Synergistic effects of enzymatic debranching and high hydrostatic pressure on the structural and digestive properties of –conjugated linoleic acid microcapsules

Abstract

This study explored the effects of enzymatic debranching (ED, 0–12 h) and high hydrostatic pressure (HHP, 300–600 MPa) on the multi-scale structure and in vitro digestibility of lotus seed starch (LS)–conjugated linoleic acid (CLA) microcapsules. XRD revealed diffraction peaks at 7°, 12°, and 20°, characteristic of V-type crystalline complexes formed between amylose and CLA. NMR showed a resonance at 103.5 ppm, while FTIR indicated masking of CLA double bonds, confirming successful complexation. Microcapsules treated at 300 MPa for 6 h exhibited the highest microcrystalline proportion (31.2%). In vitro digestion showed that the 6 h + 300 MPa group retained 25.5% resistant starch, significantly higher than the 0 h + 300 MPa group (9.6%). SEM revealed spherical resistant starch clusters, contributing to the delayed release of CLA. These findings demonstrated the potential of ED and HHP to produce stable, targeted starch-based microcapsules for heat-sensitive bioactives, offering a sustainable alternative to conventional thermal processing.

Highlights

• •Debranching and HHP synergistically enhance the structural stability of microcapsules.
• •Six hours of enzymatic debranching combined with 300 MPa HHP yields the highest microcrystalline proportion (32.09%).
• •In vitro digestion shows 25.5% RS retention, suggesting potential for colon-targeted delivery.
• •Non-thermal HHP prevents thermal degradation, providing a novel delivery strategy.

1. Introduction

Driven by economic growth and increasing health awareness, the consumption of health-functional foods has been rising. The demand for personalized diets and precision nutrition has not only created a substantial consumer market but has also provided strong momentum for the expansion of the functional food industry (Rezagholizade-shirvan et al., 2024). Within this industry, bioactive compounds—such as vitamins, probiotics, polyphenols, and polyunsaturated fatty acids—have attracted considerable attention for their health-promoting properties when incorporated into functional products (Rezagholizade-shirvan et al., 2024). However, many of these compounds are highly susceptible to degradation under environmental stresses such as oxygen, light, heat, or pH fluctuations during processing, storage, and digestion (Petkova, Mihaylova, & Desseva, 2022). Such instability reduces their bioactivity and limits their use in fortified foods. Microencapsulation technology, which protects sensitive ingredients within a protective wall matrix, has therefore emerged as a promising solution (Choudhury, Meghwal, & Das, 2021).

Microencapsulation forms a physical barrier between the core (active ingredient) and the external environment, thereby protecting labile compounds from adverse conditions and enabling controlled release in response to specific triggers (e.g., temperature, enzymatic activity, or mechanical rupture) (Pant, 2023). This technology is widely applied in the food industry across various sectors. For example, in beverages, nutrients such as vitamin C and anthocyanins are protected from oxidation, while bitterness and astringency are masked to improve flavor (Tan & McClements, 2021). In baked goods, yeast, spices, and oils are encapsulated to enhance stability and prolong shelf life (Sharma, Mulrey, Byrne, Jaiswal, & Jaiswal, 2022). In food preservation, plant essential oils delivered via microencapsulation effectively inhibit microbial growth, thereby extending the freshness of fruits, vegetables, meat, and seafood (Sun et al., 2023). Similarly, in meat products, ω-3 fatty acids are encapsulated to improve nutritional value, while in dairy products, probiotics are encapsulated to maintain their viability during processing and storage (Turek, Khachatryan, Khachatryan, & Krystyjan, 2023). These applications not only enhance the quality and functionality of food but also drive innovation and growth within the food industry. The choice of wall materials—such as polysaccharides (e.g., alginate, starch) (Díaz-Montes, 2023), proteins (e.g., whey protein, gelatin) (Choudhury et al., 2021), lipids, and synthetic polymers (Yang et al., 2025)—in combination with encapsulation techniques such as spray drying, phase separation, and extrusion, plays a critical role in determining the encapsulation efficiency, stability, and functionality of the final product (Yang et al., 2025). However, the stability of heat-sensitive components during encapsulation has been insufficiently addressed in current food microencapsulation research, thereby highlighting the urgent need to explore non-thermal encapsulation techniques.

Most microcapsules are currently produced using traditional thermal processing techniques, such as spray drying and thermal curing (Zhao et al., 2023). However, excessively high temperatures can weaken the mechanical strength of the wall material, leading to core material leakage (Zhao et al., 2023). For example, during the preparation of β-cyclodextrin microcapsules via spray drying, excessively high temperatures were found to cause adverse structural effects. Increased heat induced rapid surface shrinkage and cracking, which in turn accelerated core material deterioration under environmental stress (Yang et al., 2022).

