© Wuhan University 2025

0 Introduction

Hollow capsules, a pharmaceutical dosage form based on gelatin or polysaccharides, improve the patient experience by masking the unpleasant taste and odor of medication, whether solid or liquid^[1-2]. Capsule shells are widely used and recognized in the medical field due to their high bioavailability and ability to improve drug targeting^[3-4]. The quality of the capsule shell is directly correlated with the effectiveness and safety of the capsule in clinical applications^[5-8].Currently, animal-derived gelatin is the most commonly used material for capsule shells^[9]. However, gelatin capsules are unsuitable for vegetarians or individuals with specific religious beliefs due to their animal origin^[10]. Furthermore, gelatin has a relatively low stability when exposed to temperature and humidity fluctuations^[11-12].

The growing demand for non-animal-based hollow capsules has driven increased research into plant-derived capsule materials^[13]. Hydroxypropyl starch (HPS), a plant-based material, offers an environmentally friendly alternative that contributes to reducing the pharmaceutical industry's carbon footprint^[14]. At a given temperature, the threshold of ambient relative humidity at which a substance starts to absorb moisture significantly is known as the Critical Relative Humidity (CRH). Capsules will absorb moisture and soften if the storage environment humidity is higher than the CRH of the capsule material. This can cause distortion, adhesion, or even rupture, which could compromise the accuracy of the dosage and the integrity of the sealing of encapsulated medications. Therefore, the stability of capsule storage depends critically on a greater CRH. Yang et al^[15] contrasted the gelatin, hydroxypropyl methyl cellulose (HPMC), and starch capsules' hygroscopicity, crystallinity, and thermal properties. They discovered that while starch capsules had a greater CRH than gelatin or HPMC capsules, they were less hygroscopic and had less capacity to store water. In another study, Chen et al^[16] developed hard capsules using mung bean starch (MBS) and cellulose nanocrystals (CNC). The addition of CNC significantly enhanced the tensile strength of MBS-based capsule shells. Similarly, Song et al^[17] improved the strain capacity and toughness of hydroxypropyl starch films by incorporating hydroxypropylated crosslinked potato starch. This enhancement was achieved without compromising the film's light transmittance, owing to the high structural compatibility between hydroxypropyl starch and cross-linked hydroxypropyl starch. Zhang et al^[18] methodically examined how the performance of starch-based empty capsules was impacted by the degree substitution (DS) of HPS replacement. The findings showed that the physicochemical characteristics and pharmacopoeia compliance of the capsules were significantly influenced by the regulation of DS. The capsules' elongation at break, hydroscopicity, and disintegration time all markedly increased as the substitution degree rose, but their breaking strength declined. The capsules demonstrated ideal comprehensive qualities (zero fragility, qualified tightness, and a disintegration period of 12 min and 38 s) when the DS reached 2.3%. Likewise, research on the impact of solid content, molecular weight, and process parameters on HPS hollow capsule performance metrics can also offer some direction for the controlled preparation of HPS capsules.

The process of etherifying natural starch to produce HPS retains its inherent biodegradability and biocompatibility while improving its stability and water solubility. HPS further supports environmental sustainability by lowering carbon emissions in pharmaceutical manufacturing^[19]. Research on hollow HPS capsules has primarily focused on optimizing the production process and formulation. Hydroxypropyl groups are introduced to the amylose macromolecular chain through an effective etherification reaction, leading to the synthesis of HPS. Nuclear magnetic resonance (NMR) carbon and hydrogen spectra have been employed to confirm successful synthesis and characterize the products^[20].

A range of HPS materials with varying DS and molecular weight (MW) has been produced by controlling key etherification process parameters, including reaction temperature, pH, reaction time, and raw material feed ratio^[21]. The MW of HPS was determined using viscosity methods, while the DS was measured using the gas chromatography (GC) method specified in the 2020 edition of the Pharmacopoeia. To achieve the controllable synthesis of HPS with the appropriate DS and MW, the impact of the reaction parameters was examined. This offers an experimental basis for investigating the connection between HPS and capsule performance in more detail. Response surface methodology (RSM) was applied to investigate the influence of solid content, MW, and dipping temperature, as well as their potential interaction effects on wall thickness, moisture absorption, and disintegration time.

