Cellulose yield
Table 1 shows the yield after defatting, acid/alkaline treatments and bleaching. Defatting preserved the bulk of the biomass (86.81-94.64%) with a significantly (p<0.05) higher value in the banana leaf (Table 1). Such high retention (>85%) is in line with observations in banana pseudostem waste, where defatting by ethanol or acetone retains over 90% of the dry mass
(Bedru et al., 2024). Moreover, raw banana pseudostem naturally comprises more than 50% cellulose, with the remainder largely comprising hemicellulose and lignin; Therefore, solvent extraction removes only minor extractives, preserving the polysaccharide matrix
(Bampidis et al., 2020). This high defatting yield establishes a solid foundation for downstream delignification, maximizing the cellulose fraction exposed to chemical treatment. Previous work on non-wood biomass confirms similar value (above 88%) in banana peels, with negligible loss of cellulosic content
(Mirzaee et al., 2023).
Subsequent acid and alkaline treatments retained 84.94% to 91.26% of the material (Table 1). These values closely match typical solid recoveries reported for chemical delignification of Banana Pseudostem waste. Sulfuric acid pretreatments (25% H
2SO
4) achieved up to 82% mass retention, while sodium hydroxide delignification (25% NaOH) retains around 85% of the original biomass (
Nascimento et al., 2023).
Mohamad and Jai (2022) optimized banana stem pulping using NaOH-EDTA and reported a yield of 82.1%, indicating a similarly high solid retention under alkaline conditions. The slightly higher yields observed in this study (up to 91%) may reflect milder conditions or the inherent variability between stem and leaf tissues, with stem tissues often richer in structural polysaccharides. Such high post-delignification yields are critical as they ensure sufficient cellulose remains for subsequent bleaching and functionalization. They also demonstrate the effectiveness of the chosen acid/alkali sequence in selectively removing non-cellulosic barriers
(Olawuni et al., 2024; Odimegwu et al., 2025).
The final bleaching step recovered significantly (p<0.05) higher values in the stems compared to leaves. The values obtained were: 16.11% (plantain leaf), 17.84% (banana leaf), 20.41% (plantain stem) and 20.77% (banana stem). These figures are comparable to the 20.0% yield reported for bleached cellulose from banana pseudocore using acid/alkaline/bleaching sequences (
Amaya, 2025). In contrast, TEMPO-mediated oxidation of banana pseudostem achieved a final cellulose yield of 25.25% (dried basis), reflecting a more aggressive delignification-oxidation approach that sacrifices some crystallinity for higher extraction efficiency
(Nascimento et al., 2023). Enzymatic hydrolysis yields are lower (within 14.58%) but offer higher crystallinity and thermal stability
(Nascimento et al., 2023). Hence, the 16-21% bleaching yields observed in this study sit squarely between enzymatic and TEMPO methods, which may have balanced purity and crystalline retention.
Fourier transform infrared (FTIR) of the extracted cellulose
The FTIR spectra of both banana and plantain raw leaves and stems and their corresponding extracted cellulose are shown in Fig 3. The cellulose samples closely matched that of commercial cellulose, indicating successful isolation of cellulose. All raw leaf and stem spectra displayed a broad absorption band at approximately 3300 cm
-1, corresponding to O-H stretching vibrations of polysaccharide hydroxyl groups and adsorbed moisture. This band remains in the extracted cellulose but with diminished intensity due to removal of non-cellulosic components
(Ofoedum et al., 2025). The C-H stretching region at 2900 cm
-1 appears in both raw and purified samples, sharpening in the latter as the cellulose backbone becomes more dominant (
Hozman-Manrique et al., 2023). Comparable O-H and C-H bands have been reported in celluloses derived from other non-wood sources such as water hyacinth and
Agave americana (
Chaiwarit et al., 2022;
Krishnadev et al., 2020). Notably, the stem-derived celluloses exhibit sharper and more intense O-H and C–H bands compared to their leaf counterparts, indicating a higher degree of crystallinity and lower residual hemicellulose
(Ofoedum et al., 2025; Bedru et al., 2024).
