Food Chemistry 433 (2024) 137326

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Lecithin enhances the complexation between pea starch and fatty acids in
aqueous system, and affects the starch’s structure and enzymatic hydrolysis
Xiaoyang He a, b, Liyang Zhou a, c, Purnima Gunness b, Vicky A. Solah b, *, Qingjie Sun a, b, c, *
a
College of Food Science and Engineering, Qingdao Agricultural University, Qingdao, Shandong Province 266109, China
b
School of Medical, Molecular & Forensic Sciences, College of Environmental & Life Sciences, Murdoch 6150, Western Australia, Australia
c
Qingdao Special Food Research Institution, Qingdao 266109, China

A R T I C L E I N F O A B S T R A C T

Keywords: This paper investigates a newly found effect of lecithin on the complexation between starch and saturated fatty
Pea starch acids. Rapid visco analysis showed that adding lecithin to the pea starch-fatty acid mixtures resulted in a vis­
Fatty acids cosity peak during the setback stage of the pasting curve. Subsequent differential scanning calorimetry showed
Lecithin
that pea starch-fatty acid-lecithin systems formed more V-type structures than pea starch-fatty acid complexes. X-
Complexation
Enzymatic Hydrolysis
ray diffraction and Fourier transform infrared spectroscopy indicated the addition of lecithin developed the long-
range and short-range order of the V-complexes. Small-angle X-ray scattering showed the ternary system had a
more compact stack in nano-scale and smaller D bragg than the binary complex. In vitro enzymatic hydrolysis
revealed higher hydrolysis resistance of ternary systems compared to binary complexes. The results of this
research provide a mechanism for modifying starch-lipid complexes and contribute to scientific understanding of
food ingredient interactions.

1. Introduction (Gutiérrez, et al., 2021). Furthermore, type 5 RS has been shown to resist
digestion in the small intestine and be utilized by colonic bacteria
Starch is an essential macronutrient and is the primary energy source through fermentation. Compared to other types of RS, type 5 RS pro­
in the human diet (Chi et al., 2021). The interactions between starch and duces more butyric acid during colonic fermentation, thus having a
other food components, such as proteins, lipids, and volatile substances, prebiotic effect protecting the colonocytes against the onset of colorectal
have been shown to significantly influence starch properties (Wang cancer (Qin et al., 2021).
et al., 2020). Starch molecules, amylose in particular, can form a left- While it has been demonstrated that starch and fatty acids can form
hand single-helical structure, also called the V-type helical structure, complexes under certain conditions using chemical reagents such as
with multiple inclusion molecules such as lipids, polyphenols and al­ dimethyl sulfoxide (DMSO) or strong alkaline solutions, these conditions
cohols; these complexes are formed by non-covalent interactions are not practical or commonly encountered in food systems (Wang et al.,
involving hydrogen bonds, hydrophobic interactions, and/or van der 2020). In aqueous systems, the complexation is not optimized as it
Waals force (Deng et al., 2021; Gutiérrez & Tovar, 2021; Oyeyinka, would be with DMSO or strong alkaline conditions due to the low sol­
Singh, & Amonsou, 2021; Wang et al., 2020). ubility of fatty acids. Recent research has revealed that the complexation
Fatty acids are common guest molecules for starch complexation. between starch and fatty acids in aqueous systems is influenced by
The single helical complexes are more compact in structure; the struc­ multiple factors including heating temperature and time as well as the
ture prevents contact between digestive enzymes and susceptible sites type of fatty acids. Additionally, ingredients other than the guest mol­
within starch, thereby enhancing its resistance to hydrolysis (Deng et al., ecules also impact the formation of complexes. For instance, whey
2021; Gutiérrez & Tovar, 2021; Oyeyinka et al., 2021; Wang et al., protein and beta-lactoglobulin (Zhang, Maladen, & Hamaker, 2003,
2020). Consequently, complexes formed by starch and inclusion mole­ Zhang, Maladen, Campanella, & Hamaker, 2010), sodium chloride (Niu
cules, including fatty acids, are identified as type 5 resistant starch (RS) et al., 2019), and triglycerides (Li et al., 2020) have all been reported to

* Corresponding authors at: School of Medical, Molecular & Forensic Sciences, College of Environmental & Life Sciences, Murdoch University, Murdoch 6150,
Western Australia, Australia. (V. Solah). College of Food Science and Engineering, Qingdao Agricultural University, 266109, 700 Changcheng Road, Chengyang
District, Qingdao, China (Q. Sun).
E-mail addresses: Vicky.Solah@murdoch.edu.au (V.A. Solah), phdsun@163.com (Q. Sun).

https://doi.org/10.1016/j.foodchem.2023.137326
Received 11 April 2023; Received in revised form 12 August 2023; Accepted 28 August 2023
Available online 30 August 2023
0308-8146/© 2023 Elsevier Ltd. All rights reserved.
X. He et al. Food Chemistry 433 (2024) 137326

