Optimization and characterization of typha–lantana stem fiber composites using taguchi–marcos and ann for sustainable food packaging applications

ABSTRACT

The growing demand for biodegradable packaging materials has accelerated the exploration of natural fiber-reinforced composites as sustainable alternatives to petroleum-based plastics. This research develops and optimizes Typha latifolia and Lantana camara stem fiber composites bonded with corn starch bio-resin. Key fabrication parameters of fiber orientation (FO), NaOH treatment (NT), binder concentration (BC), and fiber-to-matrix ratio (FMR) were systematically investigated using Taguchi’s L9 orthogonal array. Mechanical performance was evaluated through tensile and flexural testing, while structural, functional, and durability characteristics of the composites were also investigated. Multi-objective optimization was carried out using an integrated decision-making approach supported by statistical analysis, while predictive validation was achieved through an ANN (4-3-1 topology), which showed strong correlation with experimental results (R² = 94.28%). The optimal composite achieved tensile and flexural strengths of 30.77 MPa and 201.27 MPa, respectively, with SEM confirming strong interfacial bonding and EDAX verifying compositional integrity. Moisture absorption studies confirmed stability under humid conditions. Antibacterial tests showed inhibition zones of 12.4 ± 0.51 mm for E. coli and 10.1 ± 0.43 mm for S. aureus, indicating effective antibacterial activity. These results establish Typha–Lantana bio-composites as viable, biodegradable, and antibacterial packaging materials, offering a scalable eco-friendly alternative to synthetic polymers.

Keywords:
Typha latifolia fiber; Lantana camara stem fiber; Antibacterial activity; Mechanical properties; ANN

1. INTRODUCTION

Natural fiber-based composites have emerged as promising materials for developing environmentally friendly and sustainable packaging solutions. Typha latifolia fiber panels of varying thickness, density and binder content were evaluated for acoustic and thermal insulation [¹]. Cowpea-derived bio-resin was used to fabricate vetiver and jute composites at 50–70 wt% fiber. Moderate hydrophobicity, biodegradability, and soil-burial stability confirmed its potential as a cost-effective alternative to synthetic resins [²]. Pineapple leaf fibers treated with alkali at varied concentrations, times, and temperatures were reinforced in tapioca-based bio-resin. FTIR confirmed chemical modifications, including disruption of hydrogen bonds and exposure of functional groups [³]. Crotalaria juncea (sunnhemp) fibers were reinforced in polyester composites with untreated and alkali-treated forms. Laminates with random, continuous, biaxial and triaxial orientations were fabricated and tested for tensile, flexural, impact, and hardness properties. Results confirmed sunnhemp as a promising reinforcement for polyester-based composites [⁴]. Crotalaria juncea (CJ) and Borassus flabellifer (BF) short fibers were hybridized in polyester composites to utilize agricultural waste.

Different extraction methods such as manual, alkali, seawater retting, and combined enzymatic–alkali treatments were studied for Typha leaf fibers. Combined treatment (NaOH + enzyme) removed impurities, raising crystallinity by 57% and cellulose to 69.8%, while sea water retting boosted tensile strength by 82%. Chemical changes confirmed the removal of hemicellulose/lignin, and enhanced the fiber quality of the composites [⁵]. Typha angustata stem fibers were extracted by alkali and enzymatic treatments, optimized using response surface methodology and desirability functions. Best results obtained from 20 ml/L enzyme and 20-day duration, improving diameter, density, yield, lignin, and tenacity. SEM showed well-structured cellulose bundles with improved crystallinity (58.47%), cellulose (66.86%), and reduced lignin (10.83%) [⁶]. Polyester composites reinforced with Typha fibers (leaf and stem) were fabricated under different treatments and structures. Maximum tensile strength reached ~25 MPa with 12.6% treated leaf fibers and 10.3% stem fibers, showing improvement from mercerization. Statistical optimization confirmed bulk chemically treated structures as the best configuration [⁷]. Aerial roots of Ficus amplissima fibers (ARFAFs) were alkali-treated (5% NaOH) and used at 5–25 wt% in polymer composites. Treated fibers improved tensile (44.76 → 46.71 MPa), flexural (59.38 → 63.12 MPa), and impact strength (85.66 → 92.22 J/m), while reducing water absorption [⁸]. The alkali (NaOH) treatment enhances the performance of natural fiber composites by improving fiber–matrix adhesion, leading to better mechanical properties such as tensile and flexural strength. Balanites aegyptiaca and Tapsi fibers showed optimal performance around 15 wt./g loading of NaOH treatment [⁹, ¹⁰], while Cordia dichotoma exhibited significant improvement with 5% NaOH treatment [¹¹].

Cordia dichotoma fiber and nano-alumina reinforced epoxy composites were fabricated by hand lay-up and optimized using Taguchi and ANN models [¹²]. Fibrovascular bundle (FVB) fibers from Salacca sumatrana fronds were alkali-treated (NaOH ± Na₂SO₃) and formed into oriented boards at 0°, 45°, and 90°. The 1% NaOH + 0.2% Na₂SO₃ treatment showed lowest swelling, highest modulus of rupture/elasticity, and strong citric acid bonding confirmed by FTIR [¹³]. Solanum nigrum stem fibers were characterized for textile and biomedical use by XRD, FTIR, TGA, SEM, tensile, and antibacterial tests. Antibacterial tests showed a 13 mm inhibition zone and 72.6% biofilm reduction, highlighted its biomedical potential [¹⁴].

Natural extracts of Calotropis gigantea, Eucalyptus globulus and Syzygium aromaticum were coated on bamboo gauze fabric for wound dressing. SEM confirmed porous structure, while antibacterial tests showed inhibition against E. coli and S. aureus [¹⁵]. Borassus fruit fiber (BF) and cigarette butt fiber (CF) reinforced polyvinyl ester composites were fabricated with eggshell powder (EP) (0–10 wt%) as filler. CRITIC–EDAS optimization with sensitivity analysis confirmed C9 (BF 40% + CF 40% + EP 5%) composites as the most effective [¹⁶]. Cigarette filter fiber (CFF) reinforced polyvinyl ester composites with eggshell powder (ESP, 0–5 wt%) were optimized using CRITIC–TODIM. ESP at 3% with 60 wt% CFF (20 mm length) yielded the best tensile, flexural and impact properties. Typha and Lantana fibers were selected due to their abundant availability, low cost, and adequate mechanical properties. Typha fibers possess good cellulose content and lightweight structure, while Lantana camara stem fibers provide adequate rigidity and reinforcement potential. Additionally, these underutilized natural fibers offer potential for sustainable reinforcement in biodegradable composite materials. Fibers are extracted from the outer fibrous bark of the stem. Corn starch bio-resin was selected due to its biodegradability, renewability, and lower environmental impact compared to conventional petroleum-based resins such as Epoxy LY556, supporting the development of eco-friendly composite materials.

Recent reviews on lignocellulosic biomass-based green composites reported that cellulose-rich natural fibers significantly improve fiber–matrix adhesion, crystallinity, biodegradability, and mechanical performance, making them promising alternatives to petroleum-based plastics for sustainable composite applications [¹⁷]. Water hyacinth reinforced thermoplastic starch composites exhibited considerable improvements in tensile and flexural strength along with enhanced biodegradability and thermal stability, demonstrating their suitability for eco-friendly packaging applications [¹⁸]. Grass fiber-reinforced composites have attracted attention due to their lightweight structure, renewability, and favorable physico-mechanical properties, supporting their utilization in sustainable packaging and structural applications [¹⁹]. Bamboo fiber reinforced composites demonstrated promising thermal and mechanical characteristics with strong potential to replace non-degradable thermoplastics in environmentally sustainable material applications [²⁰]. Luffa cylindrica fiber reinforced biodegradable composites exhibited enhanced tensile and flexural properties with reduced water absorption and complete biodegradability, indicating their effectiveness as sustainable alternatives for packaging and indoor applications [²¹]. Thermoplastic starch reinforced natural fiber composites showed improved mechanical performance and moderate hydrophobic behavior while maintaining biodegradability under soil and fungal degradation conditions, highlighting their applicability in sustainable packaging sectors [²²].

Existing studies on natural fiber composites for sustainable packaging primarily focus on single fibers or cellulose derivatives, with limited attention to hybrid systems combining Typha latifolia and Lantana camara stem fibers. Moreover, the synergistic effects of fiber orientation, alkali treatment, and binder concentration on mechanical, antibacterial, and moisture-resistant properties remain insufficiently explored. This work introduces a biodegradable hybrid composite reinforced with Typha–Lantana fibers and corn starch bio-resin, optimized through a Taguchi–CRITIC–MARCOS framework with ANN-based predictive modeling. The study establishes an integrated optimization strategy for achieving balanced mechanical strength, antibacterial activity, and moisture resistance, demonstrating the composite’s potential as a fully biodegradable and sustainable alternative to synthetic packaging materials.

2. MATERIALS AND METHODS

2.1. Materials

Typha latifolia (cattail) and Lantana camara fibers are abundant in tropical and subtropical regions, particularly in India. Typha typically grows in wetlands and irrigation canals, whereas Lantana is a widely invasive species found in forested, degraded, and semi-arid regions, making both readily available raw materials. Typha latifolia, the leaves are long (1.5–3 m) and narrow (2–3 cm), with typically 7–10 leaves per stem, and cylindrical inflorescences typically up to 1.5 m in height that turn brown when mature [²³]. Fiber identification is further confirmed through oxalate plate analysis, which reveals higher concentrations of calcium oxalate in stem fibers compared to leaf fibers. For Lantana camara, the leaves are 3–5 cm long and 2–6 cm wide with a rough texture, and the plant produces dense flower clusters in yellow to purple [²⁴]. Fibers are extracted from the outer fibrous porous pith of the stem. The fiber of Typha latifolia contains 65% cellulose, possesses a tensile strength of approximately 158 MPa, a Young’s modulus of 1600 MPa, and an elongation at break of about 7% following optimum alkali treatment [²⁵]. Conversely, Lantana camara stem fiber comprises around 34.9% cellulose, 17% hemicellulose, 17% lignin, and 2.1% moisture with holocellulose at about 66.8%; its reinforcement in composites varies according to fiber loading [²⁶]. The mean fiber diameter of both fibers was found to be approximately 15.6 µm, as measured from SEM micrographs using ImageJ software. These attributes are optimal for improving moisture-resistant and mechanical performance in environmentally sustainable composites. The biodegradable corn bio-resin was employed and sourced from locally accessible corn starch. The bio-resin demonstrates a flexible tensile property, a low moisture content (under 15%) and substantial starch content (exceeding 75%) that facilitate robust adhesion with natural fibers while preserving biodegradability and flexibility [²⁷] (Table 1).