Again, distinct peak at 1732 cm
-1, attributed to C=O stretching of acetyl and ester groups in hemicellulose and lignin, is evident in all raw leaf and stem spectra but vanishes entirely in the extracted cellulose, especially in the stems
(Harsono et al., 2021). However, only the plantain leaf cellulose retained minor absorbance near 1733 cm
-1. Another minor residual absorbance near 1620 cm
-1, associated with aromatic C=C stretching in lignin, is markedly attenuated in purified samples, further demonstrating lignin removal
(Odimegwu et al., 2025). At the finger print of the spectra, the 1420 cm
-1 band, assigned to -CH‚ scissoring of crystalline cellulose, intensifies in the extracted samples (especially for the Banana stem) relative to raw material, reflecting an increased proportion of ordered cellulose domains (
Sebayang and Sembiring, 2017). The C-H bending/CH‚ wagging band at 1370-1318 cm
-1 sharpens post-extraction, consistent with enrichment of pure cellulose. An anti-symmetric C-O-C stretching band at approximately 1160 cm
-1 remains pronounced after bleaching, confirming retention of the β-(1®4) glycosidic framework. Sharp C-O stretching bands at 1110 cm
-1 and 1055 cm
-1 in purified celluloses further indicate removal of hemicellulosic and lignin impurities
(Sophonputtanaphoca et al., 2023). The β-glycosidic linkage vibration at 894.5 cm
-1, diagnostic of cellulose, is evident in all four extracted samples, corroborating literature values for plant-derived cellulose
(Suebsuntorn et al., 2023). Similar studies on
Musa pseudostem fractions demonstrated systematic decreases in lignin-related bands (1726 cm-1 to 1042 cm
-1) from outer to inner fractions, confirming the spectral trends observed here [16, 37, 39, 41]. Plantain pseudostem fibers subjected to NaOH mercerization also exhibited broad O-H bands at 3400 cm
-1, cellulose bands at 3335-2924 cm
-1 and fingerprint peaks at 1325 cm
-1, 1033 cm
-¹ and 552 cm
-1 (
Hurtado-Figueroa et al., 2025).
Properties of the Na-CMC samples from banana and plantain leaves and stems
The Na-CMC yield, Degree of substitution and purity level are presented in Table 2. Na-CMC yield of the samples which ranged from 12.5% to 16.72% (Fig 3) which is relatively low compared to values reported by other researchers
Ogheneochuko and Jude (2023). Reported a yield of 135.2 for Na-CMC synthesized from rubber seed shells by mercerizing the shells with 30% NaOH solution and subsequent etherification with monochloroacetic acid.
El-Sakhawy (2018) reported a Na-CMC yield as high as 185.3% from sugarcane bagasse using 40% NaOH. These lower yields in the present study are largely influenced by the cellulose source and the chemical treatment conditions. According to previous reports, the Na-CMC yield increases with an increase in the concentration of NaOH and the optimum concentration is at 20-40%
(Rahman et al., 2022). As the NaOH concentration increases, the medium becomes more alkaline, replacing the hydroxyl group and increasing the total mass of Na-CMC
(Ibikunle et al., 2019).
The DS of Na-CMC refers to the average number of the hydroxyl group (-OH) in the cellulose structure that are replaced by carboxymethyl groups
(Sunardi et al., 2017). DS value is one of the most important characteristics of the Na-CMC and it can be used to predict the solubility and viscosity of the Na-CMC. According to
Suebsuntorn and Jirukkakul (2023), a higher DS correlates with greater viscosity and cation exchange capacity. Additionally, DS values in the range of 0.2 to 1.5 are accepted for commercial Na-CMCs and can be utilized in the food and pharmaceutical industries
(Bampidis et al., 2020). In this study, the DS values of the samples, which range from 0.45 to 0.57 (Fig 4), show that the Na-CMCs from the plantain and banana leaves and stems are slightly soluble in water. Thus, they can be considered a potential ingredient for various use in the food and pharmaceutical industries. The DS of Na-CMCs are usually affected by the concentration of the NaOH and the quantity of monochloroacetic acid used for methylation. The substitution degree increases with an increase in concentration of NaOH or the monochloroacetic acid up to an optimum level (
Adinugraha and Marseno, 2005).
El-Sakhawy et al. (2018) recorded the highest DS value (1.1) for sugarcane bagasse at NaOH concentration of 40% while
Ofoedu et al. (2021) reported 1.2 g as the optimum quantity of monochloroacetic acid for obtaining the highest DS value from
Eleocharis dulcis.
