Chemically tuned cellulose nanocrystals/single wall carbon nanosheet based electrodes for hybrid supercapacitors

Abstract

Adopting a green and environmentally friendly strategy requires the development of supercapacitor electrodes using sustainable, renewable, and environmentally beneficial materials. Chemically stable and renewable cellulose-based supercapacitors need high-quality carbon materials with excellent mechanical and electrical characteristics to create a three-dimensional network-based electrode. Nevertheless, using cellulose as a supercapacitor electrode with enhanced electrochemical characteristics presents a difficulty. This paper describes creating and producing electrodes for supercapacitors using nano-composites consisting of wrapped-around single-walled carbon nanotubes to improve performance. The electrode's optimal electrochemical characteristics were achieved by using a concentration of 9 wt% MACNC/CNT nanocomposites. The proposed electrode material for the MACNC-based flexible supercapacitor assembly demonstrates outstanding electrochemical stability and effective electrochemical performance. When tested in a three-electrode cell configuration, it achieves an areal capacitance of 1389.202 mF cm⁻² at a current density of 0.02 A cm⁻², with 74.6% cyclic retention after 12 000 cycles. This study effectively converted agricultural waste into high-performing supercapacitor electrodes using a simple and cost-efficient method. This innovative design and outstanding electrochemical performance show great promise in using environmentally friendly materials to improve nanocellulose-based sustainable energy storage systems.

---

1. Introduction

Life on Earth is closely interconnected with nature, which supplies the essential materials and energy needed for existence. Cellulose is a plentiful and adaptable biopolymer present in plant cell walls, functioning as both a structural element and a source of energy. Cellulose-based materials are being thoroughly studied as potential replacements for conventional materials because of their cost-effectiveness, large surface area, non-toxicity, renewability, biodegradability, and eco-friendly characteristics. Using renewable and sustainable resources is crucial for transitioning to a green approach, as it tackles energy storage challenges and promotes environmental conservation. Much work has gone into creating electrochemical energy storage (EES) devices made of sustainable and environmentally friendly materials, such as supercapacitors. The aim is to substitute fossil fuels and enable the fast growth of batteries and supercapacitors (SCs).^1,2

The electrochemical and structural properties of their electrode materials significantly impact the key features of energy storage devices. While activated carbon (AC) is commonly used as the electrode material in energy storage devices, its commercialization has been hindered by low energy and power densities caused by high equivalent series resistance.³ The energy and power densities of commercially available electrochemical capacitors based on AC electrodes are thus restricted to 1 to 10 W h kg⁻¹ and 2 to 10 kW kg⁻¹, respectively. To enhance the energy and power densities of SCs, single-walled carbon nanotubes (SWCNTs) offer a promising alternative as electrode materials. CNTs possess excellent conductivity, significant specific surface area, low mass density, and remarkable mechanical strength.⁴ Furthermore, cellulose exhibits remarkable physical/chemical/thermal properties that make it suitable for various applications in energy, sensors, medical devices, transistors, actuators, electric heaters, electromagnetic interference (EMI) shielding materials, and hybrid applications, particularly in the nanoscale domain.^5–10

Additionally, it possesses exceptional mechanical properties and abundant active sites.¹¹ These qualities make it a promising candidate for acting as a binder or active material when combined with different electrochemical constituents.^12,13 It has become a trend to utilize waste materials for high-end applications, taking the context of sustainability and economic viability into account. Such waste materials are generally rich in cellulose content, which can be utilized for versatile applications.

Utilizing biomass-derived carbon materials as binder-free, freestanding electrodes is a particularly promising approach, as it eliminates the need for additional inactive components and enables high mass loading while maintaining excellent electrochemical properties.¹⁴ Cellulose is a natural, renewable, and biodegradable material, making cellulose-based binders an environmentally friendly alternative to conventional binders like polyvinylidene difluoride (PVDF), which is essential for sustainable energy solutions by minimalizing the use of different hazardous chemicals and offering a more biocompatible alternative to conventional polymers for energy storage devices.¹⁵ Recently, Guo et al. observed a simple yet effective method for bonding binders and electrolytes, which remarkably improved the cycle stability of high-capacity anodes utilizing K⁺ ions.¹⁶

The CNF binder retains BN-Ac's micropores and improves wettability, while the PVDF binder fills micropores and hinders ion-conductive pathways. The as-synthesized BN-Ac/CNF exhibits a capacitance of 268.4 F g⁻¹ at 5 A g⁻¹, which is 1.4 times higher than that of BN-Ac/PVDF. In addition, the energy/power density improves for replacing PVDF with CNF and outstanding capacitance and decreased bulk electrolyte resistance application of CNF in energy devices.¹⁷

Their mechanical characteristics, including their flexibility and tensile strength, also contribute to the improved cyclic performance of SC electrodes. The strong hydrogen bonding network within the CNC helps maintain the electrode's structural integrity during charge–discharge cycles, leading to enhanced stability and longer lifespan.¹⁸ Han et al. observed that upon surface coating of the PANI shell, the proposed PANI@CNT–CNC/PVA–PAA nanofibrous membranes show a large porosity, excellent tensile strength of ∼54.8 MPa, and an electroconductivity of ∼0.44 S m⁻¹ and capacitance of 164.6 F g⁻¹ with 91% retention after 2000 cycles.¹⁹

When cellulose is defined as having at least one dimension less than 100 nm, it is referred to as nanocellulose (NC). It can be obtained from various sources, including bio, agricultural, and plant wastes.²⁰ Using top-down synthetic approaches involving various chemical and physical procedures, NC can be produced from cellulose-rich biomass, such as pea hull waste, pineapple peel,^12,20 areca nut husk,²¹ cotton,²² and many more.²³

The remarkable flexibility, affordable cost, renewable nature, and environmental friendliness of cellulose make it a crucial component in developing diverse electrode materials, separators, and electrolytes for flexible SCs.^9,24 As an electrode material, hydrophilic NC prevents the aggregation of active materials like CNT and enhances electrode wettability for absorbing liquid electrolytes. As a separator, NC creates a porous structure that allows electrolyte ions to move more efficiently, increasing energy storage efficiency. Islam et al. proposed NC as a separator material in their study because of its strong ion conductivity and weak electron conductivity. The transport of electrolyte ions is facilitated by the porous structure created by NC acting as a separator, increasing energy storage efficiency.²⁵

Besides serving as an electrode and separator, NC also exhibits excellent potential as an electrolyte. Recently, Chen et al. reported that hydrogel electrolytes with excellent mechanical performance, good repairability, adhesion ability, and promising electrochemical performance could be obtained using hydrophobic interactions.²⁶ Overall, cellulose is a very promising material for improving the efficiency and environmental sustainability of SCs because of its availability, porosity, mechanical strength, biocompatibility, ion conductivity, and sustainability.

