Ozonation of Sago Starch (Metroxylon sagu Rottb.): Kinetic modeling and effects of the process on viscosity, structural, and morphological properties

1.1 Kinetic modelling of Sago starch ozonation

The starch ozonation reaction mechanism involves a series of oxidation steps, where ozone interacts with hydroxyl ions (OH) to generate hydroxyl radicals (OH*) and oxygen (O₂) (Eq. 1). As a result of this process, starch molecules undergo oxidation, which produces carbonyl and carboxyl groups (Eq. 2 and 3) (Brodowska et al., 2018). The reaction can be represented as follows:

Assuming a continuous and excessive supply of ozone, the impact of ozone concentration can be disregarded, allowing the ozonation reaction equation to be expressed in a simplified form in equation (4) as follows:

if m/n is assumed to be a new variable α, then the equation (4) becomes equation (5) as follows.

Thus, the reaction kinetic equation for the formation of the carboxyl group can be expressed in equation (6) as follows

In the reaction system, hydroxyl (OH) will be transformed into hydroxyl radicals, and ultimately revert to hydroxyl, while maintaining a constant pH in the solution (Codorniu-Hernández and Kusalik, 2012). The concentration of carboxyl is represented as C_A, while the concentration of starch is denoted as C_S. If X represents the conversion value of oxidized starch to carboxylate, the reaction rate equation can be expressed in equation (7) as follows:

According to the principle of stoichiometry, C_A and C_S can be defined as a function of conversion as expressed in equations (8) and (9).

Substituting equations (8) and (9) into equation (7) and assuming that C_A0 = 0 when t = 0, equation (7) becomes equation (10) as follows.

Equation (10) can be rearranged to the following equation (11)

Equation (11) is then integrated with the boundary conditions of t = 0, X = 0, and t = t, then X = X. The integrated equation have been shown in equation (13)

Equation (13) is linearized to the following equations (14) and (15):

k_r, [OH], C_S0, and α are constants that can be substituted as k_obs, therefore, equation (15) can be written as equation (16).

Thus, the simplified form of equation (16) can also be expressed in equation (17).

The parameters of k_obs and α can be determined by linear regression from the plot of ln(1-X) vs ln t. This plot will yield a linear equation where the parameters can be determined from the slope and intercept of the linear equation.

The value of α can be calculated from slope data.

The value of k_obs can be evaluated from intercept data.

According to equation (16), k_r can be determined from the value of k_obs by rearranging equation (16) to Equation (21)

2.1 Materials

The Sago starch as the raw material for this study was obtained under the Superindo brand, produced by Kekal Jaya Sentosa Ltd, and distributed by Lion Super Indo Ltd. The starch has a total carbohydrate content of 87%. Oxygen with a purity of 99%, used as the feed gas for the ozone generator, was supplied by Samator Aneka Industrial Gas Ltd. The ozone generator used in the ozonation process was an SQ-3 model (manufactured in China), with device specifications of 140 watts of power, 220 V voltage, and an ozone output capacity of 3 g/h.

2.2 Ozonation of sago starch

The method for modifying Sago starch using the ozonation method is based on the research by Çatal & İbanoğlu, (2014), with certain adjustments made in this research. In this procedure, 70 g of Sago starch was dissolved in 130 mL of distilled water to achieve a 35% w/w concentration of dry matter. The pH levels of the suspension were set to be 7.5, 8.0, 8.5, 9.0, and 9.5 before being transferred to the ozonation reactor. The suspension was continuously stirred and subjected to ozone treatment at a gas flow rate of 1 L/min, with ozone injected at a concentration of 690 ppm. Ozonation was performed with a process duration of 5, 10, 15, 20, and 25 min. After the ozonation process, the ozonated Sago starch slurry was decanted and washed until the pH reached neutral. Ultimately, the slurry was left to dry for 24 h at room temperature (Çatal and İbanoğlu, 2014).

2.3 Carbonyl and carboxyl

Carbonyl and carboxyl contents in modified Sago starch were analyzed using established methods (Sumardiono et al., 2023). For carbonyl assessment, a sample (4 g) was dissolved in distilled water, heated to induce gel formation, and the pH was adjusted to 3.2 with 0.1 N hydrochloric acid. Hydroxylamine hydrochloride solution (25 g in 0.5 N sodium hydroxide) was added, and the mixture was incubated at 40°C for 4 h. Afterward, titration was performed with 0.1 N hydrochloric acid until it reached a pH of 3.2. (A: Volume of HCl for modified starch and B: Volume of HCl for native starch (Sumardiono et al., 2023).

