Valorization of endemic sigupai aceh rice husk as a renewable cellulose source for nanocellulose production via sulfuric acid hydrolysis

Abstract

Endemic rice varieties can create locally abundant lignocellulosic residues. This study evaluates Sigupai rice husk, an Aceh endemic biomass, as a feedstock for cellulose nanocrystal production through alkaline delignification, hypochlorite bleaching and controlled sulfuric acid hydrolysis at 1, 2 and 3 mol/L. The work tests how increasing acid concentration affects yield, residual non-cellulosic material, crystallinity, surface chemistry and particle-size distribution. The products were assessed by gravimetry, moisture analysis, scanning electron microscopy, Fourier transform infrared spectroscopy, X-ray fluorescence, powder X-ray diffraction and dynamic light scattering particle-size analysis. Increasing the acid concentration increased recovered nanocellulose yield from 18.23 to 21.30%, decreased the XRF-derived residual fraction from 3.279 to 0.253%, and increased the apparent Segal crystallinity index from 91.26 to 99.34%, while reducing moisture content and the DLS-derived mean diameter from 43.31 to 5.27 nm. The calculated complementary XRF fraction increased from 96.721 to 99.747%, but this value is not interpreted as direct cellulose purity because XRF is an elemental technique. FTIR bands supported removal of non-cellulosic components and the appearance of sulfate-ester functionality. XRD remained cellulose-I-dominated, with possible cellulose-II-like features at 3 mol/L that require more rigorous phase analysis for confirmation. Because independent replicate extractions and complementary TEM/AFM, zeta-potential and thermal analyses were not available, the concentration-dependent trends are interpreted descriptively. These results support Sigupai rice husk as a local renewable feedstock and identify acid concentration as a practical processing variable for tuning measured nanocellulose properties.

Introduction

Rice is the dominant agricultural crop in Indonesia, where production remains central to food security and rural economies. In 2024, the Central Bureau of Statistics of Indonesia reported total rice production of 52.66 million tons of milled dry grain, and approximately one fifth of the harvested mass becomes rice husk waste [1]. Although rice husk is often treated as a low-value residue and may be burned in open fields, it contains a high fraction of cellulose, lignin, hemicellulose and silica, making it a technically attractive feedstock for the recovery of biobased functional materials ([2]; Rojib Aminudin et al., [3]).

The Sigupai rice variety is endemic to Aceh and is valued for its fragrant grain, adaptability to different soils and distinctive grain characteristics, including a reported grain length of up to 6.89 mm and amylose content of 20.92% (Purwanto et al. [4]. Its husk remains underutilized despite the wider recognition that rice husk possesses high porosity, high surface area and a silica-rich lignocellulosic architecture that can be exploited in advanced material production (Rojib Aminudin et al., [3]. Converting this local residue into nanocellulose therefore links biomass valorization with regional waste management and circular bioeconomy objectives. However, the chemical composition of the untreated Sigupai rice husk, including its cellulose, hemicellulose, lignin, and silica contents, was not directly quantified in the present study. Therefore, the endemic origin of Sigupai is considered primarily as a regional feedstock and biomass-valorization rationale rather than as evidence of a chemically distinct composition compared with conventional rice husks.

To contextualize the feedstock, Table 1 summarizes reported compositional ranges for rice husks from previous studies. Because the initial chemical composition of untreated Sigupai rice husk was not directly quantified in the present study, no Sigupai-specific values are assigned or inferred from other rice varieties. The comparison is therefore provided solely to establish the typical compositional range of rice husk and to contextualize the characteristics of the feedstock used in this work.

Table 1.

Literature-reported chemical composition of rice husks compared with the compositional-data status of Sigupai rice husk

Source / feedstock	Cellulose (%)	Hemicellulose (%)	Lignin (%)	Ash / inorganic fraction (%)
Current study, untreated Sigupai rice husk	Not directly measured	Not directly measured	Not directly measured	Not directly measured
Zulkifli et al., [5], rice husk	38	25	20	17 ash (~ 94% silica in ash)
Ishida et al., [6], rice husk	26.2	17.3	22.4	20.5 ash; 13.6 extractives

Rice-derived agricultural residues are also used directly as composite reinforcements. Balasubramanian and Balasubramanian, [7] fabricated epoxy composites containing raw and alkali-treated rice-straw fibers, including hybrid reinforcement with chicken feathers, and showed how cleaning and alkali treatment can support interfacial adhesion and mechanical reinforcement. That direct fiber-reinforcement route differs from the present isolation of nanoscale cellulose, but it illustrates the broader materials-valorization pathway by which rice-derived residues can be converted from low-value waste into functional polymer-matrix reinforcement.

