Multi-Objective Optimization of Mechanical Properties of Banana Pseudostem Fibers Using Sludge Retting Pretreatment

Abstract

In this study, sludge retting was used as a pretreatment method for extracting banana pseudostem fibers. A Box–Behnken response surface design was employed to optimize the retting conditions. Three variables—Bacillus subtilis concentration, treatment time, and pH—were selected for analysis. Their effects on the mechanical properties of the fibers were systematically evaluated. Experimental data were analyzed using ANOVA in Design-Expert 13, and a regression model was established for parameter optimization. The optimal conditions were determined to be a Bacillus subtilis concentration of 1.18%, a treatment time of 20 days, and a pH of 7. Under these conditions, the tensile strength, elastic modulus, and elongation at break of the fibers reached 1161.63 MPa, 50.68 GPa, and 2.32%, respectively—representing improvements of 46.23%, 42.48%, and 34.1% compared to untreated samples. In addition, the fibers were analyzed using SEM, TGA-DTG, FTIR, and XRD to investigate changes in surface topography, thermal behavior, chemical bonding, and crystalline structure. Results showed that sludge retting effectively removed non-cellulosic components, enhanced thermal stability and crystallinity, and significantly improved the mechanical performance of the fibers. This study demonstrates that sludge retting is a green and sustainable pretreatment technique with strong potential for banana pseudostem fiber processing.

1. Introduction

With increasing pressure on environmental resources, the utilization of agricultural waste has become a key focus of green development strategies. Banana is one of the most widely cultivated tropical crops globally, and its cultivation generates a large amount of pseudostem waste [1]. It is estimated that 20–30 tons of banana pseudostems are produced per hectare of plantation, most of which are discarded without treatment. This leads to both resource waste and environmental burden [2]. In fact, banana pseudostems are rich in cellulose, accounting for approximately 60–70% of their dry weight [3]. As a natural fiber resource, they show great potential for applications in textiles, papermaking, and bio-composites [4]. Therefore, the efficient extraction and processing of banana pseudostem fibers is of both economic and environmental significance [5].

Currently, there are multiple methods for extracting and modifying banana pseudostem fibers, including mechanical, chemical, enzymatic, and microbial approaches. Mechanical extraction, though simple and low-cost, often leaves residual non-cellulosic substances that compromise fiber quality. Chemical pretreatments such as alkali or acid delignification effectively remove lignin and hemicellulose but are energy-intensive and generate hazardous waste liquids [6]. Enzymatic delignification offers high selectivity and minimizes cellulose damage, yet its industrial application is constrained by high costs and enzyme instability [7]. Microbial delignification using bacterial or fungal strains is regarded as a more environmentally friendly approach, yet this process typically requires extended delignification cycles and remains susceptible to environmental fluctuations [8,9]. Although mixed microbial communities and enzyme-assisted microbial systems have been proposed in recent years to enhance efficiency, challenges persist regarding cost, microbial stability, and large-scale implementation.

In the context of fiber pretreatment, chemical and microbial methods have been extensively investigated but both exhibit inherent limitations. Chemical pretreatments, such as alkali boiling or acid hydrolysis, are effective in removing lignin and hemicellulose; however, they involve high energy consumption, generate large volumes of effluents, and may cause cellulose degradation, thereby restricting their environmental and industrial feasibility. Microbial pretreatments, including bacterial and fungal retting, offer greater selectivity and operate under mild conditions, making them environmentally friendly. Nevertheless, these processes typically require long retting periods, precise inoculation, or costly enzyme supplementation, and the performance of pure strains often lacks stability at industrial scale. Against this backdrop, sludge retting has gained attention as an innovative “waste-to-treat-waste” bioprocess. Activated sludge inherently contains a broad range of microbial populations with cellulolytic and ligninolytic functions, enabling the synergistic degradation of hemicellulose and lignin under favorable conditions [10]. Compared with chemical–biological hybrid approaches, sludge retting operates without additional chemical inputs, thereby lowering energy demand and effluent generation while maintaining cellulose integrity and improving fiber mechanical properties [11]. Previous studies on microbial retting of bast fibers such as flax and ramie [12,13], as well as banana fibers inoculated with Bacillus strains [14], mainly relied on single microbial cultures or enzyme supplementation. These techniques often require precise inoculation, strict culture maintenance, or expensive enzyme preparations, which hinder their scalability. To date, no comprehensive study has been conducted to optimize sludge retting parameters for banana pseudostem fibers.

Therefore, this study adopts sludge retting as a pretreatment method for banana pseudostem fibers and systematically investigates its effects on fiber physical properties. A Box–Behnken response surface design was used to optimize key processing parameters, including bacterial concentration, treatment time, and pH, aiming to enhance tensile strength, elastic modulus, and elongation at break [15]. In addition, scanning electron microscopy (SEM), thermogravimetric analysis (TGA), Fourier-transform infrared spectroscopy (FTIR), and X-ray diffraction (XRD) were employed to analyze changes in fiber microstructure and composition. This research provides a green processing approach for banana pseudostem fibers and offers technical support and theoretical guidance for the resource utilization and sustainable development of agricultural waste [16].

