Synergistic Charge Transfer and Surface Defect Effects in Cellulose-PANI-ZnO Sensors for Early Lung Cancer VOC Diagnosis

D. Hannah 1

Jerrin Thangam 1

J. Jayachandiran 2

M. Navaneethakannan 1

Mohana Selvi 1

T 1

Muthuraaman B 3

D. Nedumaran 1✉ Emaildnmaran@gmail.com

Central Instrumentation & Service Laboratory University of Madras Guindy Campus 600 025 Chennai Tamilnadu India

2 Department of Nuclear Physics University of Madras Guindy Campus 600 025 Chennai Tamilnadu India

3 Department of Energy University of Madras Guindy campus 600 025 Chennai India

D. Hannah Jerrin Thangam^a, J. Jayachandiran^b, M. Navaneethakannan^a Mohana Selvi T^c Muthuraaman B^c and D. Nedumaran^a*

^a Central Instrumentation & Service Laboratory, University of Madras, Guindy Campus, Chennai 600 025, Tamilnadu, India.

^b Department of Nuclear Physics, University of Madras, Guindy Campus, Chennai 600 025, Tamilnadu, India.

^c Department of Energy, University of Madras, Guindy campus, Chennai 600 025, India.

^* Corresponding author: dnmaran@gmail.com

Abstract

Gas sensing evaluations toward clinically relevant lung cancer VOC biomarkers-including toluene, benzene, ethanol, acetone, and hexane-revealed distinct selective detection profiles. ZO1 demonstrated the highest sensitivity toward toluene and acetone, benzene showed enhanced interaction with ZO2, while ethanol and hexane exhibited optimal responses with ZO3. Overall, toluene and ethanol produced the strongest signals across the composite series, while hexane consistently presented a prominent secondary response in all sensing layers.

The observed selectivity and sensitivity trends arise from synergistic effects involving interfacial charge transfer, controlled ZnO-PANI heterojunction formation, and optimized composite stoichiometry. These results highlight the potential of cellulose-PANI-ZnO nanocomposites as promising room-temperature sensing platforms for breath-based detection of lung cancer-associated VOCs.

Graphical Abstract

Keywords:

ZnO

PANI

cellulose

lung cancer

gas sensor

VOC sensing

1. Introduction

Lung cancer remains one of the most significant contributors to cancer-related mortality worldwide, and enhanced early detection is essential for improving patient survival outcomes. Current diagnostic approaches, including advanced imaging modalities and tissue biopsies, are constrained by inherent limitations such as procedural invasiveness, high financial burden, susceptibility to false-positive results, and limited accessibility in many clinical settings [1] [2]. These challenges underscore the need for accurate, non-invasive diagnostic methods capable of reliably detecting lung cancer at its earliest stages.

Breath analysis has emerged as a promising non-invasive strategy for identifying volatile organic compounds (VOCs) associated with lung cancer [3], [4], [5], [6]. These VOCs are released in the exhaled breath of patients as a consequence of metabolic and biochemical disturbances induced by malignant transformations within lung tissue [7], [8]. Several physiological mechanisms contribute to the altered VOC profiles observed in affected individuals, including: (i) oxidative stress and lipid peroxidation, resulting in aldehyde and ketone production [9], [10] (ii) dysregulated cellular metabolism generating compounds such as acetone, ethanol, and lactate derivatives; (iii) inflammatory and immune responses producing benzene derivatives, isoprene, and nitric oxide; (iv) tumour microenvironmental changes and hypoxia leading to the formation of methane, dimethyl sulphide, and other hydrocarbons; and (v) direct secretion of aromatic compounds such as styrene, toluene, and benzene derivatives by tumour cells [11]. The detection and characterization of these VOC patterns provide a foundation for developing breath-based diagnostic systems capable of distinguishing individuals with lung cancer from healthy populations [12], [13], [14].

A range of analytical techniques has been applied for the detection of VOCs, including gas chromatography-mass spectrometry (GC–MS) [15] are gas chromatography-mass spectrometry (GC-MS) [16], [17], selected ion flow tube mass spectrometry (SIFT-MS) [18], proton transfer mass reaction spectrometry (PTR-MS), optical and spectroscopic techniques [19], colorimetric and fluorescent sensors. Although these methods demonstrate high sensitivity and analytical precision, their clinical implementation is limited by high equipment costs, the need for skilled operators, lengthy analysis times, substantial infrastructure requirements, and complex data-processing demands.

Electronic nose (e-Nose) systems have therefore emerged as a practical and efficient alternative for rapid, real-time, and non-invasive VOC detection. Recent developments in e-Nose technology focus on improving sensor performance, enabling device miniaturisation, and integrating machine learning and artificial intelligence algorithms to enhance diagnostic accuracy and monitoring capabilities.

Metal oxide semiconductor (MOS) sensors are among the most widely utilised components in e-Nose platforms for VOC sensing [20]. These sensors operate by detecting changes in electrical resistance resulting from interactions between VOC molecules and oxygen species adsorbed on the MOS surface. Such interactions facilitate redox reactions that modify the width of the surface depletion layer, producing a measurable shift in electrical resistance. The magnitude of this shift is influenced by both the concentration and nature of the VOCs present, enabling selective detection of multiple compounds [21], [22]. The resulting resistance variations are converted into electrical signals and analysed using microcontrollers and machine learning algorithms. For complex VOC mixtures, e-Nose systems employ arrays of MOS sensors, each exhibiting distinct response characteristics, thereby generating a unique chemical fingerprint for each sample. Advanced multivariate analytical techniques, such as principal component analysis (PCA) and artificial neural networks (ANNs), are then applied to classify and interpret these data patterns with high accuracy [17], [23]. Overall, MOS-based e-Nose systems represent a low-cost, portable, highly sensitive, and rapid approach to VOC detection, making them a promising tool for non-invasive lung cancer diagnostics.

Although extensive research has been conducted on metal oxide semiconductor (MOS) sensors-particularly those based on ZnO, SnO₂, and CuO-their practical implementation remains restricted by several inherent drawbacks. These limitations include high operating temperatures and power requirements, insufficient selectivity, structural instability, and restricted surface area. Conventional ZnO and SnO₂-based gas sensors generally require elevated working temperatures (200–400°C) to achieve adequate surface activation, resulting in increased energy consumption and reduced compatibility with flexible or low-power platforms. Furthermore, due to the non-specific nature of gas adsorption–desorption processes, MOS sensors often exhibit broad responses to numerous reducing gases, making discrimination between similar VOCs, such as ethanol and toluene, technically challenging. In addition, nanoparticle aggregation during synthesis or thermal cycling can diminish accessible surface area and reduce the concentration of oxygen vacancies, both of which are critical for gas sensing activity.

To address these shortcomings, the present study introduces a hybrid organic–inorganic sensing material composed of cellulose, polyaniline (PANI), and ZnO, designed as a multifunctional sensing layer for VOC detection. The integration of polymeric membranes with metal oxides has demonstrated notable improvements in the performance of MOS sensors. Cellulose microfibers (CMF) and PANI have gained substantial interest due to their complementary advantages in sensor fabrication [24]. CMF offers a sustainable, flexible, mechanically robust, and highly porous matrix capable of facilitating gas diffusion and adsorption, while PANI provides tunable electrical conductivity and strong chemiresistive response characteristics, making it suitable for gas sensing applications [25]. When combined, CMF–PANI (CMF–PA) hybrid materials display enhanced conductivity, increased effective surface area, improved stability, and the capability for sensing at room temperature [26]. Moreover, the electrical properties of PANI can be modulated through doping or co-modification with nanomaterials, and the incorporation of CMF promotes the formation of a continuous conductive framework that enhances charge transport and sensor response [27].

The addition of metal oxides such as ZnO further improves sensing performance by providing increased active sites, facilitating charge transfer, and enhancing chemical selectivity [28]. Studies have demonstrated that CMF-PANI-metal oxide composites exhibit significantly higher sensitivity and improved VOC differentiation relative to the individual components. The combination of CMF as a porous structural scaffold, PANI as a room-temperature conductive matrix with selective molecular interactions, and ZnO as a catalytically active semiconductor generates a synergistic sensing framework capable of rapid, low-temperature chemiresistive response, enhanced VOC adsorption, and improved mechanical durability. ZnO is particularly advantageous due to its high electron mobility and abundance of structural defects, such as oxygen vacancies. When integrated into a PANI matrix, ZnO can form p-n heterojunctions that induce depletion or accumulation regions, amplifying resistance changes during gas exposure and improving selectivity [26]. The ZnO phase contributes catalytic and adsorption functionality, while PANI enables charge conduction and selective interactions such as hydrogen bonding and

$\:\pi\:-\pi\:$

interactions. CMF simultaneously prevents ZnO particle aggregation, increases accessible surface area, and promotes gas diffusion pathways [29]. Numerous experimental studies have confirmed that PANI–ZnO composites provide enhanced sensitivity toward a wide range of VOCs and are capable of operating at or near ambient temperature, a key requirement for exhaled-breath applications [26], [30].

Selectivity remains one of the primary challenges in gas sensing, as sensors must accurately differentiate target analytes from other interfering vapours. CMF enhances mechanical strength and environmental stability, making CMF-PANI-metal oxide composites more robust and reliable. Its porous structure increases available adsorption sites while improving structural integrity, enabling stable film formation. The hybrid design of CMF, conductive PANI, and MOS materials therefore enables flexible, low-temperature, and easily processed sensing platforms that are well suited for disposable or wearable breath-analysis devices.

In this work, CMF-PANI-ZnO nanocomposites (denoted ZO1, ZO2, and ZO3 based on ZnO loading) were synthesised and characterised, and their chemiresistive responses to lung cancer-associated VOCs-including benzene, toluene, acetone, and hexane-were evaluated under controlled humidity. Structural and chemical properties (XRD, SEM, PL, and EIS) were systematically correlated with sensing characteristics such as sensitivity and selectivity. The results demonstrate how ZnO content and composite microstructure influence heterojunction formation and the accompanying bulk and surface resistance modulations that govern chemiresistive response. Finally, sensor array performance was assessed for VOC pattern recognition relevant to lung cancer screening, and the potential integration of the composite into low-cost, disposable breath-detection systems was discussed. Overall, the study connects material design-combining CMF as a structural scaffold, PANI as a conductive phase, and ZnO as a catalytic semiconductor-to functional performance for breath-based lung cancer diagnostics.

2. Materials and Methods

2.1. Materials

The chemicals utilized in this study included zinc nitrate hexahydrate (Zn(NO₃)₂.6H₂O), ≥ 99%), sodium hydroxide pellets (NaOH), aniline monomer (C₆H₅NH₂, ≥ 99.5%), ammonium persulfate (NH₄)₂S₂O₈,), hydrochloric acid (HCl, 37%), ethanol (≥ 99.5%), and deionized (DI) water. All reagents were obtained from Sigma Aldrich and employed as received, without additional purification.

2.1.1. Synthesis of CMF-PANI Composite

Polyaniline (PANI) was synthesized through a conventional chemical oxidative polymerization route. A 1 M hydrochloric acid (HCl) solution was prepared by combining 100 mL of deionized water with 10 mL of concentrated HCl (37% w/w, density approximately 1.19 g/mL). Subsequently, 3 mL of aniline monomer was introduced into the solution and allowed to stir continuously overnight to ensure complete mixing and protonation. A 0.01 M ammonium persulfate (APS) solution, serving as the oxidizing agent, was then added dropwise while maintaining the reaction at 10°C for 2 hours. The formation of a green emeraldine salt precipitate signified successful polymerization. The product was isolated via filtration and sequentially washed three times with 100 mL of 1 M HCl, followed by three rinsing cycles with deionized water and finally washed with acetone. The recovered solid was dried at 60°C overnight and later pulverized to obtain a fine powder.