1 Experimental

1.1 Reagents and Instruments

Reagents: High straight-chain corn starch (amylose content of 70%) (Thain Chemical Technology Shanghai Limited Company); κ-carrageenan, potassium acetate, and propylene oxide (Sarn Chemical Technology Shanghai Limited Company); hydriodic acid, iodinated isopropyl, n-octane-octylbenzene, and adipic acid (McLean Biochemistry & Technology Shanghai Limited Company); and hydroxypropyl starch (Erkang Pharmaceutical Hunan Limited Company).

Instruments: SH10A Moisture Rapid Tester (Shanghai Precision Instrument Limited Company), Ubbelohde Viscometer, Nicolet6700 Fourier Transform Infrared Spectrometer (NIGOLI Instruments, USA), SY-3D Four-in-One Tablet Disintegration Tester (Shanghai Huanghai Pharmaceutical Inspection Instrument Limited Company), and HD-3 Capsule Thickness Tester (Olabo Scientific Instrument Limited Company).

1.2 Controllable Synthesis of HPS

To prepare a starch suspension with a 25% (W/V) concentration, starch and purified water were combined in a round-bottom flask and thoroughly mixed. The pH of the suspension was adjusted using a 3.0% (W/V) NaOH solution. An appropriate amount of propylene oxide (PO) was added to the flask, which was then promptly sealed with cling film. The reaction was conducted in a thermostatically controlled water bath. At the end of the reaction, the pH of the system was reduced to 5.5 using a 6.0% (W/V) diluted sulfuric acid solution. The solvent was removed by centrifugation, and the resulting samples were washed with 80% aqueous ethanol, filtered, dried, and powdered. HPS and purified water were combined to prepare a 25% (W/W) HPS suspension in a round-bottom flask. The HPS was then degraded by adding a 6% (W/V) HCl solution. To produce HPS with varying DS, the reaction parameters, including the mass fraction of PO, reaction temperature, reaction pH, and reaction time, were systematically varied. Additionally, by varying the acid hydrolysis time and degradation temperature in a 6.0% (W/V) HCl solution, HPS samples with a variety of molar substitution (MS) values were produced.

1.3 Preparation of HPS Hollow Capsules

The required amounts of purified water, C₂H₃KO₂, carrageenan, and HPS were measured. HPS was first mixed with purified water using a stirring bar to form a uniform starch suspension. The mixture was then heated to 100 ℃ and stirred for 3 h under sealed conditions to gelatinize and dissolve the starch. Once thoroughly mixed, carrageenan and C₂H₃KO₂ were added and stirred until fully dissolved. The system temperature was gradually reduced to an optimum level and held constant for 24 h to allow defoaming. A gel solution was prepared, its temperature was maintained, and a metal mold pre-coated with release oil was dipped into the solution until its surface was uniformly coated. The mold was placed in a drying oven at a constant temperature of 35 ℃ for 3.5 h. After cooling, the capsules were trimmed, assembled, and demolded.

1.4 Testing and Characterization of HPS

Characterization: The synthesized products were characterized using a Bruker Tensor 27 spectrometer (Germany) to obtain amylose Fourier Transform Infrared (FT-IR) spectroscopic curves. The structures of amylose and HPS were further analyzed using ¹H NMR and ¹³C NMR spectroscopy with a Varian 500 spectrometer (USA).