Carboxyl content was evaluated similarly by dissolving a 3 g sample in 0.1 N hydrochloric acid, filtering, and washing until chloride ions were undetectable with silver nitrate. The slurry was then heated to gel formation, followed by titration with 0.1 N sodium hydroxide using phenolphthalein as an indicator. Comparative tests were conducted with native Sago starch to assess carboxyl formation. (A: Volume of NaOH for modified starch and B: Volume of NaOH for native starch (Sumardiono et al., 2023).

2.4 Pasting properties

The gelatinization characteristics of both native and modified Sago starch were analyzed using a Rapid Visco Analyzer (RVA, Model 4500, Perten Instruments) with the noodle flour method. In this evaluation, roughly 3 g of each sample, measured on a dry weight basis, were placed into an RVA container, to which 25 g of distilled water was added. The procedure involved both heating and cooling phases, with the mixture being stirred continuously at 160 rpm. During the heating stage, the temperature of the starch suspension was increased from 60°C to 95°C at a rate of 6°C/min, maintaining 95°C for 4 min (Geng et al., 2023).

2.5 Morphology, functional group, crystallinity profile analysis

The surface morphology, functional groups, and crystallinity of native and modified sago starch (pH 9.0 and 25 min) were analyzed using advanced techniques. A scanning electron microscope (SEM-EDX JEOL JSM-6510LA) at 2000× magnification examined granule morphology, while a Fourier Transform Infrared (FTIR) spectrometer (Shimadzu IR Prestige 21) identified functional groups. Crystallinity was assessed through X-ray diffraction (XRD) using the SHIMADZU XRD-7000, with diffraction patterns recorded over a 2θ range of 5° to 40° at a scanning speed of 2°/min. (Pojić et al., 2015).

2.6 Formulation and cooking quality of sago starch-based noodles

Noodles made from Sago starch were produced by first pregelatinizing 5% of the modified starch in distilled water at a ratio of 1:9 (w/v), which was subsequently blended with the remaining 95% starch (from a total of 500 g) to create a dough with approximately 55% moisture content. The dough was kneaded to uniformity, conditioned at 40°C, then extruded directly into hot water maintained at 95-98 °C for 50-70 s, followed by cooling, as described by Chen et al. (200). The cooking performance was evaluated by measuring the cooking time, determined by the point at which the noodle core became transparent, as well as by calculating cooking loss and water absorption based on the ratio of weight increase to dry solids. Noodle color was quantified using the CIE 1976 color space parameters L*, a*, and b*(Chen et al., 2003).

2.7 Statistical analysis

Statistical analysis was conducted with the SPSS 26 software. Initially, all data underwent normality testing. If the data were found to be normally distributed, one-way ANOVA was used for comparison of treatments, followed by the Tukey honestly significant difference (HSD) test. Samples sharing the same letter notation were considered not significantly different, whereas samples with different letter notations were deemed significantly different. All measurements were conducted in duplicate (n = 2) unless otherwise specified. (Xu et al., 2020).

3.1 Carbonyl & carboxyl formation

Fig. 1 illustrates the carbonyl content profile of Sago starch, showing a clear increase with extended ozonation time. Statistical analysis confirms significant differences in carbonyl content due to ozonation time (p < 0.05). After 5 min, the carbonyl content increases from 0.015% at pH 7.5 to 0.0369% at pH 9 and 9.5. The most prominent rise occurs after 25 min, reaching 0.247% at pH 9.5, indicating prolonged ozonation facilitates carbonyl group formation, likely due to intensified oxidation reactions.

Close

Fig. 1.

The percentage of carbonyl content in sago starch at various pH levels (7.5, 8, 8.5, 9, and 9.5) and ozonation durations (5, 10, 15, 20, and 25 min).

Export to PPT

The oxidation of starch involves two primary phases: transformation of hydroxyl groups into carbonyl and carboxyl groups depolymerization of starch by breaking α-(1→4)-D and α-(1→6)-D glycosidic bonds (Rostamabadi et al., 2022; Sumardiono et al., 2023). At higher pH, hydroxyl groups are predominantly converted into carbonyl groups. A similar pattern was observed by Maqbool et al. (2024), who studied wax rice starch during a 6-h ozonation period. They reported that fluctuations in carbonyl content, reflecting oxidation levels, were linked to incomplete interactions between starch and ozone(Maqbool et al., 2024).