For preparation of the CMF-PANI composite, 0.525 g of cellulose microfibers (CMF) were dispersed in deionized water and stirred until complete homogenization. The previously synthesized PANI was incorporated in a mass ratio of 1:2 (CMF: PANI), and the mixture was thoroughly blended to develop a uniform suspension. The resulting composite was dried at 60°C for 12 hours and subsequently ground to yield a fine powder suitable for further characterization and sensor fabrication.

2.1.2. Synthesis of ZnO Nanoparticles

ZnO nanoparticles were synthesized using precipitation method. For ZnO zinc nitrate (0.5 M Zinc oxide nanoparticles were prepared using a chemical precipitation approach. Briefly, a 0.5 M solution of zinc nitrate was prepared in 50 mL of deionized water, while a separate 1 M sodium hydroxide solution was prepared in an equivalent volume of deionized water. The alkaline solution was added slowly to the zinc precursor under continuous stirring, leading to the formation of a white precipitate. The resulting solid was collected by filtration and washed three times with deionized water followed by ethanol to remove residual impurities. The product was then dried at 80°C and subsequently calcined at 400°C for 3 hours to yield crystalline ZnO nanoparticles.

2.1.3. Preparation of CMF-PANI modified ZnO Nanocomposites

A composite film was fabricated by sequentially combining CMF (please refer [31] for CMF synthesis), PANI, and ZnO in varying mass ratios of CMF:PANI:ZnO (1:2:1 for ZO1, 1:2:2 for ZO2, and 1:2:3 for ZO3). Initially, CMF (0.525 g) was dispersed in deionized water and stirred until complete dispersion was achieved. Subsequently, PANI was introduced at a 1:2 ratio (CMF: PANI) and thoroughly mixed to produce a uniform suspension. Finally, ZnO was incorporated into the mixture according to the designated mass ratios for ZO1, ZO2, and ZO3. The resulting mixtures were continuously stirred and subjected to ultrasonic treatment for 1 hour to ensure homogeneous distribution of the components. Each suspension was then poured into a petri dish and dried in a hot air oven at 70°C to form solid composite films, as illustrated in Fig. 1.

Fig. 1

Preparation steps of ZO1, ZO2, and ZO3 nanocomposites

3. Characterization Techniques

3.1. Structural Analysis: X-ray diffraction (XRD) studies

X-ray diffraction (XRD) analysis was employed to verify the successful formation of ZnO, assess phase purity, and examine the influence of CMF and PANI incorporation on crystallinity and lattice strain, parameters that are directly linked to charge transport and gas adsorption properties. XRD patterns were obtained using a PANalytical X’Pert PRO diffractometer equipped with Cu Kα radiation (λ = 1.5406 Å), operating at 40 kV and 30 mA. Data were collected over a 2θ scanning range of 5° to 80°. From the diffraction profiles, key structural parameters including interplanar spacing, crystallite size (D), microstrain (ε), and dislocation density (δ) were calculated. In addition, the Williamson-Hall (W-H) approach was applied to distinguish the contributions of crystallite size and lattice strain to the observed peak broadening.

3.2. Functional Group Identification: Fourier-transform infrared Spectroscopy (FTIR)

Fourier-transform infrared spectroscopy (FTIR) was utilized to verify the characteristic functional groups associated with CMF (O-H and C-O), PANI (C-N and C = C), and ZnO vibrations, thereby demonstrating effective interaction and bonding between the organic and inorganic components within the hybrid structure. FTIR measurements were conducted using a Perkin-Elmer Spectrum Two spectrometer over the range of 4000 − 400 cm^− 1. The spectra displayed distinct vibrational peaks corresponding to O-H stretching of CMF, C-N and C = N stretching modes of PANI, and Zn-O lattice vibrations of ZnO, confirming the formation of the composite and the chemical interactions among its constituents.

3.3. Morphology and Elemental Composition: Scanning Electron Microscopy and energy-dispersive X-ray spectroscopy (EDS)

Scanning electron microscopy (SEM) was employed to examine particle size, porosity, and surface morphology, parameters that significantly influence gas diffusion and adsorption mechanisms. Energy-dispersive X-ray spectroscopy (EDS) was used to verify elemental composition and confirm uniform incorporation of the constituent materials within the composite. Microstructural characterization was carried out using a JEOL JSM-7610F field emission scanning electron microscope (FE-SEM) operated at an accelerating voltage of 10 to 15 kV. Samples were sputter-coated with a thin layer of gold prior to imaging to minimize charging effects. EDS analysis, performed using the detector integrated with the same instrument, confirmed the elemental presence and homogeneous spatial distribution of ZnO, PANI, and CMF in the composite matrix.

3.4. Thermal Studies: Thermogravimetric Analysis (TGA)

Thermogravimetric analysis (TGA) was conducted to assess the compositional stability and decomposition behavior of the CMF, PANI, and ZnO components, providing insights into the thermal robustness of the materials for sensor fabrication and operational conditions. Measurements were performed using a TA Instruments SDT Q600 analyzer over a temperature range of 30 to 800°C, with a controlled heating rate of 10°C min^− 1. The resulting TGA curves were analyzed to quantify moisture evaporation, polymer degradation, and residual inorganic content, thereby demonstrating the improved thermal stability of the composite system.

3.5. Optical Properties: UV-Visible Diffuse Reflectance Spectroscopy (UV-DRS)

Ultraviolet-visible diffuse reflectance spectroscopy (UV-DRS) was employed to determine the optical band gap and electronic transition characteristics, parameters that are directly associated with charge carrier excitation, light absorption behavior, and the sensing mechanism governed by electron transfer within the analyte-sensor interface. UV-DRS measurements were performed using a Shimadzu UV-2600 spectrophotometer equipped with an integrating sphere, over a spectral range of 200 to 800 nm. The obtained reflectance spectra were transformed into the Kubelka-Munk function

$\:F\left(R\right)$

, and optical band gap energies were extracted using Tauc plots.

3.6. Optical Properties: Photoluminescence Spectroscopy (PL)

Photoluminescence (PL) measurements were carried out using a Horiba Jobin Yvon spectrofluorometers with an excitation wavelength of 325 nm. The resulting PL spectra were analysed to evaluate defect-induced emission processes and charge carrier recombination dynamics within the composite materials. Variations in emission band intensity and spectral position served as indicators of oxygen vacancies and interfacial defect states, offering indirect insight into factors that can influence the gas-sensing behaviour of the system.

3.7. Electrochemical Impedance Spectroscopy

EIS is a powerful technique for probing the electrical pathways, interfacial phenomena, and defect-mediated transport processes within semiconductor-polymer composites. The analysis was performed in the frequency range of 20 Hz to 1 MHz and the resulting impedance data were represented in the form of Nyquist plots. EIS measurements were carried out using Metrohm Autolab PGSTAT 204 potentiostat. In this study, the hybrid material contains semiconductor grains (ZnO). Conducting polymer chains (PANI), and an insulating yet hydrogen-bond-rich cellulose matrix its conduction mechanism involves multiple resistive and capacitive components. EIS particularly valuable for deconvoluting these contributions, as it allows the material to be modelled using an equivalent circuit.

3.8. Electrical and Gas Sensing Measurements

The sensing behaviour of the composite films was evaluated using a custom gas-testing system designed to monitor variations in electrical resistance upon exposure to lung cancer-associated volatile organic compounds (VOCs) at different concentrations and operating conditions.

3.9. Gas Sensing Measurements

Gas sensing measurements were performed in a static test chamber, where the flexible nanocomposite films were mounted using conductive clips, as illustrated schematically in Fig. 2. Target VOCs relevant to lung cancer diagnosis-including acetone, toluene, hexane, benzene, and ethanol-were introduced into the chamber in controlled quantities corresponding to concentrations between 2 and 10 µL (the ppm conversion calculation is presented in the supplementary document). The current-voltage (I-V) characteristics of the sensor were recorded before and after VOC exposure using an Autolab PGSTAT 204 FRA32M system operated in linear sweep voltammetry (LSV) mode. Changes in electrical response enabled the assessment of sensitivity and overall sensing performance toward different analytes.

Fig. 2

Schematic of gas sensing set-up

4. Results and Discussions

4.1. X-ray diffraction (XRD)

The crystalline structure and phase composition of the synthesized materials were examined using X-ray diffraction (XRD), and the corresponding patterns are presented in Fig. 3. The diffraction profile of pure CMF shows characteristic reflections at

$\:2\theta\:$

values of approximately

$\:15.7^\circ\:,\:22.5^\circ\:$

, and

$\:34.25^\circ\:$

as illustrated in Fig. 3(a) [32]. These peaks correspond to the (100), (200), and (004) planes of cellulose type I, confirming its semicrystalline nature, consistent with earlier reports [32], [33]. The crystallinity index (CrI) of CMF was quantified using the reflected peak intensities following the Segal formula (Eq. 1) [34].

$\:CrI\left(\%\right)=\frac{{I}_{200}-{I}_{am}}{{I}_{200}}\times\:100\:\:\:\:\:\:\:\left(1\right)$

where

$\:{I}_{200}$

represents the peak intensity of the (200) reflection at

$\:22.5^\circ\:$

, and

$\:{I}_{am}$

corresponds to the minimum intensity at

$\:15.7^\circ\:$

associated with amorphous cellulose. Based on this approach, the CrI of CMF was calculated to be 63.98%, indicating a predominantly crystalline cellulose structure.

The XRD pattern of the CMF-PANI composite (1:2, CMF: PANI), shown in Fig. 3(a), displays additional reflections at approximately

$\:9.4^\circ\:$

and

$\:25.6^\circ\:$

, attributed to the quinoid and benzenoid units of polyaniline (emeraldine salt form) [35], [36], [37], [38]. The broad reflection at

$\:25.6^\circ\:$

is associated with

$\:\pi\:-\pi\:$

stacking among aromatic rings, corresponding to a d-spacing of roughly 3.5 Å [39], [40], [41]. Peaks originating from CMF at 16.7°, 23°, and 34.89° remain visible, indicating preservation of crystalline cellulose regions; however, these reflections appear broadened and reduced in intensity compared to pristine CMF. This suggests partial disruption of long-range cellulose ordering due to incorporation of amorphous PANI chains [42]. The peak at 23° demonstrated the formation of PANI on the surface of CMF having semicrystalline nature [33], [43]. Using relative intensity comparison, the CrI of CMF-PANI was determined to be 60.77%, reflecting a moderate decrease in crystallinity consistent with prior studies of cellulose–polyaniline hybrids [42].

Introducing ZnO into the CMF/CMF-PANI matrix (ZO1, ZO2, ZO3) results in distinct diffraction peaks at approximately 31.8°, 34.4°, 36.2°, 47.6°, 56.6°, and 62.9°, corresponding to the (010), (002), (011), (012), (110), and (013) planes of hexagonal wurtzite ZnO, as illustrated in Fig. 3(b). In the ZO1 composite (1:2:1), slight peak shifts were observed, which may be attributed to lattice distortion and nanoscale dispersion of ZnO within the polymer matrix. According to Bragg’s law, compressive strain results in reduced interplanar spacing

$\:\left(d\right)$

and shifts peaks to higher

$\:2\theta\:$

values, whereas tensile strain increases d and shifts reflections to lower angles [44], [45]. The calculated interplanar spacings for ZnO, ZO1, ZO2, and ZO3 are summarized in Table 1.

