The fibre bundles used (Fig. 1a) in this study were extracted from pineapple leaves using a retting/scraping process, in accordance with the method described by Betene et al. [26]. The resulting fibres were carefully combed with a metal brush to ensure their alignment, then cut to a length of 40 mm. Scanning electron microscopy (SEM) observations confirmed the bundle structure (Fig. 1b), consisting of several dozen individual fibres along with residual mucilage from the leaf.

The palm kernel oil CONFORTA used in this study (Fig. 1c) was cold-pressed from palm kernels and supplied by the company DJS (Douala, Cameroon).

The beeswax flakes (Fig. 1d) were obtained from the Pycnolab Laboratory (Douala, Cameroon).

To combine the advantages of both coating agents, a hybrid formulation with a mass ratio of oil to wax of 1:2 wt/wt was developed. Specifically, 25 g of beeswax flakes were melted at 115°C using a hot plate, and 50 g of liquid palm kernel oil were gradually incorporated into the molten wax under continuous stirring (Fig. 2a). The resulting mixture formed a drying substance (Fig. 2b), capable of auto-oxidation and polymerisation upon exposure to air, leading to the formation of a hydrophobic film.

For the coating application, pineapple fibre bundles were first dried at 30 ± 2°C for 24 h, then placed in a planetary mixer (Fig. 2c). After 30 s of low-speed mixing, the coating agent (palm kernel oil, beeswax, or their mixture), previously melted at 115°C, was gradually added over a period of 2 min and 30 s under continuous agitation. Mixing was continued for an additional 2 min and 30 s to ensure uniform distribution of the coating on the fibres. The treated fibres were then placed in a ventilated oven at 50 ± 2°C [5,27] and maintained at that temperature for 7 days to complete the drying process (Fig. 2d). Four coating-to-fibre mass ratios (CR = 0.5, 1, 1.5, and 2) were applied to evaluate the effect of the coating treatment on the properties of the fibre bundles.

Fibre bundles were examined in longitudinal and cross-sectional views using scanning electron microscopy (SEM, HITACHI TM-3000, Tokyo, Japan), operated at 15 kV and various magnifications. Prior to observation, samples were coated with a thin layer of gold/palladium to enhance conductivity.

An OMAX optical microscope operated with Top View 3.7 software was used to measure the diameter of both untreated and coated fibre bundles, following established protocols in the literature [28,29]. For each bundle, three images were captured at different positions along its length, and three diameter measurements were taken per image using Fiji software, yielding nine measurements per bundle. A total of thirty bundles were analysed, resulting in 270 measurements for each fibre type. The collected data were used to plot diameter distributions to assess variability. The probability density function was applied to estimate the likelihood of finding a diameter within a specific range. As fibre diameter distributions are often asymmetric, the data were fitted to a two-parameter Weibull distribution, described by Eq. (1): (1)P(x)=ba(xa)b−1e−(xa)bwhere a and b are the scale and shape parameters, respectively, and x is the fibre bundle diameter. These parameters were determined via linear regression fitting, based on the ranked diameter values plotted against a cumulative distribution function on a logarithmic scale.

The true density ρf (g/cm³) of the fibres was estimated following ASTM D3800-16 [30] using a pycnometer and calculated according to Eq. (2): (2)ρf=ρeth(w2−w1)(w3−w1)(w4−w2)where w1, w2, w3 and w4 represent the mass of the empty pycnometer, the pycnometer filled with fibres, with distilled water, and with fibres plus water, respectively (in g), and ρeth= 0.789 g/cm³ is the density of water at 25°C.

Functional group modifications of pineapple fibre bundles following organic treatments were characterised by ATR-FTIR spectroscopy using a Bruker Alpha spectrometer equipped with a diamond crystal and operated via Opus/Mentor software. Spectra were recorded over the range 400–4000 cm⁻¹, with 32 scans per minute at a resolution of 4 cm⁻¹, enabling the identification of changes in functional groups.