2. Materials and Methods

2.1. Materials

Banana stems used in this study were collected from plantations in Danzhou City, Hainan Province. The widely cultivated variety “Brazil Williams” was selected. Stem sections of 50–100 cm from the base of the plants were harvested. After removing leaf sheaths, petioles, and surface impurities, the samples were immediately transported to the laboratory for further processing. The sludge used for retting was obtained from Dongpo Lake at Hainan University. The microbial agent was a commercial Bacillus subtilis preparation with a declared concentration of ≥1 × 10⁹ CFU/g. Before use, it was diluted according to the experimental design and evenly mixed into the sludge system. Relevant physicochemical parameters, such as pH and concentration, were preliminarily optimized through preliminary experiments. A phosphate-buffer solution (PBS) was used to adjust and maintain the pH during the retting process. Figure 1 shows the banana stem slices before treatment, the sludge used, and the degummed banana fiber samples.

2.2. Experimental Design

2.2.1. Selection of Experimental Factors

In this study, to optimize the effect of sludge retting on the mechanical properties of banana pseudostem fiber, three main experimental factors were selected: Bacillus subtilis concentration, treatment time, and pH value, as detailed below.

Bacillus subtilis is a widely used functional microorganism for biological pretreatment of natural fibers. It can secrete cellulase, pectinase, and lignin-degrading enzymes, effectively removing non-cellulosic substances on the fiber surface, such as pectin, lignin, and hemicellulose. This improves fiber separation and mechanical properties. Changes in bacterial concentration directly affect microbial degradation activity and degumming efficiency. An appropriate increase in concentration enhances biodegradation, but excessive concentration may cause resource competition (e.g., oxygen and nutrients) or metabolic inhibition, which can reduce degumming effectiveness [17].

Treatment time is a key factor influencing sludge retting performance, determining the extent of microbial degradation of non-cellulosic components in banana pseudostem fibers. Insufficient time may lead to incomplete treatment and limit fiber performance improvement, while excessive time can cause cellulose degradation, resulting in strength loss. Therefore, proper treatment duration helps balance impurity removal and fiber integrity preservation [9].

pH significantly affects the growth and enzyme activity of Bacillus subtilis. The optimal pH usually ranges from 6.5 to 7.5, within which the microorganism maintains strong degradation ability. Deviations from this range, either acidic or alkaline, inhibit microbial metabolism and reduce degumming efficiency. Additionally, pH influences the solubility of hemicellulose and pectin, indirectly affecting fiber microstructure and mechanical properties [18].

In summary, Bacillus subtilis concentration, treatment time, and pH were selected as optimization factors due to their clear biochemical mechanisms and strong support in the literature. These factors have been repeatedly validated in microbial pretreatment studies of natural fibers.

2.2.2. Response Surface Experimental Design

The retting trials were performed in covered plastic containers at room temperature, without external aeration. Moisture was maintained by adding distilled water when necessary to prevent drying. Under these conditions, the system remained largely microaerophilic to anaerobic, which reflects the natural retting environment and favors the activity of microbial consortia in the sludge. Response Surface Methodology (RSM) is a widely applied statistical approach for modeling and optimizing experimental responses. In this study, Bacillus subtilis concentration, treatment time, and sludge pH were selected as variables. A three-factor, three-level Box–Behnken design was used to plan the experiments. The factors and their levels are listed in

Table 1. Factors and levels of experiments.

Levels	Factors
	Bacillus subtilis Concentration $\mathit{A}$/%	Time $\mathit{B}$/day	pH $\mathit{C}$
−1	0.5	7	6.5
0	1	14	7
1	1.5	21	7.5

. Experimental data were analyzed and optimized using Design-Expert 13 software. This method allows response surface analysis of the effects of sludge treatment on the tensile strength, elastic modulus, and elongation of banana pseudostem fibers. Combined with regression equations, the optimal sludge treatment conditions were determined.

2.2.3. Sludge Retting Pretreatment of Banana Stems

To ensure the repeatability and accuracy of the retting experiments, the treatment procedure was strictly standardized. Fresh banana pseudostems were longitudinally sliced into uniform sheets measuring 5 cm × 10 cm × 1 cm. The outer bark was removed as much as possible, while retaining the bast and part of the fibrous tissues. A total of 17 treatment groups were designed (see

Table 2. Sludge Pretreatment Conditions and Experimental Results for Banana pseudostem fibers.

Run	$\mathit{A}$	$\mathit{B}$	$\mathit{C}$	Tensile Strength ${\mathbf{Y}}_1$ (MPa)	Elastic Modulus ${\mathbf{Y}}_2$ (GPa)	Elongation ${\mathbf{Y}}_3(\%)$
1	1	7	6.5	635.65	44.02	1.81
2	1	14	7	1002.94	49.97	2.3
3	0.5	7	7	644.87	45.21	1.93
4	1	21	7.5	823.72	48.45	2.01
5	0.5	14	7.5	510.44	45.57	1.94
6	0.5	21	7	902.63	49.02	2.16
7	1.5	21	7	1071.95	50.64	2.3
8	1.5	14	7.5	507.91	47.65	2.13
9	1	7	7.5	640.93	44.5	1.86
10	1	14	7	1041.63	49.79	2.31
11	0.5	14	6.5	429.3	46.65	1.94
12	1	14	7	1045.83	50.15	2.3
13	1	14	7	1042.56	50.2	2.33
14	1.5	7	7	723.72	45.11	2.05
15	1	21	6.5	965.93	47.87	2.15
16	1.5	14	6.5	607.91	45.7	2.08
17	1	14	7	1040.78	50.34	2.29