Fig. 2 also highlights the rise in carboxyl content during the ozonation process. At 5 min, carboxyl content was just 0.0001% at pH 7.5, but increased sharply to 0.0384% at pH 8.5. The content rises, peaking at 0.247% after 25 min at pH 9.5. Tukey HSD analysis confirms that ozonation time significantly impacts carboxyl content (p < 0.05). At higher pH levels (9 and 9.5), carboxyl content rises more significantly, especially after 20 min of ozonation. Ozonation significantly alters starch structure, reducing amylose and amylopectin size while oxidizing hydroxyl groups into carbonyl and carboxyl groups (Pudjihastuti et al., 2018). Sumardiono et al. (2021) studied ozone modification of Sago starch in acidic, neutral, and basic suspension conditions for 25 min. Ozonation in basic suspension showed a higher increase in carboxyl compared to the other two treatments (Sumardiono et al., 2021c). Similarly, Satmalawati et al. (2020) reported that oxidation using 2 ppm dissolved ozone produced water-soluble oxidized starch, which decreased carboxyl content due to over-oxidation (Satmalawati et al., 2020). Chávez-Murillo et al. (2009) documented a similar reduction in carboxyl content for corn starch during prolonged oxidation (Cahyono et al., 2024; Chávez-Murillo et al., 2008)).

Close

Fig. 2.

The percentage of carboxyl content in sago starch at various pH levels (7.5, 8, 8.5, 9, and 9.5) and ozonation durations (5, 10, 15, 20, and 25 min).

Export to PPT

Regulatory standards ensure oxidized starch safety for consumption, limiting carboxyl content to 1.1% (EU Regulation No 231/2012) and 1.3% (EFSA Panel on Food Additives and Nutrient Sources added to Food (ANS) et al., 2017). In this study, the maximum carboxyl content observed (0.247% after 25 min at pH 9.5) is well within these limits, confirming that the oxidation process remains safe for potential food industry applications.

3.2 Kinetic modelling of sago starch ozonation

The kinetic modeling of Sago starch ozonation demonstrates that the reaction follows pseudo-first-order kinetics with respect to starch concentration, with the reaction rate increasing as the pH rises. The ozonation mechanism involves hydroxyl radicals ([OH]*) interacting with native starch ([St]_native) to form carbonyl and subsequently carboxyl groups, as detailed in reactions (2) and (3). The overall ozonation can be simplified to reaction (4), where ozone continuously converts starch to carboxylated products. This conversion is modelled by a kinetic equation (6), which describes the formation rate of carboxyl groups as a function of starch and hydroxyl concentrations.

The assumption of constant ozone supply shifts the focus to the starch and hydroxyl concentrations, with the key parameter α representing the stoichiometric ratio of starch to hydroxyl ions. This equation transforms into an integrated form (12)-(13) to relate the conversion rate (X) to reaction time (t). Plotting ln(1-X) against ln t yields a linear relationship (Fig. 3), where the slope provides the value of parameter α, and the intercept gives the observed rate constant (k_obs). From the slope equation (18), the reaction order α is calculated using the slope of the linear plot. As shown in Table 1, α values close to 1 across the pH range imply a pseudo-first-order reaction concerning starch concentration. This consistency suggests that the ozonation reaction mechanism remains stable across different pH levels, with slight variations in the reaction order.

Close

Fig. 3.

Pseudo-first-order kinetics linearization of the Sago starch ozonation.

Export to PPT

The intercept provides the observed rate constant (k_obs), which is used to calculate the true reaction rate constant (k_r) via equation (22). The increase in k_r with pH highlights the enhanced reactivity of hydroxyl radicals at higher alkalinity, which aligns with the general understanding of ozonation reactions being more effective in basic environments. At a lower pH of 7.5, the value of α drops slightly to 0.99, with a smaller reaction rate constant k=8.00×10⁻⁴ min^-1. This result suggests a slower reaction at lower pH. As the pH increases, the reaction rate improves significantly, with k reaching 3.3538 x 10^-4min^-1 for 9.0. The increasing reaction rate with rising pH can be explained by the higher availability of hydroxyl ions (OH⁻), which are crucial for generating hydroxyl radicals (OH*) needed for oxidizing the starch. The improvement in R² values from 0.901 at pH 7.5 to 0.976 at pH 9.0 indicates a better fit of the kinetic model at higher pH levels, confirming that higher alkalinity enhances the ozonation process efficiency. Higher pH levels lead to rapid decomposition of ozone (O₃), forming highly oxidant radicals such as hydroxyl radicals (^●OH), which increase the complexity and efficiency of the ozonation process (Ávila-Sierra et al., 2023)