The water absorption and moisture regain kinetics of pineapple fibre bundles were evaluated following the RILEM TC 236-BBM protocol, adapted by Page et al. [5], and ASTM D570. After drying at 30°C for 24 h, fibre bundles weighing 1 ± 0.1 g were placed in micro-perforated sachets to form test specimens. For water absorption testing, four samples were immersed in distilled water and four in seawater at 25 ± 2°C, with five replicates per condition. Mass gain was recorded at regular intervals until saturation using a precision balance (0.001 g), after removing surface water via simulated centrifugal shaking. Water absorption (WA) and its normalised form (|WA|) were calculated using Eqs. (3) and (4): (3)WA=100mt−m0m0 (4)|WA|=100mt−m0m∞−m0where m0, mt and m∞ represent the initial, time-t, and saturated masses, respectively.

For moisture regain testing, four fibre samples of 4 cm length and initial mass 1 ± 0.1 g were exposed to a saturated atmosphere (RH 91% at 26°C) generated by distilled water in a sealed container, with the fibres suspended 10 mm above the surface, with 4 replicates per condition. Moisture regain (MR) and its normalised form (|MR|) were calculated using Eqs. (5) and (6): (5)MR=100wt−w0w0

(6)|MR|=100mt−m0m∞−m0where w0, wt and w∞ denote the dry mass, time-t mass, and saturated mass, respectively.

The absorption and sorption phenomena were modelled using hygroscopic functions from the literature, including the following models:

Quasi-static tensile tests were carried out on pineapple fibre bundles using an LDW-5 universal testing machine (China) equipped with a 100 N load cell, at a constant crosshead speed of 2 mm per minute under controlled environmental conditions (25 ± 1°C and 65 ± 1 percent relative humidity). Each bundle was mounted on a kraft paper frame using a drop of UV-curable thiolene resin, as described by Betené Omgba [37]. A gauge length of 10 mm was used for all tests. Fibre bundles exhibiting abnormal fracture or failure near the clamping zones were excluded from the analysis. The average diameter of each tested fibre was determined from images captured using an optical microscope (OMAX, Guangzhou, China) [38,39], with three random diameter measurements along the length using Fiji software, assuming a circular cross-section, as adopted in studies on flax and hemp [28]. A minimum of 25 fibre samples was tested at each temperature. The mechanical properties, including tensile strength (MPa), Young’s modulus (GPa) and elongation at break (percent), were derived from the stress-strain curves. The elastic modulus (E) was determined by linear regression of the stress-strain curve within the strain range between 0.2 percent and 0.6 percent, thereby ensuring an approximation of the linear elastic response. Statistical differences between groups were assessed using one-way analysis of variance (ANOVA), preceded by Levene’s test to verify the homogeneity of variances. When significant differences were detected, Tukey’s post hoc test was applied for pairwise comparison of means. All analyses were performed using IBM SPSS Statistics 27 software (SPSS Inc., Chicago, IL, USA).

Scanning electron micrographs (Figs. 3 and 4) show that raw pineapple fibre bundles exhibit a characteristic fibrous morphology with grooves and mucilage, consistent with observations reported by Betené et al. [40]. Treatment with palm kernel oil (ACO) gradually forms a polymeric film, with bubbles and concretions appearing from a ratio of 0.5; above a ratio of 1, lateral bubbles indicate coating defects resulting from interactions between the oil and the fibre surface [5]. The palm kernel oil-beeswax blend (ACWO) produces an irregular texture, with cracks and bubbles at higher ratios, suggesting structural instability of the film due to internal stresses. Beeswax alone (ACW) results in heterogeneous coverage, delamination, and surface imperfections, despite some film compactness. The observed bubbles indicate partial absorption of the coating agent, as reported with linseed oil [5], and may help limit water diffusion within the fibres and in cement- and gypsum-based composites exposed to wet-dry cycles. In this context, palm kernel oil provides better coverage, the ACWO blend partially enhances protection, while beeswax alone reduces the specific surface area and fibre separation [27].