), each with different combinations of three variables: Bacillus subtilis concentration (0.5–1.5%), treatment duration (7–21 days), and pH level (6.5–7.5). Retting was carried out in sealed plastic containers (5 L capacity). The banana stem slices were completely buried in moist sludge inoculated with different concentrations of Bacillus subtilis, ensuring no overlapping of samples. The sludge layer was maintained at a thickness of at least 5 cm. The initial pH was adjusted using a buffer solution and was monitored daily during the retting process, with minor adjustments made to maintain stability. After retting, the samples were removed and thoroughly rinsed with water until no visible sludge residue remained. The samples were then dried in a forced-air oven at 60 °C until constant weight (mass change < 1%) and stored in a desiccator for subsequent analysis. To minimize experimental variability, five parallel samples were prepared under each condition for mechanical testing and material characterization. All experiments were performed using the same batch of sludge and microbial inoculum to ensure consistency and comparability. Temporal changes in banana pseudostems during retting are shown in Figure 2. It should be noted that the sludge used in this study was not sterilized, and no microbial inventory was performed. Thus, the retting system contained the naturally occurring microbial consortia present in the activated sludge.

2.3. Testing of Banana Pseudostem Fibers

2.3.1. Tensile Test

Tensile tests were performed using a universal testing machine with a load range of 100 N to 5 kN. UTM2000 series electronic universal testing machine (Sansi Zongheng, Shenzhen, China). The tensile speed was set to 1 mm/min, the preload was 0.01 N, and the gauge length between the upper and lower clamps was 40 mm. To evaluate the mechanical performance under various pretreatment conditions, seventeen distinct groups of banana pseudostem fibers underwent tensile testing. Each fiber specimen was securely mounted between serrated hydraulic clamps within a universal testing machine. A constant crosshead displacement rate was applied until complete fracture occurred. The recorded parameters included tensile strength, elastic modulus, and elongation at break, enabling a comparative analysis of the treatment effects. The fibers were centered and clamped tightly to prevent slippage. Each group was tested five times to obtain an average value.

2.3.2. Scanning Electron Microscopy

The surface and cross-sectional microstructures of both untreated and sludge-retted banana pseudostem fibers were characterized using a scanning electron microscope. Verios G4 UC (Thermo Fisher Scientific, Waltham, MA, USA). Prior to imaging, the fibers were sputter-coated with a thin Au-Pd layer to enhance surface conductivity. Microstructural evaluation was performed at an accelerating voltage of 5 kV to facilitate detailed observation of morphological modifications induced by the retting process.

2.3.3. Thermogravimetric Analysis

Untreated and treated banana pseudostem fiber samples underwent thermogravimetric analysis using a TG 209 F3 instrument (Netzsch, Selb, Germany). Prior to testing, the fibers were cut into 1-mm segments and placed within crucibles. Mass loss and its derivative were recorded across a temperature range of 30 °C to 500 °C. Testing was conducted under a nitrogen atmosphere with a heating rate of 10 °C/min.

2.3.4. Fourier Transform Infrared Spectroscopy

Fourier-transform infrared (FTIR) spectroscopy was conducted to characterize chemical structural changes in untreated and sludge-retted banana pseudostem fibers, using a Bruker T27 spectrometer (Bruker, Billerica, MA, USA). Finely ground fiber powder was homogenized with potassium bromide (KBr) at an appropriate ratio and compressed into transparent pellets. Spectral data were acquired in the mid-infrared region from 4000 to 400 cm⁻¹, employing a spectral resolution of 4 cm⁻¹ and accumulating a minimum of 32 scans per sample to enhance the signal-to-noise ratio. Prior to each measurement, a background spectrum of pure KBr was collected to compensate for atmospheric interference.

2.3.5. X-Ray Diffraction

The crystalline structures of untreated and sludge-retted banana pseudostem fibers were examined by X-ray diffraction (XRD) using a Rigaku Smart Lab diffractometer (Rigaku, Tokyo, Japan). Prior to measurement, the fibers were mechanically pulverized into fine powders to ensure sample homogeneity. XRD patterns were acquired within a 2θ range of 5° to 60° at a scanning speed of 5°·min⁻¹. The crystallinity index (CrI) of each sample was quantitatively evaluated from the diffraction profiles by applying the empirical Segal method [19].

$$Cr\mathrm{{\rm I}}={\displaystyle \frac{{\mathrm{{\rm I}}}_{002}-{\mathrm{{\rm I}}}_{am}}{{\mathrm{{\rm I}}}_{002}}}\times 100\%$$

(1)

The crystallinity index (CrI) was calculated according to the empirical Segal method, where CrI represents the relative degree of crystallinity, I₀₀₂ corresponds to the peak intensity of the (002) crystalline plane diffraction observed at approximately 22° 2θ, and I_am denotes the background intensity of the amorphous region measured at around 18° 2θ.

3. Results and Discussion

3.1. Tensile Test Results and Analysis

3.1.1. Model Fitting and Analysis of Variance

Based on the Box–Behnken design principle, 17 experimental runs were conducted. The test schemes and results for each run are shown in Table 2, including tensile strength, elastic modulus, and elongation used in the models. The experimental data were imported into Design-Expert 13 software for analysis. Regression models were established for tensile strength (Y₁), elastic modulus (Y₂), and elongation (Y₃) as functions of the independent variables: Bacillus subtilis concentration(A), treatment time (B), and pH value (C). Analysis of variance (ANOVA) was performed to further investigate the effects of sludge treatment and to identify which factors significantly influence the mechanical properties of banana pseudostem fibers [20,21]. The ANOVA results are presented in

Table 3. ANOVA for the Tensile Strength of Banana pseudostem fibers.