Moreover, heat-sensitive core materials, such as probiotics and polyunsaturated fatty acids, are highly susceptible to degradation and inactivation at increased temperatures (Agriopoulou, Tarapoulouzi, Varzakas, & Jafari, 2023). Thermal processes are also associated with high energy consumption and prolonged processing cycles, which raise production costs and limit their use in premium heat-sensitive products (Jia, Zheng, & Guo, 2018). In response, non-thermal processing techniques—including high hydrostatic pressure (HHP), pulsed electric fields, and cold plasma—have attracted growing research interest (Petkova et al., 2022). Among these methods, microorganisms are inactivated by HHP (100–600 MPa), while the nutritional and sensory qualities of food are largely preserved. In recent years, high-pressure processing (HPP) has demonstrated significant advantages in enhancing the nutrition, safety, and overall quality of foods, particularly in fruits, vegetables, meats, seafood, dairy products, and egg (Jia et al., 2018). In addition, HPP can effectively reduce anti-nutritional factors and allergens, thereby broadening its potential applications within the food industry. Studies have shown that under HHP conditions, starch can form complexes with guest molecules such as long-chain fatty acids and plant polyphenols, thereby enhancing its functional properties (Jia et al., 2018). Our previous work further demonstrated that debranched starch under HHP conditions improved the encapsulation efficiency and stability of conjugated linoleic acid (CLA) (B. Wang, Jia, Chen, & Guo, 2025). However, limited research has explored the molecular mechanisms underlying the synergistic effects of HHP and enzymatic debranching (ED) in enhancing CLA encapsulation and stability, which constrains the broader application of this technology in functional food development. Additionally, the protective effect of encapsulated core materials within the digestive tract and their potential for targeted delivery have yet to be validated.

This study aimed to investigate the effects of ED (0−12h) under HHP on the multi-scale structure and in vitro digestibility of P-S-CLA microcapsules. The structural characteristics of ED–HHP–treated microcapsules were analyzed using FTIR, XRD, ¹³C NMR, and SAXS. Their digestive behavior was then evaluated in a rat bionic in vitro model to assess the feasibility of colon-targeted delivery. Overall, this work provides molecular-level insights into the synergistic mechanisms of ED and HHP in starch-based microcapsules and supports the application of non-thermal processing for developing functional food delivery systems.

2. Materials and methods

2.1. Materials

Fresh lotus seeds were obtained from Green Field Food Co., Ltd. (Fujian, China), and lotus seed starch (LS) was isolated using the method of Guo et al. (2015). The starch purity was confirmed to be ≥98% with a K-TSTA-100 A starch content kit (Megazyme International Ireland Ltd., Bray, Ireland). Pullulanase (≥1000 U/mL, Product No. E2412) and CLA (≥99%, Product No. O5507) were purchased from Sigma-Aldrich (St. Louis, MO, USA). All other reagents were of analytical grade.

2.2. Construction of LS–CLA and P–LS–CLA microcapsules

Lotus seed starches with varying degrees of debranching (0, 3, 6, and 12 h) were prepared according to previous methods and designated as P-LS-0, P-LS-3, P-LS-6, and P-LS-12, respectively (Wang, Chen, Jia, & Guo, 2025). Each P-LS sample (5 g) was dispersed in deionized water at 50 °C using a magnetic stirrer at 300 rpm. CLA (0.5 g) was dissolved in 5 mL of anhydrous ethanol and then added to the starch suspension. After stirring for 10 min, the mixture was sealed in polypropylene vacuum bags with a vacuum packaging machine (DZ-400-2F, Dajiang Machinery Co., Ltd. Wenzhou China). High hydrostatic pressure (300–600 MPa) was applied for 30 min at 25 °C. Following treatment, the precipitate was collected by high-speed centrifugation (8000×g), washed three times with anhydrous ethanol and deionized water, freeze-dried, and pulverized to obtain debranched starch–conjugated linoleic acid (P-LS-CLA) microcapsules.

2.3. Fourier transform infrared (FTIR) spectroscopy

The short-range ordered structure of P-LS-CLA was analyzed using FTIR spectroscopy (Bruker, Karlsruhe, Germany). Starch samples were mixed with solid potassium bromide, ground under an infrared lamp, and compressed into thin pellets with a tablet press. Spectra were recorded in the range of 4000–400 cm⁻¹ at a resolution of 4 cm⁻¹.

2.4. X-ray diffraction (XRD)

The crystalline structure of P-LS-CLA was analyzed using XRD (Bruker, Karlsruhe, Germany). Diffraction patterns were recorded over a 2θ range of 5°–40° at a scanning speed of 3°/min. Relative crystallinity, defined as the ratio of crystalline peak area to total peak area, was calculated using PeakFit 4.12 software.

2.5. Solid-state ¹³C nuclear magnetic resonance (NMR) spectroscopy

A 300 mg sample of P-LS-CLA was weighed and placed in the solid-state ¹³C NMR chamber for analysis. Spectra were acquired at 100.62 MHz using a 4-mm probe with 1600 scans, a spinning speed of 6 kHz, and an acquisition time of 0.013 s. The NMR spectra were processed with Origin 9.0, and peak fitting was performed using PeakFit 4.0 to quantify the relative areas of different regions. Chemical shift variations were also recorded and compared for further analysis.