The data in Table 1 and Fig. 5a show a systematic increase in the mean diameter of pineapple fibre bundles with the coating ratio (CR), regardless of the treatment, due to the progressive accumulation of the protective film, as confirmed by SEM micrographs. At CR = 2, fibres treated with the palm kernel oil–beeswax blend (ACWO) reach 240 μm (+46% vs. raw fibres, 164.1 μm, AC), slightly higher than those treated with oil alone (ACO, 230 μm) or wax alone (ACW, 233 μm), revealing an oil–wax synergy. ACO and ACWO increase diameter dispersion through penetration into inter-fibre gaps and internal cavities, with ACWO further stiffening and filling these spaces up to 238.4 μm. Wax alone also increases diameter, but with greater variability. This increase enhances air entrapment for insulation [27] but reduces specific surface area and interfacial adhesion, potentially limiting mechanical performance in cement- and gypsum-based composites [41]. Diameter distributions fit a two-parameter Weibull function, with shape parameter b higher than that reported for Neuropeltis acuminata [6].

Raw pineapple fibres exhibit an average density of 1.28 ± 0.07 g/cm³, consistent with literature values (1.26 g/cm³) [26]. Coating application progressively reduces density with increasing coating ratio (CR): for the palm kernel oil-beeswax blend (ACWO), density decreases from 1.26 ± 0.02 g/cm³ at CR = 0.5 to 1.06 ± 0.03 g/cm³ at CR = 2, reflecting the accumulation of a low-density protective film [5] and depending on the nature of both fibres and coating [24]. ACWO fibres show the lowest density, followed by ACO and ACW, indicating the hybrid coating’s ability to fill interstices and stiffen the fibres. Low standard deviations confirm the uniformity of the coatings. These coated fibres provide a favourable density–mechanical property balance (Fig. 6), enhancing their potential for lightweight, durable cementitious and plaster composites, capable of withstanding humid environments and wet-dry cycling [18,19].

The FTIR spectra of pineapple fibre bundles (Fig. 7a–c) show that the untreated fibres (AC) exhibit characteristic bands of their biochemical components: a peak at 1050 cm⁻¹ for polysaccharides (cellulose and hemicellulose), a band at 1310 cm⁻¹ for carboxyl groups (C–O), a peak at 1635 cm⁻¹ for adsorbed water (H–O–H), and a peak at 3365 cm⁻¹ for hydroxyl groups (O–H) responsible for the fibres’ hydrophilicity [26]. Fibres treated with palm kernel oil (ACO, Fig. 7a), beeswax (ACW, Fig. 7b), and the oil-wax blend (ACWO, Fig. 7c) retain these bands but display additional absorption bands at 2850, 2920, 1460, and 720 cm⁻¹ corresponding to C-H stretching vibrations of methylene and methyl groups, indicating the incorporation of the coating agents. Their intensity increases with the coating ratio (CR), reflecting progressive integration of the oil or wax. Peaks at 1738 and 1160 cm⁻¹ correspond to esters from triacylglycerides and triglycerides [42,43], indicating partial oxidation. Bands between 1230 and 1150 cm⁻¹ reflect C-O vibrations of esters. Differential analysis (dashed spectra) confirms the incorporation of organic materials and coating efficacy while preserving the fundamental fibre structure. These physicochemical modifications may enhance fibre hydrophobicity, thereby supporting dimensional stability and improving performance in cementitious and plaster composites under tropical conditions [24,44].

The water absorption kinetics curves of pineapple fibre bundles exhibit a two-phase profile (Fig. 8a–c): an initial linear phase reflecting the progressive diffusion of water molecules into the fibrous structure [3,10,11], followed by a non-linear phase that stabilises upon reaching saturation. The saturation absorption rate decreases significantly with increasing coating ratio (CR), from 202.4 ± 12.42 wt% to 76.3 ± 3.9 wt%, confirming the effectiveness of hydrophobic organic treatments, in line with previous studies [5,25]. Fibres treated with palm kernel oil (ACO) reach a plateau after approximately 5000 min, with values ranging from 161.8 to 110.9 wt% (Fig. 8a,d), suggesting a reduction of at least 40 wt% due to the formation of a hydrophobic barrier. Fibres treated with the palm kernel oil-beeswax blend (ACWO) exhibit even lower water uptake (reduction > 50.3 wt%), indicating a synergistic effect that enhances water resistance (Fig. 8b,d). Samples treated with beeswax alone (ACW) show similar performance (Fig. 8c,d). These hydrophobic coatings may be particularly advantageous for gypsum matrix composites by limiting water migration at the fibre-matrix interfaces, thereby improving long-term dimensional stability.