Source	Sum of Squares	df	Mean Square	F-Value	p-Value	Significance
Model	810,200	9	90,026.16	132.65	<0.0001	**
$A$	22,498.51	1	22,498.51	33.15	0.0007	**
$B$	156,500	1	156,500	230.65	<0.0001	**
$AC$	8202.92	1	8202.92	12.09	0.0103	*
BC	5438.33	1	5438.33	8.01	0.0254	*
$A^2$	214,700	1	214,700	316.35	<0.0001	**
$B^2$	3036.82	1	3036.82	4.47	0.0722	*
$C^2$	366,500	1	366,500	540.07	<0.0001	**
Residual	4750.79	7	678.68
Lack of Fit	3471.45	3	1157.15	3.62	0.1231	not significant
Pure Error	1279.33	4	319.83
Cor Total	815,000	16			<0.0001
R2 = 0.9942				Adjusted R2 = 0.9867
Predicted R2 = 0.9294				Adequate Precision = 30.8437

Note: p-value < 0.01 (highly significant, **); p-value < 0.05 (significant, *).

,

Table 4. ANOVA for the Elastic Modulus of Banana pseudostem fibers.

Source	Sum of Squares	df	Mean Square	F-Value	p-Value	Significance
Model	85.55	9	0.0541	94.77	<0.0001	**
$A$	0.8778	1	0.0435	76.24	<0.0001	**
$B$	36.72	1	0.1176	206.08	<0.0001	**
AC	2.30	1	2.30	31.97	0.0008	**
$A^2$	6.13	1	0.0181	31.65	0.0008	**
$B^2$	8.12	1	0.0717	125.64	<0.0001	**
$C^2$	26.13	1	0.2001	350.62	<0.0001	**
Residual	0.5026	7	0.0006
Lack of Fit	0.32	3	0.001	4.46	0.0914	not significant
Pure Error	0.1826	4	0.0002
Cor Total	86.05	16
R2 = 0.9919				Adjusted R2 = 0.9814
Predicted R2 = 0.8968				Adequate Precision = 28.1757

Note: p-value < 0.01 (highly significant, **).

and

Table 5. ANOVA for the elongation of Banana pseudostem fibers.

Source	Sum of Squares	df	Mean Square	F-Value	p-Value	Significance
Model	0.4868	9	9.51	132.4	<0.0001	**
$A$	0.0435	1	0.8778	12.23	0.01	**
$B$	0.1176	1	36.72	511.48	<0.0001	**
$BC$	0.0090	1	0.0090	15.81	0.0053	**
$A^2$	0.0181	1	6.13	85.33	<0.0001	**
$B^2$	0.0717	1	8.12	113.1	<0.0001	**
$C^2$	0.2001	1	26.13	363.97	<0.0001	**
Residual	0.004	7	0.0718
Lack of Fit	0.0031	3	0.1067	2.34	0.2151	not significant
Pure Error	0.0009	4	0.0457
Cor Total	0.4908	16
R2 = 0.9942				Adjusted R2 = 0.9867
Predicted R2 = 0.9372				Adequate Precision = 31.8602

Note: p-value < 0.01 (highly significant, **).

.

The influence of each variable in the regression model was assessed through its p-value, where p < 0.05 denotes statistical significance. The ANOVA results for tensile strength are summarized in Table 3. As indicated, factors A, B, AC, BC, A², and C² significantly affected tensile strength (Y₁), as described in the regression Equation (2). The overall model for Y₁ was highly significant (p < 0.01), with a nonsignificant lack-of-fit (p > 0.05), confirming strong agreement between the experimental results and the regression prediction. The model exhibited an R² value of 0.9942, demonstrating excellent data fitting. Furthermore, the predicted R² (0.9294) was in close agreement with the adjusted R² (0.9867), with a difference below 0.2, suggesting good model reliability. The adequate precision reached 30.8437, far exceeding the threshold value of 4, which indicates a strong signal-to-noise ratio and validates the model’s suitability for exploring the design space [22]. Importantly, the effect of Bacillus subtilis concentration on fiber properties exhibited a clear nonlinear pattern. Both tensile strength and elastic modulus improved at moderate bacterial levels, but higher concentrations resulted in performance reduction. The ANOVA results confirmed a significant negative quadratic effect (A²), indicating that excessive bacterial concentration can be detrimental. This behavior may be attributed to competition for nutrients and oxygen among microbes, or to antagonistic interactions between the inoculated Bacillus and native microorganisms present in the sludge environment.

$$Y_1=1034.75+53.03A+139.88B-45.28AC-36.87BC-225.81A^2-295.05C^2$$

(2)

The ANOVA results for elastic modulus are summarized in Table 4. According to the p-values, variables A, B, C, AB, AC, A², B², and C² significantly influenced the elastic modulus (Y₂). Insignificant terms were excluded, and the final regression relationship is presented in Equation (3). The model for Y₂ was highly significant, with a nonsignificant lack-of-fit, indicating a good fit to the experimental observations. The coefficient of determination was R² = 0.9942, with predicted R² = 0.9372 and adjusted R² = 0.9867, all of which are in close agreement, confirming the robustness of the model. Furthermore, the adequate precision reached 31.8602, far above the threshold value of 4, demonstrating that the regression equation provides a strong signal-to-noise ratio and can be reliably used for design space exploration [22]. The ANOVA results highlight bacterial concentration and treatment time as significant factors for elastic modulus. This trend reflects the microbial mechanism of sludge retting, where sufficient retting duration allows microbial consortia to degrade binding components such as pectin and hemicellulose, thereby enhancing fiber stiffness. Moderate bacterial density accelerates this process while maintaining cellulose integrity, leading to higher modulus. However, the significant quadratic effects (A² and B²) suggest a nonlinear response: excessive bacterial concentration or overly prolonged retting may promote over-degradation of structural components or intensify microbial competition, ultimately reducing fiber stiffness rather than further improving it.