affect the complexation process between starch and fatty acids, despite 2.2. Rapid Visco analysis
exhibiting minimal complexation ability with starch themselves.
Notably, beta-lactoglobulin will form ternary complexes with starch and Rapid Visco Analyzer (Model 4D, Perkin Elmer, Australia) was used
fatty acids, enhancing the long-range and short-range order of the starch to examine the pasting profiles of the pea starch and different fatty acid
molecule and improving the resistance to enzymatic hydrolysis of the mixtures with or without lecithin. The method described by Cai et al.
resulting complex (Zheng et al., 2018). These findings highlight the was followed with minor modifications (Cai et al., 2021). For the
potential to further study how food ingredients affect the complexation mixture of pea starch and fatty acids, 2.0 g of pea starch (wet weight
between starch and fatty acids, leading to optimized complex formation basis) and 100 mg of the different fatty acids (Lauric acid (LA), myristic
and improved physicochemical functional properties such as acid (MA), palmitic acid (PA), and stearic acid (SA) were added into RVA
digestibility. canisters. Then 25.9 g of deionized water was weighed and added into
It is well-known that the complexation process affects the pasting the canister to make a pea starch and fatty acid mixture with a total
properties of starch (Tang & Copeland, 2007). Altered functional weight of 28.0 g. To prepare the pea starch, fatty acid, and lecithin
properties caused by the complexation process and the influence of mixture, 100 mg of lecithin was dispersed in 5.0 g of deionized water,
specific factors can be observed in pasting curves generated by a Rapid and 2.0 g of pea starch (wet weight) and 100 mg of the different fatty
Visco Analyzer (RVA). For instance, Zhang & Hamaker (2003) demon­ acids were added into the canister before being made up to a total
strated that whey protein is capable of forming a ternary complex with weight of 28.0 g with deionized water. The mixture of pea starch and
starch and fatty acids. They initially detected this interaction by lecithin was prepared in the same way without fatty acid.
observing a distinct setback stage peak in the rheological RVA pattern of All test mixtures were vigorously stirred using the paddle prior to
the ternary mixture. During the RVA experiments conducted, we found testing ensure samples were homogenous. The RVA Standard 1 (STD 1)
that a mixture of pea starch, fatty acid, and lecithin displayed a prom­ program was used to generate the pasting curves. For the first 10 s, the
inent peak at the setback stage. Conversely, pea starch, pea starch mixed speed of the paddle was set at 960 rpm before being decreased to 160
with lecithin, and pea starch mixed with fatty acids did not show such a rpm for the remainder of the mixing time. The samples were held at
peak at the setback stage. These results suggest potential effect of leci­ 50 ◦ C for 1 min and then heated to 95 ◦ C with a heating rate of 12 ◦ C/
thin on starch and fatty acid complexation. min, held at 95 ◦ C for another 3 min, and then cooled to 50 ◦ C at a rate of
Therefore, the current study constructed a novel ternary system 12 ◦ C/min. After being maintained at 50 ◦ C for 4 min, the profiles were
containing pea starch, food-grade lecithin, and fatty acids with different obtained. RVA pasting profiles of the 2.0 g (wet weight) of pea starch
chain lengths, and compared the complexes from ternary systems with was generated as control.
conventional binary systems of starch and fatty acids for understanding
how some molecules like lecithin affects the complexation process. We 2.3. Binary and ternary systems preparation
hypothesized that adding lecithin may enhance the formation of V-type
structures between pea starch and fatty acids, consequently resulting in The binary and ternary complex systems of PS-fatty acids and PS-
the appearance of a setback stage peak in the RVA curve. In this work, fatty acids-LE were prepared following the method described by
RVA, differential scanning calorimetry (DSC), X-ray diffraction (XRD), Zhang et al. (2003) with minor modifications. Pea starch (4.0 g, dry
Fourier transform infrared spectroscopy (FTIR), small-angle X-ray scat­ weight basis) was added to 80 mL of deionized water in conical flasks,
tering (SAXS), in vitro enzymatic hydrolysis experiment, and scanning thoroughly mixed and sealed with aluminum foils. Lecithin was
electron microscopy (SEM) was employed to analyze the binary and dispersed in water to attain a concentration of 10 mg/mL by stirring
ternary systems. This study aims to expand our knowledge about the until a liquid crystalline form of lecithin in water was formed. All the
interactions among food ingredients and provide a new approach for fatty acids solution were prepared by dissolution in ethanol and made to
promoting complexation and enhancing the hydrolysis resistance of a concentration of 40 mg/mL. The pea starch dispersion was pre-
starch-lipid complexes. The findings may have potential to applications gelatinized in a water bath maintained at 95 ◦ C with constant stirring
for the development of functional foods with increased RS providing at 400 rpm for 30 min and was used to prepare the binary, ternary
health benefits for individuals with metabolic syndrome or diabetes. complex systems, and control samples.
Samples for binary systems: 5 mL fatty acids/ethanol solution and 20
2. Materials and methods mL water were added into the pre-gelatinized pea starch flasks (pea
starch/fatty acid is 20:1, dry weight basis) with heating at 95 ◦ C and
2.1. Materials continued stirring for another 30 min. After cooling to room tempera­
ture, the resulting products were the starch-fatty acid complex formed in
Pea starch (PS, 12.9% moisture content, 32.2% amylose content, the binary system, which were denoted as PS-LA, PS-MA, PS-PA, and PS-
0.04% total fat, and average molecular weight of 9.086 × 107 Da, SA according to the type of fatty acid used. The aluminum foil was
method used for measuring average molecular weight (Mw) by size removed at the first 5 min of the heating and stirring stage to allow the
exclusion chromatography (SEC) is shown in the menthod section of volatilization of the ethanol.
supporting information document) was provided by Qingdao Haider Samples for ternary systems: 5 mL fatty acids/ethanol solution and
Starch Co., Ltd. (Qingdao, China). Lauric acid (LA, C12:0), myristic acid 20 mL lecithin (LE) dispersion were added into the pre-gelatinized pea
(MA, C14:0), palmitic acid (PA, C16:0), and stearic acid (SA, C18:0) starch flasks (pea starch/fatty acid/lecithin is 20:1:1, dry weight basis)
were all analytical grade. Lauric acid was purchased from Tianjin BASF with heating and continuous stirring for another 30 min. After cooling to
Chemical Co., Ltd. (Tianjin, China); myristic acid and palmitic acid were room temperature, the resulting products were the starch-fatty acid
purchased from Tianjin Bodi Chemical Co., Ltd. (Tianjin, China); and complex formed in ternary systems; these were denoted as PS-LA-LE, PS-
stearic acid was purchased from Tianjin Beilian Fine Chemicals Devel­ MA-LE, PS-PA-LE, and PS-SA-LE. The aluminum foil was removed at the
opment Co., Ltd. (Tianjin, China). Commercial lecithin from soybean first 5 min of the heating and stirring stage to allow the volatilization of
(containing 97% total phospholipids, mainly phosphatidylinositol 26%, the ethanol.
phosphatidylethanolamine 20%, and 11% phosphatidylinositol) was The pea starch control (denoted as PS) was prepared by adding 20
food-grade and purchased from Cargill Inc. Germany (Hamburg, Ger­ mL water into pre-gelatinized pea starch flask with continuous heating
many). Porcine pancreatic α-amylase and amyloglucosidase were pur­ and stirring for another 30 min. Pea starch-lecithin control (denoted as
chased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). D- PS-LE, pea starch/lecithin is 20:1, dry weight basis) was prepared by
Glucose Assay Kit (glucose oxidase/peroxide, GOPOD) was obtained adding 20 mL lecithin dispersion into the pre-gelatinized pea starch with
from Megazyme International Ireland Ltd. (Wicklow, Ireland). continuous heating and stirring for another 30 min.

2
X. He et al. Food Chemistry 433 (2024) 137326

The ratios of each compound in the ternary, binary, and control 2.6. X-ray diffraction
systems were shown in Table S2. A portion of the fresh sample pastes
were collected at 50 ◦ C during the cooling phase and immediately used The crystalline structure of the binary and ternary complexes was
in the complex index experiment (Section 2.4). The remaining samples evaluated by an X-ray diffractometer (AXS D8 ADVANCE; Bruker,
were oven-dried at 45 ◦ C for 24 h, ground into powder, washed twice Karlsruhe, Germany). The moisture content of each sample from section
with ethanol, and passed through a 150 μm sieve. The resulting fine 2.3 was equilibrated at 75% relative humidity for 48 h. The equilibration
powders were stored in a desiccator at room temperature prior analysis. was conducted by placing samples in a desiccator with a saturated so­
dium chloride solution. The samples were placed on a glass plate for
2.4. Complex index XRD detection. The scanning range was 4 − 40◦ (2θ) at a scanning rate of
2◦ /min, and a 0.01◦ step size. The voltage and current were 36 kV and
Complex index (CI) experiment was carried out as described by Tang 20 Ma respectively. The relative crystallinity (RC) was then calculated as
& Copeland’s with slight changes (Tang & Copeland, 2007). The sample the ratio of crystalline peak area (Ac) to total area (A), and the V-type
pastes prepared in section 2.3 were collected immediately when the crystallinity (RC-v) was calculated as the ratio of the V-type crystalline
temperature of the samples dropped to 50 ◦ C (measured by a mercury peak areas (Av) (peaks at 7.4◦ , 12.8◦ , and 19.8◦ ; Av) and A. The formulas
thermometer). This allowed the complexation to occur and minimized are provided below:
retrogradation. The paste (3.0 g) was then added to 15 mL deionized
RC(%) = Ac/A × 100
water at 50 ◦ C in a capped tube and vortexed. 200 μL the resulting
dispersion was further mixed with 15 mL water at 50 ◦ C and 2 mL iodine
RC − v (%) = Av/A × 100
solution containing 2.0% KI and 1.3% I2 in water. Absorbance was
measured at 690 nm by a UV–vis spectrophotometer (TU-1810; Beijing The full width at half maximum (FWHM) of V-type peaks were
Persee Co., Ltd., Beijing, China) to calculate the complex index. The measured for evaluating the crystallite size.
entire experiment was carried out within 20 min to avoid the effect of
the retrogradation. The absorbance measurements were then used to 2.7. Fourier transform infrared (FTIR) spectroscopy
calculate the complex index according to the following equation:
CI = (ABSpeastarch − ABSsamples )/ABSpeastarch The FTIR spectra of samples were recorded using an FTIR spectro­
photometer equipped with an attenuated total reflection (ATR). The
Where ABSpea starch is the absorbance of gelatinized pea starch con­ parameters of spectrophotometer were adjusted to the following set­
trol, ABSsamples is the absorbance of different samples. tings: the wavenumber range 4000 to 400 cm− 1, 64 scans at a resolution
of 4 cm− 1 were taken for each sample obtained from section 2.3. The
2.5. Differential scanning calorimetry ratio of the peak intensities at 1047 cm− 1 and 1022 cm− 1 (1047/1022)
was calculated.
The thermal properties of the samples were determined by a DSC-3,
differential scanning calorimeter (Mettler Toledo, Schwerzenbach,
2.8. Small angle X-ray scattering (SAXS)
Switzerland). Three to five milligrams samples were precisely weighed
in 40 μL aluminum pans. Deionized water was added to the pans
The SAXS analysis was carried out on a Nano-inXider SAXS system (λ
(tolerance to mid-pressure) using a micro syringe to provide a weight
= 0.154 nm, Xenocs, Sassenage, France) operated at 50 kV and 60 mA.
ratio of 1:2 (sample:water). The pans were then sealed and equilibrated
The moisture content of PS, PS-LE, PS-LA, and PS-LE-LA was adjusted to
for 12 h at room temperature. The first scan of the samples was carried
70% by directly mixing 0.09 g sample (dry weight basis) with 0.21 g
from 25 ◦ C to 125 ◦ C at a rate of 10 ◦ C/min. The samples were then
water; and equilibrated in a sealed centrifugal tube for 24 h. The mixture
cooled down from 125 ◦ C to 25 ◦ C at a rate of 10 ◦ C/min and the cooling
of samples with water was sealed in a gel/powder sample holder with
curves obtained. The same protocol was applied for the second scan of
Kapton films with a thickness of 0.01 mm. The thickness of the samples
samples after cooling to room temperature. Every test was carried out
was 1 mm. The beam size of these measurements was 400 µm, and the
with an empty pan as the reference. Onset temperature (To), peak
exposure time was 300 s.
temperature (Tp), end temperature (Te), and enthalpy changes (ΔH)
were calculated and presented in Table 1.
2.9. In vitro enzymatic hydrolysis