Nanocellulose is a family of nanoscale cellulose materials that combines low density, high specific surface area, high stiffness, abundant surface hydroxy groups and chemical tunability [8]. These properties explain its growing use in biomedical materials [9, 10], polymer nanocomposites [11–13], food systems [14–16], sustainable packaging [17, 18] and membrane-based separations ([19, 20]; Nitodas et al., [21]; [22–24]).

Cellulose is a linear homopolysaccharide composed of β-D-glucopyranose repeating units joined by β-1,4-glycosidic bonds; this terminology is used here in accordance with Cellulose journal nomenclature recommendations for polysaccharides [25, 26]. Previous work has extracted cellulose or nanocellulose from rice husk and other agricultural residues (Departamento de Física y Matemáticas et al., [27]; [28–30]), while source-dependent differences in rice-husk composition and processing behavior remain relevant when evaluating a new local feedstock.

Nanocellulose characteristics depend strongly on source, pretreatment and isolation route [31–33]. Chemical hydrolysis is widely used to prepare cellulose nanocrystals because mineral acids preferentially cleave less ordered cellulose domains and release crystalline segments. Reported acids include hydrochloric acid [34], nitric acid [35], acetic acid [36] and sulfuric acid [2, 26, 37–39]. Sulfuric acid is particularly important because it can introduce sulfate-ester groups, which improve aqueous colloidal stability through electrostatic repulsion, while excessive acid severity can depolymerize cellulose and lower recovery [40–42].

Reported preparation strategies provide useful process context. Classical rice-husk routes combine alkaline treatment and bleaching with sulfuric-acid hydrolysis [26, 37], whereas Samsalee et al., [2] coupled chemical treatment with high-pressure homogenization. The two reviewer-specified studies by Xiao et al., [43] and Xiao et al., [44] likewise used sulfuric-acid-derived cellulose nanocrystals as precursors for subsequent functionalization. In comparison, the present protocol tests 1–3 mol/L H₂SO₄ at 45 °C for 30 min after NaOH delignification and NaOCl bleaching, without high-pressure homogenization or post-hydrolysis oxidation. Its practical advantage is therefore the relatively simple sequence and moderate tested acid molarity; superiority in yield or product quality across studies is not claimed because feedstocks, acid severity and downstream processing differ.

The objective of this work was to extract nanocellulose from Sigupai rice husk, a locally endemic and previously unexplored residue, and to determine how sulfuric acid concentration controls the resulting yield, moisture content, XRF-derived residual fraction, morphology, functional groups, crystal structure and particle-size distribution. The guiding hypothesis was that increasing H₂SO₄ concentration from 1 to 3 mol/L would intensify the removal of amorphous cellulose and non-cellulosic components, increase apparent crystallinity and reduce particle size, but may also induce detectable changes in cellulose polymorphism. The study addresses this hypothesis by combining gravimetric analysis with SEM, FTIR, XRF, XRD and PSA characterization.

Materials and methods

Materials and chemicals

The Sigupai rice husk used as the raw material was collected from Southwest Aceh Regency, Aceh, Indonesia. Sodium hydroxide (NaOH, 10%, Sigma-Aldrich), sodium hypochlorite (NaOCl, 5%, Onemed) and sulfuric acid (H₂SO₄, Sigma-Aldrich) were used for delignification, bleaching and acid hydrolysis, respectively. H₂SO₄ concentrations were adjusted to 1, 2 and 3 mol/L. All reagents were analytical grade and were used without further purification.

Cellulose extraction from rice husk

Cellulose extraction was conducted using a modified protocol based on Kaewprachu et al., [45] and Samsalee et al., [2]. Rice husk was first manually cleaned to remove straw, leaves, stones and other visible impurities, washed three times with distilled water, sun-dried for 12 h and then oven-dried at 50 °C for 2 h. The dried husk was ground and sieved through a 60-mesh sieve to obtain a homogeneous powder.