Table 1
Physical and mechanical properties of Typha latifolia and Lantana camara fibers [²⁵, ²⁶].

PROPERTIES	TYPHA LATIFOLIA FIBER	LANTANA CAMARA FIBER
Cellulose content (%)	65	34.9
Tensile strength (MPa)	158	207
Crystallinity index (%)	58.47	42.89
Fiber diameter (µm)	15.6	15.6

2.2. Fabrication process

The fabrication of Typha–Lantana camara stem fiber composites was carried out using a combination of compression molding via open mold hand lay-up techniques (Figure 1). The fibers are extracted from the stem’s porous pith (spongy core-like structure) using an aqueous alkaline solution under controlled conditions at elevated temperatures (60–80°C for 2–8 hours) [²⁸]. The extracted fibers were manually cleaned, combed, and treated with sodium hydroxide (NaOH) solutions at concentrations of 1–5 wt% for 4 h at room temperature to remove hemicellulose, lignin, and other impurities. Based on the previous research [²⁹], the NaOH concentration was selected, which shows that, the lower concentrations augment surface roughness and promote interfacial bonding without inflicting significant harm to the fiber [³⁰].

Figure 1
Composite fabrication process.

The ratio of fiber to NaOH solution was maintained at 1:20. After thorough washing with distilled water, the fibers were neutralized using 1% acetic acid solution, rinsed again with distilled water, and oven dried at 60 °C for 6 h to eliminate residual moisture. The dried fibers were cut into three categories: short random (10 mm), long chopped (40 mm), and continuous strands (100 mm). Corn starch bio-resin was prepared by gelatinization. Corn starch (10 wt%) was mixed with distilled water (90 wt%) and heated at 85 °C under stirring at 450 rpm for 25 min until gelatinization occurred. Glycerol (25 wt% of starch content) was added as a plasticizer to improve flexibility. The mixture was stirred to obtain a homogeneous viscous resin and then cooled to 25 °C before being used as the composite matrix. Key parameters included a starch–water ratio of 1:9, mixing temperature of 85 °C, stirring speed of 450 rpm, and mixing duration of 25 min. Typha and Lantana fibers were incorporated in the composite at an equal weight ratio (1:1), where each fiber constituted 50% of the total fiber content within the corn starch bio-resin matrix. Fibers and resin were combined according to design parameters (fiber-to-matrix ratios of 30:70, 35:65, and 40:60; binder concentrations of 10, 15, and 20 wt%) and placed into moulds. For different fiber orientations, the fibers were first placed in the mold in the desired alignment (random, long aligned, or continuous strand). Subsequently, the bio-resin was carefully poured and pressed through the layers to ensure uniform impregnation. Compression molding (300 mm × 300 mm × 3 mm) was then carried out by applying a load of 20 kN at 100 °C for 20 minutes to facilitate resin cross-linking and gelatinization. The total cycle time, including both molding and cooling, was approximately 45 minutes. The resulting composite sheets (thickness of 3 mm) were conditioned to ambient atmosphere before testing. Specimens were cut using an abrasive water jet machining (AWJM) to ensure clean edges, minimizing defects and variability. Tests were conducted in accordance with respective ASTM testing. Table 2 shows the experimental design followed Taguchi’s L9 OA with four parameters BC, FMR, NT and FO, each varied across three levels to study their influence on mechanical and functional performance.

Table 2
Levels and parameters for composite fabrication.

PARAMETERS	NOTATIONS	UNITS	LEVEL
			1	2	3
Fiber orientation (FO)	A	–	Short random Fibre	Unidirectional Fibre	Long random Fiber
NaOH treatment (NT)	B	wt.%	1	3	5
Fibre to matrix ratio (FMR)	C	–	40:60	35:65	30:70
Binder concentration (BC)	D	wt.%	10	15	20

The Taguchi L9(3³) orthogonal array were utilized for experimental design framework for improving the efficiency and organization of the fabrication process. The Taguchi approach effectively identifies optimal process conditions with fewer experiments, reducing variability and enhancing efficiency [³¹]. The L9 matrix facilitates a comprehensive evaluation of four selected factors at three different levels, ensuring that their interactions are accurately captured and assessed statistically. The evaluated metrics are Flexural Strength (FS) and Tensile Strength (TS), and both are essential indicators. Signal-to-noise ratios facilitate the identification of optimal manufacturing conditions, providing critical insights into the influence of each parameter on mechanical qualities. Furthermore, an ANOVA were performed to assess the contribution of each parameter to overall efficiency of composites.

2.3. Mechanical testing

Mechanical properties of fabricated Typha-Lantana camara stem fiber composites were evaluated utilizing flexural strength and tensile strength tests as per subsequent ASTM standards. Tensile properties were evaluated using a Universal Testing Machine (Model UTM-100, PQR Instruments, Kolkata, India) as per ASTM D638 (Type I) standard with a dimension of 165 mm × 13 mm × 3 mm [³²]. Tests were conducted with a maximum load of 5kN at a guage length of 50 mm and a crosshead speed of 2 mm/min at room temperature. Flexural behavior was determined using the three-point bending method in accordance with ASTM D790 using the same UTM with a dimension of 127 mm × 12 mm × 3 mm [³³]. A span length of 48 mm was maintained with a crosshead speed of 2 mm/min. The span-to-thickness ratio for flexural testing was 16:1, and the flexural strength was calculated using the following Equation 1:

(1) $$\text{Flexural Strength (FS) =}\frac{3FL}{2b{d}^2}$$

where F is the load at fracture (N), L is the span length (mm), b is the width (mm), and d is the thickness (mm). These details have been added to the manuscript for clarity.For ecah test, three trials were performed per sample group, and the average was considered for analysis.

2.4. Multi objective optimization

The S/N ratio analysis were utilized to determine ideal production conditions, hence improving mechanical qualities. This Taguchi technique metric assesses performance variability affected by various manufacturing parameters and process configurations [³⁴]. The strategy employed focused on achieving elevated values in both flexural and tensile strength, ensuring that optimal process parameters minimize variability and improve mechanical performance [³⁵]. Computing the signal-noise ratio for all empirical trial identified the most significant fabrication settings. The Taguchi Signal-to-Noise ratio, adhering to the ‘Larger the Better’ criterion, is articulated in Equation 2 [³⁶].

(2) $$\raisebox{1ex}{$S$}\!\left/ \!\raisebox{-1ex}{$N$}\right.=-10\log \text{10}\left(\frac{1}{n}{\displaystyle \sum _{i=1}^n\frac{1}{y_i^2}}\right)$$

Where, y_i indicates the measured response for the i^th experiment and n signifies the number of trials. The ‘Larger- the-Better’ signal-to-noise ratio is utilized while optimizing response factors, enhancing flexural or tensile strength at composites.

2.5. CRITIC-MARCOS optimization

The CRITIC method was employed to determine objective criteria weights based on the contrast intensity and correlation among evaluation parameters, thereby minimizing subjectivity. The MARCOS method was used for ranking alternatives relative to ideal and anti-ideal solutions, providing a stable and reliable evaluation of composite performance. The combined CRITIC–MARCOS approach enables objective weighting and effective ranking of composite alternatives.

2.5.1. CRITIC method

This is an objective weighting method utilized for the determination of attribute weights through the CRITIC methodology, as outlined below [³⁷]

The CRITIC weighting approach employs decision matrix normalization by the subsequent formulas;

(3) $$\begin{array}{l}{R}_{ij}=\frac{A_{ij}-A_j^{\text{min}}}{A_j^{\text{max}}-A_j^{\text{min}}},\text{if j}\in \text{beneficial attribute}\\ R_{ij}=\frac{A_j^{\text{max}}-A_{ij}}{A_j^{\text{max}}-A_j^{\text{min}}},\text{if j}\in \text{non}-\text{beneficial attribute}\end{array}$$

The δ_j denoted as SD is calculated as follows:

(4) $${\delta }_j=\sqrt{\frac{{\displaystyle {\sum }_{i=1}^h{(R_{ij}-{\Lambda }_j)}^2}}{\hslash -1}};j=1,\cdots \text{,}\lambda$$

Here, Λ_𝑗 represents the average of the j^th property, defined as follows:

(5) $${\Lambda }_j=\frac{{\displaystyle {\sum }_{i=1}^h{R}_{ij}}}{\hslash };j=1,\text{}2,\cdots \lambda$$

Subsequently, a correlation coefficient (ρ_jk) is established for each attribute pair utilizing the following formula;

(6) $${\rho }_{jk}=\frac{{\displaystyle {\sum }_{i=1}^n(R_i{}_j-{\Lambda }_j)(R_{ik}-{\Lambda }_k)}}{\sqrt{{\displaystyle {\sum }_{i=1}^n{(R_i{}_j-{\Lambda }_j)}^2{\displaystyle {\sum }_{i=1}^n{(R_i{}_k-{\Lambda }_k)}^2}}}}$$

The average of the k^th characteristic (Λ_k) is calculated using Equation 4 for the k^th characteristics.

For each attribute, the data measure (σ_j) is calculated as follows [³⁸]:

(7) $${\sigma }_j={\delta }_j{\displaystyle \sum _{k=1}^{\lambda }(1-{\rho }_{jk})};j=1,\cdots \text{,}\lambda$$

The objective weight (ω_j) of each characteristic is determined utilizing the CRITIC methodology as outlined below.

(8) $${\omega }_j=\frac{{\sigma }_j}{{\displaystyle {\sum }_{j=1}^{\lambda }{\sigma }_j}};j=1,\cdots \text{,}\lambda$$

2.5.2. MARCOS method

The MARCOS MCDM technology was developed for the prioritization of healthcare suppliers. The MARCOS model is employed to address many decision-making techniques due to its direct computation. The subsequent steps of the MARCOS technique are as follows [³⁹].