DS test: A 20% coating concentration was applied to the packed column using a stationary solution consisting of 25% phenyl and 75% methyl polysiloxane. The chromatographic conditions included: the temperature was raised to 230 ℃ at a rate of 50 ℃ per min and sustained for 2 min after being maintained at an initial temperature of 100 ℃ for 8 min. 200 ℃ was the intake temperature and 250 ℃ was the detector temperature. A hydrogen flame ionization detector (FID) was employed. To determine the correction factor, five successive injection operations were performed after 1 mL of the control solution was injected into the gas chromatograph. Inject 2 μL each of the standard and sample solutions into the injection port of the gas chromatograph.^[22]. The DS was calculated using the following formula:

$K_{\mathrm{1}}={\zeta }_{\mathrm{1}}\times (Q_{\mathrm{1}}/A_{\mathrm{1}})\times A_{\mathrm{2}}\times (W_{\mathrm{1}}/W_{\mathrm{2}})$(1)

$K_{\mathrm{1}}=\mathrm{31}\times \frac{\mathrm{D}\mathrm{S}}{\mathrm{162}+\mathrm{14}\times \mathrm{D}\mathrm{S}}$(2)

where ζ₁ represents the ratio of the MW of hydroxypropyl to iodomethane, with a value of 0.535 2. Q₁ is the mass ratio of iodomethane to n-octane in the standard solution; A₁ is the peak area ratio of iodomethane to n-octane in the standard solution; A₂ is the peak area ratio of iodomethane to n-octane in the sample solution; W₁ is the mass (g) of n-octane in the 2 mL internal standard solution; and W₂ is the mass (g) of dry HPS in the sample solution.

Molecular weight test: The MW of HPS was determined by preparing a specific concentration of HPS solution using analytically pure dimethyl sulfoxide (DMSO) as the solvent. The retention time was measured using an Uhrig's viscometer in a water bath maintained at 25 ℃. The solution was diluted fourfold, and the retention duration was calculated for both the solution and pure DMSO^[22]. The intrinsic viscosity was computed using the Mark-Hovink equation:

$[\eta ]=K{M}^{\alpha }$(3)

where [η] is the intrinsic viscosity, m is the viscosity-average molecular weight, K=0.364×10^-5 L/g, and α=0.82.

Water content test: The capsules were broken into pieces and evenly distributed in a pre-weighed bottle. After drying in an oven at 100 ℃ for 5 h, the samples were sealed, transferred to a desiccator for 30 min, and reweighed to determine moisture content^[23].

Friability test: The HPS capsules were conditioned for 24 h at 25 ℃ in a desiccator containing saturated magnesium nitrate. Each capsule was then placed in a glass tube on a wooden surface. A cylindrical baffle was dropped from the top of the glass tube, and the number of capsules that burst was recorded^[24].

Swelling kinetics test: HPS capsules with a wall thickness of 0.14±0.005 mm were weighed and submerged in 200 mL of deionized water at 37.5 ℃. The capsules were periodically removed, wiped with filter paper to remove any remaining deionized water, and reweighed to track mass changes over time^[25].

Hygroscopicity test: 1 g of HPS capsules was placed in a Petri dish and left in a desiccator for 12 h. The dish was weighed, then stored at 25 ℃ in a drug stability tester, and reweighed periodically^[26].

Wall thickness test: The wall thickness of six capsule samples from the same batch was measured at their centers, using a high-precision capsule wall-thickness measuring device. Each capsule was tested four times to ensure that the data were reliable, and the arithmetic mean of the outcomes from each test was determined^[27].

Disintegration time test: HPS capsules filled with talc powder were placed in six glass tubes of a rotating basket system. A fully automated disintegration tester with a baffle plate was used to assess disintegration in three media at a consistent temperature of 37±0.5 ℃ ^[28].

1.5 Statistical Analyses

RSM was performed to investigate the effects of three independent variables on the response factor. The three variables considered in this study were solid content (A), MW (B) and dipping temperature (C). The low, central and high levels of each variable are designed as -1, 0 and +1, respectively, as shown in Table 1. Moisture absorption (T₁), wall thickness (T₂) and disintegration time limit (T₃) were chosen as the response parameters. By using Design Expert software, the experimental data were fitted to a second-order polynomial model.