3.3 Pasting properties

The relationship between RVA measurements and sensory attributes, such as chewiness and mouthfeel (Table 2), underscores its role in ensuring consumer satisfaction. RVA helps in predicting the texture profile that will emerge in the noodle product. Native sago starch (pH 4) exhibits the highest peak viscosity (5344 cP), suggesting excellent swelling and water absorption. Ozonation significantly reduces peak viscosity at all pH levels, reaching the lowest value (4474 cP) at pH 8.5, likely due to molecular changes from oxidation. However, at pH values above 8.5, peak viscosity begins to recover, with 5139 cP recorded at pH 9.5, indicating partial restoration of water absorption capacity due to structural modifications (Wang et al., 2020).

Trough viscosity represents the starch paste’s stability under shear and heat. Native starch shows low trough viscosity (1520 cP), signifying granule disintegration after gelatinization. Ozonated starch demonstrates increased trough viscosity, peaking at 3109 cP at pH 9.5, suggesting enhanced stability and resistance to breakdown during gelatinization. The sharp rise at higher pH levels indicates improved structural integrity (Wani et al., 2012). Breakdown viscosity, the difference between peak and trough viscosity, reflects the mechanical stability of starch pastes (Wardhani and Cahyono, 2018)(49). Native starch displays high breakdown viscosity (3824 cP), highlighting poor stability. Ozonation reduces breakdown viscosity, reaching the lowest values (1964 cP at pH 8.5), demonstrating increased granule resilience. A slight increase at pH 9.5 (2030 cP) suggests partial recovery of swelling properties.

Final viscosity measures gel-forming capacity as the paste cools. Native starch has a relatively low final viscosity (2626 cP). In contrast, ozonated samples show higher final viscosities, peaking at 4937 cP at pH 9.5, indicating superior gel strength and suitability for applications requiring firm gels, such as noodle production (33). Setback viscosity, the difference between final and trough viscosities, quantifies retrogradation tendencies. Native starch has a setback viscosity of 1106 cP, while ozonated samples show increased values, peaking at 1828 cP at pH 9.5. This suggests enhanced retrogradation at higher pH levels, beneficial for firmer gels and improved texture (Xu et al., 2020)(50).

Peak time, indicating the time to reach peak viscosity, remains consistent across all samples (5.93-6.00 min), showing ozonation does not alter gelatinization rates. Pasting temperature, marking the onset of gelatinization, is slightly higher in ozonated starch (74.83-75.88°C) but shows minimal variation, indicating that ozonation has a limited impact on thermal properties while maintaining energy requirements for gelatinization (Yang et al., 2024; Zarski et al., 2021).

3.4 Morphological analysis

The SEM images of native sago starch (Fig. 4) and ozonated sago starch (Fig. 5) illustrate significant changes in the morphology of starch granules due to the ozonation process. In the native sago starch (Fig. 4), the granules exhibit a predominantly smooth and uniform surface, with distinct and well-defined shapes. Most granules are elliptical to oval, suggesting an intact and well-preserved granular structure, typical of native starches. In contrast, the ozonated sago starch (Fig. 5) displays clear signs of surface modification and partial degradation. The granules appear slightly rougher, with some showing minor surface erosion or fragmentation, particularly on the outer layers.

Close

Fig. 4.

SEM image of native sago starch (500× magnification)

Export to PPT

Close

Fig. 5.

SEM image of ozonated sago starch (500× magnification)

Export to PPT

The roughened surface observed can be linked to the oxidative cleavage of glycosidic bonds resulting from ozonation, which leads to chemical modifications on the granule surface. Despite these alterations, the overall shape of the granules remains largely intact, suggesting that ozonation primarily affects the surface of the starch granules rather than causing complete breakdown (Du et al., 2024).

3.5 Structural analysis of FTIR

FTIR spectroscopy was used to identify functional groups and verify structural modifications of the samples (Muwal et al., 2025; Shankar et al., 2018, 2017; Singh et al., 2018). The FTIR spectra depicted in Fig. 6 highlight significant differences in the functional group characteristics between native and ozonated sago starches. These differences provide insight into the structural modifications induced by the ozonation process (Singh et al., 2020). For the native Sago starch, the spectrum shows characteristic absorption peaks associated with common starch functional groups.