Step 1: The preliminary decision matrix ([Z_ij]_p×q) is augmented by incorporating the IV and AIV, expressed as.

(9) $${\left[Z_{ij}\right]}_{p\times q}=\begin{array}{c}\begin{array}{c}{A}_{AIV}\\ A_1\end{array}\\ \begin{array}{c}\begin{array}{c}\begin{array}{c}{A}_2\\ ⋮\end{array}\\ \begin{array}{c}{A}_i\\ ⋮\end{array}\end{array}\\ \begin{array}{c}{A}_p\\ A_{IV}\end{array}\end{array}\end{array}\left[\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}{C}_1\\ Z_{11}\end{array}\\ Z_{21}\end{array}\\ ⋮\end{array}\\ Z_{i1}\end{array}\\ \begin{array}{c}\begin{array}{c}⋮\\ Z_{p1}\end{array}\\ Z_{IV1}\end{array}\end{array}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}{C}_2\swarrow \\ Z_{AV1}\end{array}\\ Z_{12}\end{array}\\ ⋮\end{array}\\ Z_{i2}\end{array}\\ \begin{array}{c}\begin{array}{c}⋮\\ Z_{p2}\end{array}\\ Z_{IV2}\end{array}\end{array}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\cdots \\ \cdots \end{array}\\ \cdots \end{array}\\ \cdots \end{array}\\ \cdots \end{array}\\ \begin{array}{c}\begin{array}{c}\cdots \\ \cdots \end{array}\\ \cdots \end{array}\end{array}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}{C}_j\nearrow \\ Z_{AVj}\end{array}\\ Z_{1j}\end{array}\\ Z_{2j}\end{array}\\ Z_{ij}\end{array}\\ \begin{array}{c}\begin{array}{c}⋮\\ Z_{pj}\end{array}\\ Z_{IVj}\end{array}\end{array}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\cdots \\ \cdots \end{array}\\ \cdots \end{array}\\ \cdots \end{array}\\ \cdots \end{array}\\ \begin{array}{c}\begin{array}{c}\cdots \\ \cdots \end{array}\\ \cdots \end{array}\end{array}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}\begin{array}{c}{C}_q\\ Z_{AVq}\end{array}\\ Z_{1q}\end{array}\\ Z_{2q}\end{array}\\ Z_{iq}\end{array}\\ \begin{array}{c}\begin{array}{c}⋮\\ Z_{pq}\end{array}\\ Z_{IVq}\end{array}\end{array}\right]$$

The IV and AIV are calculated using the subsequent equations.

(10) $$IV=\text{max(}{Z}_{ij}\text{)}\nearrow \text{if}j\in N_b,\text{and}IV=\text{min(}{Z}_{ij}\text{)}\nearrow \text{if}j\in N_{nb}$$

(11) $$AIV=\text{max(}{Z}_{ij}\text{)}\nearrow \text{if}j\in N_{nb},\text{and}AIV=\text{min(}{Z}_{ij}\text{)}\nearrow \text{if}j\in N_b$$

Step 2: The revised decision matrix is normalized within the range of 0-1 utilizing the following formulae.

(12) $$\begin{array}{l}{Z}_{ij}=\frac{Z_{ij}}{Z_v}\nearrow \text{if}j\in N_b\\ Z_{ij}=\frac{Z_v}{Z_{ij}}\nearrow \text{if}j\in N_{nb}\end{array}$$

Step 3: The weighted decision matrix is constructed by the subsequent equation.

(13) $${\varpi }_{ij}={\omega }_j\times z_{ij}$$

Step 4: Computation of the sum of matrix values utilizing Equation 13.

(14) $$k_i={\displaystyle \sum _{i=1}^p{\varpi }_{ij}}$$

Step 5: Assessment of the utility level of options. The utility degrees for IV and AIV are ascertained using the following equations.

(15) $$k_i^{+}=\frac{k_i}{k_{IV}},\text{and}{k}_i^{-}=4$$

Step 6: Regarding the AIV and IV, the utility functions were computed using the equations.

(16) $$f(k_i^{+})=\frac{k_i^{-}}{k_i^{+}+k_i^{-}},\text{and}f(k_i)=\frac{k_i^{+}}{k_i^{+}+k_i^{-}}$$

Step 7: The ultimate value of the utility function is determined using this Equation 17.

(17) $$f(k_i^{})=\frac{k_i^{+}+k_i^{-}}{1+\frac{1-f(k_i^{+})}{f(k_i^{+})}+\frac{1-f(k_i^{-})}{f(k_i^{-})}}$$

Step 8: Evaluating the options. The alternatives are sorted in descending order based on their F(K_i) values, from best to worst. Experimental trials are prioritized according to utility function values, with the highest-ranking trial indicating optimal fabrication for improved performance. Taguchi response tables, main effects plots, and ANOVA ascertain parameter influence and statistical significance.

Step 9: Validation and verification of the experiment

Upon establishing optimal settings, a validation experiment constructs the composite under ideal conditions and assesses its resultant mechanical properties. If the experimental findings correspond with the predictions of utility function (fki) the optimization method is validated as effective.

2.6. Data evaluation

An ANOVA technique is essential for finding the elements that significantly impact mechanical qualities and quantifying their relative contribution [⁴⁰]. A 95% confidence level (α = 0.05) was utilized to guarantee reliability. Furthermore, parameter interactions were analyzed to comprehend their collective influence on overall mechanical performance.

2.6.1. Prediction of mechanical results using ANN

The outcomes of the MARCOS investigation were employed in the artificial neural network for validation by training the response information utilizing MATLAB 2024b [⁴¹]. The feedforward backpropagation method was employed by the training algorithm for input and response variables. All trials, excluding for the third, sixth, and nineth, which were utilized for data validation and were trained using Levenberg-Marquardt backpropagation techniques. Concealed nodes ranging from 1 to 21 was employed for training over 1000 epochs and 10,000 iterations to achieve the final model. The MSE with low values was employed to identify the optimal model among 10,000. The models was evaluated using MAE during the validation and testing processes to identify the most successful model between them.

2.7. Characterization techniques

2.7.1. SEM and EDAX

The morphological analysis of the fractured surfaces was carried out using Gemini-ZEISS, Scanning Electron Microscopy (SEM) operated at an accelerating voltage of 10–15 kV. Samples were sputter-coated with a thin gold layer to prevent charging. SEM micrographs were used to observe fiber–matrix interfacial adhesion, dispersion of fillers, and failure mechanisms [⁴²].

2.7.2. Moisture absorption

The ideal composite was evaluated for moisture absorption in accordance with ASTM D570 to ensure the preservation of food quality, as less water uptake inhibits degradation. The moisture absorption tests were conducted under controlled conditions of 25°C and 50°C relative humidity for a duration of 24 hours. TThe moisture absorption (%) was calculated using the following Equation 18:

(18) $$\text{Moisture Absorption (\%)}=\frac{(W_t-W_0)}{W_0}\times 100$$

where W_t is the weight at time t, and W₀ is the initial dry weight.

2.7.3. Antibacterial test

The antibacterial efficacy of the optimized composite specimens (A3-B3-C2-D1) was assessed utilizing the agar well diffusion method against Escherichia coli (ATCC 8739) and Staphylococcus aureus (ATCC 6538). Circular disks with a diameter of 6 mm were excised from the cured composite sheets, sanitized using 70% ethanol, and air-dried under aseptic circumstances. Agar plates were prepared and consistently inoculated with standardized bacterial samples (0.5 McFarland). The composite disks were meticulously positioned on the inoculated agar surface, and the plates were incubated at 35 ± 1 °C for 24 hours. After incubation, the antibacterial effect was determined by measuring the clear inhibition zone formed around each specimen in millimeters [⁴³]. Post-incubation, the antibacterial efficacy was assessed by measuring the clear inhibitory zone surrounding each specimen in millimeters. All experiments were conducted in triplicate, and the mean inhibition zone values were provided alongside standard deviations.

2.7.4. XRD analysis

The crystalline structure of the optimal composite (A3–B3–C2–D1) was examined by X-ray diffraction (ADVANCED ECO Bruker AXS, Karlsruhe, Germany, Cu Kα, λ = 1.5406 Å) at 40 kV and 30 mA throughout a 2θ range of 10°–80°, employing a step size of 0.02°. The crystallinity index was determined using Segal’s technique.

The crystallinity index (Crl) was calculated using the Segal method as follows in Equation 19:

(19) $$\text{Cr}I(\%)=\frac{I_{002}-I_{am}}{I_{002}}\times 100$$

where I₀₀₂ represents the maximum intensity of the crystalline peak at around 2θ ≈ 22°, and I_am represents the intensity of the amorphous region at around 2θ ≈ 18°.

2.7.5. FTIR analysis

FTIR analysis, the samples were ground into a fine powder and placed on the ATR (Attenuated Total Reflectance). Functional groups of the optimal composites were analyzed by FTIR (PerkinElmer Spectrum Two Shimadzu Corporation, Kyoto, Japan) within the range of 4000–400 cm⁻¹, with a resolution of 4 cm⁻¹ and 32 scans/minute. Baseline correction was applied during the analysis, and peak assignments were made based on established literature and reference spectra, identifying key functional groups relevant to the fiber-based composites.

3. RESULTS AND DISCUSSION

3.1. Mechanical property enhancement through optimization

Table 3 illustrates the influence of fabrication variables on the tensile strength (TS) and flexural strength (FS) of Typha–Lantana camara stem fiber composite. Trial 8 demonstrated highest TS of 30.77 MPa, indicating that this parameter combination facilitates robust adhesion and effective load transfer between matrix and fiber. Conversely, Trial 3 had the lowest tensile strength of 10.58 MPa, possibly due to excessive NaOH treatment, orientation and binder concentration, which have caused fiber breakdown or compromised interfacial bonding.

Table 3
Results of TS and FS and S/N ratios.