For alkaline delignification, 20 g of ground husk was mixed with 10% (w/v) NaOH solution at a rice-husk-to-NaOH ratio of 1:20 (w/v). The slurry was stirred at 400 rpm and heated at 80 °C for 2 h. The treated solid was washed repeatedly with distilled water until the filtrate reached approximately neutral pH and then dried at 50 °C for 12 h. This step was designed to remove alkali-soluble lignin, hemicellulose and part of the silica-rich inorganic fraction.

The delignified fibers were bleached in 5% (w/v) NaOCl at 70 °C for 30 min under magnetic stirring at 400 rpm to further remove residual lignin, hemicellulose, pigments and inorganic residues. The bleached cellulose was washed to neutral pH, oven-dried at 50 °C for 12 h and stored in a sealed container inside a desiccator at room temperature (25 ± 3 °C) until nanocellulose preparation.

Nanocellulose extraction from Sigupai rice husk

Before hydrolysis, dried bleached cellulose was ground and sieved through a 100-mesh sieve. For each experiment, 10 g of cellulose was treated with H₂SO₄ at 1, 2–3 mol/L using a cellulose-to-acid ratio of 1:20 (w/v), following a modified method described by Bano and Negi [36]. Hydrolysis was performed at 45 °C for 30 min under stirring at 400 rpm, after which the reaction was quenched immediately by dilution with distilled water. The suspension was washed repeatedly with distilled water to approximately neutral pH (pH ~ 7). The washed material was oven-dried at 40 °C for 24 h, ground and sieved through a 500-mesh sieve to obtain dry nanocellulose powder. Samples cellulose nanocrystal (CNC) prepared using 1, 2 and 3 mol/L H₂SO₄ were designated CNC-1, CNC-2 and CNC-3, respectively. Figure 1 presents a schematic overview of the sequential processes involved in producing nanocellulose from Sigupai rice husk, including cellulose extraction and sulfuric acid hydrolysis.

Fig. 1.

Schematic illustration of cellulose extraction, sulfuric-acid hydrolysis and nanocellulose production from Sigupai rice husk

Nanocellulose yield

Nanocellulose yield was determined gravimetrically from the dry product mass relative to the initial representative mass of Sigupai rice husk used for nanocellulose preparation, using Eq. (1):

Y_n (%) = (M_f / M_{_}i) × 100 (1).

where Y_n is the nanocellulose yield, M_f is the final dry mass after processing and M_i is the initial mass of Sigupai rice husk used for nanocellulose preparation.

Moisture content

Moisture content was measured using the oven-drying method according to the Association of Official Analytical Chemists procedure cited in the original method [32]. Approximately 1 g of each dry nanocellulose powder was heated at 130 °C for 1 h, cooled in a desiccator and weighed. Moisture content was calculated using Eq. (2):

Moisture content (%) = [(m_i - m_d) / m_i] × 100 (2).

where m_i is the initial sample mass before drying and m_d is the mass after oven drying.

Scanning electron microscopy

The morphology of nanocellulose powders was observed by scanning electron microscopy (SEM). Samples were fixed on double-sided carbon tape, vacuum-dried, sputter-coated with gold and imaged at an accelerating voltage of 15–20 kV. Images were collected at nominal magnifications within the recorded 10,000× range. The acquisition-condition bands in the original SEM outputs were cropped, and 1 μm scale bars were redrawn for clarity and compliance with figure-formatting requirements.

Fourier transform infrared spectroscopy

Fourier transform infrared spectroscopy (FTIR) was used to evaluate chemical structure and functional groups using a Shimadzu IRPrestige-21 instrument. Spectra were collected over 400–4000 cm − 1. Reflectance data were obtained using the instrument Resolution Pro software (version 2.5.5) and further processed in OriginPro 2025. Because FTIR is absorption-based spectroscopy, the spectral features are described throughout the manuscript as absorption bands rather than emission peaks.

X-ray fluorescence

X-ray fluorescence (XRF) analysis was performed using an XD Mobile-XOS instrument to assess the XRF-detectable inorganic elemental residues in the nanocellulose samples. Because XRF is an elemental analytical technique, it was not used to directly quantify cellulose, hemicellulose, or lignin. For mass-balance presentation, a complementary fraction was calculated as 100 minus the XRF-derived residual fraction. This value was treated solely as a calculated remainder and not as a direct measure of cellulose purity, because the XRF-undetected fraction may contain cellulose together with other organic constituents that cannot be resolved by XRF.