2 Results and Discussion

2.1 Controllable Preparation of HPS with Different Molecular Structure Parameters

2.1.1 Preparation of HPS with different DS

HPS was synthesized with varying DS by modifying the mass percentage of PO, reaction temperature, pH, and time. The experimental conditions and corresponding DS values are summarized in Table 2.

The results from Table 2 and Fig. 1 indicate that the reaction temperature, pH, and PO mass fraction had a significant impact on the DS of HPS. As the mass fraction of PO increased, the DS initially rose before declining. At a constant pH and reaction temperature, a higher PO concentration allowed for increased hydroxyl group availability, facilitating continuous hydroxypropylation. However, as PO content increased, the gelatinization temperature of starch gradually decreased due to the higher DS. Experimental observations revealed that when the PO mass fraction exceeded 9.0%, partial or complete gelatinization occurred. This gelatinization weakened the ether bonds formed during the reaction, significantly reducing the DS. The reaction time had a limited effect on the DS. Extending the reaction time to 16 h did not result in substantial changes in DS because the etherification process reached equilibrium within approximately 16 h. The DS exhibited an initial increase followed by a decrease as the reaction temperature rose. At lower temperatures, the restricted swelling of starch macromolecules hindered NaOH from penetrating the crystalline areas of starch, reducing the availability of reactive hydroxyl groups and negatively impacting the etherification process. At higher temperatures, although starch molecules could expand and interact with NaOH, gelatinization of starch occurred, which reduced the DS due to the instability of the formed ether bonds. NaOH played a dual role in the reaction, acting as both a granule wetting agent to facilitate starch granule expansion and as a catalyst to accelerate the etherification reaction. When the pH reached 10, hydroxide ions (OH^-) reacted with free hydroxyl groups in starch to form highly reactive intermediates, which subsequently reacted with PO to increase DS. However, at the pH levels above 10, excessive swelling of starch granules led to uneven mixing and incomplete reactions. Over-swelling also caused gelatinization, resulting in irregular reactions and a reduction in the DS of the final product. The aforementioned reaction parameters were modulated to produce a variety of HSP products with different DS. In line with general requirements for hydroxypropyl concentration in relevant research, HPS with a DS of 2.3 was selected as the focus of further experimental studies.

Fig. 1 Influence of propylene oxide (PO) concentration (a), pH (b), reaction temperature (c), reaction time (d), on DS

2.1.2 Preparation of HPS with different MW

Herein, the MW of HPS was characterized by viscosity-average molecular weight ($M_{\eta }$). When HCl was used as the degrading agent, the $M_{\eta }$ of HPS decreased significantly during the first three to seven hours of degradation before stabilizing, as shown in Fig. 2(a). This behavior can be attributed to the initially high concentration of HPS and the rapid interaction of hydrochloric acid with the glycosidic linkages in the HPS molecular chain. This interaction led to a rapid decrease in the degree of polymerization. After six hours, the $M_{\eta }$ exhibited minimal changes as the HPS concentration steadily decreased and the degradation reaction approached completion. As illustrated in Fig. 2(b), increasing the degradation temperature resulted in enhanced molecular activity within the system, thereby accelerating the breakdown of $M_{\eta }$. Consequently, the $M_{\eta }$ of HPS decreased with increasing degradation temperature.

