Encapsulation of astaxanthin by citric acid corn starch ester nanoparticle‐based Pickering emulsion: Effect of amylose content

Abstract

BACKGROUND

The effects of amylose content on citric acid starch ester nanoparticles through nanoprecipitation as Pickering emulsion stabilizers remain unknown. In this study, three types of corn starch with different amylose content, namely waxy corn starch (WCS), normal corn starch (NCS), and high‐amylose corn starch (HACS), were used to prepare starch ester nanoparticles through nanoprecipitation. This research systematically studied corn starch ester nanoparticles with regards to the structural and interfacial properties at the oil–water interface and application to the delivery of astaxanthin in vitro.

RESULTS

Corn starch ester nanoparticles exhibited the V‐type crystalline pattern compared to the native starches showing either A‐type (NCS and WCS) or B‐type (HACS) patterns. Corn starch ester nanoparticles exhibited enhanced contact angle compared to those of native starches, with the order of waxy corn starch ester nanoparticle (WCSENP) > high amylose corn starch ester nanoparticle (HACSENP) > normal corn starch ester nanoparticle (NCSENP) (P ≤ 0.05). Starch ester nanoparticle‐based Pickering emulsion (PE) showed increased storage stability at 4 and 25 °C, in contrast to those stabilized by native starches. Starch ester nanoparticle‐based PE exhibited higher retention rate and bioaccessibility of astaxanthin (AST) than those of soybean oil (P ≤ 0.05). The AST retention curve of PE was fitted to the first‐order kinetic model. The AST retention rate of WCSENP‐PE remained above 70% after 8 days of storage at 25 °C, followed by HACSENP‐PE and NCSENP‐PE.

CONCLUSION

This study provided a new strategy of producing PE stabilizers with improved properties, stability, and bioacessibility of bioactive compound in vitro. © 2026 Society of Chemical Industry.

INTRODUCTION

Starch, well‐known for its abundance, biodegradability, and low cost and being generally recognized as safe (GRAS), has been widely used to stabilize Pickering emulsions (PEs). ¹ However, native starch possesses strong hydrophilicity and poor hydrophobicity due to the presence of hydroxyl groups on the molecular chains, showing poor wettability at the oil–water interface in emulsion systems, consequently leading to poor stability. Hence, native starch requires modification for stabilizing emulsions. Esterification can be used for the chemical modification of starch, in which acids, fatty acids, anhydrides, and their derivatives react with hydroxyl groups on glucose and form ester groups.2, 3, 4 Previously, extensive studies have been focused on octenyl succinic anhydride (OSA)–starch for stabilizing PEs due to the excellent physical and storage stability.5, 6 The utilization of OSA at over 30 g kg⁻¹ generates food safety concerns, ⁷ while OSA‐modified starch at or below 30 g kg⁻¹ cannot stabilize PEs well. ⁸ Thereby, an alterative natural and safe modification method for starch esterification is urgently needed.

Citric acid, naturally present in fruits, is known to be of low cost, nontoxic, and GRAS. Citric acid starch esters have been shown to have enhanced hydrophobicity and emulsifying properties,9, 10 showing certain potential. However, citric acid starch ester‐based PEs have been shown to exhibit large droplet sizes, poor creaming index, and phase separation, indicating poor stability.2, 9 Esterification can increase starch granule size, which may adversely affect emulsion stability. ¹¹ In contrast, reduced starch particle size is known to be positively correlated with the stability of PEs. Among the available approaches for nanoparticle fabrication, nanoprecipitation is widely recognized for precisely regulating the size and surface characteristics of starch nanoparticles. ¹² Nevertheless, the synergistic combination of citric acid esterification with nanoprecipitation for preparing citric acid starch ester nanoparticles as PE stabilizers has not been fully explored. In particular, the interfacial stabilization mechanisms and emulsion performance of citric acid starch ester nanoparticle‐based PEs remain unknown.

Various types and sources of starch have been used to stabilize PEs, such as buckwheat starch, ¹³ corn starch, potato starch, pea starch, ¹⁴ and others. However, the structural attributes of starch, particularly its crystalline architecture and amylose/amylopectin composition, significantly influence the hydrophobic modification. ¹⁵ As the world's highest‐yielding cereal crop, corn contributes to 85% of starch supply globally. ¹⁶ Differentiated in amylose content, three types of corn starch are well known, namely waxy corn starch (WCS), normal corn starch (NCS), and high‐amylose corn starch (HACS). To date, few studies have concentrated on the effect of corn starch with distinct starch amylose content on citric acid starch ester nanoparticle‐based PEs. The correlation between the amylose level in corn starch and the properties of PEs stabilized by citric acid starch ester nanoparticles has yet to be elucidated.