S. NO	INPUT PARAMETERS				EXPERIMENTAL RESULTS		S/N RATIOS
	A	B	C	D	TS (MPa)	FS (MPa)	TS (MPa)	FS (MPa)
1	SF	1	40:60	10	18.79	111.48	25.48	40.94
2	SF	3	35:65	15	13.95	75.72	22.89	37.58
3	SF	5	30:70	20	10.58	49.46	20.49	33.89
4	UF	1	35:65	20.	19.64	117.51	25.86	41.40
5	UF	3	30:70	10	11.41	57.69	21.15	35.22
6	UF	5	40:60	15	20.33	13.83	26.16	22.82
7	LF	1	30:70	15	16.55	94.95	24.37	39.55
8	LF	3	40:60	20	30.77	169.34	29.76	44.58
9	LF	5	35:65	10	26.81	201.27	28.57	46.08

Trial 9 attained the maximum flexural strength of 201.27 MPa with A3-B3-C2 and D1, indicating that optimal orientation and ratios enhanced resistance to bending forces. Conversely, Trial 6 demonstrates the minimal flexural strength at 13.83 MPa. This indicates that the interaction of these elements adversely affects both tensile and flexural characteristics.

The results distinctly underscore the essential function of FO in tensile and flexural strength, indicating that appropriate alignment boosts, while moderate NaOH treatment increases strength; conversely, excessive treatment reduces due to fiber breakdown. An elevated fiber-to-matrix ratio (C2) improves tensile strength, although necessitates meticulous calibration of binder concentration. Binder concentration (D1) enhanced adhesion, hence improving composite strength and structural integrity, validating its appropriateness for sustainable food packaging applications.

Table 2 demonstrates that the optimal parameter combination for attaining maximum tensile strength is A3, B3, C2 and D1. The recommended values for superior flexural properties are Long random fiber, 5 wt % NaOH treatment, 35:65 fiber to matrix ratio with 10 wt % concentration.

Table 2 presents the S/N ratios for the FS and TS of Typha-Lantana camara stem fiber composites fabricated utilizing Taguchi’s L9(3³) OA. Trial 8 exhibited the maximum S/N ratio for TS at 29.76, aligning with maximum tensile strength value presented in Table 2. The production circumstances for Trial 8 produce a highly dependable and resilient tensile result. Similarly, the maximum signal-to-noise ratio for FS (46.08) were recorded in Trial 9, corresponding with the highest FS presented in Table 2. The outcomes indicate that the selected parameters were reliable and function effectively in conjunction, suggesting that these circumstances reduce variability and improve mechanical performance. Trial 3 and 5 had lowest S/N ratios for flexural and tensile test.

3.2. Ranking of the composites

3.2.1. CRITIC weighting method

Similarly, the decision matrix for the CRITIC weighting approach was normalized. Following normalization, the data measure (σ_j) and correlation coefficient (ρ_jk) values was calculated using Equations 3 to 7. The characteristic weight of j was calculated using Equation 8 and is presented in Table 4. The mass calculation utilizing the CRITIC approach assigned the highest weight to tensile strength (0.514) followed by flexural strength (0.486) signifying that tensile strength plays a vital role in the optimization process than flexural strength, is reported in Supplementary File Table S1.

Table 4
CRITIC method result.

	TS	FS
Δi	0.333	0.315
Σj	0.083	0.078
Ωj	0.514	0.486

3.2.2. Evaluation of alternatives

The choice matrix is constructed after establishing the ideal value (IV) and anti-ideal value (AIV) using Equations 10 and 11, respectively. The organized decision matrix employed in the MARCOS analysis for the specified alternatives and criteria is presented in Table 2. According to the MARCOS technique outlined in Section 2.5.2, each value is normalized using Equations 12 and 13. Table 5 illustrates the resultant normalized decision matrix for the MARCOS study. The weights presented in Table 3, derived from the CRITIC method are subsequently employed to construct the weighted decision matrix utilizing Equation 15. Table 4 displays the normalized and weighted choice matrix.

Table 5
Normalized and weighted normalized decision matrix for MARCOS analysis.

S. NO	NORMALIZATION		WEIGHTED NORMALIZATON
	TS	FS	TS	FS
1	0.6106	0.5539	0.3137	0.2693
2	0.4534	0.3762	0.2330	0.1829
3	0.3440	0.2457	0.1767	0.1195
4	0.6383	0.5838	0.3280	0.2838
5	0.3708	0.2866	0.1905	0.1393
6	0.6607	0.0687	0.3395	0.0334
7	0.5378	0.4718	0.2763	0.2293
8	1.0000	0.8414	0.5138	0.4090
9	0.8713	1.0000	0.4477	0.4862

The utility degrees and functions of the alternatives are derived using Equations 15 to 18 of the MARCOS approach, with the conclusive results presented in Table 5. The detailed calculations from the preceding steps are provided in the supplementary files, Tables S2 to S8.

The possibilities are ranked based on the acquired f(k_i) values. Table 6 illustrates that, the composite can be organized in decreasing order as A9, A8, A4, A1, A7, A2, A6, A5, A3. According to the function values, Alternative 9 (f(ki) = 0.9290) attained rank 1, followed by Alternative 8 (f(ki) = 0.9181) in rank 2, while Alternative 3 (f(ki) = 0.2947) ranked lowest at 9.

Table 6
MARCOS results.

S. NO	KI	UTILITY DEGREES		UTILITY FUNCTIONS		F(KI)	RANK
		KI+	KI-	F (KI+)	F (KI-)
1	0.5830	0.6243	1.968	0.7592	0.2408	0.5800	4
2	0.4159	0.4453	1.404	0.7592	0.2408	0.4137	6
3	0.2962	0.3172	1.000	0.7592	0.2408	0.2947	9
4	0.6118	0.6551	2.065	0.7592	0.2408	0.6087	3
5	0.3299	0.3532	1.114	0.7592	0.2408	0.3282	8
6	0.3729	0.3993	1.259	0.7592	0.2408	0.3710	7
7	0.5057	0.5415	1.707	0.7592	0.2408	0.5031	5
8	0.9229	0.9882	3.116	0.7592	0.2408	0.9181	2
9	0.9339	1.0000	3.153	0.7592	0.2408	0.9290	1
IV	0.93387
AIV	0.29621

The ranking derived from the utility function (f(k_i)) illustrates the overall efficacy of both replies, with Trials 8 and 9 occupying the foremost positions, corresponding to the highest signal-noise ratios indicated in Table 3. This reliability illustrates the dependability of MARCOS in optimization with multiple goals for enhancing the mechanical characteristics of natural fiber. The results demonstrate the efficacy of utility function (f(k_i)) in identifying optimal parameter configurations to improve both flexural and tensile strength, providing a comprehensive approach to the multi-objective optimization.

Table 7 presents the response table for the means of the utility function (f(k_i)) on the performance analysis on the effects of BC, FMR, NT and FO. The Delta values indicate the maximum and minimum mean utility function (f(k_i)) values for each parameter. FO exhibits the most influence at 0.3539, succeeded by FMR at 0.2752 and binder concentration at 0.1832, whereas NaOH treatment demonstrates the minimal effect at 0.0323. The parameter ranking indicates that fiber orientation (Rank 1) were critical factor in improving mechanical properties, likely because to its direct impact on structural integrity and load transfer efficiency. FMR (Rank 2) significantly influences stress distribution and overall mechanical performance, highlighting the importance of fiber orientation. BC demonstrates a moderate impact, indicating that although binder concentration influences composite bonding, it is less critical than FMR and FO. NaOH Treatment (Rank 4) signifies negligible impact, indicating the alkali treatment were advantageous for surface changes [⁴⁴], as per utility function (f(k_i)). The results suggest that fiber matrix ratio should be prioritized in the manufacturing process, followed by BC and FO to ensure the excellent performance. Simultaneously, NaOH Treatment can be adjusted within practical constraints without significantly affecting the characteristics of composites.

Table 7
Response table for means of utility function f(k_i)

LEVEL	A (FO)	B (NT) (WT.%)	C (FMR)	D (BC) (WT.%)
1	0.4295	0.5639	0.6230	0.6124
2	0.4359	0.5533	0.6505	0.4292
3	0.7834	0.5316	0.3753	0.6071
Delta	0.3539	0.0323	0.2752	0.1832
Rank	1	4	2	3

Figure 2 shows the main effect plot for means substantiates the influence of BC, FMR, NT and FO on the utility function (f(k_i)), aiding in identifying optimal parameter levels. The FO parameter exhibits a notable increment, indicating that Level A3 provides the optimal utility function (f(k_i)), rendering it the superior option for enhancing mechanical qualities. NT exhibits a slight increase over the levels, suggesting that Level B1 may be the most appropriate choice, albeit with a constrained total impact. The FMR parameter exhibits a notable negative slope, C2 yielding the maximum utility function (f(k_i)). This suggests that excessive fiber content (Level 1) undermines the composite due to insufficient matrix bonding and infiltration. The BC parameter exhibits a non-linear trend, D1 providing a superior utility function (f(k_i)) compared to Level 2, indicating an optimal concentration for preserving structural integrity. The significant influence of FMR and FO underscores their critical role in augmenting the durability and strength of composites. According to these findings, the confirmation test was conducted at A3-B1-C2-D1, as this combination is expected to produce optimal mechanical performance in Typha-Lantana camara stem fiber composites. Under ideal conditions, the tensile strength attained 26.81 MPa and the flexural strength reached 201.27 MPa, nearly corresponding to Taguchi’s ninth trial. In previous study, hybrid epoxy composites with flax, vetiver, and mahogany fruit fillers (MFFs) were optimized using Taguchi L9 and CRITIC–EDAS methods. At 10 wt.% MFF, tensile strength (56.32 MPa), and flexural strength (89.65 MPa) were highest [⁴⁵].

Figure 2
Main effect plot for utility function f(ki).

3.3. Statistical analysis

Table 8 highlights the ANOVA results for FO, NT, FMR, and BC, together with their interactions, significantly affect the utility function. The R² value of 93.62% indicates that the selected parameters shows a significant portion of the changes in utility functions (f(ki)), underscoring the specified empirical factors. The adjusted R² is 48.92%, suggesting that while the model exhibits a satisfactory fit, there may be influences from unaccounted variables or variability due to experimental noise.