Fig. 2 Effect of degradation temperature (a) and degradation time (b) on molecular weights

2.2 Characterization of HPS

2.2.1 ¹H-NMR spectrum

The ¹H-NMR hydrogen spectrum of HPS is shown in Fig. 3. A solvent peak at 3.5 was observed in deuterated aqueous DMSO. The signal peaks corresponding to individual hydrogen atoms on the glucose units of the starch backbone were detected between 3.2 and 5.6. Hydrogen atoms at positions C₁, C_3/6/5, C₂, and C₄ of the glucose unit appeared at 4.66, 3.77, 3.55, and 2.48, respectively. The hydroxyl groups at the C₂, C₃, and C₆ positions at the amyloglucose unit exhibited signal peaks at 5.46, 5.40, and 5.11, respectively. A proton peak near 1.1, corresponding to the methyl group within the hydroxypropyl moiety at the C₆ position, confirmed that hydroxypropylation of the hydroxyl group successfully synthesized HPS.

Fig. 3 1H NMR hydrogen spectrum of HPS

2.2.2 ¹³C-NMR spectroscopy

The ¹³C-NMR spectrum of HPS is shown in Fig. 4. Deuterated aqueous DMSO exhibits a solvent peak at 40. The anomeric carbon C₁ exhibited a signal peak at 103, while carbon atoms at positions C₂-C₆ of the glucose ring showed signal peaks between 55 and 75. The methyl group of the hydroxypropyl moiety at C₆ exhibited a signal peak at 20.5, confirming the successful hydroxypropylation of starch to form HPS.

Fig. 4 13C NMR carbon spectrum of HPS

2.2.3 Fourier transform infrared spectroscopy

The Fourier transform infrared (FTIR) spectrum is shown in Fig. 5. A peak corresponding to O-H stretching vibrations was observed at 3 406 cm^-1. The stretching absorption peaks of the methylene group's (-CH₂-) were located between 2 906 and 2 934 cm^-1, while that of the methyl group (-CH₃) appeared at 2 941 cm^-1. Peaks corresponding to C-O and C-O-C bonds in primary and secondary hydroxyl groups were identified in the range of 1 067-1 136 cm^-1. The substitution of propylene oxide onto starch resulted in the formation of St-O-CH₂-CH(OH)-CH₃, which altered the in-plane bending vibrations of the C-O-H bond. In comparison to starch, HPS showed smaller absorption peaks in the 1 250-1 500 cm^-1 region, indicating successful hydroxypropylation.

Fig. 5 Fourier transform infrared spectrum of HPS

2.3 HPS Capsule Performance Test

2.3.1 Water content of HPS capsules

Under the experimental conditions of this study, the moisture content of HPS capsules was maintained below 9.0%, as shown in Table 3, meeting the stringent requirements outlined in the Chinese Pharmacopoeia (2020 edition). Notably, the moisture content of HPS capsules was significantly lower than that of gelatin capsules, suggesting their potential suitability for encapsulating moisture-sensitive medications.

2.3.2 Friability of different capsules

As demonstrated in Table 4, HPS capsules of all three MW satisfied the friability requirements as specified in the Pharmacopoeia of the People's Republic of China. Furthermore, the hardness and integrity of these capsules met the necessary standards, ensuring compliance with the quality requirements for drug products.

2.3.3 Swelling kinetics of HPS hollow capsules

The swelling kinetics of HPS capsules with different MW (fixed degree of substitution DS=0.16) are illustrated in Fig. 6. Capsules with lower MW achieved dissolution equilibrium more rapidly in the same solvent. This can be attributed to the larger mesh size of the polymer network, which facilitates the diffusion of water molecules into the capsule shell. The reduced chain length and fewer entanglement points between molecular chains in lower MW capsules contribute to their higher dissolution rate. As a result, HPS capsules with lower MW exhibit shorter disintegration times, softening and cleaving more quickly in aqueous solutions.

Fig. 6 Dissolution kinetics curves of HPS capsules

2.4 Results of the Box-Burman Test

Based on the factor levels listed in Table 1, a Box-Burman experimental design was conducted using Design-Expert software, with the results summarized in Table 5.