The in vitro enzyme hydrolysis rate of samples was measured using

Table 1
The temperature and enthalpy change (ΔH) of peaks of PS, PS-LE, PS-fatty acids, and PS-fatty acids-LE samples between 70 and 125 ◦ C a.
Samples First Peak (70–125 ◦ C) Second Peak (70–125 ◦ C)

To ( C)
◦
Tp ( C)
◦
Tc ( C)
◦
ΔH (J/g) To (◦ C) Tp (◦ C) Tc (◦ C) ΔH (J/g)

PS – – – – – – – –
PS-LE – – – – – – – –
ab
PS-LA 88.51 ± 0.9 97.12 ± 0.2a 102.03 ± 0.31a
− 1.66 ± 0.05e 111.39 ± 0.63a 116.205 ± 1.005a 120.175 ± 0.645a − 1.86 ± 0.13b
PS-MA 90.74 ± 1.26bc 99.08 ± 0.14b 104.96 ± 0.91b − 1.78 ± 0.06e 114.915 ± 0.745a 118.485 ± 0.585a 124.225 ± 0.105b − 0.59 ± 0.19c
PS-PA 95.14 ± 0.34d 103.12 ± 0.01d 107.62 ± 0.07 cd − 1.3 ± 0.05f – – – –
PS-SA 91.39 ± 0.55bc 101.4 ± 0.48c 106.85 ± 0.15bcd − 0.55 ± 0.15 g – – – –
PS-LA-LE 86.61 ± 1.12a 96.53 ± 0.02a 102.45 ± 0.11a − 3.35 ± 0.09b 110.49 ± 1.89a 116.315 ± 0.195a 121.705 ± 0.225a − 3.95 ± 0.14a
PS-MA-LE 91.26 ± 0.29bc 101.08 ± 0.09c 105.84 ± 0.04bc − 4.01 ± 0.05a 112.725 ± 0.785a 117.33 ± 0.48a 121.43 ± 0.62a − 0.9 ± 0.04c
PS-PA-LE 93.52 ± 1.08 cd 102.56 ± 0.6d 106.47 ± 1.41bcd − 2.84 ± 0.1c – – – –
PS-SA-LE 89.45 ± 1.39ab 102.52 ± 0.31d 108.06 ± 0.3d − 2.24 ± 0.04d – – – –
a
Values are means ± SD. Means with different letters in a column differ significantly (p＜0.05). All experiments were repeated 3 times. PS is for pea starch, PS-LE is
for pea starch and lecithin, PS-LA is for pea starch and lauric acid, PS-MA is for pea starch and myristic acid, PS-PA is for pea starch and palmitic acid, PS-SA is for pea
starch and stearic acid, PS-LA-LE is for pea starch, lauric acid, and lecithin, PS-MA-LE is for pea starch, myristic acid, and lecithin, PS-PA-LE is for pea starch, palmitic
acid, and lecithin, and PS-SA-LE is for pea starch, stearic acid, and lecithin.

3
X. He et al. Food Chemistry 433 (2024) 137326

the Englyst method (Englyst & Cummings, 1985) with minor modifi­
cations. Firstly, three grams of porcine pancreatic α-amylase were using
a vortex-mixed in 20 mL deionized water for 5 min, and centrifuged at
3000 rpm for 10 min. After centrifugation, 15 mL supernatant was
carefully transferred to another tube, and 1.1 mL amyloglucosidase was
added. After 1 min vortex-mixing, the enzyme working solution was
maintained at 37 ◦ C.
Samples were weighed (0.1 g dry weight basis) and added to 4 mL
acetate buffer (pH = 5.2) in a tube with glass beads. The tubes were held
at 37 ◦ C for 15 min before 1 mL of the enzyme working solution was
added and the tubes were incubated at 37 ◦ C for 2 h. At 0 min, 20 min,
and 120 min, 0.05 mL from each tube was subsampled. Subsamples were
added to 0.95 mL of 95% ethanol to deactivate enzymes. Glucose con­
tent was measured using the reagents and instructions of the Megazyme
D-glucose assay kit (GOPOD kit). Rapidly digestible starch (RDS), slowly
digestible starch (SDS), and resistant starch (RS) contents were calcu­
lated using the following equations:
RDS(%) = G20 × 0.9 × 100

SDS(%) = (G120 − G20) × 0.9 × 100

RS(%) = (TS − RDS − SDS) × 100

where G20 means the glucose content at 20 min, G120 means the
glucose content at 120 min, and TS is the total starch content.

2.10. Scanning electron microscopy

The microstructures of PS, PS-LE, PS-LA, and PS-LE-LA samples were

observed using a scanning electron microscope (JEOL 7500F, Japan) at
an accelerating voltage of 5 kV. All samples were adhered to the spec­
imen holder with conductive adhesive. A layer of gold particles was
homogeneously coated on the samples. The microstructure of all sam­
ples at 1000x magnification was taken by bundled software.

2.11. Statistical analysis

All experiments were conducted in triplicate. IBM SPSS statistics Fig. 1. RVA profiles of a) PS, PS-LE, and PS-fatty acids mixtures, and b) PS, PS-
version 23.0 (IBM, Armonk, NY, USA) was used to analyze the variance LE, PS-fatty acids-LE mixtures a. a PS is for pea starch, PS-LE is for pea starch
(ANOVA). At a significance level of 95% (P < 0.05), Duncan’s multiple- and lecithin, PS-LA is for pea starch and lauric acid, PS-MA is for pea starch and
range tests were employed to define the significant differences. Results myristic acid, PS-PA is for pea starch and palmitic acid, PS-SA is for pea starch
are presented as mean ± standard deviation (SD). and stearic acid, PS-LA-LE is for pea starch, lauric acid, and lecithin, PS-MA-LE
is for pea starch, myristic acid, and lecithin, PS-PA-LE is for pea starch, palmitic
3. Results and discussion acid, and lecithin, and PS-SA-LE is for pea starch, stearic acid, and lecithin.