Fig. 3

XRD patterns of a) CMF, CMF-PANI, b) ZO1, ZO2, ZO3, ZnO

Table 1

Calculated values of d for the diffraction angles of ZO, ZO1, ZO2, and ZO3
S. No	ZnO		ZO1		ZO2		ZO3
S. No	$\:2\varvec{\theta\:}$ (°)	$\:\varvec{d}$ (Å)	$\:2\varvec{\theta\:}$ (°)	$\:\varvec{d}$ (Å)	$\:2\varvec{\theta\:}$ (°)	$\:\varvec{d}$ (Å)	$\:2\varvec{\theta\:}$ (°)	$\:\varvec{d}$ (Å)
1.	31.85837	2.8067	31.3589	2.8503	31.99181	2.7953	32.23669	2.7746
2.	34.5238	2.5959	34.0064	2.6342	34.65218	2.5866	34.88555	2.5698
3.	36.34746	2.4697	35.8441	2.5032	36.46919	2.4617	36.71383	2.4459
4.	47.65178	1.9069	47.257	1.9219	47.74467	1.9034	47.98246	1.8945
5.	56.70727	1.6220	56.2573	1.6339	56.7844	1.6200	57.02224	1.6138
6.	62.9857	1.4746	62.4693	1.4855	63.05546	1.4731	63.27037	1.4686
7.	66.5074	1.4048	66.0357	1.4137	66.5632	1.4037	68.35431	1.3993
8.	68.07648	1.3762	67.5531	1.3855	68.13128	1.3752	66.79902	1.3712
9.	69.19804	1.3566	68.7203	1.3648	69.24557	1.3558	69.48127	1.3517
10.	72.68357	1.2999	72.2997	1.3058	72.74671	1.2989	72.95324	1.2957
11.	77.09228	1.2361	76.722	1.2412	77.14028	1.2355	77.36315	1.2325
12.	81.54609	1.1795	81.5676	1.1793	81.56762	1.1793	81.77803	1.1768
13.	89.75051	1.0917	89.4054	1.0951	89.79184	2.7953	90.04774	1.0889

It is evident from Fig. 3(b) that an increase in ZnO loading (ZO2 and ZO3) results in sharper and more intense diffraction peaks, indicating enhanced crystallinity and a stronger contribution of the ZnO phase within the composite. At higher ZnO concentrations, the diffraction angles observed in ZO2 and ZO3 closely correspond to those of pure ZnO, confirming effective incorporation and preservation of the ZnO crystal structure. Minimal peak shifts observed in the composites suggest the presence of interfacial interactions between ZnO and the CMF-PANI matrix, likely arising from structural compatibility and localized strain effects within the hybrid network.

Fig. 4

W-H plot of a) ZO, b) ZO1, c) ZO2, and d) ZO3

The average crystallite size of the synthesized materials was estimated using the Scherrer’s equation:

$\:D=\frac{K\lambda\:}{\beta\:\text{c}\text{o}\text{s}\theta\:}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(2\right)$

where

$\:K$

is the shape factor (taken as 0.9),

$\:\lambda\:$

is the X-ray wavelength (0.15406 nm),

$\:\beta\:$

is the full width at half maximum (FWHM) of the diffraction peak in radians, and

$\:\theta\:$

is the corresponding Bragg angle. Since the Scherrer’s formula accounts only for size-induced broadening, a more comprehensive analysis was performed using the Williamson-Hall (W-H) method to simultaneously evaluate contributions from crystallite size and microstrain in ZO1, ZO2, and ZO3.

In this approach, the total peak broadening results from the combined effects of finite crystallite size and lattice strain and is expressed as:

$\:\beta\:\text{cos}\theta\:=\frac{K\lambda\:}{D}+4ϵ\text{sin}\theta\:\:\:\:\:\:\:\:\left(3\right)$

where

$\:ϵ$

represents the microstrain. A linear fit of the W-H plot, with

$\:4\text{s}\text{i}\text{n}\theta\:$

on the x-axis and

$\:\beta\:\text{c}\text{o}\text{s}\theta\:$

on the y-axis, provides the microstrain from the slope and the crystallite size from the intercept. The W-H plots for ZO, ZO1, ZO2, and ZO3 are shown in Fig. 4. Crystallite sizes determined from the intercepts obtained through linear fitting are summarized in Table 2.

Table 2

Calculated values of microstrain, crystallite size and dislocation density
Sample	$\:\varvec{ϵ}$ $\:\times\:{10}^{-3}$	$\:\varvec{D}$ (nm)	$\:\varvec{\delta\:}$ (nm^− 2)
ZO	2.05	50.42	$\:1.81\times\:{10}^{-3}$
ZO1	0.783546	26.82	$\:1.86\times\:{10}^{-3}$
ZO2	1.3	26.36	$\:2.66\times\:{10}^{-3}$
ZO3	1.17	37.27	$\:1.58\times\:{10}^{-3}$

4.2. Fourier Transform Infra-Red Spectroscopy

The FTIR spectra of CMF, pure PANI, CMF-PANI, ZO1, ZO2, and ZO3 were recorded within the range of 4000–400 cm^− 1 to identify characteristic functional groups and evaluate molecular interactions among the system components (Fig. 5). For pristine CMF, a broad absorption band at 3437 cm^− 1 is attributed to O-H stretching vibrations associated with intra- and intermolecular hydrogen bonding. The peak at 2930 cm^− 1 corresponds to aliphatic C-H stretching from -CH and -CH₂ groups, while the band at 1633 cm^− 1 is assigned to H-O-H bending vibrations from physically adsorbed moisture [31]. The peak at 1385 cm^− 1 reflects symmetric C-H deformation, and bands located at 1164, 1119, and 1053 cm^− 1 are characteristic of C-O-C asymmetric stretching in glycosidic linkages and C-O stretching modes. The β-glycosidic linkage distinctive to CMF is observed at 896 cm^− 1

$\:\left[46\right]$

Fig. 5

FTIR spectrum of the synthesised materials CMF, PANI, CMF-PANI, ZnO, ZO1, ZO2, & ZO3

The FTIR spectrum of synthesized PANI exhibits a broad absorption band at 3440 cm^− 1 due to N-H/O-H stretching, while peaks at 2923 and 2853 cm^− 1 arise from asymmetric C-H stretching [47]. The quinoid C = C stretching mode is observed at 1629 cm^− 1, and a band at 1041 cm^− 1 is assigned to a combination of C-N stretching and in-plane aromatic C-H deformation, typically enhanced in partially doped or more ordered PANI structures. Lower-frequency signals at 604 and 567 cm^− 1 correspond to out-of-plane aromatic C-H bending. Benzenoid C = C stretching bands appear at 1487 and 1463 cm^− 1, and their coexistence with quinoid signals indicates a mixed oxidation state consistent with emeraldine-type PANI. Additional peaks at 1402 and 1292 cm^− 1 represent aromatic C-H deformation and C-N stretching (polaronic contribution), respectively. A weaker absorption at 1240 cm^− 1, associated with protonation-sensitive vibrational modes, indicates partial protonation of the polymer backbone. In-plane C-H deformation is also reflected in the band at 1091 cm^− 1.

In the CMF-PANI composite (1:2), the principal absorption bands of both CMF and PANI are retained, confirming successful integration of the two components. The O-H stretching band exhibits reduced intensity compared to the individual components, suggesting hydrogen bonding interactions between cellulose hydroxyl groups and PANI amine/imine functional sites, resulting in a lower density of free hydroxyl oscillators. Additionally, changes in the relative intensities of quinoid and benzenoid bands imply partial protonation and electronic interactions between PANI chains and the CMF matrix. This proves that the structural disturbance in XRD is caused by interfacial bonding or coordination network, not impurities. Thus, FTIR provides chemical validation of XRD results.

The low-wavenumber peaks observed at 572 cm^− 1 and approximately 427 cm^− 1 are characteristic of Zn-O lattice vibrations associated with the wurtzite phase of ZnO, confirming its crystalline identity [48]. A broad O-H stretching band cantered at 3439 cm^− 1 and an H-O-H bending absorption at 1641 cm^− 1 indicate the presence of surface hydroxyl groups and adsorbed water, commonly observed in oxide nanoparticles. The band at 2343 cm^− 1 is attributed to adsorbed CO₂, while signals at 2920, 1270, and 1042 cm^− 1 suggest trace organic residues. Additionally, the band at 1381 cm^− 1 corresponds to residual nitrate species [49].

In the ZO1 spectrum, a broad band at 3447 cm^− 1 represents O-H stretching due to surface hydroxyls and physically adsorbed moisture. Absorptions at 2920 and 2855 cm^− 1 arise from C-H stretching of residual organic species. The band at 1609 cm^− 1 corresponds to O-H bending, whereas peaks at 1446, 1376, and 1313 cm^− 1 are associated with C-O stretching vibrations. The bands at approximately 1160, 1105, and 1044 cm^− 1 originate from C-O-C or C-OH stretching. Higher-order Zn-O vibrational modes appear at 794 and 716 cm^− 1, likely resulting from lattice strain, while the fundamental Zn–O stretching frequencies at 569, 489, and 436 cm^− 1 constitute the fingerprint region of ZnO.

For the ZO2 sample, a pronounced band at 3439 cm^− 1 indicates O-H stretching from surface hydroxyls and adsorbed water. Peaks at 2921 and 2852 cm⁻¹ correspond to aliphatic C-H stretching, suggesting traces of organic residues. The absorption at 1630 cm^− 1 is attributed to O-H bending of adsorbed water. Peaks at 1457 and 1395 cm^− 1 are assigned to surface carbonates originating from atmospheric CO₂ adsorption, while the band at 1070 cm^− 1 corresponds to C-O stretching. The Zn-O stretching mode at 509 cm^− 1 confirms the presence of ZnO.

The ZO3 spectrum exhibits a broad O-H stretching band at 3432 cm^− 1 and an H-O-H bending band at 1636 cm^− 1, reflecting surface hydroxylation and moisture adsorption. A peak at 2911 cm^− 1 corresponds to aliphatic C-H stretching, and a band at 1038 cm^− 1 indicates residual organic groups. The peak at 1395 cm^− 1 arises from carbonate species resulting from CO₂ uptake. Strong absorptions at 534 and 470 cm^− 1 are attributed to Zn-O lattice vibrations, confirming the structural integrity of ZnO in the sample.

4.3. Scanning Electron Microscope

The SEM micrographs (Fig. 6) show that the cellulose microfibers form an entangled and highly interwoven fibrous network. The individual fibers appear as fine filaments, collectively producing a highly porous matrix. As illustrated in Fig. 6(b), the microfibers exhibit a degree of alignment within the cellulose framework, while Fig. 6(c) reveals layered lamellar structures and lumen-like voids, characteristic of the native plant cell wall architecture. The presence of these channel-like features indicates that the hierarchical structural integrity of cellulose is largely retained [50]. Such intrinsic porosity and high surface area are advantageous for reinforcement, adsorption, and subsequent functional surface modification.

Fig. 6

SEM image of CMF a) cellulose matrix b) channel-like features c) lumen voids

EDAX analysis of CMF (Fig. 7) indicates that the material is primarily composed of carbon (50.54 wt%) and oxygen (49.46 wt%), which is consistent with the expected elemental composition of cellulose. The absence of additional detectable elements demonstrates the effective removal of non-cellulosic constituents such as lignin, hemicellulose, and inorganic ash residues during the chemical purification process [51] confirming the high chemical purity of the extracted fibers. The nearly equimolar carbon–oxygen distribution also reflects the presence of abundant surface hydroxyl groups, which serve as active sites for hydrogen bonding, chemical modification, and subsequent composite formation.