2.4.1 Analysis of factors affecting moisture absorption

Figure 7(a) and (b) show that increasing the solid content significantly reduces the moisture absorption rate of HPS capsules. This is likely due to the higher solid content thickening the capsule wall, which reduces the surface area available for water absorption and hinders water molecules from penetrating the polymer network. Additionally, as shown in Fig. 7(a) and (c), increasing the molecular weight of HPS further limits moisture absorption by expanding the polymer network mesh size and restricting the entry of water molecules into the capsule material. The moisture absorption process can be described by the differential equation dW/dt=KA(P_A-P), where K is the hygroscopicity rate constant, A is the surface area of the solid formulation, P_A is the partial pressure of water vapor in the atmosphere, and P is the internal vapor pressure produced when the hydrophilic component absorbs water to saturation. As shown in Fig. 7(b) and (c), dipping temperature had only a slight effect on the moisture absorption rate. Table 6 displays the results of the analysis of moisture absorption rates under varying factor levels using the response surface model.

Fig. 7 Response surface plots of the effects of solid content and MW (a), solid content and dipping temperature (b), and MW and dipping temperature (c) on the hygroscopicity of the capsule

Table 6 displays that according to the results of fitting the capsule wall thickness response surface model, there is a relationship between capsule moisture absorption T₁, solid content (A), MW (B), and dipping temperature (C).

$\begin{array}{l}{T}_{\mathrm{1}}=\mathrm{9.82}-\mathrm{1.05}A-\mathrm{0.50}B+\mathrm{0.15}C+\mathrm{0.51}AB+\mathrm{0.19}AC\\ +\mathrm{0.5}BC-\mathrm{1.13}{A}^{\mathrm{2}}-\mathrm{0.99}{B}^{\mathrm{2}}-\mathrm{0.69}{C}^{\mathrm{2}}\end{array}$(4)

The variance contribution (reflected by the F-value in Table 6) indicates the extent to which each component affects the moisture absorption rate. The impact of a factor on the moisture absorption rate increases as the F-value (variance contribution) increases. Solid content > MW > dipping temperature is the order in which each factor influences the moisture absorption rate of the HPS capsules, as indicated in Table 6.

According to the coefficient of determination of the fitted equation (R²=0.880 6), the model can explain 88.06% of the variation in moisture absorption rate; accordingly, only 11.94% (1-R²=1-0.880 6) of the variation in moisture absorption could not be explained by the model. Additionally, the adjusted coefficient of determination ($R_{\mathrm{a}\mathrm{d}\mathrm{j}}^{\mathrm{2}}$=0.876 9) was close to R², further demonstrating the significance of the model. Furthermore, the model was well fitted, and there was a significant correlation between the independent variables and the response values, as evidenced by the difference between the predicted coefficient of determination ($R_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}}^{\mathrm{2}}$=0.846 4) and $R_{\mathrm{a}\mathrm{d}\mathrm{j}}^{\mathrm{2}}$ being less than 0.2.

2.4.2 Analysis of factors affecting wall thickness

Figure 8(a) and (b) illustrate that wall thickness increased with higher solid content due to an increase in polymer interchain aggregates. This reduced the average distance between polymer chains and increased interchain interaction forces, observed macroscopically as a rise in solution viscosity. Consequently, more solution adhered to the mold surface, thickening the capsule walls. Figure 8(a) and (c) show that MW also positively correlated with wall thickness. Polymers with higher MW exhibited longer molecular chains, which increased entanglement during film formation, creating a denser and more stable network structure. This enhanced the mechanical strength and thickness of the capsules. In contrast, as the dipping temperature increased, the thermal motion of HPS polymer chains intensified, reducing friction between molecular segments and lowering solution viscosity. As a result, Fig. 8(b) and (c) demonstrate that capsule wall thickness tended to decrease with higher dipping temperatures. Table 7 displays the results of a response surface model fitting investigation of the wall thickness of the HPS capsules at various factor levels.