2.6. Small-angle X-ray scattering (SAXS)

A 40 mg sample was used to prepare a 40% (w/v) microcapsule emulsion, which was centrifuged to obtain the precipitate for SAXS analysis. A Cu target served as the X-ray source with an excitation wavelength of 1.5418 Å, and the scattering vector range was set at q = 0.007–0.23 Å⁻¹. The Bragg spacing (d) was calculated using the equation d = 2π/q₀, where q = 4π sin θ / λ. The lamellar structure of the microcapsules was evaluated according to the method of Guo et al. (2015), and the crystalline-to-amorphous lamellar ratio was determined.

2.7. In vitro digestion of P-LS-CLA using the DIVRSD model

2.7.1. Simulated in vitro digestion

Artificial rat saliva, gastric juice, pancreatic juice, and bile were prepared following the method of Wu et al. (2017). Glucose concentration was measured using a glucose assay kit. Gastric mucin, α-amylase, amyloglucosidase, pancreatin, and pepsin were purchased from Sigma-Aldrich (St. Louis, MO, USA), while other reagents were obtained from Shanghai Macklin Biochemical Co., Ltd. (China). The DIVRSD model, adapted from Wu et al. (2017) for cooked starch digestion, consists of a 3D-printed rat stomach equipped with an electromechanical driving device (to simulate peristalsis), a temperature control system, and a secretion and emptying system. For digestion experiments, 200 mg of P-LS-CLA was accurately weighed, dispersed in 2.0 mL of water, and heated for 15 min. Artificial saliva (2.0 mL, 37 °C) was added to each sample to simulate oral digestion for 30 s. The samples were then transferred to the DIVRSD model for continuous digestion, with aliquots collected at 0, 10, 20, 30, 60, 90, 120, and 180 min. At each time point, a 200 μL digest was withdrawn, mixed with 800 μL of anhydrous ethanol to terminate enzymatic activity, and stored at 4 °C for subsequent analysis. In parallel, simulated digestion of LS-CLA under equivalent HHP conditions was compared with that of native lotus seed starch. Following the method of Liu et al. (2025), starch fractions were classified as rapidly digestible starch (RDS, digested within 20 min), slowly digestible starch (SDS, digested between 20 and 120 min), and resistant starch (RS, remaining after 120 min). The starch hydrolysis rate was calculated using Eq. (1), while the contents of RDS, SDS, and RS were determined using Eqs. (2), (3), (4).

$$\mathit{Starch\; hydrolysis}\left(\%\right)=\frac{\mathit{TG}}{\mathit{TS}}\times 0.9\times 100$$ (1)

$$\mathit{RDS}\left(\%\right)=\left[\frac{\left(G_{20}-G_0\right)}{\mathit{TS}}\right]\times 0.9\times 100$$ (2)

$$\mathit{SDS}\left(\%\right)=\left[\frac{\left(G_{120}-G_{20}\right)}{\mathit{TS}}\right]\times 0.9\times 100$$ (3)

$$\mathit{RS}\left(\%\right)=\left[\frac{\left(\mathit{TS}-G_{120}\right)}{\mathit{TS}}\right]\times 0.9\times 100$$ (4)

TG represents the glucose content of the sample, and TS denotes the sample mass. The equations used to calculate the content of each starch component in the microcapsules are provided below. In addition, G_t refers to the glucose content at the digestion termination time point.

2.7.2. Morphological analysis during digestion

The microstructure of the microcapsules during digestion was analyzed using environmental scanning electron microscopy (ESEM; JSM660LV, JEOL, Tokyo, Japan). After digestion, the precipitate was collected by centrifugation, dried, and mounted on a metal stage with conductive double-sided tape. Surface debris was gently removed with a rubber bulb, after which the sample was sputter-coated with gold under vacuum using an ion sputter coater and then examined by ESEM. The acceleration voltage was set to 20 kV, and observations were conducted at magnifications of 20,000× and 100,000×.

2.8. Statistical analysis

All data were plotted using Origin 2025 software. Statistical analysis and significance testing were performed with DPS 9.5 software, where p ≤ 0.05 was considered statistically significant. All experiments were conducted in triplicate.

3. Results and discussion

3.1. Fourier transform infrared (FTIR) spectroscopy

The short-range ordered structures of CLA, LS-CLA, and P-LS-CLA were analyzed using FTIR to elucidate the interactions between starch and CLA. The infrared spectra of lotus seed starch and CLA are shown in Fig. 1a and b, respectively. In the CLA spectrum (Fig. 1b), peaks at 982 cm⁻¹ and 946.9 cm⁻¹ correspond to the bending vibrations of the cis-trans and trans-cis isomers of CLA, associated with the C—C group and the out-of-plane bending of RCH=CHR, respectively (Silverstein & Kiemle, 2005). The band at 1285 cm⁻¹ was assigned to C—O stretching, while the peaks at 1709.5 cm⁻¹ and 1653 cm⁻¹ were attributed to conjugated double bonds in CLA. The absorption peak at 2673 cm⁻¹ was attributed to the -OH group of the -COOH moiety, while the peaks at 2850 cm⁻¹ and 2925 cm⁻¹ were assigned to the stretching vibrations of –C–H in the aliphatic chain and the methylene group, respectively (Silverstein & Kiemle, 2005).

Fig. 1.