Fig. 7

EDAX analysis of CMF

In the CMF-PANI composite, SEM analysis shows that PANI particles are uniformly anchored onto the CMF surfaces, forming surface-bound “decorations” rather than separate agglomerates. This morphology maintains the native fibrous structure while introducing localized conductive and functional PANI domains (Fig. 8). Such interfacial decoration enhances electrical conductivity and adsorption characteristics without substantially compromising porosity or the inherent mechanical integrity of the cellulose network [52].

The CMF-PANI composite exhibited a predominance of carbon (73.23 wt%), which reflects the polysaccharide backbone of CMF together with aromatic carbon skeleton of PANI (Fig. 8). The relatively high carbon percentage compared to pristine CMF (⁓50% C) confirms successful coating of PANI chains, which contribute additional aromatic carbons to the composite. The presence of nitrogen (7.31 wt%) is a distinctive marker of PANI incorporation, as nitrogen is absent in pure cellulose. This nitrogen content originates from the -NH and = N- groups of the emeraldine salt form of PANI, evidencing that polymerisation occurred on the CMF surface. Such nitrogen detection in EDX is widely regarded as proof of successful PANI deposition.

Fig. 8

a) PANI decorated CMF b) & c) magnified version focusing PANI particles

Fig. 9

EDAX analysis of CMF-PANI

Oxygen arises from hydroxyl groups (-OH) and glycosidic linkages of cellulose, indicating that the CMF framework is retained in the composition. The reduction of relative oxygen proportion compared to neat CMF suggests partial masking of hydroxyl groups by the PANI layer, which is consistent with interfacial hydrogen bonding or electrostatic interactions between -OH (CMF) and -NH⁺ groups (PANI). Finally, the detection of chlorine (8.72 wt%) originates from the dopant HCl used during oxidative polymerisation of aniline. The incorporation of Cl⁻ counter ions balance the positive charges on the protonated PANI backbone (ES), thereby confirming that PANI in the CMF-PANI exists in its conductive, doped state. Chlorine detection through EDX is standard indicator of HCl doping efficiency in PANI composites [53], [54].

The addition of ZnO nanoparticles to the CMF-PANI composite leads to clear morphological changes. SEM images shown in Fig. 10 show that the ZnO particles adhere onto the cellulose microfibers’ surfaces, forming a discreate or semi-continuous coating that increases surface roughness and partially masks the underlying fibrillar structure. This provides a porous, high-surface area combined with adsorption friendly open network of CMF facilitating analyte access and gas diffusion.

EDAX spectra (Fig. 11) of ZO1 composite reveal the presence of C (47.14 wt%), O (20.78 wt%), Zn (28.95 wt%) and Cl (3.13 wt%), confirming the coexistence of organic and inorganic constituents within the hybrid structure. The dominant carbon and oxygen signals originate primarily from the cellulose backbone and the conjugated PANI chains, reflecting their polysaccharide and aromatic polymeric nature. The strong Zn peak confirms the successful incorporation and surface coverage of ZnO nanoparticles on the CMF-PANI. The Zn content indicates a substantial inorganic loading, consistent with previous reports where ZnO nanoparticles were uniformly anchored on polymeric or cellulosic substrates.

Fig. 10

SEM image of a) CMF-PANI-ZnO with magnification in b), c) & d) showing ZnO nanoparticles

Fig. 11

EDAX spectra of ZO1

XRD shows that the composite contains nanocrystalline wurtzite ZnO, with nanoscale crystallites. SEM validates this by showing well-dispersed ZnO nanoparticles embedded in the CMF-PANI matrix confirming that the nano crystallinity observed in XRD correspond to the actual morphology and particle distribution. Together, they confirm that the composite truly contains nanoscale ZnO particles uniformly distributed which is important for high surface area and gas adsorption.

4.4. Thermogravimetric analysis

Thermogravimetric analysis (TGA) revealed distinct thermal decomposition characteristics for the synthesized materials (Fig. 12). The detailed TGA plot for each of the material with weight loss percentage is given in the supplementary document (Fig. SⅠ). CMF exhibited an initial mass loss of approximately 6.1% below ~ 120°C, attributed to the evaporation of physically adsorbed moisture. A major degradation step occurred between ~ 276°C and ~ 381°C, corresponding to CMF depolymerization, ultimately leaving a char residue of ~ 25.8% at 600°C. In the case of CMF-PANI, an additional early mass loss between ~ 30.1–70.1°C was observed, associated with the removal of residual dopants or volatile components. A broader primary decomposition region spanning ~ 220.6 to 365.3°C was detected, arising from the overlapping degradation of CMF and the polyaniline (PANI) backbone, consistent with literature reports [55]; The residue at 600°C increased to ~ 35.23%, indicating enhanced char formation due to the presence of PANI.

The weight loss percentage at each stage for the materials is also tabulated in table S1 of supplementary document. For the ZnO-containing composites, ZO2 showed a larger mass loss at lower temperatures and an earlier onset of decomposition, suggesting a higher level of surface-adsorbed species and increased defect-related reactivity, consistent with its smaller crystallite size and higher microstrain. Overall, ZO1, ZO2, and ZO3 demonstrated improved thermal stability, likely arising from the combined effects of enhanced microstrain and reduced crystallite dimensions within the composite matrix.

While SEM shows the well- distributed ZnO nanoparticles, The presence of the inorganic ZnO content is confirmed using the residual mass after polymer decomposition. Thus, TGA confirms SEM’s morphological interpretation.

Fig. 12

Thermogravimetric analysis of the synthesized materials CMF, CMF-PANI, ZO1, ZO2, & ZO3

4.5. UV-DRS

The UV–DRS spectra of the synthesized materials (Fig. 13) demonstrate distinct optical signatures attributable to the cellulose matrix, conducting polymer, and ZnO phases. Pure cellulose (CMF) shows a pronounced absorption edge at approximately 212 nm, accompanied by a weak shoulder near 282 nm, which corresponds to

$\:\sigma\:\to\:{\sigma\:}^{*}$

transitions of the polysaccharide backbone and minor contributions from trace aromatic groups, respectively.

Following incorporation of PANI, additional characteristic absorptions appear at 258 nm, assigned to

$\:\pi\:-\:{\pi\:}^{*}$

transitions of benzenoid units, along with lower-energy features at around 367 nm and 404 nm, and a broad band near 494 nm. These features are indicative of polaronic charge transfer associated with protonated emeraldine salt PANI, reflecting the formation of delocalized charge carriers.

In the ternary composites ZO1, ZO2, and ZO3, a strong near-band-edge absorption attributed to ZnO is observed in the range of 367 to 368 nm. Within the composite system, this ZnO-related transition overlaps with the

$\:\pi\:-\:{\pi\:}^{*}$

/charge-transfer region of PANI, producing a merged spectral feature in the same wavelength region.

Fig. 13

UV absorbance spectra of the synthesised materials

The optical band gap energy (E_g) of the prepared films was determined using the Tauc relation, which describes the dependence of the absorption coefficient (α) on the photon energy (hν) as

$\:\alpha\:h\nu\:=A{(h\nu\:-{E}_{g})}^{n}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(4\right)$

where

$\:A$

is a constant, and the exponent n depends on the nature of electronic transition. For direct allowed transitions,

$\:n=1/2$

, while for indirect allowed transitions,

$\:n=\:2$

. The optical band gap was estimated by extrapolating the linear portion of the

$\:{\left(\alpha\:h\nu\:\right)}^{n}$

versus

$\:h\nu\:$

plot (Tauc plot) to the photon energy axis. The obtained band gap values for CMF, CMF-PANI, ZO1, ZO2, and ZO3 revealed a progressive narrowing of

$\:{E}_{g}$

with ZnO incorporation, indicating enhanced charge carrier delocalization. The reduction in bandgap with ZnO loading can be attributed to the synergistic interaction between the conducting PANI chains and the semiconductor ZnO. This reduction due to polymer-induced electronic coupling confirms the structural nanoscale interactions inferred from XRD.

4.6. Photoluminescence

The room-temperature photoluminescence (PL) spectra of CMF, CMF-PANI, and the ZnO-based composites (ZO1, ZO2, and ZO3) were recorded to analyze defect-related electronic states and charge transfer mechanisms within the materials (Fig. 14). The corresponding bandgap values obtained from UV–DRS and PL measurements are summarized in Table 3. Pure CMF exhibits weak emission features at approximately 397, 439, and 469 nm.

Fig. 14

Photoluminescence emission spectra of the synthesised materials a) CMF, b) CMF-PANI, c) ZO1, d) ZO2, & e) ZO3

Given that cellulose possesses a wide bandgap (~ 4.9 to 5.0 eV) and inherently low fluorescence intensity, these emissions are attributed to trace conjugated residues, oxidative byproducts, or structural defects introduced during material processing, rather than intrinsic band-edge transitions. In the CMF-PANI composite, the emission bands are noticeably red-shifted, with peaks observed at 451, 483, and 508 nm, along with an additional broad deep-level emission near 613 nm. These are assigned to PANI

$\:\pi\:-{\pi\:}^{*}$

, polaronic, and interfacial charge transfer (CT) transitions between PANI and cellulose, with the 613 nm peak corresponding to aggregated polaron states or dopant-related deep CT levels [56]. The observed spectral shifts demonstrate the establishment of an efficient charge-transfer network between the conducting PANI domains and the cellulose matrix, a factor that plays a significant role in enhancing gas sensing response through improved electronic communication within the composite.

Table 3

Optical bandgap values obtained from UV-DRS and PL
Material	Bandgap energy (eV)		Type
	UV-DRS	PL
CMF	3.75	3.12	Direct bandgap
CMF-PANI	3.42	2.75	Indirect Bandgap
ZO1	3.13	3.1	direct Bandgap
ZO2	3.09	3.1	direct Bandgap
ZO3	3.13	2.82	direct Bandgap

For ZO1, the dominant near-band-edge (NBE) emission appeared at 399 nm (3.10 eV), corresponding to free excitonic recombination typical of ZnO nanostructures [57]. The additional peaks at 440 nm and 470 nm fall within the blue region and are attributed to shallow intrinsic defects such as zinc interstitials (

$\:{Zn}_{i})$

and surface-related donor states [58]. The broader green-yellow emissions at 531 and 566 nm arise from deep level oxygen-related defects, including oxygen vacancies (

$\:{V}_{O}$

) or oxygen interstitial complexes, which are highly sensitive to the surrounding oxygen partial pressure [59].

ZO2 exhibited similar features, with an NBE peak at 400 nm and multiple shallow defect peaks at 440, 470, 484, and 494 nm, suggesting a higher density and diversity of donor-type defects compared to ZO. The prominent green emission at 565 nm again confirms the presence of oxygen-related deep-level defects [60]. ZO3 displayed a relatively simplified profile, with a single shallow defect peak at 439 nm and a dominant green defect emission at 553 nm, indicating fewer distinct donor levels but persistent oxygen vacancy-related states [61].