Fig. 8 Response surface plots of the effects of solid content and MW (a), solid content and dipping temperature (b), MW and dipping temperature (c) on capsule wall thickness

Table 7 shows the relationship between the capsule wall thickness (T₂), solid content (A), MW (B), and dipping temperature (C) based on the findings of fitting the capsule wall thickness response surface model.

$\begin{array}{l}{T}_{\mathrm{2}}=\mathrm{0.08}+\mathrm{0.01}A+\mathrm{0.007}B-\mathrm{0.006}C-\mathrm{0.002}AB\\ +\mathrm{0.01}AC+\mathrm{0.001}BC+\mathrm{0.009}{A}^{\mathrm{2}}+\mathrm{0.013}{B}^{\mathrm{2}}+\mathrm{0.01}{C}^{\mathrm{2}}\end{array}$(5)

The variance shows the influence of each component on wall thickness. The impact of this factor on the rate of moisture absorption increased with the variance. The degree to which each element influences the moisture absorption rate of HPS capsules is prioritized as follows, as indicated in Table 7: solid content > MW > dipping temperature.

According to the coefficient of determination of the fitted equation (R²=0.883 7), only 11.63% of the variation in capsule wall thickness could not be explained by the model. Additionally, the adjusted coefficient of determination ($R_{\mathrm{a}\mathrm{d}\mathrm{j}}^{\mathrm{2}}$=0.851 2) was close to R², further demonstrating the significance of the model. Furthermore, the model was well fitted, and there was a significant correlation between the independent variables and the response values, as evidenced by the difference between the predicted coefficient of determination ($R_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}}^{\mathrm{2}}$=0.802 5) and $R_{\mathrm{a}\mathrm{d}\mathrm{j}}^{\mathrm{2}}$ being less than 0.2.

2.4.3 Analysis of factors influencing the disintegration time

As illustrated in Fig. 9(a) and (b), an increase in solid content resulted in a more compact capsule structure, slowing water molecule penetration and extending the time required for the HPS capsule to transition from a glassy to a rubbery state. Similarly, Fig. 9(a) and (c) demonstrate that the disintegration time of the HPS capsules significantly increased with increasing MW. This is attributed to the formation of a greater number of crystalline regions in the swelling polymer, composed of HPS and water molecules, as the MW increases. These crystalline regions are more cohesive and less accessible to water molecules compared to amorphous regions, making water to infiltration and disintegration more difficult.

Fig. 9 Response surface plots of the effects of solid content and MW (a), solid content and dipping temperature (b), and MW and dipping temperature (c) on the disintegration time of the capsule

Table 8 presents the relationship between disintegration time (T₃), solid content (A), MW (B), and dipping temperature (C), as determined by the response surface model. The variance analysis highlights the relative influence of each factor, with the moisture absorption rate being influenced in the following order: solid content > MW > dipping temperature.

$\begin{array}{l}{T}_{\mathrm{3}}=\mathrm{331.2}+\mathrm{105.8}A+\mathrm{73.25}B-\mathrm{3.3}C-\mathrm{29.75}AB\\ -\mathrm{24.5}AC+\mathrm{28.25}BC+\mathrm{68.9}{A}^{\mathrm{2}}+\mathrm{102.15}{B}^{\mathrm{2}}+\mathrm{88.4}{C}^{\mathrm{2}}.\end{array}$(6)

The coefficient of determination for the fitted equation (R²=0.929 8) indicates that only 7.02% of the variability in disintegration time specificity remains unexplained by the model. The adjusted coefficient of determination ($R_{\mathrm{a}\mathrm{d}\mathrm{j}}^{\mathrm{2}}$=0.899 5) closely aligns with R², further supporting the model's reliability. Moreover, the predicted coefficient of determination ($R_{\mathrm{p}\mathrm{r}\mathrm{e}\mathrm{d}}^{\mathrm{2}}$=0.872 7) is within 0.2 of $R_{\mathrm{a}\mathrm{d}\mathrm{j}}^{\mathrm{2}}$, indicating a strong association between the independent variables and response values.

