high silica content, which plays a dual role during graphene synthesis. On the one hand, silica may act as a natural rigid template, restricting the growth of carbon domains and promoting the formation of thin graphene-like layers. On the other hand, the presence of a mineral phase reduces the overall carbon yield and may increase structural defect density by hindering graphitization. The high porosity and developed surface area of rice husk further facilitate uniform heat transfer and early-stage decomposition of the organic matrix during thermal treatment.

Apricot pit shells are characterized by a more homogeneous and compact lignocellulosic structure with a relatively high lignin content. Such a composition favors the formation of a stable carbon framework during carbonization, which may enhance the ordering of aromatic domains and support the growth of graphene fragments. The absence of significant mineral inclusions makes apricot shells a comparatively “clean” carbon precursor, potentially resulting in graphene-like materials with a lower defect density.

Walnut shells exhibit the highest degree of lignification and the greatest lignin content among the studied precursors, which accounts for their enhanced thermal stability. During thermal decomposition, this characteristic can promote the formation of larger condensed aromatic structures that serve as precursors to graphene layers. However, the high density and complex cellular architecture of walnut shells may restrict the diffusion of volatile species, potentially leading to localized defect formation and structural heterogeneity in the resulting carbon matrix.

Overall, rice husk acts primarily as a silica-containing porous precursor, favoring the formation of highly defective graphene-like structures with a developed surface area. Apricot pit shells represent a balanced carbon source that promotes the formation of relatively uniform and structurally ordered graphene fragments. Walnut shells, owing to their high lignin content, appear to be a promising precursor for producing more condensed and thermally stable graphene-like structures, albeit with possible structural heterogeneity.

The present study demonstrates that KOH activation is a highly effective approach for producing graphene-containing carbon materials from natural biomass precursors, including apricot stones, rice husks, and walnut shells. Although the multistage synthesis process leads to a pronounced reduction in mass, the resulting products exhibit high structural quality, well-developed porosity, and characteristic graphitic features typical of graphene-containing carbon materials. Starting from 100 g of raw biomass, the final yield of graphene-like material was approximately 10 g for each precursor type.

The progressive mass loss reflects sequential material depletion at each processing stage. During the initial carbonization step, approximately 30% of the total weight is released as volatile degradation products of cellulose, hemicellulose, and lignin, forming gases and tar residues. The remaining solid is converted into a carbon-rich char with a partially disordered structure. An additional 10–15% mass reduction occurs during desilicification and washing, primarily due to the removal of mineral and inorganic constituents, which simultaneously enhances the purity and accessibility of the carbon framework.

The most substantial mass loss takes place during high-temperature KOH activation. In this stage, redox reactions between KOH and carbon generate metallic potassium, potassium carbonate, and gaseous CO and CO_2_, while simultaneously creating a highly porous structure. Although this process significantly increases the specific surface area, it also consumes a portion of the carbon matrix, resulting in a final product yield of only ~10% of the initial biomass mass after activation and purification.

Furthermore, the KOH activation mechanism critically influences the morphology of the resulting graphene-containing carbon materials. At temperatures above 850 °C, KOH decomposes to form potassium oxide, metallic potassium, and water vapor, which further react with carbon through a sequence of etching and redox processes. These reactions are summarized in the following simplified scheme [13]:

$$[eqn]$$

The influence of the KOH-to-biomass mass ratio on the structural organization of the carbon framework was further investigated using Raman spectroscopy. All samples exhibited distinct D- and G-bands at approximately 1350 cm^−1^ and 1580 cm^−1^, confirming the formation of sp^2^-hybridized carbon domains. However, the evolution of the 2D-band showed a clear dependence on the activation intensity. Specifically, samples prepared at KOH/biomass ratios of 1:1, 1:2, and 1:3 displayed only the D- and G-bands, while the 2D-band in the 2670–2720 cm^−1^ range was either absent or poorly resolved, indicating a limited degree of graphenic layer formation. In contrast, distinct 2D-bands emerged in materials obtained at higher activation ratios (1:4 and 1:5), suggesting the development of turbostratic graphene-like layers. This behavior is attributed to enhanced chemical etching and carbon rearrangement at elevated KOH loadings, which promote delamination and ordering of carbon sheets. Based on these observations, a KOH/biomass ratio of 1:5 was identified as optimal for achieving graphene-like structural features.

