The effect of chitosan nanoparticles on the rheo-viscoelastic properties and lipid digestibility of oil/vinegar mixtures (vinaigrettes)

Journal of Functional Foods 93 (2022) 105092 Available online 28 April 2022 1756-4646/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). The effect of chitosan nanoparticles on the rheo-viscoelastic properties and lipid digestibility of oil/vinegar mixtures (vinaigrettes) Clara A. Tovar a , Karina Oliveira Lima b , Ail´ en Alem´ an c , M. Pilar Montero c , M. Carmen G´ omezGuill´ en c , * a Department of Applied Physics, University of Vigo, As Lagoas, 32004 Ourense, Spain b School of Chemistry and Food, Federal University of Rio Grande (FURG), Rio Grande, RS, Brazil c Institute of Food Science, Technology and Nutrition (ICTAN-CSIC), Calle Jos´ e Antonio Novais, 10, 28040 Madrid, Spain ARTICLE INFO Keywords: Chitosan nanoparticles Rheological properties Vinaigrettes Lipid digestion Antioxidant activity ABSTRACT The role of chitosan nanoparticles on the rheological behaviour, antioxidant properties and lipid digestibility of vinaigrettes was investigated in comparison with a protonated chitosan polymer. Nanoparticles were produced using different concentrations of tripolyphosphate (TPP) and acetic acid. The lower ionic concentrations were used to encapsulate a fish protein hydrolysate. Rotatory tests produced shear-thinning behaviour in the different nanoparticle suspensions. The vinaigrettes with addition of both chitosan and nanoparticles exhibited a stationary shear-thinning flow with viscoelastic characteristics of weak gels. The presence of chitosan nanoparticles in the emulsified oil phase produced a three-dimensional network with higher energy stability in the case of hydrolysate-loaded nanoparticles. Both chitosan and chitosan nanoparticles increased the radical scavenging capacity of vinaigrettes, but not the ferric ion reducing power. Addition of empty or hydrolysate-loaded nanoparticles reduced lipid digestion in vinaigrettes to a greater extent than chitosan (≈45% with nanoparticles; 33% with chitosan). 1. Introduction The use of chitosan for food applications has attracted increasing interest because of its unique chemical and functional properties, as well as its biocompatibility, biodegradability and non-toxicity. Chitosan is a cationic polysaccharide composed of D-glucosamine and N-acetyl-Dglucosamine units linked by β-(1 → 4)-glycosidic bonds that becomes soluble by the protonation of amino groups, resulting in a positively charged polyelectrolyte. The property of chitosan of gelling upon contact with anions was originally used as a means to produce nanoparticles using tripolyphosphate (TPP) as a polyanion (Calvo et al., 1997). This technique has been applied for the nanoencapsulation of food bioactives (Akbari-Alavijeh et al., 2020); in particular, chitosan nanoparticles are good carriers of peptides, due to their hydrophilicity and their capacity to establish electrostatic interactions and hydrogen bonds (Calvo et al., 1997; Piras et al., 2015; Hosseini et al., 2018; Du et al., 2019). These nanosystems are also effective in maintaining the antioxidant properties of fish-derived peptides (Hosseini et al., 2018). Both chitosan concentration and chitosan:TPP mass ratio are among the most influential parameters determining the size of nanoparticles and their encapsulation efficiency (Lima et al., 2021a; Li & Huang, 2012). Li & Huang (2012) reported that the chitosan:TPP mass ratio affected the crosslinking efficiency between chitosan and TPP, whereas chitosan concentration was more relevant to the inter-particle distance, surface charges and fluid viscosity of nanoparticle suspensions. The study of the rheological properties of dilute nanoparticle suspensions, which differ from those of dissolved chitosan, could be of interest for both processing and product design in the food industry. For instance, they are especially related to the production of functional beverages, light creams or emulsion-based salad dressings, into which dilute acid nanoparticle suspensions can be directly incorporated without the need of being previously dried or concentrated. Chitosan nanoparticles can be also effective stabilizers of Pickering emulsions, which are defined as emulsions stabilized by solid particles that reduce the interfacial tension. In a previous work, concentrated chitosan nanoparticle suspensions were already found to increase the stability and rheological properties of a conventional food emulsion system based on sunflower oil and egg white (Lima et al., 2021a). At low * Corresponding author. E-mail addresses: [email protected], [email protected] (M.C. G´ omez-Guill´ en). Contents lists available at ScienceDirect Journal of Functional Foods journal homepage: www.elsevier.com/locate/jff https://doi.org/10.1016/j.jff.2022.105092 Received 1 March 2022; Received in revised form 23 April 2022; Accepted 25 April 2022 Journal of Functional Foods 93 (2022) 105092 2 pH, the stability of Pickering emulsions containing chitosan nanoparticles decreases, producing