2.4.4 Regression model prediction

The guidelines for wall thickness in "Specifications, Dimensions, and Appearance Quality of Hollow Capsules" state that capsule walls must be kept within the 0.09-0.12 mm window. Capsules' compressive strength and resistance to deformation are directly impacted by wall thickness. In addition to reducing production efficiency or compromising disintegration performance, excessively thick walls may cause capsule rupture during filling or shipment. Consequently, wall thickness must be kept within the standard range's lower bound. High moisture absorption rate capsules can soften, adhere, or distort in humid settings, which can lead to caking or content deterioration. Therefore, it is necessary to reduce the rates of moisture absorption. Drug action onset and bioavailability are directly impacted by disintegration time, which is the amount of time needed for a capsule to completely break in vivo or in vitro simulated fluids. As a result, disintegration time restrictions must also be tightly regulated at the lowest level that is acceptable.

According to the regression model, the optimal preparation conditions for HPS capsules are as follows: a solid content of 17.819%, an MW of 77.125 kDa, and a dipping temperature of 56.932 ℃. Under these conditions, the capsules achieved a wall thickness of 0.092 mm, a moisture absorption of 9.744%, and a disintegration time limit of 349.832 s. To simplify the verification process, the following optimized parameters were selected: a solid content of 18%, an MW of 77 kDa, and a dipping temperature of 57 ℃. Capsules prepared using these parameters exhibited a wall thickness of 0.086 mm, a moisture absorption of 8.594%, and a disintegration time of 328 s. After conducting three consecutive validation tests, all performance metrics met the standards of the Chinese Pharmacopoeia, with a relative error of less than 5%.

3 Conclusion

This study provides a comprehensive discussion of the process by which various reaction conditions influence the MW and DS of HPS. The starch could not be pasteurized and was completely etherified when the PO mass fraction was 9%, the reaction temperature was 45 ℃, the reaction pH was 10, and the reaction time was 16 h. HPS capsules were successfully prepared using cold-gel technology and examined for several characteristics, all of which met the standards outlined in the Chinese Pharmacopoeia. The RSM experimental design proved effective in evaluating the effects of operative variables–solid content, MW, and dipping temperature–and their interactions on capsule wall thickness, moisture absorption, and disintegration time. The final optimized process parameters were: 18% solid content, 77 kDa MW, and a dipping temperature of 57 ℃. Under these conditions, the HPS capsules exhibited a disintegration time of 328 s, a wall thickness of 0.086 mm, and a moisture absorption rate of 8.594%. These data offer a theoretical foundation and practical suggestions for the production of HPS capsules with exact moisture absorption rates, wall thicknesses, and disintegration periods. Future studies should explore the long-term stability, cost-efficiency of large-scale production, and the applicability of this methodology to other biopolymer-based capsule systems.

References

All Tables

All Figures

	Fig. 1 Influence of propylene oxide (PO) concentration (a), pH (b), reaction temperature (c), reaction time (d), on DS
In the text

	Fig. 2 Effect of degradation temperature (a) and degradation time (b) on molecular weights
In the text

	Fig. 3 1H NMR hydrogen spectrum of HPS
In the text

	Fig. 4 13C NMR carbon spectrum of HPS
In the text

	Fig. 5 Fourier transform infrared spectrum of HPS
In the text

	Fig. 6 Dissolution kinetics curves of HPS capsules
In the text

	Fig. 7 Response surface plots of the effects of solid content and MW (a), solid content and dipping temperature (b), and MW and dipping temperature (c) on the hygroscopicity of the capsule
In the text

	Fig. 8 Response surface plots of the effects of solid content and MW (a), solid content and dipping temperature (b), MW and dipping temperature (c) on capsule wall thickness
In the text

	Fig. 9 Response surface plots of the effects of solid content and MW (a), solid content and dipping temperature (b), and MW and dipping temperature (c) on the disintegration time of the capsule
In the text