Table 8
Analysis of variance.

SOURCE	DF	SEQ SS	ADJ SS	ADJ MS	F-VALUE	P-VALUE	CONTRIBUTION
Model	7	0.421949	0.421949	0.060278	2.09	0.488	93.62%
Linear	4	0.281565	0.246780	0.061695	2.14	0.468	62.47%
A	1	0.187915	0.173101	0.173101	6.02	0.246	41.69%
B	1	0.001569	0.000934	0.000934	0.03	0.887	0.35%
C	1	0.092039	0.000538	0.000538	0.02	0.913	20.42%
D	1	0.000041	0.017546	0.017546	0.61	0.578	0.01%
2-Way Interaction	3	0.140384	0.140384	0.046795	1.63	0.510	31.15%
A × B	1	0.110106	0.026198	0.026198	0.91	0.515	24.43%
A × C	1	0.000544	0.003686	0.003686	0.13	0.781	0.12%
B × C	1	0.029734	0.029734	0.029734	1.03	0.495	6.60%
Error	1	0.028776	0.028776	0.028776			6.38%
Total	8	0.450725					100.00%
S = 0.169636, R2 = 93.62, Adj. R2 = 48.92

3.3.1. Effect of individual factors

FO (A) was most significant contributor at 41.69%, followed by C at 20.42%, B at 0.35%, and D at 0.01%. Table 7 and Figure 2 demonstrate that FMR predominantly influences the utility function (f(k_i)) mainly by its impact load distribution and fiber–matrix bonding. FO showed considerable impact with A3 augmenting composite characteristics. Conversely, NT and BC exhibited negligible effects, indicating a modest influence on optimization. P-values (A = 0.246, C = 0.913, B = 0.887, D = 0.578) above 0.05, indicating a lack of statistical significance in accordance with the Pareto chart.

3.3.2. Effect of two-way relations

Bidirectional relations contributed to 31.15% of variability with A × B (24.43%) being the most significant followed by B × C (6.60%) and A × C (0.12%). The A–B interaction may improve fiber surface characteristics and mechanical interlocking. All p-values surpassed 0.05, signifying that the contributions lacked substantial statistical significance.

3.3.3. Statistical validation of the model

The model’s p-value of 2.09 and F-value of 0.488 indicate that the selected factors and their interactions do not significantly affect the utility function f(k_i) with 95%. This suggests the empirical factors explain 93.62% of variance, statistical significance were limited, either due to variability in the experimental settings or unaccounted impacts. The error contribution is 6.38%, signifying a reasonably low degree of error. This indicates experiment were well-structured; however, enhancements could involve augmenting the size of the specimen or including higher interactions.

Although the statistical significance is limited empirical engineering data distinctly highlights Fiber to Matrix Ratio and Fiber Orientation (FO) critical characteristics, as illustrated in Figure 2 and Table 7. The confirmation test conducted at A3-B1-C2-D1 as previously recommended, since these levels augment the utility function f(k_i) based on average response patterns rather than exclusively at statistical validation. Future research may enhance the reliability augmenting experimental repetitions and optimizing parameter selection to bolster statistical significance. Despite the absence of robust evidence for present validation, practical optimization and observed trends indicates prioritizing Fiber to Matrix Ratio and Fiber Orientation (FO) during composite manufacturing is essential for attaining enhanced mechanical performance.

3.3.4. ANN prediction

The MARCOS results were predicted utilizing the soft computing Artificial Neural Network (ANN) methodology. The initial stage involved normalizing the input and output data to a range of –1 to 1, ensuring equal weighting for all factors. Equation 20 is elucidated below.

(20) $$N_n=\left[\frac{N_i-N_{\text{min}}}{M_{\max }-N_{\min }}\right]$$

Where, Nn represents the normalized findings, Ni denotes the reading, Nmax and Nmin indicate the maximum and minimum readings inside the specified array. The model completed 10,000 iterations, enhancing prediction efficiency. The normalized values were transformed into MSE utilizing Equation 21 as demonstrated below.

(21) $$\text{MSE =}\frac{1}{t\ast r}{\displaystyle \sum _{x=0}^t{\displaystyle \sum _{y=1}^r{[W_y(x)-Q_y(x)]}^2}}$$

Where, ‘r’ denotes the total number of specimens, ‘t’ represents the total parameters utilized in training, ‘W’ signifies the actual data, and ‘Q’ indicates the projected response from the ANN. The optimal models were determined using MAE as per Equation 22 across multiple developed models.

(22) $$\text{MAE =}\frac{1}{t\ast r}|W_y(x)-Q_y(x)|$$

The superior model was selected based on the minimum value of MAE. The graphical depiction of concealed nodes along the X-axis and MAE along the Y-axis indicated that, the third node had the lowest MAE value (Figure 3). The 4-3-1 topology were suitable model with 4 components, 3 hidden nodes, and 1 response respectively (Figure 4).

Figure 3
Mean Absolute Error (MAE) values for Artificial Neural Network (ANN) models.

Figure 4
Adopted topology for artificial neural network model.

The efficacy of the suggested model was assessed by choosing three random values, specifically the third, eighth, and fourteenth trials. The aforementioned values was utilized for model validation, whereas the remaining trials were employed for model training. The results of the actual calculation prediction and artificial neural network exhibited only minor discrepancies (Figure 5). The readings exhibited comparable findings with minimal error, and a regression graph indicated a 94.28% R² value (Figure 6).

Figure 5
Validation of predicted and experimental outcomes for the 4-3-1 accepted topology.

Figure 6
Comparison between experimental and predicted values.

Figure 5 presents the validation of experimental and ANN-predicted utility values for the selected trials. The close agreement between the experimental and predicted results indicates that the developed ANN model effectively captures the relationship between the processing parameters and the composite performance. The small deviation between the values suggests that the model is capable of accurately predicting the system behavior. Figure 6 shows the correlation between the experimental and ANN-predicted utility values. The high coefficient of determination (R² = 0.9428) indicates a strong correlation between the predicted and experimental results. This confirms that the developed ANN model provides reliable predictions of the composite performance and validates the effectiveness of the optimization approach. The graph depicting the relationship between actual results and those computed by the ANN demonstrates a strong connection, indicating that the 4-3-1 topology is optimal for examining factors related to mechanical strength.

3.4. Microstructural features of composite surface

3.4.1. SEM analysis

Figures 7a and 7b illustrate the fractured surfaces of the tensile test specimens, while Figures 7c and 7d show the fractured surfaces of the flexural test specimens for the evaluated Typha–Lantana camara stem fiber composite intended for sustainable food-packaging applications. These micrographs offer critical insights into fiber-matrix interaction, reinforcing efficacy, and structural behavior. The failure nodes encompass matrix cracking, fiber pullout, fiber breaking, void formation and delamination underscoring the vulnerabilities of composites.

Figure 7a illustrates fiber pullout areas, indicating inadequate adhesion at the fiber-matrix interface, which may lead to premature composite failure under tensile stresses. Figure 7b illustrates substantial cracking, a consequence of brittle fracture occurring when the matrix does not deform under stress and also a fiber breakage as the primary failure mode, as demonstrated by broken fiber ends and fragmentation, which further signifies structural weakness within the composite material.

Figure 7
SEM images of Typha-Lantana camara stem fibre composites (a & b) Tensile test specimens, (c & d) Flexural test specimens.

Figure 7c illustrates delamination accompanied by fiber bridging at the fiber-matrix contact. Figure 7d illustrates fiber pullout and interfacial debonding, wherein fibers disengage from the matrix, resulting in voids and compromising the structural integrity of the composites. SEM revealed that hybridization reduced fiber pull-out and voids compared to single-fiber composites [⁴²].

3.4.2. EDAX

The EDAX analysis of Typha–Lantana camara stem fiber composite (Figure 8) indicates that carbon (85.49 wt%) and oxygen (14.51 wt%) are the primary constituents. The elevated carbon content verifies the organic composition of the composite, emanating from hemicellulose, lignin and cellulose structure found in bio-based matrix and the natural fibers. The significant amount of oxygen suggests the existence of hydroxyl groups, esters, and other oxygenated chemicals characteristic of plant-derived fibers. The lack of inorganic or metallic residues indicates material purity, suggesting potential suitability for food packaging uses. SEM observations further validate exceptional fiber-matrix adhesion, underscoring robust compatibility, structural integrity, and dependable performance of the composite.

Figure 8
EDAX of Typha-Lantana camara stem fiber composite.

3.5. Moisture absorption behaviour of the composite samples

Figure 9 illustrates the moisture absorption of the optimum Typha–Lantana camara composite (A3-B3-C2-D1), exhibiting fast uptake within the initial 60 minutes, stabilization after 95 minutes, and saturation at 317.19 g under controlled conditions. Five-minute intervals demonstrate a swift initial absorption, facilitated by the hydrophilic characteristics of natural fibers that allow water infiltration into composite. After 1 hour, absorption rate declines, signifying that the composite had attained its maximum water retention capacity, thereby preventing the absorption.

Figure 9
Moisture absorption of Typha-Lantana camara stem fibre composite.

The interaction of moisture with the composite is crucial in food packaging, as it affects the material’s ability to withstand excessive water absorption, which in turn impacts its efficacy in food storage. This material is suitable for packaging semi-dry and dry food items, such as baked products, tea leaves, grains, and nuts. It permits controlled moisture absorption, but preventing excessive water retention is necessary to maintain food quality and prevent spoilage.

3.6. Antibacterial test

The agar well diffusion method validated the antibacterial efficacy of the optimized composite specimens (A3–B3–C2–D1). The mean inhibitory zone diameter for Escherichia coli was 12.4 ± 0.51 mm, whereas for Staphylococcus aureus it was 10.1 ± 0.43 mm (Figure 10). The results unequivocally demonstrate that the composite material inhibited the proliferation of both Gram-negative and Gram-positive bacteria. The increased inhibition noted against E. coli indicates a improved susceptibility of Gram-negative organisms to the leachable constituents of the composite. Similarly, Henequen plant leaf fibers composite showed strong antibacterial activity (31 mm inhibition zone) against E. coli [⁴⁶].