Utilizing agricultural waste materials to extract NC offers a viable waste management and environmental protection approach. NC, upon its surface tuning with charged moieties, exhibits favorable characteristics that render it suitable for various electrochemical and energy storage applications. The reactive surfaces, abundant in –OH groups, offer several hybridization possibilities with other active materials for creating composite electrodes/separators/electrolytes based on nanocellulose. Hence, NC is a desirable material for flexible energy storage devices.²⁷ Its carbon content, high aspect ratios, and ease of surface modification make it an excellent starting material for producing carbon electrodes with large pore structures, surface areas, and adjustable microstructures.²⁸ On the other hand, NC has also been discovered to contribute to achieving an advanced specific capacitance of the nano-composite electrodes by reducing the stacking of nanosheets and facilitating the ion movement of electrolytes.²⁹

Additionally, it is a strong support in the arrangement of CNT sheets. Besides these qualities, the combination results in decreased electrode conductivity, requiring additional modifications to offset the loss of conductivity. To meet this requirement, it is crucial to have an ideal concentration of modified CNC and CNT, as conducting polymers (rGO) can enhance the conductivity of active materials and introduce pseudo-capacitance.³⁰ This incorporation effectively mitigates the restacking of CNT sheets, improving the interaction between the electrode and electrolyte. Consequently, such an approach enables the attainment of superior electrochemical performance.

Researchers have explored various surface modification techniques to improve CNC's conductivity and electrochemical performance, like TEMPO-mediated oxidation to introduce carboxylic groups on the CNC surface and sulfonation, acetylation, and silylation to further functionalize the CNC surface.³¹ These surface alterations enhance the compatibility and interaction between CNC and conductive polymers or other active materials in electrode composites. The novelty of the preparation method is the modification of cellulose by an MA moiety and incorporation with an SWCNT sheet, its use of a hybrid composite as an electrode composite with the CNT sheet, and the assessment of the proposed composites' electrochemical performance and thermal stability.

In this study, a novel renewable, eco-friendly, cost-effective, and agriculture-waste-based electrode, which does not contain any binders, i.e. a binder-free electrode, is fabricated for application in supercapacitors. The fabrication includes coating the SWCNT using MA-modified CNC. This combination prevents the stacking of CNT sheets, establishes a framework for arranging them, and incorporates pseudo-capacitance. The optimized concentration of binary nano-composites (9 wt% MACNC/SWCNT) had the best electrochemical properties. CNT can also raise the thermal stability and conductivity of the material.³² This study converts waste materials into efficient electrode materials and enhances the electrochemical properties of CNC-based SC electrodes derived from agricultural waste. The electrochemical performance of the produced MACNC/SWCNT ternary nano-composite was assessed as an electrode material for SCs using 1 M Li₂SO₄ electrolyte. This allows for the creation of eco-friendly and cost-effective electrodes for high-performance SCs.

2. Experimental details: materials and methods

2.1 Materials used

The agricultural waste known as sugarcane bagasse, which has a scientific nomenclature of Saccharum officinarum, was obtained from a local field in Haryana, India. The SWCNT sheet was acquired from Poland, and 99.9% pure ethanol (C₂H₅OH) was purchased from Changshu Hongsheng Fine Chemical Co. Ltd. Other chemicals used in the study, like toluene (C₇H₈), potassium hydroxide (KOH), sodium sulfite (Na₂SO₃), sodium chlorite (NaClO₂), dimethyl sulfoxide ((CH₃)₂SO), glacial acetic acid (CH₃COOH), and dialysis membrane hydrochloric acid (HCl), were procured from Himedia Pvt. Ltd. The distilled water was collected from a double distillation setup. All the analytical grade compounds from AR were used without any additional purification for this study.

2.2 Cellulose extraction from sugarcane bagasse and synthesis of cellulose nanocrystals (CNC)

A schematic representation and detailed process of synthesis of chemically purified cellulose (CPC) followed by cellulose nanocrystal (CNC) extraction from the agricultural waste Saccharum officinarum are provided in ESI section S1, and Fig. S1 and S2.†

2.2.1 Surface modification of CNC: maleic anhydride grafting on cellulose. The CNC surface can be altered using chemical modification like esterification, amination, acetylation, benzoylation, salinization, and polymer grafting, as well as physical techniques, including the addition of surfactants, compatibilizing agents, and polymers. Several paths may be taken to obtain surface-functionalized CNC.³³ The hydrophilic nature of CNC, especially sulfonated and carboxylated CNC, prevents them from dispersing in hydrophobic matrices and makes them non-reactive with high crystallinity. Consequently, to improve their physicochemical qualities and enable novel applications, appropriate groups of CNC must be comprised.³⁴ Recent challenging requirements in creating applications for energy storage devices have been addressed by the surface functionalization of CNC.³⁵ The CNC surface can be altered with a chemical method, i.e., esterification, as shown in Fig. 1.

Fig. 1 (a) Modification of CNC with maleic anhydride via a chemical process. (b) Proposed reaction scheme of maleic anhydride (MA) grafted on cellulose and MACNC.

The dry powder of the extracted CNC (1.0 g) was dissolved in 20 ml DMF using a bath sonicator to modify their surface using maleic anhydride (MA). After adding 8.0 g of MA sample to the suspension, it was stirred for about half an hour, or until the MA completely dissolved. The surface tuning was done for 7–8 h with continuous stirring in an oil bath at 110 °C. The mixture was stirred for two hours before being centrifuged for twenty minutes at 8000 rpm after the reaction, and 100 ml of ethanol was added. After recovering the sediment, it was re-diluted with 150 ml of purified ethanol. Three further centrifugal cleanses were carried out to eliminate any remaining unreacted MA and DMF. Lastly, distilled water was used to disperse the MACNC which was then freeze-dried. Fig. 1(a) depicts the suggested reaction route for MA modification, and Fig. 1(b) predicts the chemical reaction between cellulose and MA and synthesized MACNC.

2.3 Material characterization

The composite's morphology and structure were examined using a field emission scanning electron microscope (FE-SEM, MIRA3 TESCAN, USA). A high-resolution picture was produced by scanning the surface of the nanoparticles with a concentrated electron beam that was accelerated between 5 kV and 10 kV. A PerkinElmer instrument (Model No. C91158, NJ, USA) was used to perform the FTIR analysis to analyze the generation and existence of different functional groups in the material collected in the course of experimentation at a resolution of 4 cm⁻¹ per scan over an array between 4000 cm⁻¹ and 520 cm⁻¹. The samples' crystallinity index (C. I.) was assessed using Japanese Rigaku Ultima IV XRD equipment with a Cu K-alpha radiation source. Digital multimeter-15B was purchased from Sai Scientific and Traders to check the material's electrical characteristics. Atomic Force Microscopy (AFM, Bruker Dimension Icon, USA) in ScanAsyst mode was conducted to study the sample's surface morphology. To obtain diffraction curves from 5–55° (2° scale), scanning was done at 4° min⁻¹. The peak force tapping variant was calibrated to capture the high-resolution surface images. To analyze the AFM images further, Nanoscope Analysis software was employed. A Thermogravimetric Analyzer (TGA, 55, TA Instruments, USA), was used to analyze the thermal deterioration and stability of the prepared samples.

2.4 Role of MACNC in electrodes and its fabrication methods

CNC are naturally insulating materials.³⁶ To improve their conductivity or surface charge, it is necessary to modify their surface by adding conductive materials/polymers or charged moieties. In this study, we grafted cellulose with maleic anhydride through an esterification reaction and incorporated the modified cellulose onto a sheet of SWCNTs. The resulting cellulose/carbon hybrid composite material exhibited enhanced electrochemical performance, energy, and power density along with thermal, structural stability, and other conductivity/mechanical properties compared to the individual components.³⁷ This is due to the synergistic effects between the modified cellulose and the exceptional characteristics of CNTs, which are known for their high modulus, large surface area, and excellent tensile strength, resulting in a highly conductive and mechanically robust material.