FT-IR spectra of the samples. “a” represents native lotus seed starch, “b” represents conjugated linoleic acid (CLA), and 300, 500, and 600 MPa indicate the pressures used for HHP treatment. “c,” “d,” “e,” and “f” correspond to LS-CLA and P-LS-CLA prepared with starch wall materials debranched for 0, 3, 6, and 12 h, respectively.

Compared with CLA, LS-CLA exhibited a distinct infrared spectrum (Fig. 1c). The characteristic peaks at 982 cm⁻¹ and 946.9 cm⁻¹, observed in CLA, were absent in LS-CLA. In addition, the peak at 1709.5 cm⁻¹, associated with a conjugated double bond, was present in LS-CLA but not in P-LS-CLA (Figs. 1d–f). These results suggest that the binding mode of LS-CLA is mainly based on physical encapsulation, leaving the double bond structure of CLA exposed, whereas in P-LS-CLA, the conjugated double bond of CLA is effectively incorporated into the starch structure. Moreover, the infrared spectrum of lotus seed starch prior to debranching showed an intensified peak at 2925 cm⁻¹, corresponding to the methylene group, along with the emergence of a new C—H stretching vibration peak at 2850 cm⁻¹. These findings indicate that segments of the methylene and fatty acid chains of CLA encapsulated in LS-CLA and P-LS-CLA remain partially exposed outside the helical cavity.

By comparing the absorbance ratios at 1045/1022 cm⁻¹ and 995/1022 cm⁻¹, the effects of enzymatic hydrolysis and HHP treatment on the short-range ordered structure of microcapsules were evaluated (Wang et al., 2023). As shown in Table 1, the sample treated for 6 h at 300 MPa exhibited the highest degree of order (0.9759), indicating that the composite material formed under these conditions possessed the most ordered structure. This sample also showed the highest encapsulation efficiency (Wang, Jia, et al., 2025), suggesting that the self-assembly of amylose helices with CLA promotes structural order, whereas excessive debranching (12h) disrupts it. According to the degree of polymerization and branching data, it is suggested that the polymerization degree of amylose chains complexed with CLA and short linear chains should not be excessively low (Tan et al., 2025). Notably, the 0-h 300 MPa sample also exhibited a high degree of order (0.9712). This may be attributed to the cluster structure of amylopectin, which possesses higher crystallinity and readily forms lamellar crystals, whereas amylose and short linear chain segments exhibit comparatively lower crystallinity (Rindlav-Westling, Stading, & Gatenholm, 2002).

Table 1.

Structural Parameters of LS-CLA and P-LS-CLA.

Sample	1045/1022	995/1022	C1 (%)	C4(%)	C2,3,5(%)
0 h 300	0.9712	0.9679	11.222	15.458	58.081
0 h 500	0.9356	0.9362	10.874	16.981	57.787
0 h 600	0.9356	0.9449	11.453	15.400	57.597
3 h 300	0.9448	0.9451	16.542	14.477	55.980
3 h 500	0.9161	0.9450	14.423	15.322	54.498
3 h 600	0.8244	0.8236	13.030	15.066	54.157
6 h 300	0.9759	0.9769	14.430	16.360	55.565
6 h 500	0.9500	0.9377	14.395	12.754	55.366
6 h 600	0.9476	0.9359	13.23	15.690	53.519
12 h 300	0.9337	0.9332	15.511	14.133	54.534
12 h 500	0.9324	0.9615	13.342	15.151	53.990
12 h 600	0.9151	0.8972	13.504	15.80	53.476

Note:0 h, 3 h, 6 h, and 12 h represent the time of pullulanase action. 300, 500, and 600 represent different HHP pressures.

3.2. XRD

The crystalline structure of lotus seed starch CLA microcapsules was analyzed using wide-angle X-ray diffraction (XRD). As shown in Fig. 2a, the 0-h sample exhibited strong diffraction peaks at 12° and 20°, along with weaker peaks at 7°, 15°, 17°, and 25°. These patterns indicate that lotus seed starch retained a C-type crystalline structure (Guo et al., 2015). The peaks at 7°, 12°, and 20° further suggest that, under HHP conditions, a V6-type complex was formed between dissolved amylose and CLA (Figs. 2b–d) (Jia et al., 2018). In contrast, the samples subjected to 3 h and 6 h of debranching exhibited diffraction peaks at 7°, 12°, and 20°, confirming that their crystalline regions consisted exclusively of V-type crystals (Seok, Lee, Lim, & Chagam, 2019). When the debranching time was extended to 12 h, diffraction peaks at 15° and 17° reappeared, which are characteristic of A-type crystalline structure (Guo et al., 2015).

Fig. 2.

Wide-angle X-ray diffraction (XRD) patterns of LS-CLA and P-LS-CLA. 300, 500, and 600 MPa indicate the HHP treatment pressures, while “a,” “b,” “c,” and “d” correspond to LS-CLA and P-LS-CLA prepared with starch wall materials debranched for 0, 3, 6, and 12 h, respectively.