Figure 10
Inhibition zone diameters of the optimized composite against E. coli and S. aureus.

The antibacterial efficacy is due to the synergistic action of the corn bio-resin matrix and surface-treated natural fibers, which may emit trace amounts of bioactive chemicals or modify the local microenvironment surrounding the composite. The existence of clearly defined inhibition zones indicates that the mechanism is at least partially diffusion-driven, wherein active drugs permeate the agar medium and suppress bacterial growth. The findings demonstrate that the optimized composite provides both mechanical integrity and enhanced sanitary performance. This dual functionality increases its applicability in packaging, biomedical, and food-contact products necessitating antibacterial characteristics.

3.7. XRD analysis of optimum composite

XRD patterns of both Typha latifolia and Lantana camara fibers in previous study, exhibited characteristic cellulose I peaks at 2θ ≈ 16° and 22–24°, confirming their semi-crystalline nature. The crystallinity index increased significantly for Typha fiber (29.6% to 55.8% after treatment), while Lantana fiber showed a crystallinity of ~42.89%, indicating good crystalline behavior. The increase in crystallinity is attributed to the removal of amorphous components, which enhances stiffness and interfacial bonding in composites [⁴⁷, ⁴⁸, ⁴⁹]. The X-ray diffraction (XRD) pattern of the optimum Typha–Lantana camara fiber composite (A3-B3-C2-D1) revealed distinct crystalline peaks characteristic of cellulose I structure, along with broad amorphous regions originating from hemicellulose, lignin, and starch-based matrix. A sharp diffraction peak was observed at 2θ ≈ 22.86°, corresponding to the (002) crystallographic plane of cellulose, confirming the high crystalline nature of the treated fibers (Figure 11). Additional minor peaks appeared at 2θ ≈ 17.93° (101 plane) and 26.56 ° (040 plane), which are typical of natural lignocellulosic fibers. The broad hump between 17°–20° indicates the presence of amorphous starch and residual hemicellulose, which provide flexibility to the matrix. The calculated crystallinity index (CI) of the optimized composite was found to be 21.43%. This improvement in crystallinity suggests stronger interfacial bonding between the fibers and starch matrix, which correlates with the enhanced tensile strength (26.81 MPa) and flexural strength (201.27 MPa) observed in the optimum sample.

Figure 11
XRD analysis of optimum composite.

X-ray diffraction (XRD) analysis was conducted on the composite samples to assess the crystallinity changes after alkali treatment. The crystallinity index (CI) for the alkali treatment increased to 21.43%. This increase in crystallinity indicates that the alkali treatment significantly enhanced the order and structure of the cellulose in the fiber component of the composite. The improvement in crystallinity can be attributed to the alkali treatment’s ability to remove non-cellulosic impurities, such as lignin and hemicellulose, from the fibers. This results in a more exposed and aligned cellulose structure, which enhances the overall crystallinity of the composite. The increased crystallinity not only improves the fiber-matrix interfacial bonding but also strengthens the composite’s mechanical properties, such as tensile and flexural strength. Furthermore, the higher crystallinity reduces the composite’s susceptibility to moisture absorption, thereby increasing its stability and durability in various environmental conditions. These changes demonstrate that alkali treatment is an effective method for enhancing the crystallinity of natural fibers within composites, leading to improved composite performance and suitability for structural applications.

3.8. FTIR analysis of optimum composite

FTIR spectra of Typha latifolia and Lantana camara fibers showed cellulose peaks such as –OH stretching (~3400 cm⁻¹) and C–H stretching (~2900 cm⁻¹), confirming the lignocellulosic nature. Peaks around ~1735–1740 cm⁻¹ and ~1250–1514 cm⁻¹ were associated with hemicellulose and lignin, which diminished or disappeared after alkali treatment, indicating removal of amorphous constituents. This confirms enhancement of cellulose structure and potential for improved fiber–matrix interaction [⁴⁷, ⁴⁸, ⁴⁹]. The Fourier Transform Infrared (FTIR) spectrum of the optimum composite (A3-B3-C2-D1) as shown in Figure 12 confirmed the presence of functional groups associated with both the natural fibers and the starch-based resin. FTIR analysis of the alkali-treated composite fibers revealed significant changes in key functional groups and its peak assignments are shown in Table 9. The O-H stretching band at 3289 cm⁻¹ showed increased intensity, indicating the exposure of additional hydroxyl groups, which enhances the fiber-matrix bonding potential and moisture absorption capacity. The C-H stretching band at 2926 cm⁻¹ decreased slightly, reflecting the reduction of aliphatic content as non-cellulosic components like hemicellulose and lignin were removed. A reduction in the C=O stretching band at 1736 cm⁻¹ was observed, confirming the breakdown of esterified and carboxylated groups, and supporting the removal of hemicellulose and lignin. The C=C stretching band at 1410 cm⁻¹ also diminished, further indicating the effective removal of lignin and hemicellulose. Additionally, the C-O stretching band at 1231 cm⁻¹ intensified, reflecting the enhanced crystallinity of cellulose and the ordered arrangement of cellulose chains following alkali treatment. The C-O-C stretching band at 1014 cm⁻¹ was significantly reduced, indicating the effective removal of hemicellulose, which contains ether linkages. These changes in the functional groups suggest that alkali treatment effectively improves the crystallinity, structural integrity, and interfacial bonding of the fibers, enhancing their performance for use in composite materials. In prior studies, cellulose microfibers from Typha australis were extracted using NaOH pretreatment followed by hydrogen peroxide bleaching treatments increased cellulose from 43% to 75% and crystallinity from 28% to 63% confirmed by FTIR and XRD [⁵⁰].

Figure 12
FTIR analysis of optimum composite.

Table 9
FTIR peak assignments of optimum composite.

WAVENUMBER (cm−1)	FUNCTIONAL GROUP	ASSIGNMENT	ALKALI-TREATED FIBER
3289	O-H Stretching	Hydroxyl groups (cellulose)	Intensified
2926	C-H Stretching	Aliphatic CH2	Slight reduction
1736	C=O Stretching	Ester and carboxyl groups	Reduced
1522	C=C Stretching	Lignin and hemicellulose	Reduced
1410	C-H Bending	Cellulose and lignin	Slight shift
1231	C-O Stretching	C-O in cellulose	Intensified
1014	C-O-C Stretching	Hemicellulose	Significantly reduced

The interfacial bonding between the Typha–Lantana fibers and corn starch bio-resin primarily occurs through hydrogen bonding and mechanical interlocking mechanisms. The alkali treatment removed hemicellulose, lignin, wax, and surface impurities from the fiber surface, thereby exposing additional hydroxyl groups and increasing surface roughness. These exposed hydroxyl groups interact with the hydroxyl-rich starch matrix through intermolecular hydrogen bonding, which enhances fiber–matrix adhesion. Simultaneously, the increased surface roughness after NaOH treatment promotes mechanical interlocking between the fibers and matrix, improving stress transfer efficiency and overall mechanical performance of the composite.

4. Conclusion

This study fabricated and optimized Typha–Lantana stem fiber composites reinforced with corn starch bio-resin for sustainable food packaging applications. Using Taguchi–MARCOS approach, fiber orientation (41.69%) and fiber-to-matrix ratio (20.42%) were identified as the most influential parameters, with the optimum combination (A3–B3–C2–D1) achieving a tensile strength of 26.81 MPa and flexural strength of 201.27 MPa. ANN validation (4-3-1 topology) demonstrated strong predictive reliability, with a regression correlation of R² = 94.28% between experimental and predicted values.

XRD analysis of the optimum sample showed a sharp peak at 2θ ≈ 22.86° (002 plane), with additional peaks at 17.93° (101) and 26.56° (040), yielding a crystallinity index (CI) of 21.43%, which supported improved load transfer. FTIR spectra exhibited a broad O–H band at 3289 cm⁻¹, C–H stretching at 2926 cm⁻¹, C=O peak at 1736 cm⁻¹, lignin skeletal vibration at 1522 cm⁻¹, and strong cellulose C–O–C bands near 1014 cm⁻¹, confirming hemicellulose/lignin removal and enhanced fiber–matrix adhesion. SEM micrographs revealed reduced voids and improved interfacial bonding, while EDAX confirmed high organic composition carbon (85.49 wt%) and oxygen (14.51 wt%).

Moisture absorption stabilized for 95 minutes, achieving saturation weight gain of 312.79 g confirming suitability under humid conditions. Antibacterial tests showed inhibition zones of 12.4 ± 0.6 mm for E. coli and 10.1 ± 0.4 mm for S. aureus, validating hygienic performance for food-contact applications. Future scope includes expanding experimental replicates to improve statistical significance, evaluating thermal stability and long-term durability. Cytotoxicity and migration studies will be explored in future research to fully evaluate the food-contact safety of the material.

5. DATA AVAILABILITY

All the required data’s were incorporated within the manuscript.

6. SUPPLEMENTARY MATERIAL

The following online material is available for this article:

Table S1 - Detailed Calculation of CRITIC (Equation 3 to 7).

Table S2 - Constructing decision matrix and incorporating IV and AIV (Equation 9 to 11).

Table S3 - Normalized decision matrix (Equation 12).

Table S4 - Sum of matrix (K_i) (Equation 14).

Table S5 - Determining IV & AIV utility degree (Equation 15).

Table S6 - Determining utility function (Equation 16).

Table S7 - Determining ultimate value f(ki) for step 1: $\frac{1-f(k_i^{+})}{f(k_i^{+})}and\frac{1-f(k_i^{-})}{f(k_i^{-})}$ (Equation 17).

Table S8 - Determining ultimate value f(ki) for step 2: $1+\frac{1-f(k_i^{+})}{f(k_i^{+})}\text{and}\frac{1-f(k_i^{-})}{f(k_i^{-})}$ and step 3: f(K_i) (Equation 17).