Given this, WCS, NCS, and HACS were firstly esterified using citric acid to prepare citric acid corn starch esters. The starch esters were further processed as nanoparticles to stabilize PE. The correlation between the starches with different amylose content and characterizations of corn starch ester nanoparticle‐based PE was investigated. The research reported here systematically studied corn starch ester nanoparticles with regards to the structural and interfacial properties at the oil–water interface. The emulsifying characteristics of corn starch ester nanoparticle‐based PE encompass the emulsifying activity index, emulsifying stability index, creaming index, and storage stability. Furthermore, astaxanthin (AST)‐containing PE was prepared and evaluated for controlled release and bioaccessibility. The study aimed to investigate the effect of corn starches with distinct amylose content on starch ester nanoparticle‐based PEs, and their applications in the delivery of functional agents.

MATERIALS AND METHODS

Materials

WCS (~50 g kg⁻¹ amylose content) was purchased from Gansu Kunlun Biochemical Co. Ltd (Gansu, China). NCS (~230 g kg⁻¹ amylose content) and α‐amylase were purchased from Shanghai Yuanye Bio‐Technology Co. Ltd (Shanghai, China). HACS (~730 g kg⁻¹amylose content) was purchased from Quan Yin Xiang Yu Biotechnology Co. Ltd (Beijing, China). Soybean oil was purchased from Yihai Kerry Arawana Holdings Co. Ltd (Shanghai, China). Citric acid, phenolphthalein, pepsin (from porcine gastric mucosa), lipase (from the porcine pancreas), AST, porcine bile salt, pancreatin, Nile Blue A, and Nile Red were purchased from Shanghai Maclean Biochemical Technology Co. Ltd (Shanghai, China). Sodium hydroxide, hydrochloric acid, sodium chloride, sodium acetate, phenol, sulfuric acid, and ethanol were purchased from Sinopharm Chemical Reagent Co. Ltd (Shanghai, China). Deuterium oxide (D₂O) was obtained from Shanghai Aladdin Biochemical Technology Co. Ltd (Shanghai, China).

Preparation of corn starch ester nanoparticles

Corn starch esterification was performed according to a previously reported method. ¹⁷ The corn starch ester nanoparticles were prepared following a previously reported method, with modifications. ¹⁸ A 50 g L⁻¹ corn starch ester suspension was preheated in NaOH solution (0.1 mol L⁻¹) at 80 °C and stirred for 30 min. After cooling to room temperature, the suspension was subjected to ultrasonication at 25 °C using an ultrasonic processor (VCX750, Shanghai Lueshen Instrument Equipment Co. Ltd). The treatment was performed in an ice bath at an output power of 600 W with 10 s pulse intervals for a total duration of 10 min. The solution was precipitated by the addition of absolute ethanol (1:2) with stirring. After centrifugation at 3220 × g for 10 min, the precipitate was collected and freeze‐dried. Starch ester nanoparticles formulated with WCS, NCS, and HACS were designated as WCSENP, NCSENP, and HACSENP, respectively.

Characterization of corn starch ester nanoparticles

Degree of substitution (DS) and ¹H NMR spectroscopy

The DS of corn starch ester nanoparticles was determined according to a previously reported method. ² One gram (dry weight) of sample was dispersed in 100 mL of deionized water with phenolphthalein indicator. After initial titration with 0.1 mol L⁻¹ NaOH to the endpoint (red color), 10 mL of 0.5 mol L⁻¹ NaOH was added and stirred for 40 min. The solution was titrated with 0.5 mol L⁻¹ HCl to colorless endpoint. The volume of hydrochloric acid was measured with native starch as the control. The mass fraction of citric acid substituents (A, g kg⁻¹) was calculated according to Eqn (1):

$$A\left(\%\right)=\frac{\left(V_0-V_1\right)\times M\times 0.158}{W}\times 100$$ (1)

where $V_0$ and $V_1$ (mL) represent the volume of HCl consumed for the blank and esterified starch, respectively; $M$ (mol L⁻¹) is the molarity of HCl; and $W$ (g) is the dry weight of starch.

The DS value was subsequently calculated using Eqn (2):

$$\mathrm{DS}=\frac{162A}{15800-156A}$$ (2)

where 162 corresponds to the molecular weight of an anhydroglucose unit in starch and 156 represents the molecular weight of the introduced citric acid substituent.

Sample preparation for ¹H NMR spectroscopy was based on a previously reported method, with modifications. ¹⁰ A total of 2 g of starch was suspended in 30 mL of deionized water, followed by the addition of 10 mL of α‐amylase and 0.3 g of sodium acetate. The reaction was carried out at 85 °C for 2 h under continuous agitation, after which the enzyme was deactivated by boiling the mixture for 30 min. After cooling and centrifugation at 805 × g for 10 min, the supernatant was freeze‐dried. The hydrolysate (0.2 g) was dissolved in 3 mL of D₂O, allowed to stand for 2 h at room temperature, and re‐lyophilized. Finally, 0.1 g of D₂O‐exchanged starch was dissolved in 1.0 mL of D₂O. ¹H NMR spectra were recorded using a Bruker AVANCE III 400 MHz spectrometer (Bruker BioSpin, Rheinstetten, Germany).