$$Y_2=50.09+0.3313A+2.14B+0.2412C+0.4300AB+0.7575AC-1.21A^2-1.39B^2-2.49$$

(3)

Table 5 summarizes the ANOVA outcomes for elongation. The regression model for elongation was highly significant (p < 0.0001), whereas the lack-of-fit term was not significant, confirming the model’s adequacy. The corresponding regression equation is shown in Equation (4). Variables A, B, BC, A², B², and C² significantly influenced elongation (Y₃). The coefficient of determination (R² = 0.9919), predicted R² (0.8968), and adjusted R² (0.9814) were in good agreement, indicating high reliability of the model. Moreover, the adequate precision ratio was 28.1757, well above the minimum requirement of 4, which ensures that the model possesses a sufficient signal-to-noise ratio for effective exploration of the design space [22]. The significance of treatment time and pH for elongation at break may be attributed to the balance between fiber flexibility and degradation of non-cellulosic components. Prolonged retting enables removal of amorphous substances that restrict fiber slippage, increasing elongation. However, if the retting conditions are not well controlled, excessive degradation may weaken fibers. Similarly, appropriate pH ensures optimal microbial enzyme activity, preventing over-degradation and supporting improved extensibility.

$$Y_3=2.31+0.0738A+0.1213B-0.0475BC-0.0655A^2-0.1305B^2-0.2180C^2$$

(4)

To further understand the test results, the perturbation plots in Figure 3 illustrate the effects of processing conditions on tensile strength, elastic modulus, and elongation. These trends are also supported by the F-values in Table 3, Table 4 and Table 5. The relative influence of each factor on the mechanical properties, ranked from most to least significant, was: time, Bacillus subtilis concentration, and pH. Specifically, time had the most pronounced effect on tensile strength, while pH had the least. Similarly, time and Bacillus subtilis concentration exhibited greater effects on the elastic modulus compared to pH. For elongation, time again showed a stronger influence than the other two variables. Similar findings have been reported in previous studies, indicating that treatment duration plays a more critical role in improving the mechanical properties of natural fibers. This is likely due to the extended interaction time between the microbes and the banana pseudostem fibers [23]. Since no microbial inventory was conducted, the observed effects may reflect not only the activity of the inoculated Bacillus subtilis but also the influence of native microbial populations present in the sludge. These interactions, including possible competition or synergistic degradation, could partly explain the nonlinear trends observed in tensile strength and elastic modulus.

3.1.2. Effects of Interaction Factors on Tensile Strength

Enhancing the tensile strength of banana pseudostem fibers is critical for improving their performance in fiber-reinforced composites. The 3D response surface and contour plots in Figure 4 illustrate the interaction between Bacillus subtilis concentration and pH on fiber tensile strength. When the treatment time is fixed at level B = 14 days, the tensile strength varies significantly with the interaction of Bacillus subtilis concentration and pH.

At a constant Bacillus subtilis concentration, an increase in pH initially promotes tensile strength, reaching a maximum near pH 7. However, further increases in pH lead to a decline in strength, as shown clearly in the contour plot. This trend is attributed to the fact that both mildly acidic and mildly alkaline environments can inhibit microbial activity in the sludge, thereby reducing the efficiency of degumming and component breakdown in the fiber [24].

On the other hand, as the Bacillus subtilis concentration increases while pH remains constant, tensile strength also improves to a certain extent. However, excessive concentration of Bacillus subtilis has adverse effects: overgrowth of microbes can lead to resource competition (e.g., for oxygen and nutrients) and metabolic inhibition (e.g., accumulation of organic acids), ultimately decreasing decomposition efficiency. Initially, surface impurities attached to the fibers are gradually removed due to microbial synergy. Yet, over-degradation may cause cellulose chain scission, increased surface defects, and significantly reduced strength, thereby leading to a decrease in tensile strength [25].

3.1.3. Effects of Interaction Factors on Elastic Modulus

The elastic modulus is one of the most important properties of banana pseudostem fibers. The interaction between Bacillus subtilis concentration and pH significantly affects the elastic modulus. The 3D response surface and contour plots are shown in Figure 5. When the treatment time is fixed at level B = 14 days, analysis of the 3D plot in Figure 3a shows that at a constant Bacillus subtilis concentration, the elastic modulus of the banana fibers first increases and then decreases as the treatment pH rises. Similarly, at a fixed pH, the elastic modulus follows the same trend with increasing Bacillus subtilis concentration. Furthermore, the contour plot in Figure 3b indicates that the elastic modulus changes more rapidly along the Bacillus subtilis concentration axis than along the pH axis. It is noteworthy that excessively high Bacillus subtilis concentration or pH can damage the banana fibers. Once the elastic modulus reaches its maximum, any further increase in either factor results in a decline in the elastic modulus [26,27].