Powder X-ray diffraction

Powder X-ray diffraction (XRD) was used to assess crystallinity, crystallite size and cellulose polymorphism. Nanocellulose powders were mounted as flat powder specimens and analyzed using a Shimadzu XRD-7000 diffractometer equipped with a scintillation detector in reflection geometry with Cu Kα radiation operated at 40 kV and 30 mA. The diffraction system employed a divergence slit of 1.0°, scatter slit of 1.0° and receiving slit of 0.3 mm. The weighted mean wavelength was taken as λ = 1.5418 Å after Kβ suppression. Diffraction patterns were collected over 2θ = 10–80° using continuous scanning at 10° min^− 1 with a sampling pitch (step size) of 0.02°, corresponding to an approximate total scan time of 7 min. Background correction, peak inspection and comparative plotting were performed using OriginPro 2025, while phase matching and Miller-index assignment were performed using PowderX and Match! software against the International Centre for Diffraction Data database.

The average crystallite size was calculated using the Debye-Scherrer equation, Eq. (3):

D = Kλ / (β cos θ) (3).

where D is the average crystallite size, K is the shape factor (0.9), λ is the X-ray wavelength, β is the full width at half maximum in radians and θ is the Bragg angle.

Crystallinity index was calculated using the Segal peak-height method, Eq. (4):

CrI (%) = [(I₂₀₀ - I_am) / I₂₀₀] × 100 (4).

For cellulose Iβ, I₂₀₀ was taken from the 200 reflections near 22° and I_am from the intensity minimum between the 1–10/110 and 200 reflections near 18.6°, rather than from the 1–10/110 reflection itself. Diffraction labels follow the Cellulose journal convention: observed cellulose Iβ peaks are labelled 1–10/110, 200 and 004, while cellulose II peaks are labelled 1–10, 110 and 020 [46].

Particle-size analysis

Particle-size analysis (PSA) was performed with a Horiba SZ-100 instrument using dynamic light scattering (DLS) to determine the mean hydrodynamic/equivalent particle diameter and size distribution of each nanocellulose dispersion. Distributions were compared among CNC-1, CNC-2 and CNC-3 to evaluate the influence of acid concentration on dispersed-particle size and sample heterogeneity. Because DLS probes Brownian diffusion in liquid, these values are not direct measurements of dry SEM aggregate dimensions or of primary nanocrystal length and width.

Statistical treatment

Nanocellulose yield and moisture content were determined in triplicate (n = 3), and the results are reported as mean ± standard deviation (SD). The SEM, FTIR, XRF, XRD, and DLS/PSA data were obtained as representative characterization measurements for each hydrolysis condition and were therefore interpreted descriptively without inferential statistical analysis.

Results and discussion

Nanocellulose yield and moisture content

The mean nanocellulose yield showed a progressive increase with H₂SO₄ concentration, from 18.226 ± 0.043% for CNC-1 to 20.393 ± 0.027% for CNC-2 and 21.303 ± 0.014% for CNC-3 (Table 2). This pattern is consistent with acid-promoted cleavage of less ordered cellulose domains, which releases crystalline cellulose segments from the microfibrillar structure. At 1 mol/L, proton availability was sufficient to initiate hydrolysis but not to fully fragment residual amorphous regions; at 3 mol/L, the greater hydrolysis severity was consistent with more extensive cleavage of less ordered cellulose regions and a higher recovered nanocellulose yield.

Table 2.

Sample designations, processing conditions, yield and moisture content of nanocellulose obtained from Sigupai rice husk

Sample code	NaOH (%)	NaOCl (%)	H2SO4 (mol/L)	mi (g)	md (g)	Yield (%)	Moisture content (%)
CNC-1	10	5	1	1.000	0.978	18.226 ± 0.043	2.190 ± 0.002
CNC-2	10	5	2	1.000	0.978	20.392 ± 0.027	2.113 ± 0.009
CNC-3	10	5	3	1.000	0.981	21.302 ± 0.014	1.820 ± 0.002
Descriptive trend	-	-	increasing	-	-	increased	decreased

The trend agrees with reports that increasing H₂SO₄ concentration can increase cellulose nanocrystal yield when the conditions remain below the threshold for severe depolymerization. For example, higher H₂SO₄ concentration increased CNC recovery from Acacia wood kraft pulp, and optimized sulfuric-acid hydrolysis of brewer’s spent grain produced a high CNC yield under controlled severity ([47]; Zulkifli et al., [5]; [48]). In the present work, the maximum concentration of 3 mol/L remained moderate and did not produce an observable yield loss, suggesting that hydrolysis severity was sufficient for amorphous-domain removal but not excessive enough to degrade the crystalline cellulose fraction.