coalescence and creaming due to an increase in the electrostatic repulsions that hinder the adsorption of nanoparticles at the oil/water interface (Mwangi, Ho, Tey & Chan, 2016). The emulsion instability at low pH was previously proposed as an advantage to formulate reversible pH-responsive Pickering emulsions in the presence of chitosan (Liu, Wang, Zou, Wei, & Tong, 2012). This property could meet the requirements of a fluid emulsion-like salad dressing produced by simply hand-shaking an oil/vinegar mixture (vinaigrette), taking into account the occurrence of phase separation at rest, but the reversible miscibility and emulsion-like appearance recovery when shaken prior to use. The incorporation of dilute chitosan nanoparticles into vinaigrettes could improve their appeal by providing greater stability and modifying their rheological properties. Besides acting as efficient nanocarriers and texture-modifiers, chitosan nanoparticles could endorse innovative salad dressings with antioxidant properties, as well as anti-obesity and hipocholesterolemic functionalities due to the fat binding capacity of chitosan favoured by the pH changes that occur along the gastrointestinal tract (Zhang, Zhong, Tao, Wu, & Su, 2012). The aim of the present work was to characterize the rheological properties of several chitosan nanoparticle suspensions, including one with an entrapped fish protein hydrolysate as a model nanocarrier system of entire marine origin, in comparison with the corresponding chitosan protonated polymer. The second part of the study deals with the application of empty nanoparticles and hydrolysate-loaded nanoparticles to improve the properties of shaken oil/vinegar mixtures (vinaigrettes) in terms of phase separation, rheological properties and antioxidant capacity, which could reveal innovative applications in food catering services. The ability of chitosan, empty nanoparticles and hydrolysate-loaded nanoparticles to decrease lipid digestion in vinaigrettes was also evaluated after in vitro simulated digestion. 2. Materials and methods 2.1. Synthesis of nanoparticles Nanoparticles were prepared by the ionic gelation method, as reported previously (Lima et al., 2021a). Chitosan (220 kDa and 78.6% deacetylation; Guinama, Valencia, Spain) was dissolved (3 mg/mL) in acetic acid at 1% (v/v) (A1) or 3% (v/v) (A3) stirring overnight, and then the pH was adjusted to 5.0 using 5 M NaOH. These samples constituted the protonated chitosan solutions used as control samples, i. e. A1T0 and A3T0 respectively. To produce nanoparticles, 10 mL of sodium tripolyphosphate (TPP) (technical grade, 85%, Sigma-Aldrich, St. Louis, MO, USA) (T1, 1 mg/ mL) were added dropwise to 25 mL of each chitosan solution under constant stirring; after adding TPP the suspension was stirred for 30 min at 750 rpm (MM30E, Ovan, Badalona, Spain). The nanoparticle suspensions thus prepared were A1T1 and A3T1, depending on whether the chitosan was pre-solubilized in 1% or 3% acetic acid, respectively. Similarly, nanoparticles were synthesized with TPP at 3 mg/mL (T3), rendering the corresponding nanoparticle suspensions A1T3 and A3T3. A sonication step at 50% amplitude for 4 min (cycles of 1 min and 1 min stop) where TPP was added was applied to all samples, using a Q700 sonicator (Qsonica, Newton, CT, USA, max 700 W) with a 12.7 mm diameter probe. The A1T1 formulation was used to encapsulate a Stripped weakfish (Cynoscion guatucupa) hydrolysate, which was obtained using Protamex to 5% degree of hydrolysis according to Lima et al. (2021b). After overnight chitosan solubilisation and pH adjustment to 5.0, 20 mg of dried hydrolysate were added and homogenized, prior to the addition of TPP (1 mg/mL) (HA1T1). The hydrolysate was composed of 35% hydrophobic residues, 62% hydrophilic residues and 5% amphiphilic residues (Met, Tyr). Negatively charged residues (Asp, Glu) were quantified together with their uncharged amide derivatives (Asn, Gln), representing a total of ≈26%. The entrapment efficiency of the fish hydrolysate in the nanoparticles was 59.6 ±0.8% (Lima et al., 2021a). 2.2. Hydrodynamic properties and ζ potential Chitosan solutions and nanoparticle suspensions were diluted in distilled water (10-fold). Dynamic light scattering and Laser Doppler electrophoresis were used to determine the mean particle size (zaverage), polydispersity index (PdI) and ζ potential using a Zetasizer Nano ZS90 (Malver Instruments, UK). Measurements were performed at 25 ±0.1 ◦C. 2.3. Preparation of vinaigrettes (oil/vinegar mixtures) A volume of 20 mL refined sunflower oil (Hacendado, Mercadona, Valencia, Spain) was mixed with 10 mL sherry vinegar (Alipende, Vinagres Parras, S.A., Toledo, Spain) and 10 mL of nanoparticle suspension (A1T1 or HA1T1) or chitosan solution (A1T0), by vigorous hand-shaking for 30 s, vortexing for 1 min and subsequent hand-shaking for 30 s. A control sample was performed by replacing the nanoparticle suspension or chitosan solution by distilled water. The oil/vinegar mixtures were coded as follows: VC (control with water), VNP (with empty