Fourier transform infrared (FTIR) spectroscopy

The short‐order structures of corn starches and corn starch ester nanoparticles were analyzed using a FTIR spectrometer (Nicolet iS10, Thermo Fisher Scientific, USA), with spectra scanned from 400 to 4000 cm⁻¹ at a resolution of 4 cm⁻¹.

X‐ray diffraction (XRD)

The crystalline properties of corn starches and corn starch ester nanoparticles were analyzed using an X‐ray diffractometer (Dmax2500, Rigaku, Japan). The instrument was operated at 40 kV and 40 mA with Cu Kα‐filtered radiation (λ = 0.154 nm). Diffraction patterns were recorded across 2θ ranging from 4° to 40° at a scanning rate of 4° min⁻¹.

Scanning electron microscopy (SEM)

Surface morphology of native corn starches and corn starch ester nanoparticles was characterized by scanning electron microscopy (SEM; SU5000, Hitachi, Japan). The dry powder underwent a gold spraying process and was examined under an accelerating voltage of 5 kV. Images were acquired at magnifications of ×1000 and ×10 000.

Water contact angle

A contact angle meter (JY‐82B Kruss DSA; Hamburg, Germany) was employed to measure the water contact angle of native corn starches and corn starch ester nanoparticles. Each sample was compressed at 30 MPa for 2 min to create a tablet with a diameter of 15 mm and an approximate thickness of 1–2 mm. A 2 mL droplet of deionized water was subsequently added to the surface and photographs were taken.

Preparation of PE and microscopic observation and creaming index (CI) of PE during storage

PE was made according to a previously reported approach with modifications. ³ Soybean oil and corn starch ester nanoparticle dispersion (40 g L⁻¹) were combined at an oil volume fraction (φ) of 0.5, corresponding to a 1:1 (v/v) ratio between the oil and aqueous phases. The mixture was homogenized at 20 000 rpm for 2 min to obtain the PE. Samples of PE were stored at 4 and 25 °C, and photographs were taken on days 0, 1, 7, 15, and 30. PE was poured onto a microscope slide, covered with glass coverslip, and viewed under a light microscope at a magnification of ×40.

The PE was placed in glass bottle and allowed to stand. The CI was employed to indicate the stability of PE over storage durations of 1, 7, 15, and 30 days. The CI was computed using the following equation:

$$\mathrm{CI}\left(\%\right)=\frac{H_{\mathrm{c}}}{H_{\mathrm{t}}}\times 100$$ (3)

where H _c denotes the height of the emulsified layer, while H _t indicates the overall height of the emulsion.

Characterization of PE

Emulsifying activity index (EAI), emulsifying stability index (ESI), droplet size distribution, and zeta potential

EAI and ESI were measured according to a previously reported method. ¹⁹ The droplet size distribution of the PE was measured using a Mastersizer 3000 (Malvern, Worcestershire, UK). The obscuration level was maintained within 10–20%. The refractive indices were set at 1.33 for the dispersing medium and 1.47 for soybean oil. The Mastersizer was employed to ascertain the zeta potential of PE. The measurement was conducted at the equilibration time of 60 s and equilibration temperature at 25 °C.

Interfacial tension and starch adsorption rate (%) at the interface of PE

The interfacial tension of the corn starch ester nanoparticles at the oil–water interface was measured using a contact angle meter (SL250, Kino, USA). A droplet (10 μL) of the dispersion was dispensed into the oil and allowed to stabilize for 100 s. The interfacial tension was calculated using OneAttension software.

PE was centrifuged at 3580 × g for 30 min and the top layer was discarded. Starch concentration in the middle supernatant and bottom precipitate was determined via the phenol–sulfuric acid method. ²⁰ The supernatant and precipitate were diluted 30‐fold and 200‐fold, and mixed with 1 mL of water and 60 g L⁻¹ of phenol, respectively, and vortexed for 30 s. An amount of 5 mL of concentrated sulfuric acid was added to the supernatant and precipitate and vortexed. The reaction mixture was incubated for 30 min and subsequently measured for absorbance at 490 nm. The starch adsorption rate at the interface was calculated using the following equation:

$$\text{Interfacial starch adsorption rate}\left(\%\right)=\frac{C_0-C_{\mathrm{s}}}{C_0}\times 100$$ (4)

where C ₀ is the initial starch concentration for the preparation of the emulsion and C _s is the starch concentration in the clear liquid of the middle layer plus the precipitated layer at the bottom after the emulsion had been centrifuged.

Confocal laser scanning microscopy (CLSM)

The spatial distribution of corn starch ester nanoparticles and oil droplets in the PEs was visualized using CLSM (TCS SP8, Leica, Aden‐Württemberg, Germany). Nile Blue A (0.1 mg mL⁻¹ in water) and Nile Red (0.1 mg mL⁻¹ in 1,2‐propanediol) were combined at a 1:1 (v/v) ratio and stored in the dark prior to use. Subsequently, 100 μL of the dye mixture was introduced into 1 mL of emulsion and observed under a magnification of ×40. Nile Blue A (for starch labeling) and Nile Red (for oil phase labeling) were excited using a 633 nm He–Ne laser and a 488 nm Ar laser, respectively.