7. BIBLIOGRAPHY

• ^[1] KHOSRO, S.K., SOLTANI, P., SHEIKHMOZAFARI, M.J., et al, “Acoustical, thermal, and mechanical performance of Typha Latifolia fiber panels: experimental evaluation and modeling for sustainable building applications”, Journal of Building Engineering, v. 99, pp. 111579, 2025. doi: https://doi.org/10.1016/j.jobe.2024.111579.
  » https://doi.org/10.1016/j.jobe.2024.111579
• ^[2] PATTNAIK, S.S., BEHERA, D., DAS, N., et al, “Fabrication and characterization of natural fiber reinforced cowpea resin-based green composites: an approach towards agro-waste valorization”, RSC Advances, v. 14, n. 35, pp. 25728–25739, 2024. doi: https://doi.org/10.1039/D4RA03546A. PubMed PMID: 39148754.
  » https://doi.org/10.1039/D4RA03546A
• ^[3] LUVIA, E.U., OCHIAGHA, E.I., NKIRU, O.E., et al, “Effects of alkaline treatment on the properties of pineapple (Ananas comosus) leaf fiber (PALF) reinforced with tapioca-based bio resin (cassava starch)”, Progress in Color, Colorants and Coatings Journal, v. 18, n. 7, pp. 797–804, 2023.
• ^[4] KRISHNAN, T., JAYABAL, S., KRISHNA, V.N., “Tensile, flexural, impact, and hardness properties of alkaline-treated Sunnhemp fiber reinforced polyester composites”, Journal of Natural Fibers, v. 17, n. 3, pp. 326–336, 2020. doi: https://doi.org/10.1080/15440478.2018.1492488.
  » https://doi.org/10.1080/15440478.2018.1492488
• ^[5] REZIG, S., JAOUADI, M., EL OUDIANI, A., et al, “Physical properties, morphology and crystallinity of Typha leaf fibers processed by different extraction treatments”, Journal of the Textile Institute, v. 116, n. 1, pp. 21–32, 2025. doi: https://doi.org/10.1080/00405000.2024.2314261.
  » https://doi.org/10.1080/00405000.2024.2314261
• ^[6] REZIG, S., KHOFFI, F., JAOUADI, M., et al, “Optimization and characterization of combined degumming process of Typha angustata L. stem fibers”, Journal of Renewable Materials, v. 12, n. 6, pp. 1071–1086, 2024. doi: https://doi.org/10.32604/jrm.2024.049935.
  » https://doi.org/10.32604/jrm.2024.049935
• ^[7] REZIG, S., JAOUADI, M., KHOFFI, F., et al, “Typha fiber reinforced polyester composites: tensile properties and statistical analysis”, Journal of the Textile Institute, v. 114, n. 5, pp. 717–725, 2023. doi: https://doi.org/10.1080/00405000.2022.2066433.
  » https://doi.org/10.1080/00405000.2022.2066433
• ^[8] RAMESHKANNAN, G., RAMESH BABU, S., SENTHAMARAIKANNAN, P., et al, “Mechanical and wear characterization of sodium hydroxide–treated ficus amplissima natural fiber epoxy composites”, Journal of Natural Fibers, v. 22, n. 1, pp. 2503969, 2025. doi: https://doi.org/10.1080/15440478.2025.2503969.
  » https://doi.org/10.1080/15440478.2025.2503969
• ^[9] REDDY, B.M., REDDY, R.M., REDDY, P.V., et al, “Effect of alkali treatment on mechanical properties and morphology of the Balanites aegyptiaca composite”, Proceedings of the Institution of Mechanical Engineers. Part C, Journal of Mechanical Engineering Science, v. 238, n. 11, pp. 5077–5086, 2024. doi: https://doi.org/10.1177/09544062231217596.
  » https://doi.org/10.1177/09544062231217596
• ^[10] REDDY, K.H., REDDY, R.M., RAMESH, M., et al, “Impact of alkali treatment on characterization of Tapsi (Sterculia urens) natural bark fiber reinforced polymer composites”, Journal of Natural Fibers, v. 18, n. 3, pp. 378–389, 2021. doi: https://doi.org/10.1080/15440478.2019.1623747.
  » https://doi.org/10.1080/15440478.2019.1623747
• ^[11] REDDY, B.M., REDDY, Y.V.M., REDDY, B.C.M., et al, “A study on mechanical, structural, morphological, and thermal properties of raw and alkali treated Cordia dichotoma-polyester composite”, Polymer Composites, v. 42, n. 1, pp. 309–319, 2021. doi: https://doi.org/10.1002/pc.25826.
  » https://doi.org/10.1002/pc.25826
• ^[12] EDLABADKAR, A.P., MADHU, M.C., ARUNBHARATHI, R., et al, “Optimization of mechanical properties of Cordia dichotoma/alumina nanoparticle reinforced epoxy nanocomposites under alkali treatment using ANN techniques”, Journal of Environmental Nanotechnology, v. 14, n. 2, pp. 525–540, 2025. doi: https://doi.org/10.13074/jent.2025.06.2521549.
  » https://doi.org/10.13074/jent.2025.06.2521549
• ^[13] HAKIM, L., WIDYORINI, R., UMEMURA, K., et al, “The characteristics of citric acid-bonded composite boards produced from alkaline-modified fibrovascular bundle of Salacca sumatrana fronds”, Journal of Natural Fibers, v. 21, n. 1, pp. 2315505, 2024. doi: https://doi.org/10.1080/15440478.2024.2315505.
  » https://doi.org/10.1080/15440478.2024.2315505
• ^[14] SUBBIAH, G., CHANNAPPAGOUDRA, M., BHANOT, D., et al, “Bio-based innovation: assessing the suitability of Solanum nigrum stem fibers for textile and biomedical applications”, Results in Engineering, v. 26, pp. 105441, 2025. doi: https://doi.org/10.1016/j.rineng.2025.105441.
  » https://doi.org/10.1016/j.rineng.2025.105441
• ^[15] KUBERA SAMPATH KUMAR, S., PRAKASH, C., RAMESH, P., et al, “Study of wound dressing material coated with natural extracts of Calotropis gigantea, Eucalyptus globulus and buds of Syzygium aromaticum solution enhanced with rhEGF (REGEN-D^TM 60)”, Journal of Natural Fibers, v. 18, n. 12, pp. 2270–2283, 2021. doi: https://doi.org/10.1080/15440478.2020.1726239.
  » https://doi.org/10.1080/15440478.2020.1726239
• ^[16] SRIARIYANUN, M., KARUPPIAH, K., PARAMASIVAM, P., et al, “Manufacturing and mechanical characterization of alkali-treated Borassus fruit natural fiber with cellulose acetate synthetic fiber reinforced polymer composite: selection of optimum configuration using CRITIC-integrated EDAS soft computing techniques”, Journal of Material Cycles and Waste Management, v. 27, n. 2, pp. 880–897, 2025. doi: https://doi.org/10.1007/s10163-024-02142-y.
  » https://doi.org/10.1007/s10163-024-02142-y
• ^[17] BEHERA, D., PATTNAIK, S.S., MANNA, S., et al, “Sustainable valorization of lignocellulosic biomass into green composites: recent advances, applications, and future perspectives”, Sustainable Materials and Technologies, v. 46, e01748, 2025. doi: https://doi.org/10.1016/j.susmat.2025.e01748.
  » https://doi.org/10.1016/j.susmat.2025.e01748
• ^[18] BEHERA, D., PATTNAIK, S.S., PATRA, S.S., et al, “Development and characterization of water hyacinth reinforced thermoplastic starch as sustainable biocomposites”, RSC Sustainability, v. 3, n. 4, pp. 1807–1818, 2025. doi: https://doi.org/10.1039/D4SU00803K.
  » https://doi.org/10.1039/D4SU00803K
• ^[19] PATTNAIK, S.S., BEHERA, D., NANDA, D., et al, “Green chemistry approaches in materials science: physico-mechanical properties and sustainable applications of grass fiber-reinforced composites”, Green Chemistry, v. 27, n. 10, pp. 2629–2660, 2025. doi: https://doi.org/10.1039/D4GC05569A.
  » https://doi.org/10.1039/D4GC05569A
• ^[20] BEHERA, D., PATTNAIK, S.S., NANDA, D., et al, “A review on bamboo fiber reinforced composites and their potential applications”, Emergent Materials, v. 8, n. 1, pp. 157–172, 2025. doi: https://doi.org/10.1007/s42247-024-00832-9.
  » https://doi.org/10.1007/s42247-024-00832-9
• ^[21] BEHERA, D., PATTNAIK, S.S., MISHRA, P.P., et al, “Fabrication and characterization of industrial biocomposite from cellulosic fibers of Luffa cylindrica in a protein based natural matrix”, Industrial Crops and Products, v. 212, pp. 118328, 2024. doi: https://doi.org/10.1016/j.indcrop.2024.118328.
  » https://doi.org/10.1016/j.indcrop.2024.118328
• ^[22] BEHERA, A.K., MOHANTY, C., PRADHAN, S.K., et al, “Assessment of soil and fungal degradability of thermoplastic starch reinforced natural fiber composite”, Journal of Polymers and the Environment, v. 29, n. 4, pp. 1031–1039, 2021. doi: https://doi.org/10.1007/s10924-020-01944-z.
  » https://doi.org/10.1007/s10924-020-01944-z
• ^[23] YU, C.H., “A study on the extraction methods and characterizations of Typha leaf fibers,” D.Sc. Thesis, University of Manitoba, Winnipeg, 2021.
• ^[24] DHIKALE, R., GULECHA, V., ZALTE, A., “Pharmacognostic standardization of medicinally important notorious weed-Lantana camara”, International Journal of Pharmaceutical Sciences and Nanotechnology, v. 15, n. 4, pp. 6061–6071, 2022. doi: https://doi.org/10.37285/ijpsn.2022.15.4.6.
  » https://doi.org/10.37285/ijpsn.2022.15.4.6
• ^[25] HASAN, M., RAHMAN, M., CHEN, Y., et al, “Optimization of Typha fibre extraction and properties for bio-composite applications using desirability function analysis”, Polymers, v. 14, n. 9, pp. 1685, 2022. doi: https://doi.org/10.3390/polym14091685. PubMed PMID: 35566855.
  » https://doi.org/10.3390/polym14091685