The mean moisture content decreased from 2.1867 ± 0.002% for CNC-1 to 2.1133 ± 0.009% for CNC-2 and 1.8200 ± 0.002% for CNC-3. (Table 2). This decrease can be explained by two complementary effects. First, removal of amorphous regions reduces the number of disordered sorption sites that bind water. Second, sulfuric acid can esterify surface hydroxy groups to form sulfate-ester groups, thereby changing the balance between water binding, surface charge and particle packing [49, 50]. The low moisture contents (< 5%) are favorable for dry powder storage because they reduce water-mediated agglomeration and microbial susceptibility.

Values are expressed as mean ± standard deviation (SD) from triplicate measurements (n = 3).

XRF-derived residual fraction and calculated complementary fraction

XRF screening indicated a progressive decrease in the XRF-derived residual fraction from 3.279% for CNC-1 to 1.193% for CNC-2 and 0.253% for CNC-3 (Table 3). The corresponding complementary fractions, calculated as 100 minus the XRF-derived residual fraction, were 96.721%, 98.807%, and 99.747%, respectively. These complementary values are reported solely as calculated remainders and should not be interpreted as direct measurements of cellulose purity. XRF is an elemental analytical technique and does not directly quantify cellulose, lignin, or hemicellulose. Moreover, because individual elemental or oxide contributions corresponding to the residual fractions were not available, the residual values cannot be unequivocally assigned to specific inorganic species or contaminants.

Table 3.

XRF-derived residual fraction and calculated complementary fraction for nanocellulose samples

Sample code	XRF-derived residual fraction (%)	Calculated complementary fraction (%)	Interpretation
CNC-1	3.279	96.721	Highest residual fraction
CNC-2	1.193	98.807	Intermediate residual fraction
CNC-3	0.253	99.747	Lowest residual fraction
Descriptive trend	decreased	increased	Lower XRF-derived residue with increasing acid concentration

The progressive decrease in the XRF-derived residual fraction is therefore interpreted conservatively as evidence of lower levels of XRF-detectable inorganic residues in the samples obtained at increasing H₂SO₄ concentration. This trend may reflect enhanced release and subsequent removal of inorganic residues during hydrolysis and washing; however, XRF alone cannot demonstrate that the resulting nanocellulose is free of contaminants. This interpretation is also consistent with the FTIR results, in which the weak absorption band near 1640 cm⁻¹ is primarily attributed to adsorbed water, while the possible presence of trace non-cellulosic constituents cannot be completely excluded without complementary compositional analysis. Accordingly, the combined XRF and FTIR results support progressive purification of the cellulose-containing material but do not substantiate a claim of approximately 99% cellulose purity.

Similar improvements in component fractionation and inorganic-residue removal have been reported for chemically treated lignocellulosic feedstocks [6, 51, 52].

The calculated complementary fraction was obtained as 100 − XRF-derived residual fraction and is reported solely as a calculated remainder; it should not be interpreted as directly measured cellulose purity.

Morphological evolution

SEM observations reveal that hydrolysis severity controlled the degree of fibril fragmentation and agglomeration (Fig. 2). CNC-1 retained relatively large, compact microfibrillar fragments, consistent with incomplete hydrolysis of amorphous cellulose domains. The residual less ordered cellulose can act as a physical binder, allowing adjacent crystalline regions to remain aggregated. This interpretation is consistent with prior observations that mild acid conditions produce partially dissociated fibers and fibrous aggregates rather than isolated nanocrystals [32].

Fig. 2.

SEM micrographs of nanocellulose prepared at different H₂SO₄ concentrations: a CNC-1, b CNC-2 and c CNC-3. Magnification: 10,000×; scale bar: 1 μm.