The absorption kinetics data, normalised between 0 and 1 using Eq. (4), were fitted to several mathematical models (Table A1), with fitting parameters determined using MATLAB. The Czel model showed the best performance (R² = 0.980–0.996; RMSE ≤ 0.027), especially for coated fibres (ACO, ACW, ACWO), reflecting a slower and more stabilised absorption process. The Mohsenin model also demonstrated high accuracy (R² up to 0.998), particularly suitable for complex kinetics. Although the Page and Peleg models performed slightly less well (R² ≈ 0.97), they remain valid under certain treatment conditions (ACO0.5, ACWO2). These findings highlight the relevance of multi-parameter models, such as those of Czel and Mohsenin, for accurately modelling water absorption in functionalised plant fibres.

The moisture uptake curves for pineapple fibres (Fig. 9a–c,e–g) show two distinct phases: a rapid increase of around 5% in the first two hours, followed by slower uptake to saturation after eight days, in agreement with the observations of Monreal et al. [45] on linseed oil-coated beetroot crisps. Sorption induced by distilled water is systematically higher than that induced by seawater, reflecting the influence of the environment on the hygroscopic response of the fibres. This difference is explained in particular by the presence of dissolved ions (Na⁺, Cl⁻, Mg²⁺, Ca²⁺, SO₄^2-) in seawater, which lowers the chemical potential of free water and limits capillary imbibition [11,46]. Raw fibres show high absorption, stable at around 27.9% by weight in distilled water and 21.12% in seawater, reflecting their intrinsic hydrophilicity. For treated fibres, absorption decreases as the coating rate (CR) increases, reflecting the increasing effectiveness of the treatments. In distilled water (Fig. 9d), the absorption of ACO fibres decreases from ~20.76 wt% at CR = 0.5 to ~14.6 wt% at CR = 2, a reduction of 25.6%–47.7%. The ACWO and ACW fibres show greater reductions (39.4%–52% and 43%–59.4%, respectively), indicating a greater effectiveness of beeswax as a barrier. The maximum reduction rates obtained (52% and 59.4%) exceed those reported for identical fibres previously treated with caustic soda and hypochlorite [40], or for chemically treated palm nut mesocarp fibres [47], as well as for Neuropeltis acuminatas fibres [26], suggesting the superior effectiveness of organic wax or combined wax and oil treatments. In seawater (Fig. 9f), absorption also decreases: ACO decreases from ~15.84 to ~10 wt%, and ACWO/ACW reach ~8–10 wt% at CR = 2, confirming the increased efficiency of the blend in saline environments. These results underline the fact that coating, particularly with beeswax or wax-oil blends, optimises moisture resistance, especially at high CR. Tables A2 and A3 show that the Peleg and Pilosef models offer the best fit to the data, with R²s between 0.992 and 0.999 and minimum RMSEs of 0.060, guaranteeing accuracy and robustness.