The electrode was fabricated using the following procedure: synthesis of MACNC solution, electrode dip-coating, drying, and sample preparation. The first step involved the synthesis of a modified cellulose nanocrystal (MACNC) solution. A specific area of the electrode was then dipped overnight in various concentrations of the MACNC solution. The MACNC solution was subjected to constant magnetic stirring for 2 h during this process. After the overnight dipping, the electrode samples were removed from the MACNC solution and placed in an oven for drying. The dried electrode samples were then ready for further use and characterization as shown in Scheme 1, which illustrates an electrode fabrication scheme with different wt% and then evaluates their electrochemical measurements for optimization.

Scheme 1 Illustration of an electrode fabrication scheme with different composites and their electrochemical measurements.

2.5 Electrochemical measurement

On an Autolab PGSTAT302N potentiostat/galvanostat electrochemical workstation with the setup of a three-electrode system, the MACNC/CNT, CNC/CNT, and pristine CNT composites were subjected to CV, EIS, and GCD, where the working electrodes were different wt% of MACNC/CNT nano-composite (1 cm × 1 cm), Ag/AgCl was used as the reference electrode, and platinum was chosen to serve as the counter electrode. Utilizing 1 M Li₂SO₄ electrolyte solution, platinum served as the counter electrode, Ag/AgCl as the reference electrode, and CNC/CNT and MACNC/CNT as the working electrode. EIS experiments were carried out at an open circuit potential using a sinusoidal alternating voltage with a frequency range of (10⁻²–10⁵) Hz and an amplitude of 5 mV. For the synthesis of the electrode material for the energy storage device, simply take the SWCNT sheet of (1 cm × 1 cm) electrode and dip for 12 h in the solution of CNC and MACNC (3 wt%, 6 wt%, 9 wt%, 12 wt%) followed by drying in an air oven and electrochemical measurements and other structural analyses.

The voltage range for CV was −0.6 V to +0.6 V, i.e., the potential window is 1.2 V, with several scan rates used. The specific capacitance, power/energy density, and cyclic efficiency were determined by eqn (S1)–(S3)† mentioned in ESI† sections. Due to its strong ionic character, 1 M Li₂SO₄ solution was used as the electrolyte for the experimental work.

The hydrophilicity of the membranes was increased by incorporating CNC or functionalized CNTs, which could also impact the surface characteristics of the MA-grafted CNC/CNT composite. In combination with CNC-based electrode materials, further research is needed to investigate the performance of other electrolyte types, such as organic solvents, ionic liquids, and solid-state electrolytes. The search results indicate that aqueous electrolytes like Li₂SO₄, sodium sulfate (Na₂SO₄), and potassium sulfate (K₂SO₄) have been explored for CNC-based electrodes. The performance of these aqueous electrolytes is influenced by the mobility and size of the alkali metal ions, with Li⁺ exhibiting the best capacitance and charge propagation properties. Aqueous electrolytes can provide good ionic conductivity and compatibility with CNC-based electrodes, making them a suitable choice.

3. Results and discussion

3.1 Structural analysis

3.1.1 FTIR analysis or Fourier transform infrared spectroscopy. To verify the successful extraction of the cellulose nanocrystals and their chemical modification with maleic anhydride moieties, the FTIR curves of CNC and MACNC (Fig. 2(a)) were examined. At the typical wavenumbers of 1745 cm⁻¹ and 1514 cm⁻¹ in CNC, there is no change in transmittance, indicating that nanocrystals have been successfully isolated, as depicted in ESI S2.1.† Fig. S3(a)† shows the FTIR spectra of the raw sugarcane bagasse, CPC, and CNC. The transmittance change corresponds to the stretching of the ester bond and C=C between lignin and hemicellulose, respectively.³⁸ Following chemical alteration, some characteristic wavenumbers (e.g., C–O–C stretching in the pyranose ring at 1055 cm⁻¹, glycosidic ether at 1160 cm⁻¹, and C–O stretching and C–H rocking vibrations in the pyranose ring at 1030 cm⁻¹ and 1100 cm⁻¹, respectively) demonstrate the intact native cellulose structure. The broad spectra (3800–3000 cm⁻¹), with a strong peak at 3350 cm⁻¹ in both the MACNC and CNC, are caused by the H-bonded hydroxyl groups being stretched by the hydroxyl group. From CNC to MACNC, the band intensity decreased slightly due to the surface modification, which affected the H-bonding between the cellulosic chains, as shown in Fig. 2(a).

Fig. 2 (a) ATR-FTIR spectra; (b) X-ray diffraction (XRD) results of cellulose nanocrystals (CNC) and maleic anhydride cellulose nanocrystals (MACNC).

The intensity of the anomeric carbon (C–H) vibration of β-glycosidic links between the cellulose units at 897 cm⁻¹ reduces from CNC to MACNC after the chemical alteration, and it does not overlap, in contrast to C–H_n (n = 1, 2) stretching. This is attributed to the sensitivity of the gluco pyranose ring to change in pendant groups, which indicates the chemical structural change on modification. Indeed, a strong appearance of intense change in transmittance at 1720 cm⁻¹ in MACNC corresponds to the carboxyl group (–O–C=O), establishing the formation of ester linkage on the surface of cellulose molecules.³⁹ Additionally, upon modification with MA, the vinyl group (C=C) is responsible for the peak at 1634 cm⁻¹. These findings show that MA may be successfully grafted onto the CNC surface.

3.1.2 X-ray diffraction (XRD) analysis. An XRD instrument equipped with a Cu K-alpha X-ray source (40 mA, 40 kV, 4° min⁻¹, and 2θ angle ranging from 5° to 55°) was used to examine the phase structure and crystallinity of the prepared samples. The XRD measurements of the CNC and MACNC samples are presented in Fig. 2(b). All the samples show the crystalline structure of the cellulose-based materials CPC, CNC, and MACNC with three diffraction peaks (Fig. S3(b)† in the ESI shows the XRD patterns of raw, CPC, and CNC). The typical peaks of the cellulose-I type structure, which correspond to the lattice planes 110, 200, and 004, were identified at 2θ = 16.45°, 22.55°, and 35.46° in the crystalline cellulose XRD patterns.⁴⁰ The obtained XRD results validate the successful formation of CNC and modification of CNC nanomaterials with a high degree of purity. However, the C. I. for MA-modified cellulose nanocrystals (MACNC) (56% C. I.) was lower than that of the CNC, i.e. 61% C. I., computed using the Segal method.⁴¹ This suggested that during modification, the cellulose chain packing and crystallinity were partially destroyed. The reduction in C. I. of MACNC from CNC resulting from the disruption in the surface H-bonding of CNC has been proven by amorphous subtraction.