At 3 mol/L, CNC-3 exhibited a more disrupted surface and finer nanoscale features, indicating more complete removal of amorphous segments and a higher degree of particle release. Sulfate-ester formation during H₂SO₄ hydrolysis can also increase negative surface charge, improving dispersion in aqueous suspension and limiting hydrogen-bond-driven agglomeration [53]. The morphology is qualitatively consistent with the yield and particle-size trends: increasing acid concentration promoted fragmentation and deaggregation while maintaining crystalline cellulose domains.

Functional-group analysis

FTIR spectra provide chemical evidence for the conversion of rice husk cellulose into purified nanocellulose (Fig. 3). Raw rice husk typically contains absorption bands associated with cellulose, hemicellulose, lignin and silica, including O-H stretching at 3200–3400 cm^{− 1}, C-H stretching near 2890 cm^{− 1}, carbonyl C = O bands near 1730 cm^{− 1} and aromatic lignin bands around 1600 –1500 cm^{− 1} [52, 54, 55]. After delignification, bleaching and hydrolysis, the carbonyl and aromatic lignin bands were strongly reduced or absent, confirming substantial removal of hemicellulose and lignin.

Fig. 3.

FTIR absorption spectra of CNC-1, CNC-2 and CNC-3

The spectra of the CNC samples retained the characteristic cellulose absorption bands: broad O-H stretching at approximately 3300–3400 cm^{− 1}, C-H stretching near 2890 cm^{− 1}, C-O-C glycosidic linkage bands around 1160 cm^{− 1} and C-O stretching bands in the 1110 –1050 cm^{− 1} range. The band near 897 cm^{− 1} is consistent with β-glycosidic linkages in cellulose [56]. The weak band around 1640 cm^{− 1} is more conservatively assigned to adsorbed water bending rather than residual lignin, although minor overlap with traces of non-cellulosic material cannot be fully excluded without complementary analysis [57].

Bands near 1205 cm^{− 1} and around 800 cm^{− 1} indicate sulfate-ester vibrations associated with S = O and C-O-S bonding, respectively, confirming that sulfuric acid treatment not only cleaved amorphous domains but also introduced surface sulfate-ester groups [58]. The terminology has been revised to refer to hydroxy groups, sulfate-ester groups and absorption bands, in accordance with IUPAC terminology and FTIR spectroscopy conventions.

Crystallinity and polymorphic transition

XRD patterns of CNC-1, CNC-2 and CNC-3 show cellulose-derived diffraction features and a concentration-dependent increase in peak sharpness (Fig. 4). Following Cellulose journal convention, the principal cellulose Iβ reflections are described as 1–10/110, 200 and 004 rather than using parenthesized Miller-index labels for observed peaks [46]. The amorphous contribution used in the Segal crystallinity calculation was assigned to the intensity minimum between the 1–10/110 and 200 reflections, near 18.6°, rather than to the 1–10/110 reflection itself.

Fig. 4.

Powder XRD patterns of nanocellulose prepared at different H₂SO₄ concentrations. The Segal-method intensities I₂₀₀ (intensity at ~ 22^o) and I_am (intensity at ~ 18.6^o) used for crystallinity-index calculation are indicated

The apparent Segal crystallinity index increased from 91.26% for CNC-1 to 99.34% for CNC-3, while the Scherrer crystallite size increased from 10.39 to 12.15 nm (Table 4). This trend is consistent with preferential removal of disordered domains, which increases the relative contribution of ordered scattering to the diffraction pattern. Comparable increases during acid hydrolysis have been reported for rice husk and other lignocellulosic feedstocks [48, 59]. The unusually high 99.34% value requires caution: the Segal peak-height method can approach 100% when the selected amorphous minimum, I_am, is very low relative to I₂₀₀, and the result is sensitive to background/baseline selection, peak overlap, preferred orientation and other profile effects [60]. Therefore, 99.34% is interpreted here as an apparent, method-dependent Segal index rather than an absolute crystalline fraction or evidence of near-perfect crystallinity. Without peak deconvolution or Rietveld-type profile analysis and replicate diffraction measurements, the magnitude should not be overinterpreted.

Table 4.