The purity levels of the Na-CMCs obtained from this study ranged from 96.56% to 99.3% and showed a significant difference. These values are considerably high when compared to reports from other researchers (
Adinugraha and Marseno, 2005). However, a little more purification is necessary to obtained a food grade purity level of 99.5% as recommended by the Food and Agriculture Organization
(Alabi et al., 2020). This further purification can be accomplished by treating them with methanol or acetone to enhance the separation of Sodium chloride and Sodium glycolate (
Amaya, 2025).
FTIR of the Na-CMC samples
The FTIR spectra of Na-CMC derived from banana leaf, banana stem, plantain leaf and plantain stem closely resemble that of commercial Na-CMC, confirming their structural similarity (Fig 3). However, minor differences in peak intensities or slight shifts may reflect variations in the degree of substitution (DS) or impurities from the source materials. The spectra of each Na-CMC sample exhibited characteristic asymmetric -COO- stretching bands around 1600 cm
-1 and symmetric -COO- bands at around 1420 cm
-1, indicating effective etherification
(Gieroba et al., 2023). The intensity of the asymmetric -COO- stretching bands varied among the samples, with banana stem Na-CMC showing the strongest absorption, followed by plantain leaf, banana leaf and plantain stem. This suggests differences in the degree of substitution (DS) and local substitution patterns
(Riaz et al., 2018). The broad O-H stretching band at 3330-3345 cm
-1 was most attenuated in plantain stem Na-CMC, indicating highest hydroxyl substitution
(Sebayang et al., 2017).
The variations in -COO- band intensity and position correlate with known relationships between FTIR peak characteristics and functional performance. Higher asymmetric -COO- intensities are associated with increased DS and enhanced water-binding capacity
(Ren et al., 2024). Thus, banana stem Na-CMC, with its pronounced asymmetric -COO- band, is expected to deliver superior viscosity enhancement and moisture retention
(Diem et al., 2023). In contrast, banana leaf Na-CMC’s slightly lower -COO- intensity may favour film-forming and sensory attributes, making it a promising candidate for surface-coating applications (
Ren et al., 2024).
Again, the broad O-H stretching vibration band between 3200 and 3400 cm
-1, the C-H stretching vibration peak in the 2900 cm
-1 region and the strong C-O-C and C-O stretching bands in the 1200-1000 cm
-1 region were all consistent with the cellulose structure. These findings suggest that the cellulose backbone was retained in all Na-CMC samples, despite differences in source material and carboxymethylation, which is crucial for maintaining the structural and functional properties of the Na-CMC samples.
Scanning electron microscopy of the CMC samples from banana and plantain leaves and stems
The Scanning Electron Microscopy (SEM) micrographs of the samples as shown in Fig 5 illustrate homogenous films with fewer exposed fibril edges, indicating intensive substitution of surface hydroxyls by hydrophilic carboxymethyl groups (
Sophonputtanaphoca, 2023). Banana leaf Na-CMC presents continuous sheet-like structures with occasional micron-scale wrinkles, suggesting strong interchain hydrogen bonding among carboxymethylated chains
(Mirzaee et al., 2023). In contrast, banana stem Na-CMC exhibits irregular, porous aggregates of fused particles, reflecting partial preservation of underlying microfibrillar bundles entwined in CMC clusters
(Meraj et al., 2025). Plantain leaf Na-CMC shows a hybrid morphology: slender residual fibrils embedded in a continuous CMC matrix, implying incomplete shielding of the cellulose backbone and potential for enhanced moisture retention through capillary action (
Saberi et al., 2023). Finally, plantain stem Na-CMC retains distinct microfibril bundles bridged by amorphous CMC gel layers, which may provide a balance of structural reinforcement and hydration capacity in food systems
(Meraj et al., 2025; Riaz et al., 2018).
These morphological differences have direct implications for functionality in food systems. Smoother, film-forming CMC surfaces enhance solubility and reduce crystallinity, factors known to increase water-binding capacity and viscosity. Porous, aggregated morphologies, as seen in banana stem Na-CMC, offer greater specific surface area, promoting stronger interactions with other food microstructures
(Diem et al., 2023; Ardila et al., 2024). The hybrid fibril-matrix architectures in plantain leaf and stem Na-CMC suggest dual functionality: residual fibrils provide mechanical support, while the surrounding CMC network delivers moisture retention and viscosity control
(Cukrowicz et al., 2020).