3.2 Morphological analysis

3.2.1 SEM and SEM-EDX analysis. The previous section established the structural arrangements and the changes in the structure of CNC during the modification to MACNC. Apart from the structure of the molecules, the morphological arrangements will also play an important role in governing their properties. Taking this concern into account, the effect of morphology has been investigated using various characterization methods.⁴² SEM-EDX images confirmed the surface morphologies and dispersion of CNC and MACNC on the surface of the carbon nanosheet. The system was operated at an accelerated voltage of 10.0 kV at room temperature. The FESEM images of raw, CPC, CNC, and CNT sheets and cross-section SEM of the proposed material, i.e., 9 wt% AMCNC/CNT electrode, have been shown in Fig. S4† in the ESI.†Fig. 3 shows the SEM morphology of the cellulose/CNT nano-composite, shows a coating of modified cellulose over the MWCNTs, and confirms the excellent dispersion. The rough surfaces caused by the nanoparticles in the cellulose/CNT hybrid nano-composites help to promote the faradaic reaction and pore formation by creating a miniature transport route for electrolytes.⁴² The HCl used dissolved disordered regions while attacking orderly portions rather quickly, resulting in poor yield and smaller dimensional nanocrystals. CNC and MACNC have average diameters of 47.5 nm and 36.3 nm, respectively. By effectively grafting MA onto the nanocrystals, the diameter was decreased.

Fig. 3 FESEM images: (a) CNC, (b) MACNC, (c–c′′′) CNT, (d) CNC/CNT, (e) MACNC/CNT; (f) EDX images: CNC/CNT, (g) CNT, (h) MACNC/CNT electrodes; AFM images: (i) CNC, (j) MACNC, (k and k′) SWCNT, (l and l′) CNC/CNT and (m and m′) 9 wt% MACNC/CNT composite (2D and 3D) electrode materials.

The CNT sample images showed that the CNT assembled to form a 3D network with high porosity and excellent conductivity. To make it sustainable and sustain its porosity, the FESEM images of the CNC/CNT composite are also presented in Fig. 3. For the MACNC/CNT composite, better wrapping of MACNC over CNT sheets can be observed, which confirms the uniform and homogeneous interaction between them. It shows a unique 3D network topology with high porosity and excellent physical and mechanical durability. The high level of MACNC adhesion on the CNT's surface makes it promising for practical applications. According to the findings, CNTs' hydrophobic and poor solubility characteristics cause them to clump together (aggregate) and entangle.⁴³ The entangled nanotubes in these CNT aggregates create vacant regions that are difficult for a viscous fluid to disperse. The CNC/CNT composites reveal that due to the incorporation of cellulose on the CNT surface, they are not fully covered. Still, in the composite with 9 wt% MACNC, as presented in Fig. 3(a–e), the nanosheets are covered entirely and uniformly with MACNC. Further addition of MACNC may become extra layers on the already attached cellulose on the CNT surface, disturbing the surface homogeneity.

The components that are present in the samples have been identified through the use of energy-dispersive X-ray (EDX) analysis. The cellulose and composite sheet's EDX spectra reveal two prominent O and C peaks (as seen in Fig. 3(f–h)).

3.2.2 Atomic force microscopy (AFM) analysis. An AFM instrument was used to examine the surface topography of the CNC or MACNC on the CNT sheet using the ScanAsyst imaging mode. This technique is beneficial to determine the surface roughness of the prepared samples before and after the treatment. Using Nanoscope Analysis software, several significant characteristics from the height image were evaluated and summarized (as shown in Fig. 3 and Table S1†), including kurtosis value (R_ku), surface roughness (R_a), and skewness (R_sk). The average roughness value was determined to be 296 nm and 106 nm for CNT and MACNC/CNT, respectively. The skewness values represent the symmetry of the surface heights across the mean plane of MACNC and CNC. In addition, it was discovered that the kurtosis value was 3.33 and 2.51 for CNT and MACNC/CNT, respectively. The kurtosis values provided insight into the surface type of CNC/CNT and MACNC/CNT, with R_ku > 3 denoting a jagged, spiky surface and R_ku < 3 denoting a rugged surface. The results suggest that the surface roughness of the treated sample decreases from the CNT to MACNC. Nevertheless, the MACNC/CNT sample's surface roughness is notably lower than that of the others, as shown in Fig. 3(i–m′). This behavior highlights the favorable wetting properties of the fabricated electrode and effective interaction with the aq. Li₂SO₄ electrolytes. The AFM analysis also supports the uniformity and homogeneity of the RMS (Root Mean Square) over the sample surface, and Fig. 3 and Table S1† in the supplementary section provide information about the roughness, skewness (R_sk), and kurtosis values (R_ku) of the prepared electrode materials. For CNC and MACNC, the average roughness values were found to be 4.97 nm and 2.54 nm, respectively. This suggests that the surfaces of MACNC are smoother than those of CNC. The skewness values for CNC and MACNC were found to be 0.297 and −0.087, respectively. The skewness values show how symmetrically CNC and MACNC surface heights are distributed across the mean plane. This somewhat negative skewness score indicates that valleys dominate the MACNC surface instead of peaks. Additionally, the kurtosis values for MACNC and CNC were discovered to be 3.98 and 3.23, respectively. The optimized cellulose loading and dispersion inside the composite is responsible for the smoother surface of the MA-grafted CNC/CNT in comparison to CNC/CNT. Furthermore, the grafted chains lessen the reactivity of the –OH groups on the CNC, which might affect the composite's overall surface properties. This enhanced dispersion makes better contact between the CNTs and the surrounding matrix possible, which produces a smoother surface.⁴⁴ The reactivity of –OH groups on the CNC surface is also decreased by the presence of grafted chains, which might affect the composite's overall surface properties.

4. Characterization of electrochemical performance

By providing distinct routes for charge carriers, CNT can increase the charge storage capacity of SCs. They facilitate the movement of ions and electrons, allow for better electrolyte accessibility, improving overall performance. The results of the electrochemical experiments showed that the MACNC/CNT-based composite outperforms the pristine CNT electrode and CNC/CNT electrode, as shown in Fig. 4(a–c).

Fig. 4 (a) Electrochemical impedance spectroscopy (EIS), (b) cyclic voltammetry (CV), and (c) galvanostatic charge–discharge (GCD) curves of the CNC/CNT sheet and MACNC/CNT electrodes; (d) electrochemical impedance spectroscopy (EIS), (e) cyclic voltammetry (CV), and (f) galvanostatic charge–discharge (GCD) curves and (g) graph of scan rate (V s⁻¹) vs. energy density (mW h cm⁻²) of pristine CNT, CNC/CNT, and MACNC/CNT electrodes.

The low mass/volume energy density, poor cycle stability, and limited capacitance of CNC/CNT electrode materials make them unsuitable for large-scale applications. As a result, several research studies have effectively enhanced the electrochemical efficiency of the SC by creating richer pore architectures, raising the specific surface area, and combining it with pseudo-capacitive materials. Furthermore, adding conductive materials can improve the electrolyte's adsorption capacity and lower the interface resistance between the electrode and the electrolyte, improving the electrode's electrochemical stability, rate, and cycle stability.⁴²

Thus, the modification of cellulose with MA is required to improve performance. We attribute the MACNC/CNT-based electrode's superior capacitance performance to (i) high active-material mass loadings, (ii) the freestanding electrode's high active surface area, which offers more redox sites for charge storage, (iii) the increased interfacial contact between the porous surface and the electrolyte, (iv) the excellent wettability of the electrode, and (v) the high electrochemical performance, which facilitated quick charge transport. The amount of material deposited on the CNT film was evaluated by measuring the weight of the SWCNT sheet film before and after the MACNC deposition. The weight percentage of MACNC (W_MACNC%) was estimated at ∼70%, and mass loading is 25.34 mg on the surface of the (1 cm × 1 cm) CNT sheet, as discussed in the ESI† of the revised manuscript. Further, the thickness of active material loading on the proposed 9 wt% MACNC/CNT films is 20–30 μm and typically measured using cross-sectional analysis techniques, such as SEM, as shown in Fig. S4† in the ESI,†i.e., SEM cross-section of the 9 wt% MACNC/CNT proposed electrode material.