According to Pfannemüller (1987), starch chains with a degree of polymerization (DP) below 10 do not form crystals. Chains with DP values between 10 and 12 may form A-type crystals, whereas those with DP greater than 12 are more likely to form B-type crystals. Low-DP chains are therefore less favorable for complexation with CLA and are prone to polymerization and interchain rearrangement. These observations suggest that the crystalline structure of the microcapsules consists of a mixture of V-type and A-type crystals. This mixed pattern can be attributed to the 12-h debranching treatment, which significantly increased the proportion of A chains (DP ≤ 12) in the LS sample (Wang, Jia, et al., 2025), thereby hindering the formation of V-type complexes. Table 2 presents the crystallinity of LS–CLA microcapsules prepared under different conditions. According to Zheng et al. (2020), starch diffraction patterns can be divided into microcrystalline, semicrystalline, and amorphous regions, with their relative proportions quantified accordingly. As shown in Table 2, variations in microcapsule crystallinity mainly reflect differences in the proportions of microcrystalline and semicrystalline regions. In LS-CLA samples, the microcrystalline regions accounted for 5.6%, 8.7%, and 8.08%, while the semicrystalline regions accounted for 41.76%, 42.26%, and 40.91%. With increasing processing pressure, some lotus seed starch double helices are unfolded by HHP treatment, leading to the dissolution of amylose single chains, which subsequently form compact complexes with CLA, thereby increasing the proportion of the microcrystalline structure.

Table 2.

Crystallization Distribution of LS-CLA and P-LS-CLA.

Sample	Microcrystalline (%)	Semi-crystalline (%)	Crystalline (%)	Amorphous (%)
0 h 300	5.6	41.76	47.36	52.64
0 h500	8.70	42.26	50.96	49.04
0 h 600	8.08	40.91	48.99	51.01
3 h 300	29.40	19.21	48.61	51.39
3 h 500	24.52	24.90	49.42	50.58
3 h 600	28.94	19.10	48.04	51.96
6 h 300	32.09	24.91	57.00	43.00
6 h 500	26.47	20.65	47.12	52.88
6 h 600	30.69	16.26	46.95	53.05
12 h 300	19.69	23.97	43.66	56.34
12 h 500	23.47	20.10	43.57	56.43
12 h 600	29.49	20.30	49.79	50.21

Note:0 h, 3 h, 6 h, and 12 h represent the time of pullulanase action. 300, 500, and 600 represent different HHP pressures.

With increasing debranching time, the proportion of microcrystalline regions in the microcapsules gradually increased, reaching a maximum of 32.9% after enzymatic hydrolysis at 300 MPa for 6 h. In contrast, the semicrystalline region decreased to 24.91%. This change is attributed to the leaching of amylose from lotus seed starch during HHP treatment, which promotes the formation of a V6-type crystalline complex with CLA, which is characterized by a left-handed helical structure (Zheng et al., 2020). At this stage, some unbranched amylopectin remains in a partially depolymerized double-helix state, retaining a semicrystalline structure, while most amylopectin is cleaved during debranching, producing linear short chains. This promotes interactions between the nonpolar regions of CLA in solution, facilitating the formation of additional amylose single-helix–CLA complexes (Pengfei et al., 2019). The resulting compact and ordered structure of these complexes contributes to an increased proportion of microcrystalline regions.

3.3. NMR

The ¹³C nuclear magnetic resonance (NMR) spectra of LS–CLA microcapsules under various treatment conditions are shown in Fig. 3. Starch glucose units are six‑carbon cyclic compounds, and under ¹³C NMR, carbon atoms at different positions exhibit distinct chemical shifts. The spectra were processed using Origin 2025, and the proportions of the C₁, C_2,3,5, and C₄ regions were quantified (Table 1). According to Gidley (1985), the 96–106 ppm range corresponds to the C₁ region of starch. The 70–79 ppm range represents the C_2,3,5 regions, reflecting the mobility of amylose, while the 80–84 ppm range corresponds to the C4 region, mainly indicative of the amorphous characteristics of starch (Michael & Stephen, 1988) The chemical shift in the C₁ region can be used to identify the crystalline type of starch. As shown in Fig. 3, the 0-h samples exhibited prominent peaks at 103.5 ppm, along with weaker peaks at 101 and 100 ppm, indicating the retention of some C-type crystalline structure (Jia et al., 2018). At this stage, increased signal intensity was observed in the C_2,3,5 region, mainly owing to the presence of both amylose and amylopectin. In this mixed state, the mobility of amylose is constrained by the clustered structure of amylopectin (Jeong, Lee, & Chung, 2021). Moreover, a notable chemical shift at 30.5 ppm was assigned to the carbon chain of CLA, suggesting that the CLA chain had not yet fully entered the left-handed helical cavity of LS at this stage (Schahl, Lemassu, Jolibois, & Réat, 2022). Consequently, the CLA carbon chain demonstrates a relatively high degree of freedom, in agreement with the results obtained from FTIR spectroscopy.

Fig. 3.

¹³C NMR spectra of LS-CLA and P-LS-CLA prepared under different degrees of debranching and high hydrostatic pressure (HHP). 300, 500, and 600 MPa indicate the HHP processing pressures, while 0, 3, 6, and 12 h represent enzymatic debranching (ED) times.