• ^[26] GILLELA, S., YADAV, S.M., SIHAG, K., et al, “Effects of addition of nanoclay in phenol-formaldehyde resins on the properties of Lantana camara fibre composites”, Journal of Tropical Forest Science, v. 36, n. 1, pp. 40–50, 2024. doi: https://doi.org/10.26525/jtfs2024.36.1.40.
  » https://doi.org/10.26525/jtfs2024.36.1.40
• ^[27] KANDAVALLI, S.R., SUNDRAMURTHY, V.P., KRISHNASAMY, V.D., et al, “Bioresin based hybrid green composite preparation using Holoptelea integrifolia fibers reinforced by Ziziphus jujuba seed particles: a fuzzy logic assisted optimization of mechanical behaviour”, Zeitschrift für Physikalische Chemie, v. 239, n. 5, pp. 747–772, 2025. doi: https://doi.org/10.1515/zpch-2024-0837.
  » https://doi.org/10.1515/zpch-2024-0837
• ^[28] RAHMAN, M., CICEK, N., CHAKMA, K., “The optimum parameters for fibre yield (%) and characterization of Typha latifolia L. fibres for textile applications”, Fibers and Polymers, v. 22, n. 6, pp. 1543–1555, 2021. doi: https://doi.org/10.1007/s12221-021-0194-8.
  » https://doi.org/10.1007/s12221-021-0194-8
• ^[29] ARAVINDH, A., SELVARAJ, S., GOKULKUMAR, G., et al, “Taguchi fuzzy multi-response optimization of process parameters in compression molding of natural hybrid composite”, Iranian Polymer Journal, v. 32, n. 7, pp. 811–828, 2023. doi: https://doi.org/10.1007/s13726-023-01168-7.
  » https://doi.org/10.1007/s13726-023-01168-7
• ^[30] AARTHIPRIYA, V., UMARANI, C., “Studies on the characteristic features of Abutilon indicum fibre with chemical treatments for the potential application in cement composites”, Biomass Conversion and Biorefinery, v. 15, n. 4, pp. 6131–6147, 2025. doi: https://doi.org/10.1007/s13399-024-05496-3.
  » https://doi.org/10.1007/s13399-024-05496-3
• ^[31] BINOJ, J.S., MARIATTI, M., BRAILSON MANSINGH, B.B., et al, “Viscoelastic and heat-resistant behavior of surface-treated areca fruit husk fiber-reinforced polymer composites”, Iranian Polymer Journal, v. 32, n. 7, pp. 841–853, 2023. doi: https://doi.org/10.1007/s13726-023-01169-6.
  » https://doi.org/10.1007/s13726-023-01169-6
• ^[32] GUPTA, R.K., SELVARAJU, S., VIJAYARAGHAVALU, S., et al, “Multi-objective optimization of bamboo-madar fiber reinforced composites for lightweight automotive applications using machine learning and genetic algorithms”, Journal of Environmental Nanotechnology, v. 13, n. 4, pp. 358–376, 2024. doi: https://doi.org/10.13074/jent.2024.12.244978.
  » https://doi.org/10.13074/jent.2024.12.244978
• ^[33] NAVEENA, H.S., SUNIL, S., KAKKERI, S., et al, “Development and mechanical testing of natural fibre reinforced polypropylene resin hybrid composite”, Advances in Materials and Processing Technologies, v. 8, n. 2, pp. 790–801, 2022.
• ^[34] MUNDE, Y., SHINDE, A.S., SIVA, I., et al, “Assessment of physical and vibration damping characteristics of sisal/PLA biodegradable composite”, Materials Physics and Mechanics, v. 51, n. 7, pp. 107–116, 2023.
• ^[35] KARUMURI, S., HALDAR, B., PRADEEP, A., et al, “Multi-objective optimization using Taguchi based grey relational analysis in friction stir welding for dissimilar aluminium alloy”, International Journal on Interactive Design and Manufacturing, v. 18, n. 3, pp. 1–18, 2024. doi: https://doi.org/10.1007/s12008-023-01529-9.
  » https://doi.org/10.1007/s12008-023-01529-9
• ^[36] ARIVENDAN, A., CHEN, X., ZHANG, Y.F., et al, “Assessing the structural, mechanical, and thermal performance of Linum usitatissimum/carbon fiber composites by using novel WASPAS and CRITIC methods”, Industrial Crops and Products, v. 224, pp. 120378, 2025. doi: https://doi.org/10.1016/j.indcrop.2024.120378.
  » https://doi.org/10.1016/j.indcrop.2024.120378
• ^[37] VIJAYANANTH, V., PARAMASIVAM, B., RAJU, R., et al, “A CRITIC integrated WASPAS approach for selection of natural and synthetic fibers embedded hybrid polymer composite configuration”, Multiscale and Multidisciplinary Modeling, Experiments and Design, v. 7, n. 3, pp. 1721–1736, 2024. doi: https://doi.org/10.1007/s41939-023-00301-6.
  » https://doi.org/10.1007/s41939-023-00301-6
• ^[38] PATIL, S., RAMESH, B., SATHISH, T., et al, “Evaluation and optimization of mechanical properties of laterized concrete containing fly ash and steel fiber using Taguchi robust design method”, Alexandria Engineering Journal, v. 87, pp. 682–706, 2024. doi: https://doi.org/10.1016/j.aej.2023.12.009.
  » https://doi.org/10.1016/j.aej.2023.12.009
• ^[39] SINGH, T., GEHLEN, G.S., SINGH, V., et al, “Selection of automotive brake friction composites reinforced by agro-waste and natural fiber: an integrated multi-criteria decision-making approach”, Results in Engineering, v. 22, pp. 102030, 2024. doi: https://doi.org/10.1016/j.rineng.2024.102030.
  » https://doi.org/10.1016/j.rineng.2024.102030
• ^[40] ARUN, M., EFREMOV, C., NGUYEN, V.N., et al, “Fuzzy logic-supported building design for low-energy consumption in urban environments”, Case Studies in Thermal Engineering, v. 64, pp. 105384, 2024. doi: https://doi.org/10.1016/j.csite.2024.105384.
  » https://doi.org/10.1016/j.csite.2024.105384
• ^[41] CHAUDHARI, S., SUNHERIYA, N., GIRI, J., et al, “Qualitative enhancement in machining efficiency of SNCM8 alloy through hybrid ANN-Taguchi optimization approach”, Applied Engineering Letters, v. 10, n. 1, pp. 48–61, 2025. doi: https://doi.org/10.46793/aeletters.2025.10.1.5.
  » https://doi.org/10.46793/aeletters.2025.10.1.5
• ^[42] SUMESH, K.R., KANTHAVEL, K., “Grey relational optimization for factors influencing tensile, flexural, and impact properties of hybrid sisal banana fiber epoxy composites”, Journal of Industrial Textiles, v. 51, n. 3, pp. 4441–4459, 2022. doi: https://doi.org/10.1177/1528083720928501.
  » https://doi.org/10.1177/1528083720928501
• ^[43] RAJA, T., ALI, M.N., SHASHI KUMAR, M.E., et al, “Study on the characteristics of novel natural fiber extracted from Coccinia grandis plant vine: a novel material for engineering applications”, International Journal of Advanced Manufacturing Technology, 2024. In press. doi: https://doi.org/10.1007/s00170-024-13967-0.
  » https://doi.org/10.1007/s00170-024-13967-0
• ^[44] SENGUPTA, A., PATTNAIK, S., SUTAR, M.K., “Investigation of physico-mechanical properties of Mimosa pudica micro-fibrils as an effective reinforcement in polymer composite”, Proceedings of the Institution of Mechanical Engineers. Part C, Journal of Mechanical Engineering Science, v. 239, n. 6, pp. 2004–2013, 2025. doi: https://doi.org/10.1177/09544062241298542.
  » https://doi.org/10.1177/09544062241298542
• ^[45] NAVIN, M., RAMAKRISHNAN, T., BALAJI, D., et al, “CRITIC–EDAS approach for evaluating mechanical properties of flax/vetiver/MFF hybrid composites”, Polymers, v. 17, n. 13, pp. 1790, 2025. doi: https://doi.org/10.3390/polym17131790. PubMed PMID: 40647800.
  » https://doi.org/10.3390/polym17131790
• ^[46] RAJA, T., “Exploring a natural fiber extracted from henequen plant leaves: a sustainable innovation for reinforcement in composite materials”, Results in Engineering, v. 25, pp. 104015, 2025. doi: https://doi.org/10.1016/j.rineng.2025.104015.
  » https://doi.org/10.1016/j.rineng.2025.104015
• ^[47] RIZAL, S., IKRAMULLAH, GOPAKUMAR, D.A., et al, “Tailoring the effective properties of Typha fiber reinforced polymer composite via alkali treatment”, BioResources, v. 14, n. 3, pp. 5630–5645, 2019. doi: https://doi.org/10.15376/biores.14.3.5630-5645.
  » https://doi.org/10.15376/biores.14.3.5630-5645
• ^[48] SOMASUNDARAM, R., SENTHAMARAIKANNAN, P., SUYAMBULINGAM, I., et al, “Greening the composite industry: evaluating Lantana camara Verbenaceae fiber as a promising substitute for lightweight polymer matrix composites”, Industrial Crops and Products, v. 220, pp. 119265, 2024. doi: https://doi.org/10.1016/j.indcrop.2024.119265.
  » https://doi.org/10.1016/j.indcrop.2024.119265
• ^[49] IBRAHIM, B.M., FAKHRE, N.A., “Crown ether modification of starch for adsorption of heavy metals from synthetic wastewater”, International Journal of Biological Macromolecules, v. 123, pp. 70–80, 2019. doi: https://doi.org/10.1016/j.ijbiomac.2018.11.058. PubMed PMID: 30439424.
  » https://doi.org/10.1016/j.ijbiomac.2018.11.058
• ^[50] KAMALI MOGHADDAM, M., “Structural and thermal properties of cellulose microfiber isolated from Typha australis by sequential alkali-oxidative treatment”, Journal of Natural Fibers, v. 19, n. 15, pp. 10526–10538, 2022. doi: https://doi.org/10.1080/15440478.2021.1994094.
  » https://doi.org/10.1080/15440478.2021.1994094

Publication Dates

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

History

• Received
  24 Nov 2025
• Accepted
  26 May 2026