Collectively, the PL spectra demonstrate a strong correlation between defect chemistry and the optical bandgap determined from UV-DRS measurements. The observed discrepancies between optical bandgap energies and emission transitions indicate that radiative recombination in these materials predominantly proceeds through localized defect or interfacial states rather than direct band-to-band processes. In the ZnO-containing composites, a higher green/yellow-to-UV emission intensity ratio signifies an increased concentration of oxygen vacancies, which serve as active adsorption centers for VOC detection [62]. In contrast, the CMF-PANI sample shows broad CT-related emissions, confirming efficient electronic interaction between PANI and cellulose matrix. The CT from ZnO to PANI indicated the bonding interactions between ZnO and CMF-PANI which is confirmed in FTIR. This also clearly suggests that bonding affects electronic behaviour. These results establish that both oxygen vacancies in ZnO and CT states in PANI play complementary roles in the sensing mechanism by facilitating adsorption and enhancing charge carrier transport.

4.7. Electrochemical Impedance Spectroscopy:

The impedance analysis of the prepared film reveals that conduction behaviour evolves systematically with composition. Pure cellulose is an insulating polymer with high resistivity. The presence of a Warburg element indicates slow ionic diffusion of adsorbed water molecules through the porous structure. The fitted constant phase element (CPE), representing non-ideal double layer due to surface roughness and porosity. For CMF, the equivalent circuit given in the inset of Fig. 15(a) yields the bulk resistance (Table 4) of

$\:2.59\times\:{10}^{4}\:\varOmega\:$

Fig. 15

EIS of the synthesised materials along with fitting a) CMF b) CMF-PANI c) ZO1 d) ZO2 & d) ZO3

The incorporation of PANI in CMF-PANI lowered the bulk resistance to

$\:1.89\times\:{10}^{2}\:\varOmega\:$

, enabling electronic conduction pathways through the film. The inductive element arises due to the pseudocapacitive redox nature of PANI [63]. Upon further inclusion of ZnO, the circuits for ZO1, ZO2, and ZO3 takes the same form (Fig. 15c–e, insets) indicating complex interfacial interactions between semiconducting ZnO and conductive PANI. ZO1 shows a bulk resistance of

$\:2.79\times\:{10}^{4}\:\varOmega\:$

, suggesting efficient percolation pathways, whereas ZO2 exhibits a much higher resistance of

$\:1.15\times\:{10}^{6}\:\varOmega\:$

due to nanoparticle aggregation can increase grain-boundary resistance. The ZnO surface defects (oxygen vacancies) and resulting trap states can produce interfacial energy barriers at PANI-ZnO systems, conductivity rise [64]. ZO3 shows an intermediate bulk resistance compared to ZO1 and ZO2 (

$\:1.77\times\:{10}^{5}\:\varOmega\:)$

reflecting moderately connected conductive pathways, but cellulose domains still act as insulating barriers.

The oxygen vacancies interpreted using PL and narrowed bandgap using UV-Vis influences the conductive pathway interpreted using EIS. Overall, these observations confirm that compositional tuning effectively modulates charge transport behaviour, thereby influencing the gas-sensing characteristics of the hybrid films.

Table 4

Calculated values of bulk resistance and conductivity using equivalent circuit model
Sample	$\:{\mathbf{R}}_{\mathbf{a}\mathbf{r}\mathbf{e}\mathbf{a}}$ ( $\:\varOmega\:·{cm}^{2}$ )	$\:{\mathbf{R}}_{\text{bulk}}$ ( $\:\varOmega\:$ )	Conductivity, $\:\varvec{\sigma\:}$ ( $\:S/cm$ )
CMF	$\:8.13\times\:{10}^{4}$	$\:2.59\times\:{10}^{4}$	$\:9.84\times\:{10}^{-8}$
CMF-PANI	$\:5.948\times\:{10}^{2}$	$\:1.89\times\:{10}^{2}$	$\:1.34\times\:{10}^{-5}$
ZO1	$\:8.766\times\:{10}^{4}$	$\:2.79\times\:{10}^{4}$	$\:9.13\times\:{10}^{-8}$
ZO2	$\:3.618\times\:{10}^{6}$	$\:1.15\times\:{10}^{6}$	$\:2.21\times\:{10}^{-9}$
ZO3	$\:5.568\times\:{10}^{5}$	$\:1.77\times\:{10}^{5}$	$\:1.44\times\:{10}^{-8}$

4.8. Gas Sensing Performance:

The sensing performance of the fabricated ZO1, ZO2 and ZO3 nanocomposite sensors were evaluated based on the variation in the electrical resistance upon exposure to different concentrations of VOCs. The sensor response (R%) was determined from the change in resistance between air and target gas environments. The response was calculated using the relation [65]:

$\:R\left(\%\right)=\left|\frac{{R}_{g}-{R}_{a}}{{R}_{g}}\right|\times\:100\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(5\right)$

where

$\:{R}_{a}$

and

$\:{R}_{g}$

represent the electrical resistance of the sensor in air and in the presence of target gas, respectively.

Fig. 16

Sensing response in ZO1, ZO2, and ZO3 of a) toluene b) acetone c) benzene d) hexane e) ethanol

It is found from the Fig. 16 that each VOCs responded differently for each of the sensing layer (ZO1, ZO2, ZO3). The response of toluene (Fig. 16) which is a reducing, aromatic, and non-polar molecule. It favours ZO1 the most than the other sensing material via strong

$\:\pi\:-\pi\:$

and hydrophobic interaction with PANI with a good absorption in the ZnO vacancy density in ZO1. Whereas, acetone being polar aprotic having carbonyl lone pairs yields the highest response in with ZO2 by adsorbing into the oxygen vacancy sites. Benzene is aromatic but slightly different surface affinity than toluene shows the highest response in ZO2 due to more exposed ZnO facets seeking more hydrophobic surface. Hexane being non-polar, relatively large and linear and diffuses adsorbs more in ZO3 leading to higher response. Ethanol is polar protic molecule, and ZO3 offers more accessible active sites for ethanol adsorption through H-bonding thus exhibits more response. The sensing mechanism for various VOC adsorption in the composites is discussed in detail in the next section.

4.9. Gas Sensing Mechanism

Based on the macroscopic mechanism of gas sensing, the adsorption/desorption explains the physical or chemical interaction mechanism of gas and the sensing surface. Upon exposure to the gas, based on the type of gas, the concentration of the charge carriers’ changes which in turn alters the resistance of the sensing material.

Toluene is largely non-polar but has small dipole moment because of the methyl substituent; it readily participates in

$\:\pi\:-\pi\:$

interactions (Fig. 17) with aromatic backbone of PANI [66]and is slightly more reactive than benzene toward surface oxygen species. In the ZO1 composite, which is rich in PANI, the sensing behaviour of toluene is primarily governed by the strong interaction between the aromatic gas molecule and the conjugated polymer matrix. Since PANI is in its emeraldine salt form and act as a p-type semiconductor and n-type ZnO in the ZO1 composite leads to the formation of a p-n heterojunction. The methyl group in toluene acts as a weak electron-donating substituent, which can partially increase local electron density near the protonated imine sites (-NH⁺) of PANI. This localised electron donation decreases the hole density, resulting in an increase in the interfacial barrier at the p-n heterojunction between PANI and ZnO [67], [68], [69].

Toluene molecules can also physically adsorb onto the ZnO surface through van der Waals forces or interact weakly with pre-adsorbed oxygen species, but the dominant process occurs at the PANI surface and reduces the hole concentration. This, therefore widens the depletion region in the adjacent ZnO and increases the built-in potential of the heterojunction. Thus, the significant reduction in the charge carrier mobility across the interface, thereby increases the overall resistance of the sensing material. The strong

$\:\pi\:-\pi\:$

interactions of toluene with PANI explain the highest response observed in ZO1 (PANI rich) compared to ZO2 and ZO3, where the surface coverage of PANI is lower and ZnO becomes more exposed. This aligns well with previous reports on PANI-based composite gas sensors showing selective detection of aromatic VOCs due to

$\:\pi\:-\pi\:$

interactions with the polymer matrix [70].

Fig. 17

Sensing mechanism of toluene with PANI

In the case of acetone sensing (polar aprotic molecule), ZO2 exhibits the highest response, followed by ZO1 and ZO3. When acetone molecules adsorb on to the sensor surface, they initially react with chemisorbed oxygen species (see eq), generating electrons.

$\:{O}_{2}\left(gas\right)+{e}^{-}\to\:\:{O}_{2}^{-}\left(ads\right)\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(6\right)$

$\:{CH}_{3}CO{CH}_{3}+{O}_{2}^{-}\left(ads\right)\to\:{CH}_{3}CO+{OH}^{-}\left(ads\right)\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(7\right)$

However, these electrons are rapidly captured by hole rich PANI, widening the depletion region [27] in ZnO and increasing the interfacial potential barrier, thereby suppressing the electron transport. Since ZO2 has a moderate ZnO surface area, the heterojunction-controlled mechanism dominates resulting a rise in resistance upon acetone exposure, consistent with previous reports on heterojunction-based acetone sensors[26], [71].

Upon ethanol adsorption there is formation of H-bond with PANI which causes swelling and reduced polaron mobility. Several experimental reports on PANI films [72], [73] show that alcohol vapours increase resistance through hydrogen bonding and swelling of the polymer network [74]. The oxygen vacancies act as primary anchoring sites, therefore upon exposure to ethanol vapours, there occurs the formation of ethoxy groups (-O-CH₂CH₃) [75], [76]. This ethoxy-induced surface states can trap electrons, reducing the number of free carriers in the conduction band [77].

Benzene is a non-polar, highly stable aromatic molecule with a delocalised π-electron cloud. Its interaction with the sensing surface primarily occurs through

$\:\pi\:-\pi\:$

interactions (Fig. 18) with PANI and weak van der Waals interactions with ZnO, making its oxidation at room temperature kinetically difficult. The reaction with adsorbed oxygen follows:

$\:{C}_{6}{H}_{6}+15{O}^{-}\left(ads\right)\to\:6{CO}_{2}+3{H}_{2}O+15{e}^{-}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(8\right)$

In the ZO2, the balanced ratio of PANI and ZnO provides both PANI sites for

$\:\pi\:-\pi\:$

adsorption of benzene, and ZnO sites with surface oxygen for partial oxidation. This dual interaction promotes stronger modulation of the p-n heterojunction, enhancing the sensor’s response. ZO3, with higher ZnO content, has more adsorption sites but lacks sufficient PANI to efficiently capture benzene via

$\:\pi\:-\pi\:$

interactions, leading to a moderate response. ZO1 has high PANI coverage but low availability of active oxygen species, making benzene oxidation highly limited, resulting in the lowest response.

Fig. 18

Sensing mechanism of benzene with PANI

Hexane is a non-polar alkane molecule with very low reactivity. Its adsorption on the sensor surface occurs mainly through physical physisorption driven by van der Waals forces. The presence of surface defects and oxygen vacancies that act as adsorption traps. Upon hexane adsorption, a weak electron donation effect occurs, slightly increasing the free electron concentration in ZnO. This reduces the depletion width, decreasing the resistance. In ZO1, partial PANI coverage limits the accessible ZnO surface area, while in ZO2, lower porosity and fewer oxygen vacancies result in the weakest response.

In ZO1, the high PANI coverage produces an extended PANI surface that provides abundant

$\:\pi\:-\pi\:$

interactions for toluene and efficiently perturbs the interfacial charge balance and amplifies the Chemiresistive signal [68], [69], [70], [78]. Because toluene’s methyl substituent donates a small amount of electron density relative to other aromatic VOC like benzene, adsorbed toluene more effectively modulates PANI’s hole density and the junction barrier than benzene, producing the largest response in ZO1 (Fig. 19).