After CLA encapsulation, the weak peaks at 101 and 100 ppm disappeared, and a single peak emerged at 103.5 ppm, indicating that the sample exhibited a V-type crystalline structure (Ge, Shi, Wu, Wei, & Cao, 2023). As shown in Table 1, compared with LS-CLA, the signal intensity of the C₁ region in P-LS-CLA was significantly increased, reflected by a higher area ratio. In contrast, the signal intensity of the C_2,3,5 region decreased, accompanied by a reduced degree of freedom of amylose. These results suggest that during microcapsule formation, the linear short-chain starch of P-LS was encapsulated by CLA, forming a compact complex (Ge et al., 2023). Furthermore, when HHP was raised from 300 MPa to 600 MPa, the proportion of the C1 region in P-LS-CLA decreased. This is attributed to excessive pressure causing the lotus seed amylose-CLA complex system to unwind, preventing some CLA and P-LS complexes from forming stable intermolecular hydrogen bonds and thereby disrupting the crystalline structure (Wang, Chen, et al., 2025). This observation aligns with the XRD results. Notably, the ¹³C NMR spectrum of P-LS-CLA exhibits a significant chemical shift at 31.6 ppm, distinct from the 30.5 ppm observed for LS-CLA. The peak at 31.6 ppm corresponds to the methyl group signal (Zhou, Guo, Gladden, Contreras, & Kong, 2022), indicating that the fatty acid chain of CLA is fully encapsulated by the amylose helix, with the methyl group at the chain's terminal end exposed at the helix's extremity. This finding underscores a fundamental difference in the encapsulation mechanisms of LS-CLA and P-LS-CLA with respect to CLA, supporting the conclusions drawn from FTIR spectroscopy.

3.4. Small-angle X-ray scattering (SAXS)

Most starch granules consist of alternating amorphous and crystalline layers, with thicknesses ranging from 100 to 400 nm. These “growth rings” form as a result of cumulative effects during starch biosynthesis. Typically, the structure exhibits a periodicity of 9–10 nm, arising from the radial arrangement of amylopectin side chains (Magallanes-Cruz, Flores-Silva, & Bello-Perez, 2017). The crystalline and amorphous layers play a key role in forming amylopectin side chain branches. In this study, small-angle X-ray scattering (SAXS) was employed to investigate the crystalline lamellar structures of LS-CLA and P-LS-CLA, as shown in Fig. 4a. Previous research by our group reported a prominent scattering peak at q = 0.068 Å⁻¹ in lotus seed starch granules, attributed to the repeating unit of the microcrystalline structure of LS (Jia et al., 2018). However, following debranching, HHP treatment, and CLA incorporation, this scattering peak disappeared, indicating disruption of the microcrystalline structure of native LS (Jia et al., 2018). Further analysis using the Bragg equation allowed determination of the repetition distance of the microcrystalline unit in LS. Notably, the absence of a repeating unit in LS-CLA (Table 3) indicates an amorphous structure, suggesting that LS-CLA lacks a regularly arranged crystalline lattice (Goderis, Dries, Nivelle, & Delcour, 2022). This loss of crystallinity is attributed to the presence of amylopectin, which inhibited recrystallization (Goderis et al., 2022). In P-LS-CLA, all samples exhibited scattering peaks within the range of 0.03–0.06 Å⁻¹, indicating a semicrystalline lamellar structure. To further assess sample density and surface roughness, the power-law equation (I ∼ q⁻ᵅ) was employed to calculate the fractal dimension (D_m) of samples prepared under different conditions (Jia et al., 2018). The α value was obtained by logarithmically transforming the scattering data and performing linear fitting. The fractal dimension D_m is positively correlated with sample density and reflects the complexity and order of the internal structure. As shown in Table 3, the D_m value of P-LS-CLA was consistently lower than that of LS-CLA, with a decrease in fractal dimension typically indicating a more ordered or regular internal structure (Seok et al., 2019). Debranching treatment removes branched chains from starch molecules, allowing amylose to form a compact complex with CLA and thereby enhancing structural regularity (Seok et al., 2019). The density of P-LS-CLA was lower than that of LS-CLA, consistent with scanning electron microscopy (SEM) observations, which is attributed to the network-like structure of branched starch.

Fig. 4.

Small-angle X-ray scattering (SAXS) of LS-CLA and P-LS-CLA. (a) Scattering log-transformed curves; (b) Fourier-transformed linear plots. 300, 500, and 600 MPa indicate HHP processing pressures, while 0, 3, 6, and 12 h represent enzymatic debranching (ED) times.

Table 3.

Lamellar Structure Parameters of LS-CLA and P-LS-CLA.