Rheological properties

Rheological analysis was conducted using a rheometer (TA Instruments, New Castle, USA) with 40 mm parallel plate (1 mm gap), according to a previously reported method. ²¹ The apparent viscosity was evaluated at shear rate from 0.1 to 100 s⁻¹. The elastic modulus (G′) and loss modulus (G″) were obtained by scanning at an oscillation frequency from 0.1 to 100 rad s⁻¹ at fixed strain of 0.5%. The viscosity, G′, and G″ were plotted in logarithmic scale.

AST‐loaded PE

AST was dissolved in soybean oil to make a concentration of 0.5 mg mL⁻¹. To ensure the complete dissolution of AST in soybean oil, the mixture was stirred overnight in the dark. The AST‐loaded oil phase was then mixed with corn starch ester nanoparticle dispersion (40 g L⁻¹) at oil‐to‐aqueous phase ratio of 1:1 (v/v), corresponding to an oil volume fraction (φ) of 0.5, consistent with the section above on the preparation of PE. The mixture was homogenized under identical conditions to prepare AST‐loaded PE. Soybean oil containing AST was used as the control.

Evaluation of AST in PE

AST retention rate

The AST retention rate in PE was measured according to a previously reported method. ²² An aliquot of 0.2 mL of PE was extracted with 4.8 mL of a dichloromethane–methanol mixture (2:1 v/v). The PE was subjected to centrifugation at 1130 × g for 20 min, after which the dichloromethane phase was separated and its absorbance was measured at 480 nm. AST was determined using the standard curve (y = 0.1101x + 0.0215; R ² = 0.9988), and the retention rate was determined using the following equation:

$$\mathrm{AR}\left(\%\right)=\frac{A}{A_0}\times 100$$ (5)

where AR is the retention rate of AST; A is the content of AST embedded in PE (mg); and A ₀ is the initial content of added AST (mg).

In vitro digestion of AST‐PE and release of free fatty acid (FFA)

The in vitro digestion of AST‐PE was performed based on a previously reported method, with modifications. ²³ Simulated gastric fluid (SGF; pH 1.2) contained 2 g L⁻¹ NaCl and 3.2 g L⁻¹ pepsin. Simulated intestinal fluid (SIF) was comprised of 32.8 g L⁻¹ NaCl, 4.15 g L⁻¹ CaCl₂, 6 g L⁻¹ porcine lipase, 6 g L⁻¹ pancreatin, and 18.9 g L⁻¹ bile salts in 0.05 mol L⁻¹ PBS. Before the test, SGF, SIF, and AST‐PE were preheated to 37 °C. To simulate gastric digestion, AST‐PE was supplemented with 5 mL of SGF and incubated in a 37 °C shaking water bath (200 rpm) for 2 h. Then, the pH was adjusted to 7.0 with 1 mol L⁻¹ NaOH. For simulating intestinal digestion, 10 mL of the gastric‐digested mixture was combined with 10 mL of SIF and incubated under the same condition of 37 °C at 200 rpm for another 2 h. During the simulated intestinal digestion process, the volume of consumed NaOH to maintain the pH of the mixture at 7.0 was recorded. The pH was kept at 7.0 through the manual addition of 0.25 mol L⁻¹ NaOH solution at 20 min intervals, specifically at 20, 40, 60, 80, 100, and 120 min. Release of FFA was quantified in terms of consumed NaOH volume according to the following equation:

$$\mathrm{FFA}\left(\%\right)=\frac{V_{\text{NaOH}}\times m_{\text{NaOH}}\times M_{\text{lipid}}}{2W_{\text{lipid}}}\times 100$$ (6)

where m _NaOH is the molar concentration of consumed NaOH; V _NaOH is the volume of the consumed NaOH; M _lipid represents the average molecular weight of soybean oil; and W _lipid denotes the total mass of the initial soybean oil utilized in the digesting process.

Bioaccessibility of AST

Following the simulated digestion process, 10 mL of the digests underwent centrifugation at 5590 × g for 20 min at 25 °C. After centrifugation, the AST content in the micellar phase was quantified using the following equation:

$$\text{Bioaccessibility}\left(\%\right)=\frac{C_{\text{micelle}}}{C_{\text{digest}}}\times 100$$ (7)

where C _micelle represents the AST concentration in the micellar fraction and C _digest denotes the AST content in the sample following intestinal digestion.

Statistical analysis

The data were analyzed for statistically significant differences using SPSS statistical software (V25, IBM, New York, USA). One‐way analysis of variance (ANOVA) followed by Tukey's honestly significant difference test were employed to assess significant difference (P ≤ 0.05). Two‐way ANOVA was employed to analyze the effect of interaction between storage time and temperature on CI of PE. All tests were repeated for three replicates. The mean ± standard deviation was used to represent the results.