3.1.4. Effects of Interaction Factors on Elongation

The elongation of banana pseudostem fibers is moderate compared to most other natural fibers, and it is significantly affected by the treatment process. Appropriate sludge treatment can further increase the elongation, making the fibers more flexible. Consequently, the bending performance of composites made from these fibers is improved. Elongation also plays a vital role in flexible materials.

The interaction effects of Bacillus subtilis concentration (A), treatment time (B), and pH (C) on the elongation of banana fibers are shown in Figure 6. When one factor is fixed at level zero, the effects of the other two factors on elongation are analyzed. Combined with the F-values of interaction terms in Table 5, the interaction between treatment time (B) and pH (C) has the most significant impact on elongation, followed by the interaction between Bacillus subtilis concentration (A) and pH (C).

The 3D response surface plot and contour plot in Figure 6a,b illustrate the effect of the interaction between time and pH on elongation. When the Bacillus subtilis concentration is fixed at level zero (A = 1%), elongation increases with increasing treatment time. Additionally, at a constant treatment time, elongation first slightly increases and then decreases with rising pH [28].

3.1.5. Parameter Optimization of Treatment Conditions

Figure 7a–c illustrate the correlation between the experimental and predicted values for tensile strength, elastic modulus, and elongation. The data points align closely with the diagonal line, confirming the strong agreement between measured and predicted results and demonstrating the robustness of the models. These findings indicate that the developed models are reliable. To further enhance the performance of banana pseudostem fibers, it is necessary to determine suitable sludge retting conditions. Therefore, tensile strength, elastic modulus, and elongation were selected as the optimization targets. In this work, a multi-objective optimization approach was applied to identify the optimal combination of Bacillus subtilis concentration, treatment time, and pH. The optimization process was guided by regression equations, and the resulting mathematical model is presented in Equation (5).

$$\left\{\begin{array}{c}max{Y}_1\left(A,B,C\right)\\ max{Y}_2\left(A,B,C\right)\\ max{Y}_3\left(A,B,C\right)\end{array}\right.$$

(5)

The model was optimized and analyzed using Design-Expert 13, with sufficient parameter ranges set to ensure the optimum falls within the specified limits. As shown in Figure 8, the optimal sludge treatment conditions are 1.21% Bacillus subtilis concentration, 20 days of treatment time, and a pH of 7. Under these conditions, the tensile strength reaches 1166.52 MPa, the elastic modulus is 50.95 GPa, and the elongation is 2.34%.

3.1.6. Validation Experiments

To confirm the predictive capability of the established models, validation experiments were performed under the optimized parameters. The mean tensile strength, elastic modulus, and elongation obtained from five replicates were 1141.63 MPa, 50.68 GPa, and 2.32%, respectively. A comparison of experimental and predicted values is presented in

Table 6. Validation testing results.

Contrast Items	Tensile Strength (MPa)	Elastic Modulus (GPa)	Elongation (%)
Measured value	1161.63	50.68	2.32
Predicted value	1166.52	50.95	2.34
Error (%)	0.42	0.53	0.85

. The small deviations of 0.42%, 0.53%, and 0.85% for tensile strength, modulus, and elongation demonstrate the reliability of the optimization model.

Figure 9 compares the mechanical properties of untreated banana pseudostem fibers (UBF) and those treated under optimal conditions (OBF). The mechanical properties of untreated fibers are averaged from five samples. As shown in the figure, after treatment under optimal conditions, the tensile strength of banana pseudostem fibers increased from 794.41 MPa to 1161.63 MPa, an improvement of 46.23%. The elastic modulus increased from 35.57 GPa to 50.68 GPa, a rise of 42.48%. The elongation increased from 1.73% to 2.32%, a 34.1% enhancement. These changes are consistent with the experimental principles. Compared with other bast fibers, the sludge-retted banana pseudostem fibers exhibited competitive mechanical properties. For instance, flax fibers typically show tensile strengths of 800–1500 MPa and elastic moduli of 60–80 GPa after enzymatic or microbial retting; hemp fibers range from 550–900 MPa in tensile strength with moduli of 30–70 GPa; and jute fibers are relatively lower, with tensile strengths of 400–800 MPa and moduli of 20–30 GPa. In this study, the mechanical properties of sludge-retted banana fibers were lower than those of flax and hemp but were comparable to, or even higher than, those of jute. This indicates that sludge retting markedly enhances the competitiveness of banana pseudostem fibers, making them a promising sustainable alternative for textile and composite applications. In this study, by optimizing the sludge treatment conditions for tensile strength, elastic modulus, and elongation simultaneously, a more balanced improvement in the mechanical performance of banana pseudostem fibers was achieved.

3.2. SEM Observation and Analysis

The surface microstructure of banana pseudostem fibers before and after sludge retting treatment was characterized using scanning electron microscopy (SEM), as shown in Figure 10. The surface of untreated banana fibers (UBF) was covered with abundant non-cellulosic components, including pectin, hemicellulose, and lignin, forming a distinct gel-like coating. The low-magnification image in Figure 10a shows large and film-like impurities adhered to the fiber surface, with unclear longitudinal grooves, resulting in a rough and uneven surface with a disordered structure. The high-magnification image in Figure 10b further reveals that these non-cellulosic substances accumulate densely in layered forms, hindering fiber bundle separation and reducing the interfacial bonding between fibers and the matrix in composites [29].

In contrast, fibers treated with sludge retting (OBF) exhibited significant improvements in surface structure. As shown in Figure 10c, the fiber surface became smoother and cleaner, with clearly visible longitudinal alignment and pronounced groove structures, indicating effective removal of external impurities such as pectin and lignin. The high-magnification image in Figure 10d shows almost no large impurities remaining on the fiber surface, with only a few small particles distributed, suggesting thorough degradation of non-cellulosic components.