3.1. Rapid viscosity analysis of binary and ternary mixtures conditions. These conditions are known to induce saponification of the
fatty acids and improve their dispersion as well as promote complexa­
The RVA profiles of binary and ternary mixtures are shown in Fig. 1 tion (Niu et al., 2020). In addition, peak viscosity occured at the setback
(a) and (b) respectively. The values of peak viscosity and occurrence stage when RVA pasting was conducted with starch mixed with mono­
time of peaks that appeared at the setback stage are presented in glycerides; monoglycerides complex more readily with starch with
Table S1. Addition of the different fatty acids as well as lecithin resulted heating and stirring as their emulsifying ability improves dispersion in
in changes in the viscosity curves when compared to PS. Adding lecithin water (Blazek, Gilbert, & Copeland, 2011; Tang & Copeland, 2007).
increased viscosity. PS mixed with shorter chain fatty acids (lauric acid Sufficient V-complex formation is likely to lead to a high viscosity peak
and myristic acid) exhibited higher viscosity than pea starch itself, in the setback stage of the pasting curve. Peak viscosities of a ternary
which could be due to the formation of V-complexes. The mechanism by system with saturated fatty acids decreased as the chain length of fatty
which the V-complexes increase viscosity maybe that larger spaces were acids increased (from lauric acid to stearic acid) (Table S1). This may be
created by the gel cross linkages as compared to native starch; an in­ due to shorter chain fatty acids more easily forming V-complexes in the
crease in final viscosity has also been reported by Tang & Copeland neutral water phase (Zheng et al., 2018). Tang & Copeland (2007) re­
(2007). The results suggest that shorter chain fatty acids (LA and MA) ported that lauric acid forms more V-complexes with starch using the
might have formed more complexes with PS in aqueous systems. Standard 1 profile of the RVA in water than longer chain fatty acids; this
Interestingly, all ternary systems displayed a significant peak at the may be related to the water solubility and a critical micellar concen­
setback stage of the RVA pasting curve (Fig. 1b). Furthermore, the vis­ tration of the lipids (Tang & Copeland, 2007).
cosity of the setback stage peaks was higher than that those of the peaks
formed during the heating process. Similarly, peak viscosity during 3.2. Complex index of binary and ternary systems
setback was observed in a recent study that conducted Rapid Visco
Analysis of starch and fatty acid mixtures under high pH (over 10) The complex index (Iodine Value) of the samples prepared as

4
X. He et al. Food Chemistry 433 (2024) 137326

described in section 2.3 is shown in Table 2. The complexation index can promoting their complexation with starch. The chain length of fatty
be used to evaluate the complexation process because the starch-iodine acids showed a similar pattern on the ternary systems compared to the
complex is colored. As the number of complexed starch molecules in­ binary systems. Systems containing fatty acids with longer chain lengths
creases, fewer are available to form iodine complexes. The Iodine Value exhibited lower complex index values.
may also be reduced by the retrogradation process. In this work, retro­
gradation was minimized as described in section 2.4. 3.3. Thermal properties and self-assembly behavior of binary and ternary
As shown in Table 2, lecithin alone could not be complexed with systems
starch. Early research found the complexation ability of lecithin with
native starch in the aqueous phase was very weak due to the steric DSC was employed to understand the thermal properties and self-
hindrance of the double aliphatic chain (Eliasson, 1986; Krog, 1971). All assembly behavior of PS-fatty acid and PS-fatty acid-LE systems.
samples from the binary systems showed a complex index lower than Fig. 2a and 2b show the DSC curves of PS, PS-LE, PS-fatty acid, and PS-
45%, which may be due to the limited reaction time, relatively low fatty acid-LE systems. As mentioned in section 2.3, the PS control was
dispersion ability of the fatty acids, and/or the relatively high gelatini­ gelatinized and therefore no significant peaks appeared during the
zation temperature of pea starch limiting the release of available scanning process. DSC patterns of PS-LA, PS-MA, PS-PA, PS-SA, (Fig. 2a)
amylose. Furthermore, the complex index of samples from the binary and PS-LA-LE, PS-MA-LE, PS-PA-LE, PS-SA-LE (Fig. 2b) showed narrow
systems reduced as the chain length of the fatty acids decreased. In the endothermic peaks at 40–65 ◦ C, which were attributed to the melting of
aqueous phase, starch complexation could be preferential with shorter free fatty acids (Lu et al., 2021). Both binary complexes of PS-LA, PS-
chain fatty acids. This can be explained by previous studies that sug­ MA, PS-PA, PS-SA and ternary systems of PS-LA-LE, PS-MA-LE, PS-PA-
gested that fatty acids with shorter hydrophobic chains showed higher LE, and PS-SA-LE samples exhibited significant peaks between 85 and
critical micelle concentration and solubility (Tang & Copeland, 2007). 125 ◦ C. Typically, peaks at this temperature range suggest the dissoci­
The binary system results were consistent with many previous studies on ation of the V-type helical structures (Zheng et al., 2018). These results
starch-fatty acid complexation in the water phase and finite mixing time confirmed the evidence of V-complexes formed between starch and the
(Kawai et al., 2016; Soong, Goh, & Henry, 2013; S. Wang, Wang, Yu, & fatty acids. Interestingly, the PS-LA, PS-MA, PS-LA-LE, and PS-MA-LE
Wang, 2016). curves contained double peaks between 85 ◦ C and 125 ◦ C, which indi­
Although the ternary systems also displayed higher complex index cated the formation of two types of single-helical structures (I type and II
values with the shorter chain length fatty acids, compared to the cor­ type V-complex); only one peak was found in the pattern of PS-PA, PS-
responding binary systems, adding lecithin significantly increased the SA, PS-PA-LE, and PS-SA-LE (VI type complex) (Lu et al., 2021). The type
complex index. These results indicate the formation of more complexes of V-complex formed in different systems could be affected by many
in the ternary systems compared to the binary systems; lecithin may factors, including the ratio of amylose and guest molecules (Biliaderis &
enhance the formation of V-complex between starch and fatty acids, Galloway, 1989), the heating temperature and period (Karkalas et al.,
irrespective of their chain lengths, resulting in the peak during the 1995) and the type of guest molecules (Biliaderis et al., 1985). Since the
setback stage of the RVA pasting curve. The reason for this significant addition of lecithin in the system did not induce the change of temper­
enhancement may be due to the amphipathic and emulsifying properties ature and number of peaks caused by the dissociation of V-complexes, it
of lecithin promoting the dispersion of fatty acids in the system, and then could be concluded that lecithin did not affect the type of V-crystals and
their thermal stability (Table 1). The first peaks in Fig. 2b (85–110 ◦ C) of
the ternary systems of PS-LA-LE and PS-MA-LE and the binary systems of
Table 2
PS-LA and PS-MA displayed the formation of a type I V-complex between
The total relative crystallinity (RC) and V-type relative crystallinity (RC-v),
PS and fatty acids. The second peak (110–125 ◦ C) of PS-LA, PS-MA, PS-
complex index, and ratio of 1047/1022 of FTIR curves of PS, PS-LE, PS-fatty
acids, and PS-fatty acids-LE samples a.
LA-LE, and PS-MA-LE revealed the existence of type II V-complex
structure. In addition, the intensity of peaks in ternary systems was
Samples RC (%) RC-v (%) Ratio of 1047/ Complex Index
significantly stronger than in binary systems with the same fatty acids
1022 (%) (%)
and was strengthened with the shorter fatty acids chain. Conversely, the
PS 13.89 ± 56.06 ± 0.12a
– –
curves of PS-LE displayed no peaks from 70 ◦ C to 125 ◦ C. These results
0.74f
PS-LE 28.38 ± 3.47 ± 58.55 ± 0.64b
–
indicate that it was difficult for PS-LE to form complexes in this system
2.34bc 0.20a by simply mixing, heating, and stirring. The enthalpy changes of the PS
PS-LA 23.26 ± 14.78 ± 60.88 ± 1.16 cd 43.51 ± 0.86d control, PS-LE, PS-fatty acid, and PS-fatty acid-LE samples were calcu­
2.48a 1.42d lated and are shown in Table 1. Generally, the enthalpy change reflects
PS-MA 25.8 ± 9.67 ± 58.76 ± 0.27bc 39.94 ± 0.34c
the formation of the amount of ordered structure. The larger the value of
0.54ab 0.19bc
PS-PA 29.97 ± 5.229 ± 59.09 ± 0.42bc 27.47 ± 0.14b enthalpy change, the more ordered helixes are formed (Wang et al.,
1.6bc 0.15ab 2020; Wang et al., 2017; Zheng et al., 2018). The enthalpy changes of
PS-SA 35.08 ± 6.58 ± 59.37 ± 0.18bcd 10.07 ± 1.60a PS-fatty acid and PS-fatty acid-LE complex systems decreased with the
1.34d 0.24b
increase in the chain length of the fatty acids. This suggests that shorter-
PS-LA-LE 47.15 ± 34.14 ± 63.11 ± 0.66f 74.52 ± 0.37 h
1.78e 2.28f
chain fatty acids could form more ordered structures in binary and
PS-MA- 36.18 ± 30.37 ± 61.32 ± 1.27de 71.40 ± 0.84 g ternary systems than longer-chain fatty acids, which was consistent with
LE 0.73d 0.97e the CI result. Furthermore, the enthalpy changes of the ternary systems
PS-PA-LE 35.17 ± 32.21 ± 61.53 ± 0.26de 64.65 ± 0.41f were greater than those of the corresponding binary systems. For
0.86d 0.57ef
example, compared to PS-LA (− 3.515 J g− 1), the enthalpy change of PS-
PS-SA-LE 42.34 ± 34.13 ± 61.44 ± 1.3de 60.22 ± 0.33e
0.32e 0.48f LA-LE（− 7.30 J g− 1）was doubled (Table 1). Similarly, the other
a
ternary systems with myristic acid, palmitic acid, and stearic acid
Values are means ± SD. Means with different letters in a column differ
showed at least a doubled enthalpy change compared to the corre­
significantly (p＜0.05). All experiments were repeated 3 times. PS is for pea
sponding binary system. In general, enthalpy changes in the range of
starch, PS-LE is for pea starch and lecithin, PS-LA is for pea starch and lauric
acid, PS-MA is for pea starch and myristic acid, PS-PA is for pea starch and 85–120 ◦ C are caused by the melting of V-type crystalline structure; this
palmitic acid, PS-SA is for pea starch and stearic acid, PS-LA-LE is for pea starch, result suggests that the PS-fatty acid-LE ternary systems could form more
lauric acid, and lecithin, PS-MA-LE is for pea starch, myristic acid, and lecithin, ordered structures than the PS-fatty acid binary systems. Consequently,
PS-PA-LE is for pea starch, palmitic acid, and lecithin, and PS-SA-LE is for pea higher enthalpy in ternary systems supports that lecithin enhanced the
starch, stearic acid, and lecithin. complexation between starch and fatty acid.