RESULTS AND DISCUSSION

Characterization of starch ester nanoparticles

DS, ¹H NMR, FTIR, and XRD

Based on our previous research, differences in molecular composition, chain conformation, and granular semicrystalline structure among corn starches with different amylose contents influenced the accessibility and reactivity of hydroxyl groups during the citric acid esterification process. DS of starch esters was in the order of WCS ester (0.1631) > HACSester (0.1552) > NCS ester (0.1199). ²⁴ After treatment using the nanoprecipitation method, the DS value was decreased (Table 1). This reduction may be attributed to reactions between citric acid and starch occurring mainly in the amorphous regions, which were partially depolymerized during ultrasonication ²⁵ , leading to a decrease in DS; however, the overall trend of DS remained as WCSENP > HACSENP > NCSENP.

Table 1.

DS of starch ester nanoparticles

Treatment	DS
WCSENP	0.039 ± 0.003a
NCSENP	0.013 ± 0.003b
HACSENP	0.033 ± 0.002a

N = 3. Different letters in the same column indicate significant differences (P ≤ 0.05). Results expressed as mean ± standard deviation.

The ¹H NMR spectra of native starch and starch ester nanoparticles are presented in Fig. 1(a),(b). For native starch, characteristic resonances were observed at 5.38 and 5.20 ppm, corresponding to the anomeric protons of α‐1,4 and α‐1,6 glycosidic linkages, respectively. The proton signals distributed in the region of 3.0–4.0 ppm were assigned to H2–H5 protons of the glucopyranosyl ring, which represented the typical backbone structure of starch. ²⁴ After esterification, new resonance signals appeared in the range of 2.45–2.68 ppm, which were not present for native starch. These peaks were attributed to the methylene protons adjacent to the carbonyl groups of the introduced citric acid moieties. The emergence of these signals, combined with the preserved glycosidic proton resonances, confirmed that ester groups were successfully grafted on the starch backbone.

Figure 1.

Native starch and starch ester nanoparticles: ¹H NMR spectra (a, b); FTIR spectra (c), XRD patterns (d), SEM images at ×1000 (e) and at ×10 000 (f), and water contact angle (g).

The FTIR spectra are shown in Fig. 1(c). The absorption at 3300 cm⁻¹ corresponded to the intramolecular and intermolecular O—H stretching vibrations, and the absorption at 2924 cm⁻¹ was assigned to the asymmetric stretching vibration of C—H. ¹⁵ The absorption peak at 997 cm⁻¹ was ascribed to the C—O—H bending mode. ³ A peak at around 1722 cm⁻¹ was identified in the spectra of citric acid starch ester nanoparticles, attributed to the production of the ester bond.² The peak intensity of ester groups was in the order of WCSENP > HACSENP > NCSENP, which was in line with the results of DS (Table 1). The absorption peak observed at 1574 cm⁻¹ was assigned to the asymmetric stretching vibration of the carboxylate (COO⁻) group. ³

The XRD patterns and relative crystallinity (RC) of native starch and starch ester nanoparticles are shown in Fig. 1(d). NCS and WCS have typical A‐type crystalline structure, and HACS has typical B‐type crystalline structure. ²⁶ In native starch, RC showed the order of WCS > NCS > HACS. WCS generally exhibits higher RC than HACS and NCS due to amylopectin being predominant in the crystalline regions. ²⁷ However, after the esterification of starch and nanoparticle production, the types of starch ester nanoparticles showed V‐type pattern with new peaks at approximately 7°, 13°, and 20°, ²⁸ revealing the structural alterations in starch granules. Wherein, HACSENP displayed increased intensity of distinctive peaks and RC, then followed by NCSENP and WCSENP. Dong ²⁹ stated that amylose in starch was more predisposed to form compact V‐type single helices compared to amylopectin. Furthermore, the elevated amylose content in starch was not always the main reason for formation of V‐type pattern, which is significantly affected by the melting of B crystals for enhanced amylose release. ³⁰

SEM analysis

The morphological characteristics of starch ester nanoparticles and native starch granules were characterized by SEM (Fig. 1(e),(f)). WCS and NCS exhibited typical polygonal morphologies, with larger particle size and complete structure. In contrast, HACS mainly presented elliptical and irregular shapes, with relatively smaller particle size (Fig. 1(e)). Citric acid esterification may result in the swelling of starch granules accompanied by structural collapse, leading to surface erosion and damage of the granules. ² In addition, during the preparation process by nanoprecipitation, the crystalline regions of starch were destroyed, resulting in the depolymerization of the complete particle structure. ³¹ Therefore, the three types of starch ester nanoparticles showed structural disruption and reduced particle size compared to native starch (Fig. 1(e),(f)). In addition, starch ester nanoparticles exhibited aggregation due to the increased specific surface area and decreased particle size. ⁵