This phenomenon can be attributed to the microbial community in active sludge during the retting process. Microorganisms such as Bacillus subtilis and Pseudomonas species synergistically secrete various degrading enzymes, including cellulase, pectinase, and ligninase. These enzymes selectively degrade non-cellulosic substances surrounding the fiber bundles, enabling effective degumming while preserving the integrity of the cellulose backbone.

Overall, the SEM images clearly demonstrate the effectiveness of sludge retting pretreatment in removing surface impurities and restoring the native microstructure of banana pseudostem fibers. Before treatment, the fiber surfaces were heavily contaminated with disordered structures. After treatment, surface cleanliness was significantly improved, with well-aligned and intact surface features. These findings are consistent with previous studies on the bio-degumming of natural fibers [30,31,32,33], further confirming that this method not only improves fiber morphology but also enhances stress transfer and uniform stress distribution, contributing to higher tensile strength and elastic modulus.

3.3. Thermogravimetric Analysis Results

To evaluate the effect of sludge retting pretreatment on the thermal stability of banana pseudostem fibers, thermogravimetric analysis (TGA) and derivative thermogravimetry (DTG) were performed on untreated fibers (UBF) and treated fibers (OBF). The thermal decomposition curves are shown in Figure 11. Significant differences in mass loss patterns and thermal decomposition peaks were observed between the two groups, indicating that the pretreatment altered the thermal degradation behavior and thermal stability of the fibers.

As shown in Figure 11a, the thermal decomposition process of banana pseudostem fibers can be divided into three main stages. The first stage (30–120 °C) corresponds to the evaporation of physically adsorbed water, with mass losses of 9.82% and 9.74% for UBF and OBF, respectively, indicating that the pretreatment had minimal impact on fiber moisture content and hydrophilicity. The second stage (approximately 180–270 °C) is mainly associated with the degradation of pectin and thermally unstable hemicellulose. UBF showed a mass loss of 15.56% in this stage, while OBF exhibited no significant mass loss, suggesting effective removal of low-thermal-stability non-cellulosic components by the sludge retting treatment. The third stage (approximately 270–390 °C) corresponds to the main decomposition of cellulose and residual hemicellulose, with mass losses of 46.44% for UBF and 60.11% for OBF. The higher decomposition rate in OBF may be attributed to the increased relative content of cellulose due to the reduction in non-cellulosic substances [34]. In addition, lignin, with its complex structure and high thermal resistance, decomposed slowly during the process [35]. At 500 °C, the residual mass (ash content) of OBF increased slightly from 28.97% to 31.09%, possibly due to the accumulation of inorganic components or structural reorganization during microbial treatment.

The DTG curves in Figure 11b further reveal the differences in decomposition rates. UBF exhibited three distinct peaks at 59.23 °C, 292.67 °C, and 324.96 °C, corresponding to the three TGA stages. In contrast, OBF showed only two peaks, at 54.82 °C and 332.96 °C, lacking the second-stage degradation peak. More importantly, the main decomposition peak of OBF shifted from 324.96 °C to 332.96 °C, with a slightly increased rate and a sharper, narrower profile. This suggests a more uniform component distribution and a more concentrated degradation behavior after pretreatment, which is beneficial for enhancing thermal processability.

In summary, the TGA and DTG results confirm that sludge retting pretreatment effectively removed thermally unstable non-cellulosic components, increased the onset temperature of thermal decomposition, and improved decomposition uniformity, thereby enhancing the overall thermal stability of the fibers. This improvement contributes to better load-bearing capacity and toughness, and provides a thermal performance basis for applications involving thermoplastic processing and hot-press molding. The observed thermal behavior is consistent with that of natural fibers subjected to biological pretreatments.

3.4. Fourier Transform Infrared (FTIR) Spectroscopy Analysis

To further investigate the effect of sludge retting pretreatment on the chemical composition of banana pseudostem fibers, FTIR analysis was performed on untreated (UBF) and optimally treated (OBF) fibers. The absorption spectra are shown in Figure 12, covering the wavenumber range from 400 to 4000 cm⁻¹. The results reveal that sludge retting significantly altered the fiber’s chemical structure, as evidenced by notable changes in several absorption peaks. The peak positions and corresponding functional groups are listed in

Table 7. Peak value of change in FTIR spectrum.

Wavenumbers (cm−1)		Functional Groups	Source
UBF	OBF
2921	2921	C-H	Cellulose
1734	1734	C=O	Hemicellulose
1508	1508	C=C	Lignin characteristic peak
1373	1373	C-H	Cellulose and hemicellulose
1250	1250	C-O-C	Hemicellulose
898	898	C-H	Cellulose

.

The intensity of the absorption peak at 898 cm⁻¹ increased after treatment, attributed to C–H rocking vibration in cellulose. This indicates a higher relative cellulose content in OBF, confirming the preservation of cellulose during pretreatment [36]. The band at 1250 cm⁻¹ is attributed to C–O stretching in acetyl groups of hemicellulose, and its weakening suggests deacetylation and the partial removal of hemicellulose- and lignin-associated structures in OBF compared to UBF [37]. The absorption near 1373 cm⁻¹ is assigned to C–H bending vibrations in cellulose and hemicellulose; its slight decrease reflects partial degradation of hemicellulose while cellulose remained largely preserved [38]. The peak at 1508 cm⁻¹, characteristic of aromatic C=C stretching in lignin, was also notably reduced, further confirming lignin removal. Additionally, the slight decrease in peak intensity at 1734 cm⁻¹, assigned to C=O stretching in hemicellulose, indicates partial degradation of hemicellulose during treatment. The presence of the peak at 2921 cm⁻¹ in both UBF and OBF, attributed to CH stretching in cellulose, confirms that the cellulose backbone remained intact [39,40].