5
X. He et al. Food Chemistry 433 (2024) 137326

Fig. 2. The DSC curves of a) PS, PS-LE, and PS-fatty acids, b) PS-fatty acids-LE, c) rescan curves of PS-fatty acids and PS-fatty acids-LE, and cooling curves of d) PS-
fatty acids and e) PS-fatty acids-LE samples. a. a PS is for pea starch, PS-LE is for pea starch and lecithin, PS-LA is for pea starch and lauric acid, PS-MA is for pea starch
and myristic acid, PS-PA is for pea starch and palmitic acid, PS-SA is for pea starch and stearic acid, PS-LA-LE is for pea starch, lauric acid, and lecithin, PS-MA-LE is
for pea starch, myristic acid, and lecithin, PS-PA-LE is for pea starch, palmitic acid, and lecithin, and PS-SA-LE is for pea starch, stearic acid, and lecithin.

6
X. He et al. Food Chemistry 433 (2024) 137326

The DSC graphs of the rescanning heating process (Fig. 2c) showed The type of fatty acids also significantly influenced the crystalline
that the peaks that appeared in the first scan did not disappear in the structure in the different systems studied. For binary systems, PS and
second scan. This indicates that the complexes formed by pea starch and fatty acids with lower chain lengths (lauric acid and myristic acid)
saturated fatty acids with or without LE could be reassembled after preferentially formed V + B crystal structures dominated by V-type.
heating. The DSC curves at the cooling stage after the first heating Conversely, PS and fatty acids with a chain length longer than 16 were
process were also obtained (Fig. 2d and 2e). The exothermic peaks at 65 found to form fewer V-type crystal structures. Compared to PS-SA and
to 85 ◦ C indicated that the reformation of PS-fatty acids or PS-fatty PS-PA systems, PS-SA-LE and PS-PA-LE systems exhibited a distinct XRD
acids-LE complexes occurred during the cooling process. Peaks below pattern, in which the dominating peaks changed from B-type (5.8, 17.2,
65 ◦ C were attributed to the phase transition of free fatty acid. Fig. 2 22.4, and 23.8◦ ) to V-type peaks (7.4, 12.9, and 19.8◦ ). This may suggest
shows that the intensity of the peaks in the ternary system was higher that lecithin promoted the formation of a V-type crystalline structure.
than that in the binary system, which verified that the ternary system Total relative crystallinity (RC) is calculated to understand the long-
formed a more ordered structure. range order of the samples and the results are shown in Table 2. PS-
saturated fatty acid-LE displayed significant improvement in the RC
compared to the binary systems, which indicates that lecithin could
3.4. Crystalline structure of binary and ternary systems improve the crystalline structure formation and enhance the long-range
order of PS and saturated fatty acid systems. RC-v values were also
XRD was employed to determine the crystalline structure and long- calculated (Table 2) to analyze the effect of the single-helical structure
range ordered structure of PS control, PS-LE, PS-fatty acid, and PS- on the crystallinity of the binary and ternary systems. Ternary systems
fatty acid-LE samples. The XRD results are shown in Fig. 3. The with saturated fatty acids showed a remarkable increase in RC-v values
diffraction peaks at 7.4◦ , 12.9◦ , and 19.8◦ were due to the formation of a compared to the corresponding binary complexes. The addition of
V-type crystalline structure. The diffraction peaks close to 5.8, 17.2, lecithin increased the RC-v from 14.82% to 34.14%, 9.67% to 30.37%,
22.4, and 23.8 indicate the B-type crystalline structure resulting from 6.32% to 32.47%, and 6.58% to 34.13% for PS-LA, PS-MA, PS-PA, and
the retrogradation process of PS (Ma, Ma, Zhou, Li, & Hu, 2019). The PS PS-SA systems. The higher RC-v values of ternary systems confirmed that
treated by gelatinization and oven drying displayed a B-type crystal lecithin showed a strong ability to induce a V-type crystal structure
structure with 5.8◦ , 17.2◦ , 22.4◦ , and 23.8◦ . This could be explained by formation. The rise of RC-v was more significant in samples with longer-
retrogradation during the drying process. PS-LE sample displayed an chain fatty acids.
almost B-type crystalline structure, which proved that LE would not Interestingly, compared to the binary complexes that contained a
efficiently form a V-type crystal with pea starch. All binary and ternary mainly V-type crystalline structure (PS-LA and PS-MA), corresponding
samples displayed a combination of B + V type crystalline structures.