Water contact angle

The contact angle serves as a significant parameter indicating the hydrophilicity and lipophilicity of particles, facilitating the classification of emulsion types, including oil in water (O/W) and water in oil. The contact angle of native starch and starch ester nanoparticles was less than 90° (Fig. 1(g)), indicating the emulsions were of O/W type. Citric acid starch ester nanoparticles showed increased contact angle relative to that of native starch, with the order of WCSENP > HACSENP > NCSENP > HACS > NCS > WCS. The increased contact angle of starch ester nanoparticles was due to the substitution of the ester groups on starch molecular chains which enhanced the hydrophobicity of native starch. The production of nanoparticles changed the crystalline structure of native starch, transforming both A‐type and B‐type into V‐type, in which the hydrophobic cavity in the single helices hindered water molecules, thus enhancing the hydrophobicity of the starch ester nanoparticles. ³² The results were aligned with the XRD findings. Generally, the contact angle of solid particles close to 90° reflects a strong capacity to stabilize PE. ³³

Evaluation of PE

Visual appearance, microscopic observation, and CI of PE

Separation of oil and water layers of PE stabilized by native starch was observed after fabrication of PE after 1 h, indicating poor stability. Starch ester nanoparticle‐based PE remained stable for 1 day, while the transparent layer in the emulsion gradually appeared up to day 7 (Fig. 2(a)). Starch ester nanoparticle‐based PE did not show any separation of oil layer after the storage over 30 days regardless of stored temperature at 4 and 25 °C, suggesting the improved storage stability.

Figure 2.

Visual appearance of PE (a); optical microscopy images of PE (b); CI of PE during storage at 4 and 25 °C (c). Different uppercase letters were applied to label the significant differences (P ≤ 0.05) between different samples for the same time and temperature, and lowercase letters were applied to label the significant differences (P ≤ 0.05) of same sample across different times and temperatures.

Optical microscopy images of PEs are presented in Fig. 2(b). Native starch granules tended to precipitate, and the number of droplets that could be effectively stabilized in the emulsion was relatively limited. Optical micrographs revealed a large number of free starch granules that were not adsorbed at the oil–water interface. This observation was consistent with the morphological characteristics of the different starch granules presented in the SEM images, further indicating that native starch possessed limited interfacial stabilization capacity in the emulsion system. WCSENP‐PE showed the smallest droplet size compared to those of NCSENP‐PE and HACSENP‐PE during the storage period. The smaller droplet size of WCSENP‐PE was attributed to its higher DS and larger contact angle compared with HACSENP and NCSENP. The increased DS enhanced the hydrophobic modification of the particles, while the larger contact angle indicated more favorable wettability at the oil–water interface. These factors likely promoted stronger interfacial adsorption and more efficient surface coverage during homogenization, thereby reducing droplet coalescence and resulting in smaller droplets. Droplet size is a critical parameter influencing the stability of PEs. During storage, emulsion droplets underwent aggregation and coalescence, resulting in a rise in droplet size and ultimately instability. Compared to larger droplets, smaller droplets exhibited smaller discrepancy between buoyant and gravitational forces, resulting in a slower sedimentation rate and thereby enhancing the overall stability of the emulsion system. ³⁴

The CI was calculated to evaluate emulsion stability by measuring droplet aggregation. ³⁵ The droplets were generally dispersed uniformly due to the effect of interfacial layer, electrostatic repulsion, and steric hindrance in the stable emulsion system, without aggregation. ³⁶ CI of all PEs was decreased as the storage time increased from 1 h to 30 days. NCSENP‐PE and HACSENP‐PE showed significantly lower CI than that of WCSENP‐PE (P ≤ 0.05), regardless of storage temperature of 4 or 25 °C (Fig. 2(c)). In addition, CI of the three types of PE stored at 4 °C tended to be higher than that of PEs stored at 25 °C. After 30 days, CI of WCSENP‐PE retained over 90% at storage temperatures of 4 and 25 °C, indicating the high storage stability. A two‐way ANOVA with an interaction between storage temperature and day was employed to analyze the data. Factors of temperature, time, and their interaction statistically significantly affected the storage stability of PE (Fig. 2(c)). This indicated that storage at 4 °C delayed the creaming process and induced substantial alterations in the interfacial film structure, ultimately leading to a higher CI.

Droplet size, zeta potential, EAI, and ESI

According to Stokes' law (F = 6πηrv; η: fluid viscosity, r: particle radius, v: fluid flow velocity relative to the particle), large droplet size of an emulsion is a major factor affecting emulsion stability. The droplet size D50 and D[4,3] of PE showed the order of NCSENP > HACSENP > WCSENP (P ≤ 0.05), indicating the best stability of WCSENP‐PE (Table 2). These results are consistent with the microscopic observations.

Table 2.