Apparent Segal crystallinity index and average Scherrer crystallite size of nanocellulose samples

Sample code	Apparent Segal crystallinity index (%)	Average crystallite size (nm)
CNC-1	91.26	10.39
CNC-2	95.37	11.46
CNC-3	99.34	12.15
Descriptive trend	increased	increased

Peak matching is consistent with cellulose Iβ-dominated patterns for CNC-1 and CNC-2, while CNC-3 shows additional reflections that are compatible with a possible partial crystalline rearrangement and limited cellulose-II-like contribution (Fig. 5; Table 5). Cellulose II is associated with antiparallel chain packing and a monoclinic lattice structure [61–63]. However, the present assignment is based on peak matching only. Because no Rietveld refinement, quantitative phase analysis or peak deconvolution was performed, the data do not constitute definitive evidence of a cellulose I-to-II transition. The CNC-3 pattern is therefore described conservatively as showing possible or partial structural rearrangement that should be confirmed by more rigorous diffraction analysis.

Fig. 5.

Peak matching of nanocellulose samples. Assignments are tentative and are not equivalent to quantitative phase refinement

Table 5.

Diffraction peak matching and tentative structural assignment of nanocellulose samples using PowderX, Match! and ICDD reference data

Sample	Diffraction peak label	2θ (°)	d-spacing (Å)	Structural assignment/ Remarks	ICDD ref.
CNC-1	1–10/110	18.84	4.706	Cellulose Iβ diffraction region	00-056-1718
	200	22.22	3.998	Dominant crystalline cellulose reflection	00-056-1718
	004	39.42	2.284	Higher-order cellulose reflection	00-056-1718
	44	43.94	2.0590
	64	64.32	1.447	Residual inorganic crystalline contribution	-
	68	68.66	1.366	Minor crystalline impurity	-
	77	77.44	1.231	Weak higher-angle crystalline reflection	-
CNC-2	1–10/110	15.68–18.04	5.647–4.913	Cellulose Iβ diffraction region	00-056-1718
	200	22.09	4.021	Dominant crystalline cellulose reflection	00-056-1718
	004	39.46	2.282	Higher-order cellulose reflection	00-056-1718
	44	43.91	2.0604
	64	64.27	1.448	Residual inorganic crystalline contribution	-
	68	68.63	1.366	Minor crystalline impurity	-
	77	77.37	1.232	Weak higher-angle crystalline reflection	-
CNC-3	1–10/110	14.78–17.88	5.989–4.957	Cellulose Iβ diffraction region	00-056-1717
	200	20.38–22.28	4.354–3.987	Dominant crystalline cellulose reflection	00-056-1717
	004	34.23–35.75	2.617–2.510	Higher-order cellulose reflection	00-056-1717
	44	43.91	2.063
	64	64.27	1.4481	Residual inorganic crystalline contribution	-
	68	64.27	1.448	Minor crystalline impurity	-
	77	77.45	1.232	Weak higher-angle crystalline reflection	-

Particle-size distribution

DLS particle-size analysis showed a concentration-dependent change in the dispersed-particle distribution (Fig. 6). CNC-1 had the largest reported mean apparent hydrodynamic diameter (43.31 nm), whereas CNC-2 and CNC-3 had smaller reported means of 7.15 and 5.27 nm, respectively. These values describe the dispersed state measured by DLS and should not be read as direct SEM dimensions or as primary nanocrystal length and width.

Fig. 6.

Particle-size distributions of CNC-1, CNC-2 and CNC-3

The apparent discrepancy between micron-scale structures in SEM and the 5.27 nm DLS mean for CNC-3 arises because the methods interrogate different physical states and size metrics. SEM images a dried, sputter-coated powder in which cellulose particles can form secondary micron-scale aggregates through hydrogen bonding and drying/capillary forces. DLS instead estimates an apparent hydrodynamic diameter from Brownian motion in a liquid dispersion. A dispersed primary particle or small association can therefore yield a nanometer-scale DLS value even when the same material forms much larger dry aggregates. The two measurements are thus not directly interchangeable. TEM or AFM would be required to establish primary nanocrystal length and width and to reconcile morphology quantitatively.

Mechanistically, H+ ions catalyze cleavage of glycosidic linkages in the less ordered regions of cellulose. As acid concentration increases, the probability of bond cleavage in these regions rises, resulting in shorter crystalline fragments and more uniform nanoscale particles. Similar decreases in nanocellulose particle size with increased acid concentration have been reported for other biomass-derived nanocellulose systems [32, 64, 65]. Among the investigated conditions, 3 mol/L H₂SO₄ produced the smallest DLS-derived mean size and the highest apparent Segal index among the tested conditions, while also highlighting the need to monitor possible structural rearrangement at higher acid severity. Because independent replicate preparations were not available, this condition is not presented as a statistically optimized treatment.