Fig. 3. a) the xrd results of ps, ps-le, ps-fatty acids, and ps-fatty acids-le samples and b) SEM micrographs showing microstructure images of of PS, PS-LE, PS-LA, and
PS-LE-LA. a. a PS is for pea starch, PS-LE is for pea starch and lecithin, PS-LA is for pea starch and lauric acid, PS-MA is for pea starch and myristic acid, PS-PA is for
pea starch and palmitic acid, PS-SA is for pea starch and stearic acid, PS-LA-LE is for pea starch, lauric acid, and lecithin, PS-MA-LE is for pea starch, myristic acid,
and lecithin, PS-PA-LE is for pea starch, palmitic acid, and lecithin, and PS-SA-LE is for pea starch, stearic acid, and lecithin.

7
X. He et al. Food Chemistry 433 (2024) 137326

ternary systems (PS-LA-LE and PS-MA-LE) showed broader V-type 0.26 ± 0.006 nm− 1 and 0.23 ± 0.010 nm− 1, respectively). The pattern
diffraction peaks on the XRD pattern. The value of FWHM is inversely of PS-LA and PS-LA-LE displayed a more significant shoulder peak at q =
proportional to the crystallite size of the sample according to Scherrer 0.34 ± 0.024 nm− 1 and q = 0.40 ± 0.021 nm− 1, respectively. According
equation (Villas-Boas, Facchinatto, Colnago, Volanti, & Franco, 2020). to Bragg’s law, the Bragg distance (D bragg) was calculated as D bragg =
Therefore, these broader V-type peaks indicate smaller crystallite size in 2π/q (Zhang et al., 2020). As shown in Table S4, the D bragg of PS and
the ternary systems. The FWHM values of PS-LA and PS-MA’s V-type PS-LE was higher than PS-LA and PS-LA-LE. This suggests that the
peaks were significantly smaller than PS-LA-LE and PS-MA-LE complexation between PS and LA lowers the aggregate size compared to
(Table S3), which indicate that the addition of LE in the starch fatty PS itself, which is in line with recent studies (Lu, Shi, Zhu, Li, & Huang,
acid systems significantly reduced the crystallite size of the resulting 2019; Zhang et al., 2020). Moreover, compared to PS-LA, adding lecithin
complex. further reduced the D bragg of the complex aggregates. This may explain
the primary difference in the ternary and binary complex nano-scale
3.5. Fourier transform infrared (FTIR) spectroscopic analysis of binary structure.
and ternary systems The fractal dimension for the structure analysis of samples is
observed in the slope of the linear part of the SAXS curve (double log
The FTIR spectra of all samples are shown in Fig. S1. There was no graph), which is the power-law exponent (I ~ q-α) and shows the geo­
significant difference in the moisture content of samples, which negates metric self-similarity of the scattering objects. In this case, PS, PS-LE, PS-
the impact of moisture content on the FTIR results. PS-LE displayed no LA, and PS-LA-LE samples displayed mass dimension (Dm) (1＜α＜3,
additional absorption bands compared to the original pea starch control. Dm = α) in SAXS curves (Li, Zhu, Mo, & Hemar, 2019). In general, a
Two unique absorption bands at 2849 nm− 1 and 1698 nm− 1 appeared in higher Dm indicates a more compact structure in the sample (Yang et al.,
all PS and fatty acid systems with or without lecithin. The band at 2849 2016). According to the data shown in Fig. S2, PS-LA-LE displayed the
nm− 1 is assigned to the asymmetric stretching vibration of C–H groups most compact structure (Dm = 2.56 ± 0. 12), followed by PS-LA (Dm =
in fatty acids, and the band at around 1698 nm− 1 is attributed to the 2.32 ± 0.06). These results suggest that the addition of lecithin during
carbonyl groups in the polar head of fatty acids (Koca, Rodriguez-Saona, the complexation of PS and LA led to a more compact nano-scale stack of
Harper, & Alvarez, 2007; S. Wang et al., 2017; M. Zheng et al., 2018). complex molecules.
For all PS-fatty acids-LE samples, the relative intensity of the bands at
around 1698 nm− 1 almost disappeared. This change could be explained
3.7. In vitro enzymatic hydrolysis of ternary and binary systems
by the carbonyl group in the polar head of fatty acids interacting with
the positively charged groups in lecithin thereby the signal being
The in vitro enzymatic hydrolysis of PS control, PS-fatty acid with or
covered by this interaction. Previous studies have shown that the three-
without lecithin samples, is displayed as the contents of RDS, SDS, and
way interactions among starch, fatty acid, and protein resulted in the
RS (Table 3). The gelatinized and oven-dried PS’s RDS, SDS, and RS
attenuation of the infrared absorption of carboxyl groups on fatty acids
contents were 73.96%, 10.99%, and 15.05%, respectively. Adding
(Zheng et al., 2018).
lecithin to PS decreased the content of RDS and increased the content of
The short-range order of starch samples is reflected in the ratio of
RS and SDS in PS. This could be because the lecithin coats the surface of
1047/1022, which is the intensity ratio of decomposed peaks at 1047
gelatinized starch. Complexation with fatty acids dramatically reduced
and 1022 cm− 1 (Table 2) (Lin et al., 2020). The ratio of 1047/1022 of
the content of RDS in all PS-fatty acid systems and increased the RS
gelatinized PS control was 56.06%. Complexing with fatty acids
content in samples. The formation of a V-type single-helical structure in
increased the short-range order of PS, although there was no significant
the PS-fatty acid systems may have provided a more compact starch
difference between samples obtained from the different fatty acids.
structure, thereby inhibiting the binding between enzymes and active
Adding lecithin to the systems increased the short-range order in PS-LA,
sites on the starch molecule (Ai, Hasjim, & Jane, 2013; Hasjim et al.,
PS-MA, and PS-PA samples. The highest development of short-range
2010; Wang et al., 2020). In this study, the chain length of fatty acid did
order was found in PS-LA-LE, which sample also showed the highest
not have a pronounced effect on the hydrolysis of the binary system. A
complex index. The difference in the number of complexes in the system
previous study suggested that the effect of fatty acid chain length on the
could affect the short-range order of the final product. The other three
enzymatic hydrolysis of the binary complex was inconclusive and highly
samples formed in ternary systems did not show significant differences.
The results suggest that the short-range order was not sensitive to the
Table 3
chain length of fatty acids.
The content of RDS, SDS, and RS in PS, PS-LE, PS-fatty acids, and PS-fatty acids-
LE samples a.
3.6. Small angle X-ray scattering (SAXS) analysis
Samples RDS (%) SDS (%) RS (%)

SAXS is a powerful technique for understanding the nano-scale PS 73.96 ± 0.85f 10.99 ± 1.34a 15.05 ± 0.49a
structure of starch and its complexes. This technique has been widely PS-LE 35.35 ± 0.58a 43.03 ± 1.56e 21.62 ± 2.14b
PS-LA 35.96 ± 2.95a 41.63 ± 2.94e 22.41 ± 1.27b
used to explore the lamellar structure of alternate crystalline and
PS-MA 64.71 ± 2.68e 11.71 ± 1.88a 23.58 ± 0.80b
amorphous regions in starch granules and the fractal structures of ret­ PS-PA 48.04 ± 0.40c 28.96 ± 0.09 cd 23.00 ± 0.49b
rograded starch or starch inclusion complexes. A major consideration in PS-SA 44.64 ± 1.92bc 30.83 ± 4.47d 24.52 ± 2.55bc
SAXS analysis is the retrogradation process, which would lead to a PS-LA-LE 39.58 ± 3.40ab 21.97 ± 5.62bc 38.45 ± 2.22f
PS-MA-LE 55.37 ± 0.67d 13.09 ± 0.22a 31.54 ± 0.89e
shoulder-like peak in the SAXS pattern (Zhang, Li, Janaswamy, Chen, &
PS-PA-LE 37.18 ± 2.32a 34.28 ± 0.76de 28.55 ± 1.56cde
Chi, 2020). To avoid the effect of sample retrogradation, PS-LA and PS- PS-SA-LE 38.97 ± 1.43ab 36.02 ± 1.97e 25.01 ± 0.54bc
LA-LE were chosen to compare the difference of nano-scale fractal a
structures between binary and ternary systems because they have RDS, SDS, and RS mean rapidly digestible starch, slowly digestible starch,
and resistant starch, respectively. Values are means ± SD. Means with different
similar peaks of XRD pattern (dominant by V-type peaks) and complex
letters in a column differ significantly (p＜0.05). All experiments were repeated
index (~75%, not significantly different). The results of SAXS patterns of
3 times. PS is for pea starch, PS-LE if for pea starch and lecithin, PS-LA is for pea
PS, PS-LE, PS-LA, and PS-LA-LE and their Lorentz corrected pattern are starch and lauric acid, PS-MA is for pea starch and myristic acid, PS-PA is for pea
displayed in Fig S2. During storage of gelatinized starch, this shoulder- starch and palmitic acid, PS-SA is for pea starch and stearic acid, PS-LA-LE is for
like peak could shift to the further lower q range and the intensity of the pea starch, lauric acid, and lecithin, PS-MA-LE is for pea starch, myristic acid,
peak increased due to retrogradation (Zhang et al., 2020). The PS and and lecithin, PS-PA-LE is for pea starch, palmitic acid, and lecithin, and PS-SA-LE
PS-LE patterns exhibited weak shoulder-like peaks at low q region (q = is for pea starch, stearic acid, and lecithin.