In addition to the structural modifications induced by KOH activation, the resulting graphene-containing carbon materials retained a certain fraction of inorganic residues originating from both the activating reagent and the native mineral content of the biomass. Elemental analysis revealed up to 0–17 wt.% of elements such as K, Fe, and Si. Although present at relatively low concentrations, these species significantly influence the physicochemical properties of the final materials. Residual potassium compounds (predominantly K_2_CO_3_ and K_2_O) can enhance electrical conductivity by promoting electron delocalization and graphitic ordering; however, excessive potassium content may partially block micropores, reducing adsorption capacity [14,15]. Trace iron species may catalyze localized graphitization, while simultaneously introducing magnetic behavior and affecting electrochemical stability. Silicon-containing phases such as SiO_2_, primarily derived from the natural mineral fraction of biomass, tend to form inert domains that obstruct pores and decrease accessible surface area.

As reported in Ref. [16], these effects can be mitigated by acid washing or thermal post-treatment, which effectively remove inorganic residues and improve the textural properties of the carbon framework. Such treatments have been shown to reduce metal and silicate content to below 2 wt.% while increasing BET surface area and adsorption capacity. Although acid purification was not applied in the present study, the literature underscores the importance of post-synthesis cleaning for optimizing the functional performance of biomass-derived graphene-containing carbon materials.

Despite the widespread use of KOH activation as an efficient strategy for generating porosity in carbon materials, the underlying mechanism remains incompletely understood, largely due to the strong dependence of the process on experimental parameters and the chemical nature of the precursor. In general, activation is assumed to begin with solid-phase reactions between KOH and the carbon matrix, which progressively transition to solid–liquid interactions as temperature increases. These processes involve the reduction of potassium compounds to metallic potassium, oxidation of carbon to oxides and carbonates, and a sequence of intermediate reactions among reactive species.

Consequently, the efficiency of KOH activation is governed not only by operational conditions—such as activator loading, temperature, and activation time—but also by the intrinsic properties of the precursor. The combined effect of these factors accounts for the wide variation in structural and textural characteristics observed in the resulting nanomaterials, including pore size distribution and specific surface area (Table 5).

SEM observations reveal that all biomass-derived graphene–carbon ceramic materials exhibit a well-developed surface morphology. The preparation routes employed do not markedly alter the macroscopic structure of the materials, which largely retains the original architecture of the biomass precursor. Nevertheless, both carbonization and subsequent chemical activation significantly enhance porosity, consistent with previously reported studies [7,12].

Raman spectroscopy was employed to assess the number of graphene layers based on the intensity ratio of the G and 2D bands, the two most characteristic features in the Raman spectrum of graphene. The spectra of the activated biomass-derived graphene materials display clear peaks corresponding to ordered graphitic layers. For greater reliability, the I_G/I_2D ratio was determined only in regions where the contribution of amorphous carbon was minimal. In both GrWSh and GrAS samples, this ratio was approximately 1.5–2, indicating a multilayer graphene structure. Moreover, spectra exhibiting pronounced 2D bands consistently displayed highly symmetrical line shapes, typically associated with few-layer graphene. Although the exact number of layers requires further verification, both the I_G/I_2D ratios and the spectral profiles (Figure 5) suggest a similar nanostructural organization in GrWSh and GrAS. The symmetric and relatively sharp 2D bands further confirm the presence of ordered graphitic domains and partial stacking of graphene layers. These observations are in good agreement with prior studies on biomass-derived graphene, where comparable activation and exfoliation processes yielded similar structural characteristics [22,23,24].