The proposed electrode material shows good capacitive behavior and electrochemical performance compared to CNC/CNT and CNT electrodes, EIS, CV, and GCD, and the graph between scan rate and energy density is shown in Fig. 4. The graph between the scan rate and specific capacitance of all the prepared electrode samples is displayed in Fig. S5† in the ESI.†

In addition, the Nyquist plot was plotted using EIS to examine the prepared nano-composite ion diffusion and charge transport behavior. As the Nyquist plot reveals, the X-axis indicates the resistive behavior, while the slope represents the excellent capacitance performance of the electrodes. Here, the observed plot illustrates that the as-prepared MACNC/CNT composite shows outstanding charge transfer ability in comparison to the two other samples of CNC, as shown in Fig. 4. Furthermore, at low frequencies, the sloping line of the prepared electrode suggests an approximately ideal capacitive behavior. The lower resistance and more significant slope indicated that the composite electrode possesses superior electrochemical performance than the others. The electrochemical properties, including areal capacitance and energy density, of CNT, CNC/CNT, and MACNC at a scan rate of 0.002 V s⁻¹ are represented in Table 1.

Table 1 The areal capacitance and energy density of samples MACNC/CNT, CNC/CNT, and CNT at a 0.02 V s⁻¹ scan rate

Electrode sample	Scan rate (V s^−1)	Areal capacitance (mF cm^−2)	Energy density (μW h cm^−2)
CNT	0.02	441.798	88.3596
CNC/CNT	0.02	320.88	64.176
MACNC/CNT	0.02	644.46	128.892

---

For the CV curves, all the curves showed analogous rectangle shapes, indicating that the composite (MACNC/CNT) sample exhibited superior pseudo-capacitance in comparison to others, which can also be proved by CV, GCD, and EIS curves (Fig. 4). In addition, this composite also provides a noticeable areal capacitance and energy density of 270.8903 mF cm⁻² and 54.1780 mW h cm⁻² at a current density of 0.005 A cm⁻² respectively. In contrast, the CNC/CNT electrode shows the areal capacitance of 138.9813 mW h cm⁻² and 64.176 μW cm⁻² at the same current density. This result illustrates that the proposed MACNC/CNT composite shows the best electrochemical performance for SC applications.

Finally, the outstanding electrochemical characteristics of the proposed MACNC/CNT electrode may be attributed to the combined effect of several reasons like the superficial electron transportation due to reversible redox reactions of SWCNT and MACNC, the removal of the restacking tendency of SWCNT by incorporating cellulose, resulting in a conductive and porous 3D network microstructure based composite electrode, because cellulose-based materials produce a suitable diffusion pathway for the ions of electrolyte, resulting in overall superior electrochemical performance of the supercapacitive composite electrode.

4.1 Multimeter test

A multimeter (also known as a volt–ohm–milliammeter or volt–ohmmeter) is a measuring device that can be used for various electrical characteristics. The electrode was checked with the help of a multimeter test by making small pellets of the sample. The results revealed that the CNC pellet showed the lowest electrical behavior comparable to that of the modified cellulose. Multimeter testing also revealed that the MACNC/CNT electrode exhibited higher capacitance or least resistance and energy density compared to the CNC/CNT or CNT-based electrode. However, a significant change was noticed in the case of MACNC over the SWCNT sheet, as shown in Fig. S6.† The multimeter shows the electrical characteristics of CNC, CNT, MACNC, and MACNC/CNT sheet (see the ESI†).

4.2 Optimization of the electrode material

We investigated the performance of electrode materials with different wt% (3 wt% to 12 wt%) of MACNC/CNT and tested their electrochemical properties. In the case of 3 wt% MACNC, the electrochemical behavior was observed to be least favorable compared to other wt%. This can be attributed to insufficient interaction between CNTs and lowest wt% of modified cellulose (MACNC). The interaction may not be strong enough at low cellulose content to form a stable and uniform distribution. The 3 wt% cellulose content may not optimize the porosity and surface area of the nano-composite (inadequate porosity and surface area) and inefficient charge transfer, which are crucial factors for achieving high capacitive behavior and energy storage capabilities.⁴⁵ The different weight percentages of modified cellulose and its incorporation on the SWCNT sheet are shown in Fig. 5(a). Optimizing the electrode material, electrolyte composition, and device design for low resistance and high surface area is crucial for achieving high power performance at large current loads. At very high current densities, heat generation can become an issue. Proper thermal management is required to prevent performance degradation.⁴⁶ Several studies focus on further improving power capabilities for such high-current-density applications.

Fig. 5 CV curve of the (a) 3 wt%, (b) 6 wt%, (c) 9 wt%, and (d) 12 wt% MACNC/CNT at different scan rates, namely 0.01 V s⁻¹, 0.02 V s⁻¹, 0.05 V s⁻¹, 0.1 V s⁻¹, 0.2 V s⁻¹ and 0.5 V s⁻¹.

The further addition of MACNC varied from 3 wt% to 12 wt%. The decrement of electrochemical performance beyond 9 wt% and excellent areal and specific capacitance comes at optimum 9 wt% MACNC/CNT, as displayed in Fig. 5(a–d).

The fabrication of 9 wt% MACNC to CNT surfaces has been observed to exhibit exceptional electrochemical behavior due to factors such as strong interaction modified CNC and CNTs leading to a more stable and uniform distribution of CNTs within the MACNC matrix,⁴⁷ enhanced mechanical/thermal properties of the CNT-based nano-composites,^47–49 optimized porosity and high surface area,⁵⁰ which facilitates the adsorption of electrolyte ions thereby enhancing the performance.⁵¹ These findings suggest that the 9 wt% cellulose content may be an optimal choice, and the optimal charge transfer can be attributed to the synergistic effect of MACNC and CNTs, which can lead to enhanced electrochemical performance, as shown in Fig. 5(c).

The detailed calculation in tabulated form as area, capacitance, and energy density of other wt% has been provided in Tables S2–S5 of the ESI section S3.† The electrochemical performance of 9 wt% MACNC/NCT was examined by CV/GCD curves of these samples, which were recorded at different scan rates and current density, respectively, using an aqueous 1 M Li₂SO₄ electrolyte, as depicted in Fig. 6(a and b). The graph between power density and current density, the relation between energy density and power density (Ragone plot), and the graph of energy density with current density and scan rate, as shown in Fig. 6(c–f) and Table 2, show the calculated value of areal capacitance and energy density from cyclic voltammetry. Notably, all the analyzed samples exhibited symmetrical and analogous rectangular curves, which indicate the possession of fast reaction kinetics. Direct mixing of CNT and MACNC forms a porous matrix that facilitates electrolyte infiltration and improves electronic conduction.