Overall, the FTIR results demonstrate that sludge retting pretreatment effectively removed non-cellulosic substances such as lignin, hemicellulose, and pectin, without damaging the cellulose backbone. The decreased intensities at 1734, 1508, and 1250–1373 cm⁻¹ reflect the removal of non-cellulosic structures, while the enhanced peak at 898 cm⁻¹ indicates relative cellulose enrichment. The selective preservation of cellulose improves load transfer efficiency and enhances both strength and flexibility. These findings are consistent with SEM and TGA analyses.

3.5. X-Ray Diffraction Analysis

To investigate the structural changes in banana pseudostem fibers after treatment, X-ray diffraction (XRD) analysis was performed on both untreated (UBF) and optimally treated (OBF) samples, as shown in Figure 13. The XRD patterns of both samples exhibited similar profiles, with prominent diffraction peaks appearing near 16.5° and 22.5°, corresponding to the (110) and (200) crystallographic planes of cellulose I, respectively. This indicates that the crystalline structure and cellulose polymorph remained unchanged after the sludge retting treatment.

However, the intensity of the diffraction peaks, particularly the main peak at approximately 22.5° (200 plane), was significantly higher for the OBF sample, suggesting an increase in crystallinity. This enhancement is primarily attributed to the removal of non-cellulosic components such as hemicellulose, lignin, and pectin, thereby exposing a more complete and orderly cellulose crystalline region.

To quantify this improvement, the crystallinity index (CrI) was calculated using the Segal empirical method. The CrI increased from 43.38% in UBF to 58.20% in OBF, indicating a substantial enhancement in crystallinity [41]. This result supports the hypothesis that during sludge retting, microbial enzymes selectively degrade non-crystalline regions, such as those containing pectin and lignin, while causing minimal damage to the crystalline cellulose domains. The removal of amorphous materials facilitates the rearrangement and ordering of cellulose chains, thus increasing crystallinity.

In summary, XRD analysis confirms that sludge retting pretreatment significantly enhances the crystallinity of banana pseudostem fibers, as evidenced by intensified (110) and (200) diffraction peaks and a higher crystallinity index. This increased molecular ordering contributes to improved mechanical properties such as tensile strength and elastic modulus. These findings are consistent with the results obtained from FTIR, TGA, and SEM analyses, further validating the effectiveness of the biological pretreatment in modifying fiber structure at the molecular level.

4. Conclusions

This study demonstrates that sludge impregnation serves as an effective pretreatment method for banana pseudostem fibers, with process optimization achieved through a Box–Behnken design. By adjusting the concentration of Bacillus subtilis, treatment duration, and pH value, the tensile strength, elastic modulus, and elongation of the fibers were significantly enhanced compared to untreated samples. These enhancements were corroborated by scanning electron microscopy (SEM), thermogravimetric–differential thermal analysis (TGA–DTG), Fourier transform infrared spectroscopy (FTIR), and X-ray diffraction (XRD), confirming the removal of non-cellulosic components, improved thermal stability, and increased crystallinity.

Compared to conventional pretreatments (such as alkali or enzymatic degumming), sludge degumming offers significant advantages in sustainability and cost-effectiveness by integrating microbe-rich sludge into a “waste-to-treat-waste” process system. This not only reduces chemical inputs and energy consumption but also preserves cellulose integrity, highlighting the novelty and environmental relevance of the method and demonstrating its potential to advance natural fiber processing technologies.

The findings indicate that sludge-degummed banana fibers hold considerable potential in diverse industrial applications. Their improved mechanical properties make them suitable for use as reinforcement in bio-based composites, biodegradable packaging, geotextiles, and specialty paper products. In particular, their renewable origin and reduced environmental footprint position them as promising alternatives to traditional bast fibers such as jute or flax in sustainable materials engineering. Additionally, the process shows potential compatibility with existing wastewater treatment infrastructure, offering opportunities for integration into circular bioeconomy models, particularly in banana-producing regions where pseudostem residues are abundant.

From a scalability perspective, several challenges and opportunities exist. The sludge retting process is inherently compatible with continuous-flow bioreactor designs, suggesting feasibility for industrial-scale fiber production. However, successful scale-up will require optimization of process parameters for large volumes, improved control of microbial activity, and standardized sludge pretreatment to ensure reproducibility. Future research will also need to address downstream considerations, such as wastewater recycling, by-product valorization, and techno-economic assessments, to enable the development of commercially viable biorefinery systems.

Despite these promising outcomes, this study has certain limitations. The work mainly focused on optimizing retting parameters and evaluating fiber properties, while detailed analysis of microbial community dynamics and comprehensive monitoring of retting environmental factors were not conducted. These aspects are critical for a deeper understanding of the retting mechanism and ensuring reproducibility under industrial conditions. Future research will address these limitations by investigating microbial population behavior, developing robust process control strategies, and integrating sludge retting into scalable, closed-loop production systems to fully realize its industrial and environmental potential.