Another notable feature is the slight shift in the positions of the graphene-related Raman bands. In GrWSh, the D, G, and 2D (G′) bands were observed at 1356, 1573, and 2708 cm^−1^, respectively, whereas in GrAS they appeared at 1359, 1582, and 2712 cm^−1^. These shifts suggest that activation of GrAS may have led to a higher proportion of graphene-like domains compared to GrWSh. Defects in both samples are primarily associated with edge sites of graphene sheets and distortions within the basal planes, reflecting the combined influence of porosity development, fragmentation, and structural disorder.

A comparative analysis with previously reported biomass-derived graphene materials, summarized in Table 6 (including BET surface area, I_G/I_2D ratio, and carbon content), highlights the superior structural and textural characteristics of the materials developed in this study.

As shown in Table 6, the developed graphene-like materials exhibit superior performance in key parameters, including an optimal I_G/I_2D ratio indicative of a highly ordered graphene structure and a high carbon content reflecting excellent material purity. These values significantly exceed those reported in previous studies, highlighting the efficiency of the proposed synthesis route and the exceptional quality of the resulting carbon nanomaterials. For comparison, biomass-derived graphene from pine powder (1018 m^2^/g) [28], sweet corn husks (1370 m^2^/g) [29], and coffee grounds (1250 m^2^/g) [30] demonstrated considerably lower specific surface areas than the materials obtained in the present study.

Despite these favorable structural and textural characteristics, the overall yield of graphene-like material was approximately 10% of the initial biomass mass, which is typical for thermochemical approaches employing natural carbon-containing precursors. Nevertheless, this relatively low yield does not diminish the scientific value of the materials: their morphology, degree of crystallinity, and elemental composition confirm that high-quality carbon nanomaterials can be produced via a sustainable and cost-effective route. Therefore, the primary focus of this study is on material quality rather than quantitative yield. The combination of high surface area, favorable I_G/I_2D ratio, and substantial carbon content collectively underscores the potential of this approach for producing functional graphene materials from renewable feedstocks.

The developed synthesis strategy also offers notable environmental and economic advantages compared with conventional chemical vapor deposition (CVD) or mechanical exfoliation methods. The estimated energy consumption during pre-carbonization (523–573 K) and activation (1123 K) stages is 2.4–2.8 kWh per 100 g of biomass, approximately 40–50% lower than CVD-based graphene production. Moreover, the activation process allows recovery of 70–75% of KOH through washing and neutralization, while residual byproducts—primarily silica and carbonates—can be repurposed as adsorbents or soil conditioners. The use of inexpensive agricultural residues as precursors further reduces production costs to 0.1–0.2 USD per gram of graphene-like material, substantially lower than costs associated with synthetic precursors. These factors collectively demonstrate the sustainability of the proposed method through reduced energy consumption, reagent recovery, and waste valorization.

Considering the physicochemical characteristics of the synthesized graphene—high surface area, hierarchical porosity, and the presence of oxygen-containing functional groups—the material shows strong potential for applications in energy storage and water purification. Although experimental performance testing was not conducted in the present study, comparable KOH-activated biomass-derived graphene materials have demonstrated specific capacitances of 120–200 F/g in supercapacitors and adsorption capacities of 150–300 mg/g for heavy metal ions. These reported results suggest that the produced material is well-suited for such applications, and future work will focus on experimentally evaluating its electrochemical and adsorption performance to confirm practical applicability.

The salt adsorption behavior of the graphene-containing carbon materials is governed by the interplay between surface chemistry and textural characteristics rather than by any single structural parameter. A high density of accessible adsorption sites, combined with a hierarchically porous structure, facilitates efficient ion transport and interaction with the carbon surface. Micropores primarily contribute to ion confinement and charge accumulation, while mesopores ensure rapid diffusion of hydrated ions, reducing mass-transfer limitations and enhancing overall adsorption efficiency.

Surface functional groups play a decisive role in determining adsorption performance. Oxygen- and sulfur-containing functionalities introduced during post-treatment increase surface polarity and create chemically active sites capable of electrostatic interactions, ion–dipole attraction, and surface complexation with both monovalent and divalent ions. In addition, structural defects and edge sites inherent to few-layer graphene act as anchoring points for these functional groups, further enhancing adsorption affinity. The combined effect of favorable pore accessibility and chemically active surfaces provides a consistent mechanistic explanation for the high salt removal efficiencies observed, highlighting the importance of simultaneously optimizing surface chemistry and pore architecture for adsorption-based water treatment applications.