Based on water and moisture sorption criteria, pineapple fibres treated with beeswax or a hybrid palm kernel oil/beeswax blend (1:2 wt/wt) exhibit greater moisture resistance than those treated with palm kernel oil alone. To enable broader adoption of this method without relying exclusively on beeswax, the hybrid treatment is recommended. To assess its impact on mechanical properties, tensile tests were conducted on untreated (AC) and hybrid-coated (ACWO) fibres. The raw AC fibres exhibited a tensile strength of 415.2 MPa, a Young’s modulus of 12.3 GPa, and an elongation at break of 1.6%, in agreement with the values reported by Betené et al. [6] (413.9 MPa; 11.2 GPa; 2.0%) under similar testing conditions. These results position pineapple fibres as a credible alternative to jute (393 MPa; 1.5%–1.8%) and sisal (511–635 MPa; 9.4–22 GPa; 2.0%–2.5%) [48], both widely used in commercial biocomposites. However, the low elongation suggests a brittle failure behaviour, comparable to that of Rosa hybrida (352 MPa; 1.8%) [49] and Heteropogon contortus (476 ± 11.6 MPa; 1.6 ± 0.06%) [50]. The evolution of mechanical performance with coating ratio (CR), shown in Fig. 10, indicates an increase in tensile strength up to CR = 1 (460.8 MPa, i.e., +11%), which can be attributed to the plasticising and cohesive effects of the hybrid wax/oil layer [45,51]. Beyond this threshold, excessive coating disrupts the fibre-coating interface, reducing load transfer efficiency and leading to a decline in strength. This trend is accompanied by increased variability (Fig. 10b), likely resulting from differences in fibre bundle diameter, the heterogeneous nature of plant fibres, uneven porosity and chemical composition [2,28], and localized peeling of the coating layer, as observed in SEM micrographs (Fig. 3). Both factors can affect coating uniformity and mechanical performance.

The Young’s modulus of pineapple fibres gradually decreases from 12.3 to 10 GPa as the coating ratio (CR) increases from 0 to 2 (Fig. 10c), indicating a reduction in stiffness likely due to the increased elasticity of the organic film formed around the fibres and the infiltration of the coating agent into their porous structure, which alters inter-fibrillar interactions. Similar trends have been reported for alkaline and thermal treatments applied to other lignocellulosic fibres [51–53]. Conversely, the elongation at break increases significantly from 1.6% to 2.6% with rising CR (Fig. 10d), suggesting enhanced ductility, particularly evident from CR = 1, although this is accompanied by greater data variability, as reflected in the coefficients of variation. Levene’s test confirmed the homogeneity of variances for Young’s modulus and elongation, justifying the use of analysis of variance (ANOVA) [54], whereas the variance heterogeneity observed for tensile strength required the application of the robust Brown-Forsythe test. ANOVA revealed a statistically significant effect of the organic coating (CR) on elongation at break (p<0.05), with no significant effects on tensile strength or Young’s modulus. Post-hoc tests (Tukey for modulus and elongation; Dunnett T3 for tensile strength; see Table 2) confirmed that CR significantly enhances fibres deformability without compromising key mechanical properties, providing a clear advantage for structural applications that demand both flexibility and integrity.

Organic surface treatments using palm kernel oil, beeswax, and their hybrid blend (ACWO, 1:2) provide an effective means of reducing the hydrophilicity of pineapple leaf fibres (PALF) while maintaining their mechanical integrity. The radar plot (Fig. 11) illustrates the correlation between these properties and the coating ratio (CR = 0–2), with CR = 1 emerging as the optimal condition, showing a tensile strength of 460.8 MPa, an elongation at break of 2.5%, and a Young’s modulus of 10.9 GPa. Hygroscopically, water absorption and moisture regain decrease from 202.4% to 98.7% and from 27.8% to 14.8%, representing respective improvements of 51.2% and 46.8%. These performances surpass those obtained with linseed oil on flax fibres [5,23] and reflect a synergy between the oil’s fluid penetration into the fibre wall matrix and the formation of a continuous hydrophobic wax film that limits water diffusion without collapsing the cell structure. Unlike hornification, which induces irreversible wall collapse [18,19,21,22], the ACWO coating reproduces its densifying effect while maintaining flexibility and internal porosity. Compared to alkaline treatments, often corrosive and energy-intensive [17,40,47], this approach offers a green, low-impact alternative, free from harmful effluents and without degradation of the cellulosic network. No significant mechanical degradation was observed up to CR = 1, whereas higher coating levels (CR ≥ 1.5) caused partial surface stripping of the coating film, confirmed by SEM (Figs. 3 and 4). The variability in properties (standard deviation ≤ 52%) is attributed to morphological heterogeneity and uneven or locally stripped coatings. This study demonstrates, for the first time, the synergistic potential of palm kernel oil-beeswax coatings on PALF, providing a bio-inspired and scalable strategy to enhance the durability of cementitious and gypsum-based composites subjected to wet–dry cycling [23,25].