ZO2 and ZO3 are ZnO rich or vacancy rich samples therefore present many active adsorption sites and produce a strong modulation at the junction. Ethanol is polar and protic (-OH), so it produces a strong hydrogen bond to the ZnO surface and to the oxygen vacancy sites and leads to significant change in the ZnO-PANI surface charge.

In all the three composites, hexane adsorption is facilitated because of the mesopores allow diffusion and retention. As hexane is a large non-polar alkane that adsorbs primarily by van-der-Waals physisorption; its sensing is therefore controlled by accessible surface area, mesopores rather than by a strong redox chemistry.

Fig. 19

Bar-graph of the response comparison of the tested VOCs in ZO1, ZO2, and ZO3

The literature directly reporting cellulose-PANI-ZnO triple component sensor is limited. The most directly relevant reports are those using cellulose derivatives combined with PANI and ZnO. Several studies report PANI-ZnO hybrids or ZnO on cellulose/paper – these are included (Table 5) for performance and mechanism comparison (heterojunction) because they show contribution of each component (ZnO, PANI, cellulose substrate) to sensitivity and operating temperature.

Table 5

Comparative performance of cellulose-PANI-ZnO and related VOC sensors
S. No	Material/Substrate	Target VOC(s)	Sensor Type	Response (%)/ Sensitivity (units)			Operating Temperature	Ref.
1.	Cellulose acetate/PANI/ZnO	NH₃ (also tested VOCs in extended work)	QCM (mass loading)	4.54 (Hz/ppm)			RT	[79]
2.	PANI nanoparticles encapsulated by hydroxypropyl methyl cellulose (HPMC; cellulose derivative)	Acetone	Chemiresistive	4.2 (ppm^− 1)			RT	[27]
3.	Biomass aerogel (grapefruit peel; cellulose-rich) decorated with PANI-ZnO nanohybrids	Formaldehyde (also ethanol, acetic acid, and ammonia)	Chemiresistive	0.134% ppm^− 1			RT	[74]
4.	PANI/ZnO	Ethanol, acetone, ammonia, methanol, formaldehyde	Chemiresistive				RT	[26]
5.	PANI/ZnO	Methanol, ethanol, acetone	Chemiresistive				90°C	[80]
6.	CMF/PANI/ZnO in the ratio 1:2:1 – ZO1 1:2:2 – ZO2 1:2:3 – ZO3	Acetone	Chemiresistive	24.75	4.039	5.59	RT	PW
		Benzene		4.65	22.28	20.47
		Ethanol		17.35	47.71	85.85
		Hexane		32.35	21.49	63.26
		Toluene		89.58	12.68	9.42

5. Conclusions

The ZO1, ZO2, and ZO3 nanocomposites were successfully synthesized and comprehensively characterized using XRD, UV-DRS, PL, SEM, and EIS analyses. In this study, the combined structural, chemical and functional characterisation demonstrated that nanoscale wurtzite ZnO is homogeneously embedded within a porous cellulose-PANI matrix. XRD and SEM confirm nanocrystalline ZnO dispersion; FTIR reveals interfacial interactions between ZNO, PNI and CMF; UV-Vis/Tauc and PL reveal an effective bandgap with enhanced defect and oxygen vacancy states; TGA corroborates the ZnO loading; and EIS shows charge-transfer pathways sensitive to VOC adsorption. Together these results establish that the ZnO-PANI heterointerfaces and increased surface defect sites provides a robust mechanistic explanation for the composite’s sensing performance.

Gas sensing studies demonstrated that each composite displayed preferential detection toward specific VOCs: ZO1 showed the highest response to toluene, while acetone, benzene, hexane, and ethanol exhibited enhanced interaction with ZO2 or ZO3 depending on surface composition and heterojunction characteristics. Among all VOCs, toluene produced the strongest response in ZO1, whereas ethanol showed superior sensitivity in both ZO2 and ZO3. Notably, hexane consistently ranked as the second most responsive analyte in all sensing layers, indicating a persistent and effective interaction with the composite surfaces.

These selectivity and sensitivity trends are attributed to synergistic effects arising from engineered surface morphology, interfacial charge modulation at the PANI-ZnO heterojunctions, and optimized polymer-to-metal oxide ratios. The results highlight the promising potential of cellulose-PANI-ZnO hybrid sensing materials for reliable VOC detection, particularly in applications in real time diagnostics of lung cancer.

Conflict of Interest

The authors have no relevant financial or non-financial interests to disclose.

Data Availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request

Acknowledgement

This work was financially supported by MHRD funded RUSA 2.0 Project under Research Innovation and Quality Improvement, (T2 P3, PF8 ) Theme-2 “Functional Synthetic Material for Biomedical Applications” Sub Theme-4 “Theory, Simulation and Algorithm Development (Development of Novel Materials for Biomedical Applications)”.

Authors’ contribution statement

All authors contributed to the study conception and design. Design, Conceptualisation, Material preparation, analysis, data collection, visualization, and investigation were carried out by D. Hannah Jerrin Thangam who also prepared the first draft of the manuscript. J. Jayachandiran contributed to the design, conceptualisation, CMF preparation and in revising the manuscript. M. Navaneethakannan, and Mohana Selvi T contributed to analysis and data collection by assisting with sensor measurements and EIS experiments. Muthuraaman B provided technical input, revised the manuscript, and gave access to the necessary facilities for conducting the sensor measurements. D. Nedumaran supervised the work, reviewed the manuscript, and approved the final version for publication. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Electronic Supplementary Material

Below is the link to the electronic supplementary material

Supplementary Material 1

Author Contribution

All authors contributed to the study conception and design. Design, Conceptualisation, Material preparation, analysis, data collection, visualization, and investigation were carried out by **D. Hannah Jerrin Thangam** who also prepared the first draft of the manuscript. **J. Jayachandiran** contributed to the design, conceptualisation, CMF preparation and in revising the manuscript. **M. Navaneethakannan** , and **Mohana Selvi T** contributed to analysis and data collection by assisting with sensor measurements and EIS experiments. **Muthuraaman B** provided technical input, revised the manuscript, and gave access to the necessary facilities for conducting the sensor measurements. **D. Nedumaran** supervised the work, reviewed the manuscript, and approved the final version for publication. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

References

R.E. Ledda, G.C. Funk, N. Sverzellati, The pros and cons of lung cancer screening. Jan 01 2024 Springer Sci. Bus. Media Deutschland GmbH 10.1007/s00330-024-10939-6

D.L. Tárnoki et al., Lung imaging methods: indications, strengths and limitations, Oct. 01, 2024, European Respiratory Society. 10.1183/20734735.0127-2023

R.J. Keogh, J.C. Riches, The Use of Breath Analysis in the Management of Lung Cancer: Is It Ready for Primetime? Oct. 01, 2022, MDPI. 10.3390/curroncol29100578

Z. Jia, V. Thavasi, T. Venkatesan, P. Lee, Breath Analysis for Lung Cancer Early Detection—A Clinical Study. Metabolites. 13(12) (Dec. 2023). 10.3390/metabo13121197

N. Peled, V. Fuchs, E.H. Kestenbaum, E. Oscar, R. Bitran, An update on the use of exhaled breath analysis for the early detection of lung cancer. Dove Med. Press. Ltd. (2021). 10.2147/LCTT.S320493

A.G. Dent, T.G. Sutedja, P.V. Zimmerman, Exhaled breath analysis for lung cancer. Pioneer. Bioscience Publishing. (2013). 10.3978/j.issn.2072-1439.2013.08.44

S. Janfaza, B. Khorsand, M. Nikkhah, J. Zahiri, Digging deeper into volatile organic compounds associated with cancer, Dec. 27, 2019, Oxford University Press. 10.1093/biomethods/bpz014

E. Janssens, J.P. van Meerbeeck, K. Lamote, Volatile organic compounds in human matrices as lung cancer biomarkers: A systematic review, Sep. 01, 2020, Elsevier Ireland Ltd. 10.1016/j.critrevonc.2020.103037

Y. Gao, B. Chen, X. Cheng, S. LiuD, Q. Li, M. Xi, Volatile organic compounds in exhaled breath: Applications in cancer diagnosis and predicting treatment efficacy. Chin. Med. Association. (2025). 10.1016/j.cpt.2024.12.004

10.

F. Schmidt et al., Mapping the landscape of lung cancer breath analysis: A scoping review (ELCABA), Jan. 01, 2023, Elsevier Ireland Ltd. 10.1016/j.lungcan.2022.12.003

11.

T. Furuhashi, K. Toda, W. Weckwerth, Review of cancer cell volatile organic compounds: their metabolism and evolution. Front. Media SA (2024). 10.3389/fmolb.2024.1499104

12.

W. Lv et al., Identification of volatile biomarkers for lung cancer from different histological sources: A comprehensive study. Anal. Biochem. 690 (Jul. 2024). 10.1016/j.ab.2024.115527

13.

P.C. Moura, M. Raposo, V. Vassilenko, Breath volatile organic compounds (VOCs) as biomarkers for the diagnosis of pathological conditions: A review. Aug 01 2023 Elsevier B V 10.1016/j.bj.2023.100623

14.

X. Fan et al., Exhaled VOC detection in lung cancer screening: a comprehensive meta-analysis. Dec. 01 2024 BioMed Cent. Ltd 10.1186/s12885-024-12537-7

15.

R. Epping, M. Koch, On-Site Detection of Volatile Organic Compounds (VOCs), Feb. 01, 2023, MDPI. 10.3390/molecules28041598

16.

M. Phillips et al., Volatile organic compounds in breath as markers of lung cancer: a cross-sectional study. Lancet. 353(9168), 1930–1933 (Jun. 1999). 10.1016/S0140-6736(98)07552-7

17.

Y.M. Lim, V. Swamy, N. Ramakrishnan, E.S. Chan, H.P. Kesuma, Volatile organic compounds (VOCs) in wastewater: Recent advances in detection and quantification. Dec. 01 2023 Elsevier Inc 10.1016/j.microc.2023.109537

18.

C. Hastie, A. Thompson, M. Perkins, V.S. Langford, M. Eddleston, N.Z.M. Homer, Selected ion flow tube-mass spectrometry (sift-ms) as an alternative to gas chromatography/mass spectrometry (gc/ms) for the analysis of cyclohexanone and cyclohexanol in plasma, ACS Omega, vol. 6, no. 48, pp. 32818–32822, Dec. 2021, 10.1021/acsomega.1c03827

19.

J. Hue et al., Benzene and xylene detection by absorbance in the range of 10–100 ppb application: Quality of indoor air. Sens. Actuators B Chem. 189, 194–198 (2013). 10.1016/j.snb.2013.03.047

20.

T. Lin, X. Lv, Z. Hu, A. Xu, C. Feng, Semiconductor metal oxides as chemoresistive sensors for detecting volatile organic compounds. Jan 02 2019 MDPI AG 10.3390/s19020233

21.

F. Vajhadin, M. Mazloum-Ardakani, A. Amini, Metal oxide‐based gas sensors for the detection of exhaled breath markers. Med. Devices Sens. 4(1) (Feb. 2021). 10.1002/mds3.10161

22.

G. Li, X. Zhu, J. Liu, S. Li, X. Liu, Metal Oxide Semiconductor Gas Sensors for Lung Cancer Diagnosis. Apr 01 2023 Multidisciplinary Digit. Publishing Inst. (MDPI) 10.3390/chemosensors11040251

23.

S. Rahman et al., Wireless E-Nose sensors to detect volatile organic gases through multivariate analysis. Micromachines (Basel). 11(6), 597 (Jun. 2020). 10.3390/MI11060597

24.