Sample	d (nm)	dc (nm)	da (nm)	ϕ (nm)	Dm
0 h 300		–	–	–	2.08
0 h 500	–	–	–	–	2.05
0 h 600		–	–	–	2.22
3 h 300		4.82 ± 0.060d	7.07 ± 0.045a	59.47 ± 0.149 a	1.89
3 h 500	11.88	4.86 ± 0.064cd	6.95 ± 0.105a	58.83 ± 0.050 b	2.05
3 h 600		5.00 ± 0.051ab	6.87 ± 0.097a	57.85 ± 0.160 c	1.81
6 h 300		5.10 ± 0.055b	6.39 ± 0.129bc	55.62 ± 0.254 g	1.84
6 h 500	11.45	5.07 ± 0.040b	6.40 ± 0.070bc	55.80 ± 0.075 fg	1.91
6 h 600		4.92 ± 0.055cd	6.57 ± 0.085b	57.19 ± 0.066 d	1.81
12 h 300		4.95 ± 0.045bc	6.27 ± 0.071c	55.89 ± 0.057 f	2.02
12 h 500	11.23	4.87 ± 0.056cd	6.40 ± 0.062bc	56.80 ± 0.050 e	1.91
12 h 600		4.85 ± 0.062cd	6.39 ± 0.061bc	56.84 ± 0.150 e	1.81

Note: Different lowercase letters within the same column indicate significant differences (p < 0.05). In the table, 3 h, 6 h, and 12 h correspond to different debranching times, while 300, 500, and 600 indicate treatment pressures in MPa. The symbols are defined as follows: d—average thickness of the semicrystalline lamellae; da—average thickness of the amorphous lamellae; dc—average thickness of the crystalline lamellae; ϕ—proportion of amorphous lamellae within the semicrystalline lamellae; and Dm—mass fractal dimension.

The distance between the amorphous and crystalline lamellae of the microcapsules was further analyzed using Fourier transformation of the scattering patterns (Fig. 4b). The analysis showed that LS-CLA lacked a repeating unit, indicating the absence of an ordered crystalline region. In contrast, P-LS-CLA subjected to debranching treatments for 3, 6, and 12 h exhibited microcrystalline repeating units of 11.88, 11.45, and 11.23 nm, respectively. The size of the repeating unit decreased with increasing debranching time, a trend attributed to enhanced linearization of starch molecular chains and strengthened intermolecular interactions, resulting in a more compact microcrystalline structure (Zhang, Li, Janaswamy, Chen, & Chi, 2020). This compact arrangement reduces the spacing of microcrystalline repeating units, as reflected by the narrowing of diffraction peaks. Furthermore, debranching treatment may promote a more uniform size and spacing of starch crystalline layers, further contributing to peak narrowing. In contrast, varying HHP does not significantly affect the spacing of microcrystalline repeating units (Zhang et al., 2020).

3.5. In vitro digestion

3.5.1. Digestion curve

As shown in Fig. 5a, all three samples exhibited high digestibility within the first 60 min, with the 6-h 300 MPa group showing the highest glucose content (80.28 mg/mL) during the 0–20-min period. This behavior is attributed to the linear short chains (DP ≤ 10) that remained free after debranching treatment and that did not form microcapsules with CLA through self-assembly. These short chains, existing in an amorphous single-helix state, were more readily digested and broken down into monosaccharides by digestive enzymes (Zhu, Liu, & Gilbert, 2024). Furthermore, the P-LS-CLA surface exhibited a uniform mesh structure, which increased the specific surface area, enhanced enzyme accessibility, and improved enzymatic hydrolysis efficiency within a short period (Gong et al., 2024). At 120 min of digestion, the glucose content in the supernatant reached its peak, indicating that the digestion process was nearly complete. Among the samples, native lotus seed starch showed the highest glucose concentration (178.59 mg/mL), followed by the 0-h 300 MPa sample (172.52 mg/mL), while the 6-h 300 MPa group exhibited the lowest concentration (116.96 mg/mL). The digestibility of the 0-h 300 MPa group did not differ significantly from that of native starch, suggesting that under this condition, most glucose release was due to physical entrapment within the dense amylopectin structure rather than the limited formation of V-type complexes between amylose and CLA. This finding aligns with the FTIR spectroscopy results. In contrast, the 6-h 300 MPa group experienced rapid digestion and breakdown of amorphous linear short chains within the first 20 min due to enzymatic activity. Consequently, the remaining semicrystalline structures showed a slower digestion rate (Guo et al., 2024), leading to the lowest glucose content during the 120–180-min period.

Fig. 5.

In vitro digestion of LS, LS-CLA, and P-LS-CLA. (a) Digestion curves; (b) Proportions of starch fractions. 300 MPa indicates HHP processing pressure, while 0 and 6 h represent enzymatic debranching (ED) times.

Further analysis of the starch microcapsule components revealed notable differences among the three groups—native starch, 0-h 300 MPa, and 6-h 300 MPa—as shown in Fig. 5b. Compared with native starch and the 0-h 300 MPa group, the 6-h 300 MPa group exhibited the highest contents of RS and RDS, at 25.5% and 42.2%, respectively, while the SDS content was the lowest at 32.3%. These differences help explain the observed variations in digestibility during the initial 0–20-min digestion phase. The increase in RS is attributed to the compact complex formed between amylose in LS and CLA. Notably, the RS content in the 0-h 300 MPa group was 9.61%, representing a 2.6% increase over that in native LS. This suggests that, aside from a small fraction of dissolved amylopectin forming complexes with CLA, the majority of CLA was physically mixed with the starch-based wall material, consistent with the conclusions from FTIR spectroscopy. According to these findings, subsequent studies further examined the morphological changes of the microcapsules during digestion.