8
X. He et al. Food Chemistry 433 (2024) 137326

dependent on experimental conditions (Wang et al., 2020). Lecithin enzymatic susceptibility than the binary systems. These results build on
played a vital role in the in vitro enzymatic hydrolysis of PS-fatty acid existing knowledge of the interactions among food ingredients. Results
complexes. Samples with lecithin had a higher content of RS than PS- also provide a new methodology to modify starch-lipid complexes for
fatty acid samples, except for the PS-SA system. The differences in SDS potentially developing novel RS food products. Further research is
and RS contents of PS-fatty acids samples and PS-fatty acids-LE samples needed on the formation mechanism of the ternary system and possible
could be explained by more complexes formed due to the enhancement applications in producing high resistant starch food products.
of lecithin. The XRD test, which revealed long-range order in the sam­
ples, suggested a more ordered structure formed in systems with leci­ CRediT authorship contribution statement
thin. Previous works claimed that starch with higher crystallinity could
exhibit slower hydrolysis of starch (Ao et al., 2007). Similarly, the more Xiaoyang He: Investigation, Formal analysis, Writing – original
ordered the starch structure, the higher the hydrolysis resistance (Wang draft. Liyang Zhou: Software, Data curation. Purnima Gunness:
et al., 2020). Writing – review & editing. Vicky A. Solah: Supervision. Qingjie Sun:
The RS contents of PS-SA and PS-SA-LE samples were not signifi­ Supervision, Project administration, Conceptualization, Funding
cantly different, although complex indexes, XRD, and DSC results indi­ acquisition.
cated that the PS-SA-LE system formed more complexes than PS-SA. This
could be also attributed to PS-SA that showed a B + V crystalline
structure dominate by B-type, providing evidence of retrogradation in Declaration of Competing Interest
the sample. Although PS-SA-LE showed more V-type crystal formed, due
to the high content of retrograded starch, PS-SA displayed similar RS The authors declare that they have no known competing financial
content with PS-SA-LE. Complex index and DSC results indicate that the interests or personal relationships that could have appeared to influence
PS-SA-LE sample showed a relatively lower complexation degree than the work reported in this paper.
other ternary systems. Compared to the retrogradation in the PS-SA
sample, the complexes formed in PS-SA-LE may be not sufficient to Data availability
show a higher resistant starch content. In this study, ternary systems
formed by shorter-chain fatty acids (lauric acid and myristic acid) No data was used for the research described in the article.
exhibited a higher hydrolysis resistance than stearic acid and palmitic
acid systems. This could be due to ternary systems with shorter-chain Acknowledgements
fatty acids forming more complexes than ternary systems with longer-
chain fatty acids. The study was supported by the National Natural Science Foundation
of China [No. 32272349], Key R & D plan of Shandong Province [No.
3.8. Morphology of binary and ternary systems 2022CXGC010604], Science & Technology Specific Projects in Agri­
cultural High-tech Industrial Demonstration Area of the Yellow River
The SEM images of PS, PS-LE, PS-LA, and PS-LE-LA are depicted in Delta [No. 2022SZX27], Special Funds for Taishan Scholars Project of
Fig. 3 (b). The microstructure of all samples exhibited an absence of Shandong Province, and Qingdao Health Food Innovation and Entre­
granular form, signifying the complete gelatinization and disruption of preneurship Community [22–7-5-gtt-2-gx]. We acknowledge Dr. Wendy
starch granules during sample preparation. The PS sample matrix Hunt (Murdoch University) for her assistance with language checking
exhibited a densely compacted structure. The multi-layered surface and polishing.
structure indicated retrogradation during cooling and storage, which
was supported by the XRD result (Lian et al., 2013). The fortuitous Appendix A. Supplementary data
dispersion of fine particles on the PS-LE surface might be attributed to
lecithin aggregation. This supports the notion that the comparatively Supplementary data to this article can be found online at https://doi.
lower RDS content of PS-LE in the enzymatic hydrolysis experiments org/10.1016/j.foodchem.2023.137326.
could be ascribed to lecithin’s physical barrier effect on enzymes. PS-LA
and PS-LA-LE samples displayed protruding spherical or elongated
References
lamellar-like structures on their surfaces, displaying the appearance of
starch-fatty acid complex samples subjected to oven drying. Similar Ai, Y., Hasjim, J., & Jane, J. L. (2013). Effects of lipids on enzymatic hydrolysis and
observations were reported in a prior study (Marinopoulou et al., 2016). physical properties of starch. Carbohydrate Polymers, 92(1), 120–127.
The rougher surface of the PS-LA-LE sample might indicate the forma­ Ao, Z., Simsek, S., Zhang, G., Venkatachalam, M., Reuhs, B. L., & Hamaker, B. R. (2007).
Starch with a Slow Digestion Property Produced by Altering Its Chain Length, Branch
tion of more V-type complexes in the ternary system involving lecithin. Density, and Crystalline Structure. Journal of Agricultural and Food Chemistry, 55(11),
This phenomenon supports the higher crystallinity observed in the PS- 4540–4547.
LA-LE samples compared to PS-LA in XRD experiments, which might Biliaderis, C. G., & Galloway, G. (1989). Crystallization behavior of amylose-V
complexes: Structure-property relationships. Carbohydrate research, 189, 31–48.
intuitively demonstrate the promoting effect of lecithin on the starch- Biliaderis, C. G., Page, C. M., Slade, L., & Sirett, R. R. (1985). Thermal behavior of
lipid complexation process. amylose-lipid complexes. Carbohydrate Polymers, 5(5), 367–389.
Blazek, J., Gilbert, E. P., & Copeland, L. (2011). Effects of monoglycerides on pasting
properties of wheat starch after repeated heating and cooling. Journal of Cereal
4. Conclusion Science, 54(1), 151–159.
Cai, J., Chao, C., Niu, B., Copeland, L., Yu, J., Wang, S., & Wang, S. (2021). New insight
In conclusion, lecithin can significantly enhance the complexation into the interactions among starch, lipid and protein in model systems with different
starches. Food Hydrocolloids, 112.
process between starch and fatty acids and in turn affect the physico­ Chi, C., Li, X., Huang, S., Chen, L., Zhang, Y., Li, L., & Miao, S. (2021). Basic principles in
chemical properties and enzymatic susceptibility of the binary V-com­ starch multi-scale structuration to mitigate digestibility: A review. Trends in Food
plex structure of starch and fatty acid. The notable viscosity peaks Science & Technology, 109, 154–168.
Deng, N., Deng, Z., Tang, C., Liu, C., Luo, S., Chen, T., & Hu, X. (2021). Formation,
during the setback stage in the RVA pasting curves of PS-fatty acid-LE
structure and properties of the starch-polyphenol inclusion complex: A review.
systems showed that lecithin did affect the complexation between starch Trends in Food Science & Technology, 112, 667–675.
and fatty acids. The results of DSC, XRD and FTIR supported that lecithin Eliasson, A. C. (1986). On the effects of surface active agents on the gelatinization of
enhanced the V-type complexes formation in the starch-fatty acids- starch — a calorimetric investigation. Carbohydrate Polymers, 6(6), 463–476.
Englyst, H. N., & Cummings, J. H. (1985). Digestion of the polysaccharides of some cereal
water system. Ternary systems containing lecithin showed higher crys­ foods in the human small intestine. The American Journal of Clinical Nutrition, 42(5),
tallinity, short-range order, a more compact nano-scale stack and lower 778–787.