Fig. 6 The graphical illustration of the study demonstrates the (a) cyclic voltammetry, (b) galvanostatic charge–discharge of MACNC/CNT sheet at different scan rates (V s⁻¹), (c) current density (A cm⁻²) vs. power density (μW cm⁻²), (d) Ragone plot, (e) current density vs. energy density (μW h cm⁻²), and (f) scan rate vs. energy density.

Table 2 The calculated value from cyclic voltammetry of areal capacitance and energy density from the CV of the proposed electrode

9 wt% CV scan rate (V s^−1)	Area	Capacitance (mF cm^−2)	Energy density (μW h cm^−2)
0.005	0.00579	1389.202	277.84032
0.01	0.00532	638.9299	127.78598
0.02	0.00906	543.44243	108.68849
0.05	0.01264	303.46244	60.69249
0.1	0.01408	169.00158	33.80032
0.2	0.01508	90.46439	18.09288

---

Furthermore, the redox peaks observed in CV curves indicate the co-existence of electric double-layer capacitive behavior and pseudo-capacitance. The proposed electrodes showed a specific capacitance of 62 F g⁻¹ at a current density of 0.005 A g⁻¹, and the gravimetric power density and energy density are 819 W kg⁻¹ and 14.8 W h kg⁻¹, respectively. Further, the electrode exhibited a maximum areal capacitance of 1389.202 mF cm⁻² with a voltage window of 1.2 V calculated from curves.

It has been observed that 9 wt% MACNC/CNT gave the best electrochemical performance due to the synergistic effect of the carboxylic group on the surface of cellulose (MACNC) and high surface area of CNT, enhancing the overall faradaic performance of the electrode material.

The presence of the –COOH group can improve the wettability of the electrode and promote better ion transport. At the same time, the CNT sheet can provide a conducive environment for efficient electron transport in the system. The intercalation of Li⁺ ions with the composite from the 1 M Li₂SO₄ electrolyte can also influence the faradaic mechanism. Fig. 7(a and b) illustrate the chemical structure of MACNC and adsorption mechanisms involving Li ions on the material surface and the carboxylic acid group of MACNC. The possible proposed mechanism of the carboxylic group present on the surface of the cellulose nanocrystals and CNTs with the Li₂SO₄ electrolyte is shown in Fig. 7.

Fig. 7 The proposed mechanism and chemical structure of (a) MACNC; (b) adsorption mechanisms involving Li ions on the material surface and carboxylic acid group of MACNC; (I) chelation, (II) bridging.

Following the GCD curves, all the samples displayed almost symmetrical triangular shapes, indicating the materials' reversible electrochemical process. This is a clear indication of excellent capacitance behavior. The capacitance and power density at 0.02 A cm⁻² current density were calculated from GCD curves to be 34.88167 mF cm⁻² and 12 000 μW cm⁻², respectively, and at a current density of 0.05 A cm⁻², capacitance and power density were 68.72 mF cm⁻² and 30 000 μW cm⁻² respectively. Furthermore, the charge/discharge curve also suggested the highly reversible redox reaction-based faradaic pseudo-capacitance behavior, which originated from the addition of MACNC.

The above section shows the graph obtained from the above-mentioned table, which shows the energy density of the proposed 9 wt% MACNC/CNT electrode at various scan rates.

The development of high-performance energy storage materials is crucial for the advancement of renewable energy technologies. Modified CNC has emerged as a promising candidate for enhancing the electrochemical properties of CNT or other composites. This research study investigates the optimal ratio of MACNC/CNT for electrochemical performance. It discusses the results in the context of existing literature, as shown in Table S6† (in the ESI). Table S6† and Fig. 8 present a comparison of the electrode materials used, as well as their areal capacitance, energy, and power density at several scan rates and current densities, with the present study and other materials that have been previously described.

Fig. 8 Radar (spider plot) comparison plot of the MACNC/CNT hybrid sheet composite with other works in the literature.

An essential way to assess an electrode's electrochemical performance and appropriateness for SC usage is through cycle performance analysis. Using GCD experiments at a current density of 0.02 A cm⁻², the cycling stability of 9 wt% MACNC/CNT was assessed. As depicted in Fig. 9, the capacitance efficiency, retention ratio, or coulombic efficiency (%) versus cycle number for the 9 wt% MACNC/CNT electrode was 74.6% after 12 000 cycles. In this regard, cellulose, a naturally occurring, renewable, nontoxic polymer, has offered a cohesive, durable, and mechanically stable electrode structure with less electrode cracking or delamination during cycling.

Fig. 9 Cycling stability of the 9 wt% MACNC/CNT proposed electrode.

5. Thermal behavior: thermal gravimetric analysis (TGA)

The thermal stability of CNC, CNT, CNC/CNT, and MACNC/CNT nano-composites was analyzed using TGA and Derivative Thermogravimetric (DTA) curves, as shown in Fig. S7 and S8† in the ESI.

The initial ∼2 wt% loss between 140 and 240 °C can be attributed to the evaporation of tightly bound moisture to the carboxyl groups and the decarboxylation of substituted maleic anhydride in the disordered regions,^52,53 which can be corroborated by the decrement of the crystallinity index (C. I.) of MACNC from the XRD analysis. The DTA curve reached a steady state between 240 and 250 °C before the accelerated decomposition began at 250 °C. The maximum weight loss between 250 and 400 °C corresponds to the depolymerization and cleavage of glycosidic linkages of cellulose with a maximum degradation temperature of 320 °C. The single ultimately synthesized MACNC and deprivation of CNC indicate a controlled degree of substitution and narrow size distribution of nanocellulose, as displayed in Fig. S7† thermal stability – TGA and DTG curves of CNC and MACNC and Fig. S8† thermal stability – TGA and DTG curves of CNT (blue color), CNC/CNT (pink color), and MACNC/CNT (green color), in the ESI.† Using traditional H₂SO₄ improves the suspension stability at the expense of thermal stability, differentiating between the ordered and amorphous regions, leading towards a wide distribution spread, which constrained the nanocrystals for thermal applications. With HCl, the nanocrystals are not distinguished into two areas of thermal degradation and exhibit a higher C. I. with a narrow distribution.^54,55 Somseemee et al. modified the CNC with maleic anhydride, which was extracted from H₂SO₄, which displayed a two-stage degradation between 220 and 300 °C and 330 and 400 °C with two DTA peaks at 270 and 350 °C, which is in agreement with the statements mentioned above.⁵⁶

The extracted CNC has exhibited a single-step degradation, unlike the CNC isolated through H₂SO₄, which displays a multi-stage degradation profile.⁵⁷ The isolated CNC via HCl showed a relatively higher onset temperature (250 °C) than CNC extracted through H₂SO₄. Unlike CNC from H₂SO₄, the maximum degradation temperature (351 °C) with a maximum weight loss of 86 wt% was in the region between 250 and 400 °C. This can be attributed to the absence of sulfate groups (–OSO₃⁻), which decreases activation energy for pyrolysis. However, CNC modified with the MA moiety displayed a multi-step degradation with two peaks in the DTA curve at 144 °C and 322 °C. Eventually, the residual char of MACNC and CNC is 2 wt% and 7 wt%, respectively, and the DTA peak at 555 °C could correspond to the degradation of interior nanocrystals, which are not modified.