3.5.2. Digestive morphology

Digestion products were collected at 20, 120, and 180 min, and the precipitates obtained after centrifugation were analyzed using SEM, with the results shown in Fig. 6. Fig. 6a illustrates the microstructure of native starch during digestion. During the initial stage, RDS on the microcapsule surface was degraded by digestive enzymes, leading to fragmentation, while a smooth, dense layer remained on the surface. As digestion progressed, pores appeared, and clustered spherical crystals formed. By the end of the 180-min digestion period, the initially smooth and dense structure exhibited numerous enzymatic hydrolysis lines and pores, resulting in the accumulation of flaky material.

Fig. 6.

Scanning electron microscopy (SEM) images of LS, LS-CLA, and P-LS-CLA at different digestion stages. Magnifications are 20,000× and 100,000×. (a) LS; (b) 0-h 300 MPa group; (c) 6-h 300 MPa group.

Fig. 6b and c show the microstructural changes during digestion of LS-CLA and debranched P-LS-CLA microcapsules, respectively (Gong et al., 2024). The observed changes are attributed to the degradation of RDS on the microcapsule surface (Ouyang et al., 2024), which exposes the underlying SDS and RS framework. This explains the higher digestion rate observed in the 6-h 300 MPa group during the initial stages. Specifically, Fig. 6c shows that P-LS-CLA exhibited fewer surface pores and a dough-like structure after 20 min of digestion, a phenomenon attributed to the adhesion of low-molecular-weight linear amylose short chains to the microcapsule surface (Shen et al., 2023). As digestion progressed to 120 min, the samples exhibited fragmentation and increased pore size, indicating that the amorphous RDS had been fully degraded, leaving semicrystalline structures interspersed with microcrystalline regions. By the end of the 180-min digestion period, SEM revealed that the final digested product consisted of uniform granular crystals (Fig. 6c). Post-digestion, the complexes reorganized through intermolecular forces, including hydrogen bonds and hydrophobic interactions, forming a stable structure with minimal energy. Consequently, the spherical morphology emerged as the most stable configuration due to its minimal surface area and lowest energy state (Sang, Xu, Zhu, & Narsimhan, 2021).

Previous studies have shown that starch spherulite formation is influenced by amylose chain length. High-molecular-weight amylose facilitates the development of numerous well-formed spherulites, whereas low-molecular-weight amylose inhibits spherulite formation (Hu, Julian McClements, Wang, & Li, 2022). This effect is attributed to the inability of short-chain or low-molecular-weight amylose to readily adopt chain-folded lamellar structures, which are essential for spherulite formation (Nordmark & Ziegler, 2002). Left-handed helical complexes formed between CLA and high-molecular-weight LS amylose provided the structural framework necessary for spherulite development (Jia et al., 2018). Previous studies have shown that linear amylose chains are more likely to form crystalline lamellar structures within spherulites, and spherulite development is influenced by the concentration of long-chain amylose (Mandelkern, Go, Peiffer, & Stein, 1977). As the degree of ED increases, spherulites become smaller and less well-formed, often exhibiting non-spherical morphologies. This effect is likely owing to the shorter chain length of amylose, which limits its extensibility and restricts spherulite growth, resulting in irregular morphology (Luo et al., 2020). In complexes of long-chain amylose with fatty acids, the fatty acid molecules are more uniformly incorporated within the amylose helical structure, forming a more compact complex that promotes regular spherulite formation. In contrast, the helical structure of short-chain amylose is less stable, leading to uneven distribution of fatty acid molecule distribution and irregular spherulite morphology (Nordmark & Ziegler, 2002). This helps explain the observed reduction in crystallinity and structural order of microcapsules following 12 h of ED and complexation with CLA.

4. Conclusion

This study explored the synergistic effects of ED and HHP on the multi-scale structure and in vitro digestibility of LS-CLA microcapsules to enhance the stability of heat-sensitive bioactives. Optimal conditions (6 h ED at 300 MPa) produced V-type complexes, as evidenced by FTIR (masked CLA double bonds), XRD (peaks at 7°, 12°, and 20°; 32.09% microcrystalline content), ¹³C NMR (strong C1 signals at 96–106 ppm and a peak at 31.6 ppm), and SAXS (semicrystalline lamellae with an 11.45 nm repeat distance) analyses. In vitro digestion using the DIVRSD model indicated 25.5% RS, significantly higher than the untreated group's 9.61%, while SEM revealed spherical crystalline residues, suggesting potential for colon-targeted CLA delivery. This non-thermal ED-HHP approach outperforms conventional thermal methods, supporting the need for in vivo validation and assessment of scalability for broader applications in sustainable functional foods.

CRediT authorship contribution statement

Bailong Wang: Writing – original draft, Software, Methodology, Data curation, Conceptualization. Ru Jia: Investigation, Data curation. Wenjing Chen: Software, Data curation. Zebin Guo: Writing – review & editing, Supervision, Resources, Project administration.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.