9
X. He et al. Food Chemistry 433 (2024) 137326

Gutiérrez, T. J., & Tovar, J. (2021). Update of the concept of type 5 resistant starch Oyeyinka, S. A., Singh, S., & Amonsou, E. O. (2021). A review on structural, digestibility
(RS5): Self-assembled starch V-type complexes. Trends in Food Science & Technology, and physicochemical properties of legume starch-lipid complexes. Food Chemistry,
109, 711–724. 349, Article 129165.
Hasjim, J., Lee, S.-O., Hendrich, S., Setiawan, S., Ai, Y., & Jane, J.-L. (2010). Qin, R., Wang, J., Chao, C., Yu, J., Copeland, L., Wang, S., & Wang, S. (2021). RS5
Characterization of a Novel Resistant-Starch and Its Effects on Postprandial Plasma- Produced More Butyric Acid through Regulating the Microbial Community of
Glucose and Insulin Responses. Cereal Chemistry, 87(4), 257–262. Human Gut Microbiota. Journal of Agricultural and Food Chemistry, 69(10),
Karkalas, J., Ma, S., Morrison, W. R., & Pethrick, R. A. (1995). Some factors determining 3209–3218.
the thermal properties of amylose inclusion complexes with fatty acids. Carbohydrate Soong, Y. Y., Goh, H. J., & Henry, C. J. (2013). The influence of saturated fatty acids on
Research, 268(2), 233–247. complex index and in vitro digestibility of rice starch. International Journal of Food
Kawai, K., Takato, S., Ueda, M., Ohnishi, N., Viriyarattanasak, C., & Kajiwara, K. (2016). Sciences and Nutrition, 64(5), 641–647.
Effects of fatty acid and emulsifier on the complex formation andin vitrodigestibility Tang, M. C., & Copeland, L. (2007). Analysis of complexes between lipids and wheat
of gelatinized potato starch. International Journal of Food Properties, 20(7), starch. Carbohydrate Polymers, 67(1), 80–85.
1500–1510. Villas-Boas, F., Facchinatto, W. M., Colnago, L. A., Volanti, D. P., & Franco, C. M. L.
Koca, N., Rodriguez-Saona, L. E., Harper, W. J., & Alvarez, V. B. (2007). Application of (2020). Effect of amylolysis on the formation, the molecular, crystalline and thermal
Fourier transform infrared spectroscopy for monitoring short-chain free fatty acids in characteristics and the digestibility of retrograded starches. International Journal of
Swiss cheese. Journal of Dairy Science, 90(8), 3596–3603. Biological Macromolecules, 163, 1333–1343.
Krog, N. (1971). Amylose Complexing Effect of Food Grade Emulsifiers. Starch - Stärke, Wang, S., Chao, C., Cai, J., Niu, B., Copeland, L., & Wang, S. (2020). Starch–lipid and
23(6), 206–210. starch–lipid–protein complexes: A comprehensive review. Comprehensive Reviews in
Li, X., Luo, S., Hou, Y., Liu, Y., Hu, X., & Liu, C. (2020). Effect of triglyceride on Food Science and Food Safety, 19(3), 1056–1079.
complexation between starch and fatty acid. International Journal of Biological Wang, S., Wang, J., Yu, J., & Wang, S. (2016). Effect of fatty acids on functional
Macromolecules, 155, 1069–1074. properties of normal wheat and waxy wheat starches: A structural basis. Food
Li, G., Zhu, F., Mo, G., & Hemar, Y. (2019). Supramolecular structure of high hydrostatic Chemistry, 190, 285–292.
pressure treated quinoa and maize starches. Food Hydrocolloids, 92, 276–284. Wang, S., Zheng, M., Yu, J., Wang, S., & Copeland, L. (2017). Insights into the Formation
Lian, X., Liu, L., Guo, J., Li, L., & Wu, C. (2013). Screening of seeds prepared from and Structures of Starch-Protein-Lipid Complexes. Journal of Agricultural and Food
retrograded potato starch to increase retrogradation rate of maize starch. Chemistry, 65(9), 1960–1966.
International journal of biological macromolecules, 60, 181–185. Yang, Z., Swedlund, P., Hemar, Y., Mo, G., Wei, Y., Li, Z., & Wu, Z. (2016). Effect of high
Lin, L., Yang, H., Chi, C., & Ma, X. (2020). Effect of protein types on structure and hydrostatic pressure on the supramolecular structure of corn starch with different
digestibility of starch-protein-lipids complexes. Lwt, 134, Article 110175. amylose contents. International Journal of Biological Macromolecules, 85, 604–614.
Lu, H., Yang, Z., Yu, M., Ji, N., Dai, L., Dong, X., … Sun, Q. (2021). Characterization of Zhang, G., Maladen, M., Campanella, O. H., & Hamaker, B. R. (2010). Free fatty acids
complexes formed between debranched starch and fatty acids having different electronically bridge the self-assembly of a three-component nanocomplex consisting
carbon chain lengths. International Journal of Biological Macromolecules, 167, of amylose, protein, and free fatty acids. Journal of Agricultural and Food Chemistry,
595–604. 58(16), 9164–9170.
Lu, X., Shi, C., Zhu, J., Li, Y., & Huang, Q. (2019). Structure of starch-fatty acid Zhang, G., Maladen, M. D., & Hamaker, B. R. (2003). Detection of a novel three
complexes produced via hydrothermal treatment. Food Hydrocolloids, 88, 58–67. component complex consisting of starch, protein, and free fatty acids. Journal of
Ma, Z., Ma, M., Zhou, D., Li, X., & Hu, X. (2019). The retrogradation characteristics of Agricultural and Food Chemistry, 51(9), 2801–2805.
pullulanase debranched field pea starch: Effects of storage time and temperature. Zhang, L., Li, X., Janaswamy, S., Chen, L., & Chi, C. (2020). Further insights into the
International Journal of Biological Macromolecules, 134, 984–992. evolution of starch assembly during retrogradation using SAXS. International Journal
Marinopoulou, A., Papastergiadis, E., Raphaelides, S. N., & Kontominas, M. G. (2016). of Biological Macromolecules, 154, 521–527.
Morphological characteristics, oxidative stability and enzymic hydrolysis of Zheng, M., Chao, C., Yu, J., Copeland, L., Wang, S., & Wang, S. (2018). Effects of Chain
amylose-fatty acid complexes. Carbohydrate Polymers, 141, 106–115. Length and Degree of Unsaturation of Fatty Acids on Structure and in Vitro
Niu, B., Chao, C., Cai, J., Yan, Y., Copeland, L., Yu, J., … Wang, S. (2020). Effect of pH on Digestibility of Starch-Protein-Fatty Acid Complexes. Journal of Agricultural and Food
formation of starch complexes with lauric acid and β-lactoglobulin. Lwt, 132. Chemistry, 66(8), 1872–1880.

10