CNC-3 was the most noticeably different of the three samples based on the measured measurements: it had the highest descriptive yield (21.302%), lowest moisture content (1.824%), lowest XRF-derived residual fraction (0.253%), highest apparent Segal crystallinity index (99.34%), largest Scherrer crystallite size (12.15 nm) and smallest DLS-derived mean diameter (5.27 nm). Its XRD pattern also showed the additional reflections discussed above as possible cellulose-II-like features. These differences are descriptive rather than statistically significant because replicate extractions were not available.

Complementary characterization was not performed in the present study. In particular, TEM or AFM would provide primary-particle dimensions, zeta-potential measurements would directly test the colloidal effect of sulfate esterification, and TGA/DTG would quantify thermal stability. The present FTIR, SEM, XRD and DLS interpretations are therefore limited to the evidence provided by those methods, and the additional measurements are identified as priorities for future validation.

Conclusion

Sigupai rice husk, an endemic agro-residue from Aceh, was converted into cellulose-rich nanocellulose by sequential alkaline delignification, bleaching and sulfuric acid hydrolysis. The resulting materials exhibited FTIR bands characteristic of cellulose with sulfate-ester functionality, concentration-dependent DLS distributions, and cellulose-derived XRD patterns. Among the tested conditions, 3 mol/L H₂SO₄ gave the highest mean yield (21.30%), the lowest XRF-derived residual fraction (0.253%), the highest apparent Segal crystallinity index (99.34%), and the smallest DLS-derived apparent hydrodynamic diameter (5.27 nm). The complementary XRF fraction is not interpreted as direct cellulose purity.

Increasing acid concentration was associated with a lower XRF-derived residual fraction, higher apparent Segal index, and smaller DLS-derived size, while peak matching for CNC-3 also showed possible cellulose-II-like features. Because this structural interpretation was not supported by Rietveld refinement or peak deconvolution, it is considered provisional. Important limitations include the absence of direct initial chemical-composition data for untreated Sigupai husk, replicate characterization measurements, individual XRF elemental speciation, and complementary TEM/AFM, zeta-potential, and TGA/DTG measurements. Although the nanocellulose preparations were independently repeated at least three times for yield and moisture-content determination, the characterization results were obtained from representative samples and are therefore interpreted descriptively. These limitations preclude claims of statistically optimized processing, 99% cellulose purity, or definitive polymorphic conversion. Overall, the work supports the valorization of Sigupai rice husk as a local lignocellulosic feedstock and provides a basis for future studies that directly quantify composition, primary nanocrystal morphology, surface charge, and thermal behavior.

Author contributions

Wafiq Alni Dzulhijjah: Conceptualization, Methodology, Investigation, Formal analysis, Data curation, Visualization, Writing – original draft. Sri Aprilia: Supervision, Conceptualization, Methodology, Validation, Writing – review & editing. Nasrul Arahman: Supervision, Validation, Writing – review & editing. Sri Mulyati: Supervision, Validation, Writing – review & editing. Muhammad Roil Bilad: Supervision, Validation, Writing – review & editing. Erda Deng: Supervision, Validation, Writing – review & editing. Muhammad Prayogie Aulia: Validation, Writing – review & editing. Keizo Nakagawa: Validation, Writing – review & editing. Eiji Kamio: Validation, Writing – review & editing. Ryosuke Takagi: Validation, Writing – review & editing. Hideto Matsuyama: Supervision, Validation, Writing – review & editing.

Funding

This research was funded by the Ministry of Higher Education, Science, and Technology of the Republic of Indonesia through the PMDSU (Pendidikan Magister menuju Doktor untuk Sarjana Unggul) research grant, in accordance with the Assignment Agreement Letter for the Implementation of PMDSU Research (Contract No. 1164/UN11.L1/PG.01.03/DPPM/2025), as well as the Enhancing International Publication (EIP) - Improving the Quality of International Publications (PKPI) program 2025.

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethical approval

Not applicable.

Consent to participate

Not applicable.

Consent to publish

Not applicable.

Competing interests

The authors declare no competing interests.