4. Materials and Methods

Potassium hydroxide (KOH) and sodium hydroxide (NaOH) were procured from LABHIMPROM (Chemical and Analytical Trading Company, Almaty, Kazakhstan) in accordance with TU 24363-80 specifications. Apricot stones and walnut shells were collected from agricultural regions in the Karasai district of the Almaty region, whereas rice husks were sourced from the village of Bakanas in the Balkhash district of the Almaty region.

Initially, the biomass precursors were repeatedly rinsed with distilled water to remove external impurities and subsequently dried at 383 K for 1 h. The dried materials were mechanically milled and classified into defined particle-size fractions using sieves with mesh openings of 0.1 mm, 1 mm, and 1 cm.

Pre-carbonization and Desilication. Preliminary carbonization (pre-carbonization) was performed in a tubular reactor at 523–573 K for 45 min under oxygen-limited conditions to obtain carbon-rich char. The resulting char materials were then subjected to desilication by treatment in 3 L of 1 M NaOH at 353 K for 3 h to remove silica. After settling, the supernatant containing dissolved sodium silicate was decanted, and the solid residue was washed repeatedly with distilled water (5–7 heating–settling–decanting cycles) until neutral pH (≈7) was reached. Purified samples were dried in an air oven at 383 K for at least 2 h.

Chemical Activation. The desilicified biomass was mixed with powdered KOH at a 1:5 mass ratio, compacted in an iron crucible, and subjected to thermal activation at 1123 K for 2 h under an argon flow of 5 sccm to prevent oxidation. Following activation, samples were washed several times with distilled water until neutrality (pH ≈ 7) and dried at 373 K for 24 h.

Exfoliation and Functionalization. Exfoliation was performed by treating the activated materials with 37% hydrogen peroxide (H_2_O_2_) for 48 h to remove amorphous carbon. Post-exfoliation, the samples were washed and dried as described above.

To enhance adsorption performance, sulfonic functional groups were introduced via acid functionalization. For this purpose, 0.3 g of each graphene sample was dispersed in 10 mL of 15 M H_2_SO_4_ in a round-bottom flask and maintained at 80 °C with continuous stirring for 60 min or 24 h. The suspension was centrifuged at 4000 rpm for 10 min, and the resulting solid phase was repeatedly washed with bidistilled water until neutral pH (≈7) was reached. The material was then filtered through a PVDF membrane and dried at 105 °C for 12 h.

Characterization. Structural and chemical properties of the synthesized graphene-containing carbon materials were analyzed using scanning electron microscopy (SEM), high-resolution transmission electron microscopy (HRTEM), Raman spectroscopy, Fourier-transform infrared spectroscopy (FTIR), elemental analysis, and nitrogen adsorption–desorption (BET) measurements. Additionally, adsorption and ion-removal experiments were performed using atomic absorption flame emission spectrophotometry (Shimadzu AA-6200, Kyoto, Japan). Analyses were conducted at the NNLOT Laboratories of Al-Farabi Kazakh National University (Almaty, Kazakhstan), the Center for Physico-Chemical Research and Analysis Methods (Almaty, Kazakhstan), the Institute of Combustion Problems (Almaty, Kazakhstan), the Laboratory of Analytical Chemistry for the Environment at the University of Naples Federico II (Italy), and the Institute for Combustion Research (IRC-CNR, Naples, Italy).

Carbonization and Activation Apparatus. A custom laboratory device for carbonization and thermochemical activation was designed and fabricated by LLP “Bes Saiman Group.” The system consists of a vertical steel reactor (height 50 cm, diameter 89 mm) installed in a tubular electric furnace with an integrated thermostat. The apparatus allows controlled heating up to 1200 °C at a rate of 7.5 °C/min and precise regulation of the gas flow using a digital controller. The schematic layout and components of the device are shown in Figure 8.