Droplet size including D10, D50, D90, D[3,2], and D[4,3], interfacial tension, interfacial starch absorption rate, zeta potential, EAI, and ESI of PE

Treatment	Droplet size (μm)					Interfacial tension (mN m−1)	Interfacial starch absorption rate (%)	Zeta potential (mV)	EAI (m2 g−1)	ESI (min)
	D10	D50	D90	D[3,2]	D[4,3]
WCSENP	7.86 ± 0.04b	29.63 ± 0.10c	61.10 ± 0.10c	10.48 ± 0.03b	32.39 ± 0.08c	3.03 ± 0.03c	62.17 ± 0.47a	−48.86 ± 1.99b	1.12 ± 0.007a	49.86 ± 0.83a
NCSENP	7.09 ± 0.03c	38.34 ± 0.02a	74.28 ± 0.43a	10.69 ± 0.01a	40.76 ± 0.13a	10.53 ± 0.31a	39.66 ± 0.48c	−36.77 ± 1.9a	1.07 ± 0.002c	44.04 ± 1.72b
HACSENP	8.09 ± 0.02a	32.20 ± 0.08b	68.81 ± 0.24b	10.67 ± 0.02a	35.83 ± 0.10b	8.06 ± 0.11b	41.45 ± 0.96b	−38.28 ± 1.27a	1.10 ± 0.008b	46.37 ± 1.79b

N = 3. Different letters in the same column indicate significant differences (P ≤ 0.05). Results expressed as mean ± standard deviation. D10, D50, and D90 correspond to the particle size below which 10%, 50%, and 90% of the sample particles fall, respectively. D[4, 3] and D[3, 2] refer to the volume‐weighted particle diameter and surface‐weighted particle diameter, respectively.

High absolute value of zeta potential of PE generally indicates strong electrostatic repulsion between the droplets which contributes to maintaining the stability of PE. ³⁷ The absolute value of zeta potential of starch ester nanoparticle‐based PE was over 30 mV (Table 2), suggesting the good stability of PE. ³⁸ The absolute value of zeta potential of PE showed the order of WCSENP > HACSENP > NCSENP. WCSENP‐PE exhibited higher EAI and ESI than NCSENP‐PE and HACSENP‐PE (P ≤ 0.05) (Table 2). This result was consistent with the microscopic findings (Fig. 2(b)). The increase in droplet size reduced droplet mobility, which can be explained by the Stokes–Einstein equation (D = kT/6πηr), wherein the diffusion coefficient (D) is inversely proportional to the droplet radius (r). As the droplet size increased, the diffusion coefficient decreased, leading to weaker Brownian motion and reduced collision frequency, thereby influencing interfacial rearrangement and adsorption efficiency. ³⁹

Interfacial properties

Interfacial tension reflects the surface activity and adsorption ability of solid particles. A decrease in interfacial tension implies enhanced surface activity, stabilizing the oil–water interface. ⁴⁰ Interfacial tensions of different starch ester nanoparticles were in the order WCSENP < NCSENP < HACSENP (P ≤ 0.05) (Table 2). WCSENP exhibited the lowest interfacial tension, showing the most stable emulsions. Esterification improved O/W interfacial wettability, reducing tension and strengthening the droplet barrier to delay phase separation. ⁴¹ In this study, WCSENP with the highest DS showed the lowest interfacial tension. This finding was consistent with previous reports ⁴² showing that esterified starch exhibited reduced dynamic interfacial tension compared with native starch, and this reduction effect became more pronounced with increasing DS.

The interfacial starch adsorption rate serves as an important parameter to characterize the degree of starch accumulation at the oil–water interface within the emulsion system. The adsorption layer functioned as a physical barrier, successfully inhibiting the collision and coalescence of oil droplets. ⁴³ WCSENP‐PE exhibited the highest interfacial starch adsorption rate of 62.17% (Table 2), indicating that WCSENP interfacial film was more compact, exhibiting the highest stability. ⁴⁴

CLSM analysis

CLSM was further employed to characterize the microstructure of the emulsions. Oil phase and water phase were stained with Nile Red and Nile Blue appearing as green and red colors, respectively (Fig. 3). The green oil droplets were surrounded by the continuous red aqueous phase, with numerous densely packed droplets of varying sizes. These results indicate that all emulsions exhibited a typical O/W‐type structure. The observations are consistent with the contact angle data of corn starch esters, and the approximate droplet sizes also corresponded with analysis obtained from optical microscopy.

Figure 3.

CLSM images of PEs.

Rheological properties

PE exhibited decreased viscosity as the shear rate increased, showing the shear‐thinning phenomenon, which was typical non‐Newtonian pseudoplastic behavior ⁴⁵ (Fig. 4(a)). The applied shear force resulted in the decomposition or rearrangement of droplets in emulsions due to the weak network structure. ⁴⁶ WCSENP‐PE showed higher shear‐resistance stability than NCSENP‐PE and HACSENP‐PE. This was due to the smallest droplet size of WCSENP‐PE, showing short inter‐droplet distance, leading to increased interaction force and collision frequency. ³ The storage modulus (G′) and loss modulus (G″) of PE represent the solid‐elastic and fluid‐viscous characteristics, respectively (Fig. 4(b)). G' of PE was greater than G″, implying the typical elastic gel‐like network structure of PE. ⁴⁷ The increased G′ was attributed to the increased interactions with adjacent droplets due to the small droplets possessing large surface area. ⁴⁸ G′ and G″ gradually increased as the frequency increased due to the weakened internal structure of PE under high shear force. ⁴⁹

Figure 4.