Close

Fig. 6.

FT-IR spectra of (a) native sago starch, (b) ozonated sago starch.

Export to PPT

O–H stretching was observed in the band range of 3200-3600 cm⁻¹, indicating the presence of hydroxyl groups (Shankar et al., 2020, 2014). This peak is slightly sharper in native starch than ozonated starch, indicating a more intact hydrogen bonding network in the native form. The peaks near 2900 cm⁻¹ can be attributed to C–H stretching vibrations, which are indicative of the polysaccharide backbone of starch molecules. In the case of ozonated sago starch, several notable changes are observed. The intensity of the O–H stretching band decreases, implying that ozonation disrupts the hydrogen bonds within the starch granules, possibly due to oxidative cleavage or chemical modification of hydroxyl groups. Additionally, the peak broadening in this region can suggest the formation of new functional groups, such as carboxyl or carbonyl groups, generated during the oxidative process.

The fingerprint region (900-1200 cm⁻¹) displays key differences between the native and ozonated starch. The absorption bands in this region, associated with C–O and C–C stretching, show a reduction in intensity in the ozonated starch, suggesting partial degradation or structural alteration of the starch molecules. This reduction is likely due to the breakdown of glycosidic linkages and other structural changes caused by the oxidative effect of ozone (Brodowska et al., 2018; Singh et al., 2019).

Additionally, a slight shift in the peaks around 1600 cm⁻¹, associated with water molecules and possibly carboxyl groups, suggests the potential formation of new oxidative degradation products in the ozonated starch. This can be associated with the introduction of carbonyl groups (C=O) as a result of the ozonation process, which is consistent with the known effects of ozone on polysaccharides (Pandiselvam et al., 2019).

3.6 Crystalline structure and thermal properties

The XRD patterns reveal shifts and intensity variations in the diffraction peaks of starch samples. Both native and ozonated sago starch exhibit characteristic peaks between 13° and 23° (2θ), reflecting their semicrystalline nature (Fig. 7). Notably, the ozonated starch shows higher peak intensities around 17° and 23°, suggesting an enhanced degree of crystallinity due to molecular rearrangements from oxidative treatment (Dome et al., 2020). The sharper, more intense peaks in the ozonated starch indicate a more ordered crystalline structure, likely resulting from selective degradation of amorphous regions during ozonation. In contrast, native Sago starch displays broader, less intense peaks, indicative of a lower crystallinity and a higher proportion of amorphous content (Yang et al., 2024). Structural changes induced by ozonation, including partial oxidation of amorphous regions, expose or reinforce crystalline domains, explaining the higher diffraction intensities observed in ozonated starch (Zhou et al., 2016). Minimal peak shifts further suggest subtle reorganization and tighter packing of crystallites within starch granules.

Close

Fig. 7.

The XRD patterns of (a) native sago starch and (b) ozonated sago starch.

Export to PPT

3.7 Cooking quality of noodles

In terms of cooking quality (Table 3), NS noodles showed the shortest cooking time (5 ± 0.74 min), moderate water absorption (87.18 ± 3.28%), and the highest cooking loss (20.47 ± 0.57%), with a color profile of 51.2/-0.5/2.7. OS noodles appeared darker, with slightly higher water absorption (88.61 ± 3.49%), longer cooking time (7 ± 0.31 min), and a marked reduction in cooking loss (16.98 ± 0.21%) compared to NS. Wheat-based noodles exhibited the brightest appearance, the longest cooking time (10 ± 0.56 minutes), and the highest water absorption (109.3 ± 3.62%), while also showing the lowest cooking loss (14.92 ± 0.22%). These findings suggest that ozonation and wheat flour substitution improved noodle cooking stability by reducing starch solubility during cooking. However, wheat flour provided a superior water absorption capacity and minimized cooking loss more effectively than ozonation and wheat flour substitution.

These observations can be attributed to structural changes induced by starch modification and the presence of gluten in wheat flour. The longer cooking time and higher water absorption observed in wheat noodles reflect the hydration and strong water-binding capacity of gluten, which also contributes to their superior elasticity and cohesiveness. Meanwhile, the decreased cooking loss in OS and W noodles indicates that ozonation and gluten create a more stable gel structure that resists dissolution during cooking (Qu et al., 2024).