A.K. Rana, F. Scarpa, V.K. Thakur, Cellulose/polyaniline hybrid nanocomposites: Design, fabrication, and emerging multidimensional applications. Nov 01 2022 Elsevier B V 10.1016/j.indcrop.2022.115356

25.

M.R. Chia, S.W. Phang, I. Ahmad, Emerging Applications of Versatile Polyaniline-Based Polymers in the Food Industry, Dec. 01, 2022, MDPI. 10.3390/polym14235168

26.

J.R.F. Foronda, L.G. Aryaswara, G.N.C. Santos, S.N.V. Raghu, M.A. Muflikhun, Broad-class volatile organic compounds (VOCs) detection via polyaniline/zinc oxide (PANI/ZnO) composite materials as gas sensor application. Heliyon. 9(2) (Feb. 2023). 10.1016/j.heliyon.2023.e13544

27.

J.S. Kim, J.H. Byeon, S. Kang, J.Y. Kim, A high sensitivity acetone gas sensor based on polyaniline-hydroxypropyl methylcellulose core-shell-shaped nanoparticles, Nanoscale Adv, vol. 4, no. 24, pp. 5312–5319, Nov. 2022, 10.1039/d2na00647b

28.

R. Souissi, B. Bouricha, N. Bouguila, L. El Mir, A. Labidi, M. Abderrabba, Chemical VOC sensing mechanism of sol-gel ZnO pellets and linear discriminant analysis for instantaneous selectivity, RSC Adv, vol. 13, no. 30, pp. 20651–20662, Jul. 2023, 10.1039/d3ra03042c

29.

H. Xu et al., Zinc Oxide-Loaded Cellulose-Based Carbon Gas Sensor for Selective Detection of Ammonia. Nanomaterials. 13(24) (Dec. 2023). 10.3390/nano13243151

30.

R.S. Andre, R.C. Sanfelice, A. Pavinatto, L.H.C. Mattoso, D.S. Correa, Hybrid nanomaterials designed for volatile organic compounds sensors: A review, Oct. 15, 2018, Elsevier Ltd. 10.1016/j.matdes.2018.06.041

31.

J. Jayachandiran, S. Vajravijayan, N. Nandhagopal, K. Gunasekaran, D. Nedumaran, Fabrication and characterization of ZnO incorporated cellulose microfiber film: structural, morphological and functional investigations, Journal of Materials Science: Materials in Electronics, vol. 30, no. 6, pp. 6037–6049, Mar. 2019, 10.1007/s10854-019-00904-1

32.

M.M. El-Sheekh, W.E. Yousuf, E.R. Kenawy, T.M. Mohamed, Biosynthesis of cellulose from Ulva lactuca, manufacture of nanocellulose and its application as antimicrobial polymer. Sci. Rep. 13(1) (Dec. 2023). 10.1038/s41598-023-37287-7

33.

R.L. Razalli, M.M. Abdi, P.M. Tahir, A. Moradbak, Y. Sulaiman, L.Y. Heng, Polyaniline-modified nanocellulose prepared from Semantan bamboo by chemical polymerization: Preparation and characterization. RSC Adv. 7(41), 25191–25198 (2017). 10.1039/c7ra03379f

34.

L. Segal, J.J. Creely, A.E. Martin, C.M. Conrad, opportunity for new developments in all phases of textile manufacturing (’ Literature Cited An Empirical Method for Estimating the Degree of Crystallinity of Native Cellulose Using the X-Ray Diffractometer, 1952)

35.

V.K. T, S. Belagali, Characterization of Polyaniline for Optical and Electrical Properties. IOSR J. Appl. Chem. (IOSR-JAC 8(1), 53–56 (2015). 10.9790/5736-08125356

36.

S. Boudjelida et al., Jun., Synthesis and Characterization of Polyaniline Emeraldine Salt (PANI-ES) Colloids Using Potato Starch as a Stabilizer to Enhance the Physicochemical Properties and Processability, Materials, vol. 17, no. 12, 2024, 10.3390/ma17122941

37.

H. Assefa, S. Singh, N. Shehata, N.A. Khan, F.E. Olu, P.C. Ramamurthy, Green synthesis and characterization of CuO/PANI nanocomposite for efficient Pb (II) adsorption from contaminated water, Sci Rep, vol. 14, no. 1, Dec. 2024, 10.1038/s41598-024-81970-2

38.

A. Hsini et al., Development of a novel PANI@WO3 hybrid composite and its application as a promising adsorbent for Cr(VI) ions removal. J. Environ. Chem. Eng. 9(5) (Oct. 2021). 10.1016/j.jece.2021.105885

39.

S. Aldwais, A.A. Al-Muntaser, C. Chen, J. Robles, A. Pal, J.T. Abiade, Enhancing the Adhesion of Polyaniline on Steel Substrates Without a Binding Agent: Evaluated by ASTM D 3359 Tape Test and Sodium Chloride (NaCl) Exposure, Polymers (Basel), vol. 17, no. 8, Apr. 2025, 10.3390/polym17081082

40.

F. Erden, S.C. Lai, H. Chi, F. Wang, C. He, Tailoring the diameters of polyaniline nanofibers for sensor application. ACS Omega. 2(10), 6506–6513 (Oct. 2017). 10.1021/acsomega.7b00544

41.

M.P. Sokolova et al., Interaction of Polyaniline with Surface of Carbon Steel. Int. J. Polym. Sci. 2017 (2017). 10.1155/2017/6904862

42.

L.S. da Silva et al., Semiconducting nanocomposite based on the incorporation of polyaniline on the cellulose extracted from Bambusa vulgaris: structural, thermal and electrical properties. Chem. Pap. 76(1), 309–322 (Jan. 2022). 10.1007/s11696-021-01844-z

43.

G. Kumar et al., Synergistic performance of epoxy modified cellulose/polyaniline/ternary metal oxide nanocomposites. Next Mater. 2 (Jan. 2024). 10.1016/j.nxmate.2024.100141

44.

S. Kumar, K. Asokan, R.K. Singh, S. Chatterjee, D. Kanjilal, A.K. Ghosh, Investigations on structural and optical properties of ZnO and ZnO:Co nanoparticles under dense electronic excitations. RSC Adv. 4(107), 62123–62131 (2014). 10.1039/c4ra09937k

45.

A. Khorsand Zak, W.H. Abd, M.E. Majid, Abrishami, R. Yousefi, X-ray analysis of ZnO nanoparticles by Williamson-Hall and size-strain plot methods. Solid State Sci. 13(1), 251–256 (Jan. 2011). 10.1016/j.solidstatesciences.2010.11.024

46.

R. Vârban et al., Sep., Comparative ft-ir prospecting for cellulose in stems of some fiber plants: Flax, velvet leaf, hemp and jute, Applied Sciences (Switzerland), vol. 11, no. 18, 2021, 10.3390/app11188570

47.

48.

N. Jayarambabu, Ȧ. Nj., B. Siva Kumari, Ȧ.K. Venkateswara Rao Ḃ, and, Y. Prabhu, Ḃ Ȧ, Germination and Growth Characteristics of Mungbean Seeds (Vigna radiata L.) affected by Synthesized Zinc Oxide Nanoparticles, 2014. [Online]. Available: http://inpressco.com/category/ijcet

49.

A. Bhunia, T. Kamilya, S. Saha, Optical and Structural Properties of Protein Capped ZnO Nanoparticles and Its Antimicrobial Activity. J. Adv. Biol. Biotechnol. 10(1), 1–9 (Jan. 2016). 10.9734/jabb/2016/29626

50.

R.J. Moon, A. Martini, J. Nairn, J. Simonsen, J. Youngblood, Cellulose nanomaterials review: Structure, properties and nanocomposites, Chem Soc Rev, vol. 40, no. 7, pp. 3941–3994, Jun. 2011, 10.1039/c0cs00108b

51.

M.K. Mohamad Haafiz, S.J. Eichhorn, A. Hassan, M. Jawaid, Isolation and characterization of microcrystalline cellulose from oil palm biomass residue, Carbohydr Polym, vol. 93, no. 2, pp. 628–634, Apr. 2013, 10.1016/j.carbpol.2013.01.035

52.

F.V. Avelar Dutra, B.C. Pires, T.A. Nascimento, V. Mano, K.B. Borges, Polyaniline-deposited cellulose fiber composite prepared via in situ polymerization: enhancing adsorption properties for removal of meloxicam from aqueous media. RSC Adv. 7(21), 12639–12649 (2017). 10.1039/c6ra27019k

53.

M. Beygisangchin, S.A. Rashid, S. Shafie, A.R. Sadrolhosseini, Polyaniline synthesized by different dopants for fluorene detection via photoluminescence spectroscopy, Materials, vol. 14, no. 23, Dec. 2021, 10.3390/ma14237382

54.

Y. Li et al., Green synthesis of free standing cellulose/graphene oxide/polyaniline aerogel electrode for high-performance flexible all-solid-state supercapacitors. Nanomaterials. 10(8), 1–18 (Aug. 2020). 10.3390/nano10081546

55.

A. John, S.K. Mahadeva, J. Kim, The preparation, characterization and actuation behavior of polyaniline and cellulose blended electro-active paper. Smart Mater. Struct. 19(4) (2010). 10.1088/0964-1726/19/4/045011

56.

A. Samanta, R. Gangopadhyay, C.K. Ghosh, M. Ray, Enhanced photoluminescence from gold nanoparticle decorated polyaniline nanowire bundles. RSC Adv. 7(44), 27473–27479 (2017). 10.1039/c7ra01124e

57.

A. Tashiro, Y. Adachi, T. Uchino, Excitonic processes and lasing in ZnO thin films and micro/nanostructures. J. Appl. Phys. 133(22) (Jun. 2023). 10.1063/5.0142719

58.

B. Huang, G. He, H. Yang, Effects of annealing on the crystal structures and blue emission properties of sputtered ZnO films, Physica B Condens Matter, vol. 405, no. 19, pp. 4101–4104, Oct. 2010, 10.1016/j.physb.2010.06.060

59.

Investigation of Oxygen Vacancy, and Interstitial Oxygen Defects in ZnO Films by Photoluminescence and X-Ray Photoelectron Spectroscopy. [Online]. Available: http://iopscience.iop.org/0256-307X/24/7/089

60.

C. Tuc Altaf et al., Impact on the Photocatalytic Dye Degradation of Morphology and Annealing-Induced Defects in Zinc Oxide Nanostructures. ACS Omega. 8, 14952–14964 (May 2023). 10.1021/acsomega.2c07412

61.

S. Das, U.K. Ghorai, R. Dey, C.K. Ghosh, M. Pal, Effect of annealing on the defect-mediated blue phosphorescence in ZnO nanocrystals, RSC Adv, vol. 11, no. 1, pp. 335–348, Dec. 2020, 10.1039/d0ra04766j

62.

H. Yang, M.M. Zhao, L. Li, L.X. Zhang, ZnO Nanorods with Doubly Positive Oxygen Vacancies for Efficient Xylene Sensing, ACS Appl Nano Mater, vol. 5, no. 3, pp. 3512–3520, Mar. 2022, 10.1021/acsanm.1c04146

63.

W.-C. Chen, T.-C. Wen, A. Gopalan, Negative capacitance for polyaniline: an analysis via electrochemical impedance spectroscopy.

64.

T.J. Penfold et al., Revealing hole trapping in zinc oxide nanoparticles by time-resolved X-ray spectroscopy. Nat. Commun. 9(1) (Dec. 2018). 10.1038/s41467-018-02870-4

65.