6. Conclusion and future directions of the work

In summary, modified NC and CNT-based materials were utilized for the creation of an electrode material for supercapacitors with improved electrochemical performance. This work involved grafting cellulose with maleic anhydride via an esterification reaction to increase the cellulose's surface charge (by tuned –OH groups with a conjugated –COOH group). The modified cellulose was then incorporated onto an SWCNT sheet to form composites with varying weight percentages (3 wt% to 12 wt%) and their electrochemical measurements were then conducted for optimization. MACNC may be evenly anchored on the surface of CNT with plenty of oxygen-containing functional groups confirmed by SEM and AFM testing during the chemical process for the suggested proposed electrode of 9 wt% MACNC/SWCNT. The proposed electrode demonstrates superior electrochemical performance due to the synergistic effects of CNC grafted with the maleic anhydride moiety and conductive carbon nanosheets. For the negative electrode of porous carbon, interconnected carbon frameworks with hierarchical porous structures can be achieved by using 1 M Li₂SO₄ electrolytes, which play a significant role and show significantly enhanced electrochemical properties. For the supercapacitor, the proposed electrode, i.e., 9 wt% MACNC/CNT, can reach a relatively high areal capacitance of 1389.202 mF cm⁻² and energy density of 277.84032 μW h cm⁻² and specific capacitance of 62 F g⁻¹ at a current density of 0.005 A g⁻¹ and the gravimetric power density and energy density are 819 W kg⁻¹ and 14.8 W h kg⁻¹ respectively with 76.4% retention after 12 000 cycles and favorable thermal stability. This approach offers new opportunities for producing high-performance electrodes for energy storage systems utilizing sustainable resources in a simple, economical, and widely feasible way. The benefits of the suggested composite materials in energy storage, artificial skin, biometric devices, biomedicine, and aerospace as well as sensing and environmental protection will make them more useful in engineering applications.

Abbreviations

MACNC	Maleic anhydride-tuned cellulose nanocrystals
SWCNT	Single-walled carbon nanotubes
EES	Electrochemical energy storage
SC	Supercapacitors
AC	Activated carbon
EMI	Electromagnetic interference
NC	Nanocellulose
CNT	Carbon nanotubes
rGO	Reduced graphene oxide
CNC	Cellulose nanocrystals
MA	Maleic anhydride
EIS	Electrochemical Impedance Spectroscopy
HCl	Hydrochloric acid
CPC	Chemically purified cellulose
DMF	Dimethyl formamide
CV	Cyclic voltammetry
GCD	Galvanostatic charge–discharge
FE-SEM	Field emission scanning electron microscopy
C. I.	Crystallinity index
AFM	Atomic force microscopy
TGA	Thermal gravimetric analysis
RMS	Root Mean Square
ATR-FTIR	Attenuated total reflectance-Fourier-transform infrared spectroscopy
Ag/AgCl	Silver/silver chloride
XRD	X-ray diffraction
R sk	Skewness
R ku	Kurtosis value
EDX	Energy dispersive X-ray spectroscopy
DTA	Derivative thermogravimetry
C s	Specific capacitance
PVDF	Polyvinylidene difluoride
BN-Ac/CNF	Binders for biochar/cellulose nanofiber
PVA–PAA-PANI	Poly(vinyl alcohol)/poly(acrylic acid)

Data availability

The authors confirm that the data supporting the findings of this study are available within the article and in its ESI.† Furthermore, they will be shared on request.

Author contributions

Nitesh Choudhary: original draft preparation, conceptualization, data collection, investigation, methodology, writing, data curation. Shiva Singh: writing – original draft preparation, data curation. Gaurav Malik: writing and calculation. Shakshi Bhardwaj: data collection and reviewing. Siddharth Sharma: reviewing and writing. Akshay Tomar: data collection and correction. Sheetal Issar: correction and data collection. Ramesh Chandra: supervision, data collection, conceptualization, reviewing. Pradip K. Maji: supervision, conceptualization, data collection, reviewing.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

Advanced Material Research Laboratory, IIT Roorkee, and PMRF (Prime Minister Research Fellowship), Government of India (GoI), are warmly acknowledged by the author for their financial assistance.