Rheological properties of PEs: apparent viscosity (a); storage modulus (G′) and loss modulus (G″) (b).

Retention rate of AST, in vitro digestion of AST‐PE and bioaccessibility

The retention rate of AST released by PE showed the order of WCSENP > HACSENP > NCSENP > soybean oil (Fig. 5(a)). The AST retention curve was fitted to the first‐order kinetic model. The corresponding first‐order regression equation and the rate constant (k) were calculated (Table 3). WCSENP‐PE exhibited the smallest k value of 0.1154 indicating slowest AST degradation rate and improved retention rate. This outcome was consistent with the PE stability. At the emulsion interface, WCSENP‐PE created a thicker interfacial layer that inhibited the diffusion of oxygen, oxidants, and free radicals. ⁵⁰

Figure 5.

AST‐PE stabilized by corn starch ester nanoparticles with different amylose contents. Retention rate of AST (a); microstructural changes of AST‐PE during the digestion process (b); FFA release rate of AST (c); bioaccessibility of AST (d).

Table 3.

First‐order kinetic simulation of AST retention rate

Treatment	First‐order response regression equation	k	R 2
Oil	y = −0.2398x + 4.0413	0.2398	0.9968
WCSENP	y = −0.1154x + 3.6372	0.1154	0.9696
NCSENP	y = −0.1297x + 4.0141	0.1297	0.9893
HACSENP	y = −0.129x + 3.8785	0.129	0.9961

Under simulated gastric conditions, microscopic observations revealed marked flocculation within the emulsion system, showing droplets aggregating into larger particles (Fig. 5(b)). This aggregation behavior was primarily attributed to the acidic environment and elevated ionic strength characteristic of the gastric phase, which weakened the electrostatic repulsion between droplets and thereby promoted flocculation and aggregation. ⁵¹ Following subsequent intestinal digestion, the original emulsion droplets were progressively hydrolyzed by lipase, leading to a reduction in the oil phase volume fraction. As lipid digestion proceeded, the generated digestion products, including FFAs and monoacylglycerols, underwent competitive adsorption and interfacial rearrangement. This process partially displaced the original interfacial structure. ⁵² Meanwhile, in the presence of bile salts and calcium ions, various colloidal structures such as mixed micelles, vesicles, and insoluble calcium soaps were formed. ⁵¹ The formation of these new self‐assembled aggregates, coupled with interfacial reconstruction, resulted in increased droplet size. This process gradually transformed the system from a single emulsion system to a more complex multiphase dispersed system. ⁵³

The FFA release kinetics is critical for assessing digestion of PE. FFA release rate of PE showed the order of WCSENP > HACSENP > NCSENP > soybean oil (Fig. 5(c)), due to the smaller droplets of WCSENP providing more surface area for lipase which accelerated lipid digestion. ⁵⁴ The bioaccessibility of AST showed a similar order of WCSENP‐PE > HACSENP‐PE > NCSENP‐PE > soybean oil (P ≤ 0.05) (Fig. 5(d)). Starch ester nanoparticle‐based PE showed distinct contact angle, interfacial tension, droplet size, and others, influencing pepsin and lipase digestion, further affecting the release of bioactive compounds. ⁵⁵ Smaller droplets of WCSENP‐PE generated more FFAs, providing additional loading sites for AST and, consequently, enhancing the bioaccessibility. ²¹ The lowest bioaccessibility of AST loaded in soybean oil was probably attributed to the small surface area of soybean oil. ²²

CONCLUSION

Three types of corn starch with distinct amylose content, WCS, NCS, and HACS, were esterified using citric acid, and subjected to nanoprecipitation to prepare starch ester nanoparticles. The formation of ester groups on starch ester nanoparticles was confirmed by FTIR and ¹H NMR analyses. XRD results demonstrated that starch ester nanoparticles showed the V‐type crystalline structure. Starch ester nanoparticles exhibited enhanced contact angle, zeta potential, EAI, ESI, and CI compared to native starch. WCSENP‐PE showed the best emulsifying performance, then HACSENP‐PE followed by NCSENP‐PE. PE was further evaluated for its protective effect on AST in terms of retention rate, released FFAs, and bioaccessibility, showing the trend of WCSENP‐PE > HACSENP‐PE > NCSENP‐PE > soybean oil. Taken together, citric acid starch ester nanoparticles showed great potential for stabilizing PEs and encapsulating biological agents as carriers. WCSENP exhibited the best performance in stabilizing PE, followed by HACSENP and NCSENP.

FUNDING INFORMATION

This work was supported by Shanghai Ocean University.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.