High-Temperature Vertical Tube Furnace. The high-temperature vertical tube furnace employed in this study has a maximum operating temperature of 1200 °C, which can be reached in less than one hour, and operates reliably over a temperature range of 40–1200 °C. The furnace has a total power consumption of 6 kW, provided by three heating wires rated at 2 kW each (220 V, R = 24.2 Ω).

The furnace chamber is constructed from aluminosilicate ceramics and has a working volume of 8138 cm^3^, with internal dimensions of OD 220 × ID 120 × 720 mm. The heating element consists of HRE resistance wire with a diameter of 1.5 mm and a total length of 29.48 m. Temperature regulation is achieved using an automatic controller block, which incorporates a short-circuit protection switch, an electromagnetic starter for switching the heating elements, a dual START–STOP manual control, a programmable temperature controller for setting and maintaining the desired temperature, and a thyristor-based power control section for smooth modulation of heating power. Temperature measurements are performed with a type-K (chromel–alumel) thermocouple.

Evaluation of Membrane Performance. The performance of membranes fabricated from biomass-derived graphene was evaluated by measuring the flux of deionized (DI) water and 100 ppm salt water using a laboratory vacuum filtration system at a trans-membrane pressure of −0.8 bar. The water and salt water fluxes (J) were calculated in units of liters per square meter per hour (L·m^−2^·h^−1^, LMH) according to the following equation:

$$[eqn]$$

where V is the volume of permeated water (mL), t is time (min), and A is the active membrane surface area (14.6 cm^2^). The salt concentration of the feed, permeate and filtrate solution was measured using graphene-based membranes for the salt water before and after permeation by the following equation:

$$[eqn]$$

where R is the salt filtration and C0 and CP are the feed and permeate concentrations (ppm).

5. Conclusions

In this study, graphene-containing carbon materials were successfully synthesized from agricultural waste precursors, including apricot stones, walnut shells, and rice husks, using a multi-step KOH activation route. Despite a relatively low overall yield of approximately 10–13% of the initial biomass mass (RHdGr: 11%, ASdGr: 12%, WShdGr: 13%), the resulting materials exhibited high structural quality, well-developed porosity, and graphitic characteristics.

Elemental analysis indicated an increase in carbon content after H_2_SO_4_ functionalization (RHdGr: 80.6 wt.%, WShdGr: 86.3 wt.%, ASdGr: 66.9 wt.%), accompanied by moderate hydrogen content (1.8–3.7 wt.%) consistent with the formation of hydroxyl and sulfonic groups, while nitrogen remained at low concentrations, confirming material stability.

SEM and TEM analyses revealed sheet-like morphologies with wrinkled and folded regions. ASdGr exhibited large, thin, and minimally stacked graphene domains; WShdGr showed partially stacked layers; RHdGr displayed more defective and fragmented structures. Raman spectroscopy confirmed D, G, and 2D bands, with the lowest D-band intensity in ASdGr (indicating higher structural order) and higher D-band intensity in RHdGr (indicating defects). The IG/I2D ratios of WShdGr and ASdGr ranged from 1.5 to 2, confirming a multilayer graphene structure.

BET analysis demonstrated that RHdGr possessed the highest specific surface area (1809.8 m^2^/g), largest pore volume (1.69 cm^3^/g), and average pore size of 26.1 nm, while ASdGr and WShdGr exhibited lower surface areas (1458.7 and 1378.5 m^2^/g) and pore volumes (0.06 and 0.09 cm^3^/g). These textural properties suggest that RHdGr is the most promising material for adsorption and electrochemical applications.

Membrane performance tests confirmed that RHdGr-based membranes achieved the highest ion rejection efficiency, correlating with their high surface area and pore volume. ASdGr and WShdGr membranes exhibited stable filtration but lower salt removal, reflecting their less developed porosity.

In summary, the proposed multi-step KOH activation method provides an efficient, cost-effective, and sustainable route to obtain high-quality graphene-like carbons from inexpensive agricultural residues. The combination of high surface area, hierarchical porosity, and functionalized surfaces renders the materials, particularly RHdGr, highly suitable for water purification, adsorption, and energy storage applications.