References

1. R. R. baby, Adv. Energy Mater., 2012, 2(3), 381–389.
2. S. Dutta, A. Bhaumik and K. C. W. Wu, Energy Environ. Sci., 2014, 7, 3574–3592.
3. S. Sharma, R. Adalati, N. Choudhary, B. S. Unnikrishnan, M. Sharma, P. Gopinath and R. Chandra, J. Alloys Compd., 2023, 960, 170749.
4. E. Frackowiak and F. Béguin, Carbon, 2001, 39, 937–950.
5. A. Izadi-Najafabadi, S. Yasuda, K. Kobashi, T. Yamada, D. N. Futaba, H. Hatori, M. Yumura, S. Iijima and K. Hata, Adv. Mater., 2010, 22(35), E235–E241.
6. A. B. Dalton, S. Collins, E. Muñoz, J. M. Razal, V. H. Ebron, J. P. Ferraris, J. N. Coleman, B. G. Kim and R. H. Baughman, Nature, 2003, 423, 703.
7. J. Chen, A. I. Minett, Y. Liu, C. Lynam, P. Sherrell, C. Wang and G. G. Wallace, Adv. Mater., 2008, 20, 566–570.
8. A. Aggarwal, E. G. Strangas and J. Agapiou, 2019 IEEE International Electric Machines and Drives Conference, IEMDC 2019, 2019, pp. 351–358.
9. N. Choudhary, A. Kumar, H. Jaiswal and R. Dharmaraju, AIP Conf. Proc., 2024, 2995, 020082.
10. S. Singh, S. Bhardwaj, P. Tiwari, K. Dev, K. Ghosh and P. K. Maji, Mater. Adv., 2024, 5, 2622–2654.
11. Y. Liu, Q. Fu, J. Mo, Y. Lu, C. Cai, B. Luo and S. Nie, Nano Energy, 2021, 89, 106369.
12. J.-H. Kim, D. Lee, Y.-H. Lee, W. Chen, S.-Y. Lee, J.-H. Kim, D. Lee, Y.-H. Lee, S.-Y. Lee and W. Chen, Adv. Mater., 2019, 31, 1804826.
13. J. A. Ajani Lakmini Jayarathna, S. Hajra, S. Panda, E. Chamanehpour, I. Sulania, M. Singh Goyat, S. H. Hsu, H. Joon Kim, T. Treeratanaphitak and Y. Kumar Mishra, Mater. Lett., 2024, 363, 13625.
14. L. Zhang, Y. Zhang, S. Jiao, J. Zhang, X. Zhao, H. Chen and J. Jiang, ChemSusChem, 2023, 16(7), e202202393.
15. H. Chen, M. Ling, L. Hencz, H. Y. Ling, G. Li, Z. Lin, G. Liu and S. Zhang, Chem. Rev., 2018, 118, 8936–8982.
16. J. Wu, Q. Zhang, S. Liu, J. Long, Z. Wu, W. Zhang, W. K. Pang, V. Sencadas, R. Song, W. Song, J. Mao and Z. Guo, Nano Energy, 2020, 77, 105118.
17. M. M. Mian, I. M. L. Kamana, X. An, S. C. Abbas, M. S. Ahommed, Z. He and Y. Ni, Carbohydr. Polym., 2023, 301, 120353.
18. A. Durairaj, M. Maruthapandi, A. Saravanan, J. H. T. Luong and A. Gedanken, Nanomaterials, 2022, 12(11), 1828.
19. J. Han, S. Wang, S. Zhu, C. Huang, Y. Yue, C. Mei, X. Xu and C. Xia, ACS Appl. Mater. Interfaces, 2019, 11, 44624–44635.
20. P. Phanthong, P. Reubroycharoen, X. Hao, G. Xu, A. Abudula and G. Guan, Carbon Resour. Convers., 2018, 1, 32–43.
21. S. Sasi, C. A. Krishna, S. K. Sugunan, A. Chandran, P. R. Nair, K. R. V. Subramanian and S. Mathew, RSC Adv., 2021, 11, 29564–29575.
22. S. Chutia and K. Phukan, J. Mater. Sci., 2024, 59, 2466–2482.
23. S. Bhardwaj, S. Singh, P. Gupta, N. Choudhary and P. K. Maji, Int. J. Green Energy, 2023, 1–22.
24. H. Zhuo, Y. Hu, Z. Chen and L. Zhong, Carbohydr. Polym., 2019, 215, 322–329.
25. M. A. Islam, H. L. Ong, A. R. Villagracia, K. A. Khairul, A. B. Ganganboina and R. A. Doong, Ind. Crops Prod., 2021, 170, 113694.
26. N. Chen, L. Tao, X. Lu, M. M. Hassan, R. Yang, J. Liu and Q. Lu, Giant, 2024, 17, 100230.
27. N. Choudhary, S. Singh, S. Bhardwaj, S. Gupta, U. Nandi, R. Chandra and P. K. Maji, Carbohydr. Polym. Technol. Appl., 2024, 7, 100416.
28. G. Malik, P. Tiwari, N. Choudhary, J. Hirpara, A. Kumar and R. Chandra, Surf. Coat. Technol., 2023, 466(15), 129660.
29. A. Bharti, R. Thangavel and R. Natarajan, J. Energy Storage, 2020, 32, 101969.
30. U. Saha, R. Jaiswal, T. H. Goswami and P. K. Maji, Electrochim. Acta, 2022, 435, 141400.
31. Z. Tang, X. Lin, M. Yu, A. K. Mondal and H. Wu, Int. J. Biol. Macromol., 2024, 259, 129081.
32. Z. Sun, H. Qi, M. Chen, S. Guo, Z. Huang, S. Maganti, V. Murugadoss, M. Huang and Z. Guo, Eng. Sci., 2022, 18, 59–74.
33. S. Eyley and W. Thielemans, Nanoscale, 2014, 6, 7764–7779.
34. A. Durairaj, M. Maruthapandi, A. Saravanan, J. H. T. Luong and A. Gedanken, Nanomaterials, 2022, 12, 1828.
35. H. F. Li, H. Li, X. Zhong, X. Li, M. E. Gibril, Y. Zhang, K. Han and M. Yu, Adv. Mat. Res., 2012, 581–582, 287–291.
36. P. Lopez Hurtado, A. Rouilly, V. Vandenbossche and C. Raynaud, Build. Environ., 2016, 96, 170–177.
37. M. F. Yu, O. Lourie, M. J. Dyer, K. Moloni, T. F. Kelly and R. S. Ruoff, Science, 2000, 287, 637–640.
38. Reference FT-IR spectra for cellulose, hemicelluloses, and lignin… | Download Scientific Diagram, https://www.researchgate.net/figure/Reference-FT-IR-spectra-for-cellulose-hemicelluloses-and-lignin-y-axis-of-the-spectra_fig1_276833689, accessed 18 February 2024.
39. H. F. Li, H. Li, X. Zhong, X. Li, M. E. Gibril, Y. Zhang, K. Han and M. Yu, Adv. Mat. Res., 2012, 581–582, 287–291.
40. S. Singh, G. L. Dhakar, B. P. Kapgate, P. K. Maji, C. Verma, M. Chhajed, K. Rajkumar and C. Das, Polym. Adv. Technol., 2020, 31, 3059–3069.
41. M. Beroual, L. Boumaza, O. Mehelli, D. Trache, A. F. Tarchoun and K. Khimeche, J. Polym. Environ., 2021, 29, 130–142.
42. S. Ramesh, S. Khandelwal, K. Y. Rhee and D. Hui, Composites, Part B, 2018, 138, 45–54.
43. D. Lu, Y. Huo, Z. Jiang and J. Zhong, Carbon, 2023, 206, 340–350.
44. J. Astruc, P. Cousin, G. Laroche, M. Robert and S. Elkoun, Mater. Sci. Appl., 2020, 11, 744–756.
45. Y. Liu, X. Fan, Z. Zhang, H. H. Wu, D. Liu, A. Dou, M. Su, Q. Zhang and D. Chu, ACS Sustain. Chem. Eng., 2019, 7, 2225–2235.
46. H. Tariq, S. U. Awan, D. Hussain, S. Rizwan, S. A. Shah, S. Zainab and M. B. Riaz, Sci. Rep., 2023, 21116.
47. U. Mostafa, M. J. Rahman, T. Mieno, M. D. Abu and H. Bhuiyan, Bull. Mater. Sci., 2020, 43, 142.
48. Nitesh, A. Kumar, S. Saini, K. L. Yadav, P. K. Ghosh and A. Rathi, Mater. Chem. Phys., 2022, 277, 125336.
49. S. Bhardwaj, S. Singh, K. Dev, M. Chhajed and P. K. Maji, ACS Appl. Mater. Interfaces, 2024, 16, 18075–18089.
50. L. Deng, R. J. Young, I. A. Kinloch, A. M. Abdelkader, S. M. Holmes, D. A. De Haro-Del Rio and S. J. Eichhorn, ACS Appl. Mater. Interfaces, 2013, 5, 9983–9990.
51. L. A. N. El-Din, A. El-Gendi, N. Ismail, K. A. Abed and A. I. Ahmed, J. Ind. Eng. Chem., 2015, 26, 259–264.
52. F. Martínez, E. Uribe and A. F. Olea, J. Macromol. Sci., Part A: Pure Appl.Chem., 2005, 42A, 1063–1072.
53. Y. Zuo, J. Gu, L. Yang, Z. Qiao, H. Tan and Y. Zhang, Int. J. Biol. Macromol., 2013, 62, 241–247.
54. M. Cheng, Z. Qin, Y. Chen, S. Hu, Z. Ren and M. Zhu, ACS Sustain. Chem. Eng., 2017, 5, 4656–4664.
55. H. Yu, Z. Qin, B. Liang, N. Liu, Z. Zhou and L. Chen, J. Mater. Chem. A, 2013, 1, 3938–3944.
56. O. Somseemee, P. Saeoui, F. T. Schevenels and C. Siriwong, Sci. Rep., 2022, 12, 1–13.
57. J. S. Chagas, J. N. S. Almeida, A. C. L. Pereira, N. F. I. Silva, R. A. Raimundo, E. S. Medeiros, B. A. S. G. Lima, L. S. Galvão, A. S. F. Santos and F. I. Silva, Cellulose, 2023, 30, 11455–11472.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4se00286e

---