S. Goutham, K.K. Sadasivuni, D.S. Kumar, K.V. Rao, Flexible ultra-sensitive and resistive NO2 gas sensor based on nanostructured Zn(X)Fe(1- x)2O4. RSC Adv. 8(6), 3243–3249 (2018). 10.1039/c7ra10478b

66.

A. Abdelraheem, A.H. El-Shazly, M. Elkady, Comparable investigation of polyaniline behavior towards gaseous ammonia and toluene adsorption. Environ. Sci. Pollut. Res. 26(4), 3991–3999 (Feb. 2019). 10.1007/s11356-018-3877-y

67.

N. Indarit, Y.H. Kim, N. Petchsang, R. Jaisutti, Highly sensitive polyaniline-coated fiber gas sensors for real-time monitoring of ammonia gas. RSC Adv. 9(46), 26773–26779 (2019). 10.1039/c9ra04005f

68.

M. Parmar, C. Balamurugan, D.W. Lee, PANI and graphene/PANI nanocomposite films-comparative toluene gas sensing behavior, Sensors (Switzerland), vol. 13, no. 12, pp. 16611–16624, Dec. 2013, 10.3390/s131216611

69.

R. Kashyap, R. Kumar, M. Kumar, S. Tyagi, D. Kumar, Polyaniline nanofibers based gas sensor for detection of volatile organic compounds at room temperature. Mater. Res. Express. 6(11) (Oct. 2019). 10.1088/2053-1591/ab4e43

70.

T. Sen, S. Mishra, S.S. Sonawane, N.G. Shimpi, Polyaniline/zinc oxide nanocomposite as room-temperature sensing layer for methane, Polym Eng Sci, vol. 58, no. 8, pp. 1438–1445, Aug. 2018, 10.1002/pen.24740

71.

D. Zhang, X. Fan, X. Hao, G. Dong, Facile Fabrication of Polyaniline Nanocapsule Modified Zinc Oxide Hexagonal Microdiscs for H2S Gas Sensing Applications. Ind. Eng. Chem. Res. 58(5), 1906–1913 (Feb. 2019). 10.1021/acs.iecr.8b04869

72.

P. Fadojutimi et al., Room temperature sensing of primary alcohols via polyaniline/zirconium disulphide. Heliyon. 9(5) (May 2023). 10.1016/j.heliyon.2023.e16216

73.

A. Verma, R. Gupta, A.S. Verma, T. Kumar, A review of composite conducting polymer-based sensors for detection of industrial waste gases. Jun 01 2023 Elsevier B V 10.1016/j.snr.2023.100143

74.

C. Wang, H.Y. Yu, Z. Miao, D. Ge, S.Y.H. Abdalkarim, J. Yao, Interface Growth of PANI-ZnO Nanohybrids on a Self-Formed Grapefruit Peel Aerogel to Construct a Quick Self-Restored Gas Sensor. ACS Sustain. Chem. Eng. 10(20), 6573–6583 (May 2022). 10.1021/acssuschemeng.1c08195

75.

K. Geunjae, Y. Kijung, Adsorption and reaction of ethanol on ZnO nanowires, Journal of Physical Chemistry C, vol. 112, no. 8, pp. 3036–3041, Feb. 2008, 10.1021/jp7103819

76.

J. Podobiński, J. Datka, Characterisation of Basic Sites on Ga2O3, MgO, and ZnO with Preadsorbed Ethanol and Ammonia—IR Study. Molecules. 29(13) (Jul. 2024). 10.3390/molecules29133070

77.

A. Kathalingam, J.K. Rhee, Hysteretic I-V nature of ethanol adsorbed ZnO nanorods. Mater. Lett. 106, 122–124 (2013). 10.1016/j.matlet.2013.04.103

78.

H. Peng et al., Highly Sensitive and Selective Toluene Gas Sensors Based on ZnO Nanoflowers Decorated with Bimetallic AuPt. Molecules. 29(7) (Apr. 2024). 10.3390/molecules29071657

79.

D. Zhou et al., High sensitivity ammonia QCM sensor based on ZnO nanoflower assisted cellulose acetate-polyaniline composite nanofibers. Sens. Actuators B Chem. 392 (Oct. 2023). 10.1016/j.snb.2023.134072

80.

J. Huang, T. Yang, Y. Kang, Y. Wang, S. Wang, Gas sensing performance of polyaniline/ZnO organic-inorganic hybrids for detecting VOCs at low temperature. J. Nat. Gas Chem. 20(5), 515–519 (Sep. 2011). 10.1016/S1003-9953(10)60230-7

Table 4. Calculated values of bulk resistance and conductivity using equivalent circuit model

Sample	$\:{\mathbf{R}}_{\mathbf{a}\mathbf{r}\mathbf{e}\mathbf{a}}$ ( $\:\varOmega\:·{cm}^{2}$ )	$\:{\mathbf{R}}_{\text{bulk}}$ ( $\:\varOmega\:$ )	Conductivity, $\:\varvec{\sigma\:}$ ( $\:S/cm$ )
CMF	$\:8.13\times\:{10}^{4}$	$\:2.59\times\:{10}^{4}$	$\:9.84\times\:{10}^{-8}$
CMF-PANI	$\:5.948\times\:{10}^{2}$	$\:1.89\times\:{10}^{2}$	$\:1.34\times\:{10}^{-5}$
ZO1	$\:8.766\times\:{10}^{4}$	$\:2.79\times\:{10}^{4}$	$\:9.13\times\:{10}^{-8}$
ZO2	$\:3.618\times\:{10}^{6}$	$\:1.15\times\:{10}^{6}$	$\:2.21\times\:{10}^{-9}$
ZO3	$\:5.568\times\:{10}^{5}$	$\:1.77\times\:{10}^{5}$	$\:1.44\times\:{10}^{-8}$

Table 1. Calculated values of d for the diffraction angles of ZO, ZO1, ZO2, and ZO3

S. No	ZnO		ZO1		ZO2		ZO3
S. No	$\:2\varvec{\theta\:}$ (°)	$\:\varvec{d}$ (Å)	$\:2\varvec{\theta\:}$ (°)	$\:\varvec{d}$ (Å)	$\:2\varvec{\theta\:}$ (°)	$\:\varvec{d}$ (Å)	$\:2\varvec{\theta\:}$ (°)	$\:\varvec{d}$ (Å)
14.	31.85837	2.8067	31.3589	2.8503	31.99181	2.7953	32.23669	2.7746
15.	34.5238	2.5959	34.0064	2.6342	34.65218	2.5866	34.88555	2.5698
16.	36.34746	2.4697	35.8441	2.5032	36.46919	2.4617	36.71383	2.4459
17.	47.65178	1.9069	47.257	1.9219	47.74467	1.9034	47.98246	1.8945
18.	56.70727	1.6220	56.2573	1.6339	56.7844	1.6200	57.02224	1.6138
19.	62.9857	1.4746	62.4693	1.4855	63.05546	1.4731	63.27037	1.4686
20.	66.5074	1.4048	66.0357	1.4137	66.5632	1.4037	68.35431	1.3993
21.	68.07648	1.3762	67.5531	1.3855	68.13128	1.3752	66.79902	1.3712
22.	69.19804	1.3566	68.7203	1.3648	69.24557	1.3558	69.48127	1.3517
23.	72.68357	1.2999	72.2997	1.3058	72.74671	1.2989	72.95324	1.2957
24.	77.09228	1.2361	76.722	1.2412	77.14028	1.2355	77.36315	1.2325
25.	81.54609	1.1795	81.5676	1.1793	81.56762	1.1793	81.77803	1.1768
26.	89.75051	1.0917	89.4054	1.0951	89.79184	2.7953	90.04774	1.0889

Table 2. Calculated values of microstrain, crystallite size and dislocation density

Sample	$\:\varvec{ϵ}$ $\:\times\:{10}^{-3}$	$\:\varvec{D}$ (nm)	$\:\varvec{\delta\:}$ (nm^− 2)
ZO	2.05	50.42	$\:1.81\times\:{10}^{-3}$
ZO1	0.783546	26.82	$\:1.86\times\:{10}^{-3}$
ZO2	1.3	26.36	$\:2.66\times\:{10}^{-3}$
ZO3	1.17	37.27	$\:1.58\times\:{10}^{-3}$

Table 3. Optical bandgap values obtained from UV-DRS and PL

Material	Bandgap energy (eV)		Type
	UV-DRS	PL
CMF	3.75	3.12	Direct bandgap
CMF-PANI	3.42	2.75	Indirect Bandgap
ZO1	3.13	3.1	direct Bandgap
ZO2	3.09	3.1	direct Bandgap
ZO3	3.13	2.82	direct Bandgap

Table 5. Comparative performance of cellulose-PANI-ZnO and related VOC sensors

S. No	Material/Substrate	Target VOC(s)	Sensor Type	Response (%)/ Sensitivity (units)			Operating Temperature	Ref.
7.	Cellulose acetate/PANI/ZnO	NH₃ (also tested VOCs in extended work)	QCM (mass loading)	4.54 (Hz/ppm)			RT	[79]
8.	PANI nanoparticles encapsulated by hydroxypropyl methyl cellulose (HPMC; cellulose derivative)	Acetone	Chemiresistive	4.2 (ppm^− 1)			RT	[27]
9.	Biomass aerogel (grapefruit peel; cellulose-rich) decorated with PANI-ZnO nanohybrids	Formaldehyde (also ethanol, acetic acid, and ammonia)	Chemiresistive	0.134% ppm^− 1			RT	[74]
10.	PANI/ZnO	Ethanol, acetone, ammonia, methanol, formaldehyde	Chemiresistive				RT	[26]
11.	PANI/ZnO	Methanol, ethanol, acetone	Chemiresistive				90°C	[80]
12.	CMF/PANI/ZnO in the ratio 1:2:1 – ZO1 1:2:2 – ZO2 1:2:3 – ZO3	Acetone	Chemiresistive	24.75	4.039	5.59	RT	PW
		Benzene		4.65	22.28	20.47
		Ethanol		17.35	47.71	85.85
		Hexane		32.35	21.49	63.26
		Toluene		89.58	12.68	9.42

Yes

Abstract

Non-invasive breath analysis of volatile organic compounds (VOCs) represents a promising pathway for early-stage lung cancer detection. In this study, cellulose-PANI-ZnO nanocomposites (ZO1, ZO2, and ZO3) with component ratios of 1:2:1, 1:2:2, and 1:2:3 were synthesized and systematically characterized using XRD, UV-DRS, PL, SEM, and EIS techniques. The analyses confirmed the successful incorporation of crystalline ZnO along with strong interfacial interactions that promoted enhanced charge transport across the p-n heterojunction and facilitated defect-mediated gas adsorption. Gas sensing evaluations toward clinically relevant lung cancer VOC biomarkers-including toluene, benzene, ethanol, acetone, and hexane-revealed distinct selective detection profiles. ZO1 demonstrated the highest sensitivity toward toluene and acetone, benzene showed enhanced interaction with ZO2, while ethanol and hexane exhibited optimal responses with ZO3. Overall, toluene and ethanol produced the strongest signals across the composite series, while hexane consistently presented a prominent secondary response in all sensing layers. The observed selectivity and sensitivity trends arise from synergistic effects involving interfacial charge transfer, controlled ZnO-PANI heterojunction formation, and optimized composite stoichiometry. These results highlight the potential of cellulose-PANI-ZnO nanocomposites as promising room-temperature sensing platforms for breath-based detection of lung cancer-associated VOCs.