nanoparticles), VNPH (with hydrolysate-loaded nanoparticles) and VCS (with chitosan solution). Homogeneous mixtures (vinaigrettes) were formed after shaking, and almost complete phase separation (emulsified oil phase and clear aqueous-vinegar phase) took place within about 15–30 min. The final chitosan concentration in the corresponding vinaigrettes was 0.05% (w/v). The measured pH in the different samples, namely VC, VNP, VNPH and VCS was 3.3, 3.6, 3.6 and 3.2, respectively. The upper phase (emulsified oil phase), coded in the respective studied samples as EC, ENP, ENPH and ECS, was also used for rheology analyses. The creaming index (CI) in the newly prepared vinaigrettes after complete phase separation was determined in graduated centrifugal tubes according to Eq. (1): CI (%) =Hv/He ×100 (1) where, Hv is the height in mL of the vinegar phase at the bottom (liquid phase) and He is the initial total height (40 mL) of the mixture. 2.4. Rheological measurements Rheological tests were performed using a Discovery HR10 rheometer (Waters TA Instruments, Milford, MA, USA). The measurement system was the cone-plate geometry with a diameter of 40 mm, a cone angle of 2◦and 0.053 mm gap. The temperature in the lower plate was 20.0 ± 0.1 ◦C. Preliminary trials of shear rates (up and down ramps), time intervals and pre-shearing rates were performed to optimize data reproducibility. Tests were carried out first, for both chitosan solutions and nanoparticle suspensions. Flow curves were plotted using four intervals: 1) pre-shear phase (100 s −1 , 300 s) to homogenize the solution without breaking the structure; 2) upward shear-rate ramp (from 100 s −1 to 800 s −1 ; 600 s); 3) high shear-rate phase (800 s −1 , 150 s.); and 4) downward shear-rate ramp (from 800 s −1 to 100 s −1 ; 600 s). Possible thixotropy (timedependence of flow) was analysed to obtain the area of the hysteresis loop that corresponds to the area between the up and down curves. Flow ramps for vinaigrettes were performed using four intervals: 1) pre-shear phase (100 s −1 , 40 s), 2) upward shear-rate ramp (from 0.1 s −1 to 1000 s −1 , 24 min); 3) high shear-rate phase (1000 s −1 , 40 s.); and 4) downward shear-rate ramp (from 1000 s −1 to 0.1 s −1 , 24 min). Viscosity as a function of shear rate was plotted for both upward and downward ramps. Frequency sweeps were carried out over a range of 10 to 0.1 Hz with an oscillation strain of 1%, selected from the linear viscoelastic region C.A. Tovar et al. Journal of Functional Foods 93 (2022) 105092 3 (LVER). The storage modulus (G’), loss modulus (G’’), and loss factor (tanδ) were plotted as a function of frequency (Hz). The rheological tests were carried out at least in triplicate. 2.5. Antioxidant properties Radical scavenging capacity (ABTS test) and ferric iron reducing power (FRAP test) were determined on shaken vinaigrette samples before the phase separation took place, following the procedure described by Alem´ an et al. (2011). The ABTS and FRAP results were expressed as µmol Trolox Eq./L vinaigrette and µmol Fe 2+ Eq./L vinaigrette, respectively. 2.6. In vitro simulated digestion An aliquot of 2 g of vinaigrettes was accurately weighed and subjected to an in vitro gastrointestinal digestion (GID) assay, following the harmonized INFOGEST protocol (Minekus et al., 2014), which considers consecutive simulated oral (pH 7), gastric (pH 3) and intestinal (pH 7) digestions. At the end of digestion, the enzymes were inactivated by heating at 90 ◦C for 5 min and the samples were centrifuged at 50000 rpm for 60 min using a Sorvall Combi Plus ultracentrifuge (Thermo Fisher Scientific, Massachusetts, USA). After centrifugation, the digested samples were separated into an oil phase, a highly emulsified aqueous phase and a precipitated pellet phase. The aqueous phase, which included micellar structure and corresponds to the “bioaccesible” fraction, was accurately weighed and kept for further analyses. The experiments were carried out in duplicate. 2.7. Lipid digestibility Lipid digestibility was determined by measuring the amount of free fatty acids released during gastrointestinal digestion, which was determined by the AOAC (1995) method with some modifications. Ten grams of samples were weighed and mixed with 10 mL of ethanol previously neutralized and 1 mL of phenolphthalein (1% dissolved in ethanol). Free fatty acids, expressed as oleic acid, were titrated with 0.05 N NaOH and were calculated by the equation: Free fatty acid (mg/g) =V ×N ×MW / SW. where V: volume (mL) of NaOH solution; N: concentration (mol/L) of NaOH (0.05), MW: molecular weight (g/mol) of oleic acid (282.4) and SW: sample weight (g). 2.8. Optical microscopy The digested vinaigrettes were placed between glass slide and cover slip, and observed with a Nikon Optiphot microscope (Nikon Europe BV, Amsterdam, Netherlands). 2.9. Statistical analysis Particle properties, antioxidant properties and lipid digestibility data were evaluated by analysis of variance (ANOVA) and means were compared by the Tukey test at 5% significance level (p <0.05). 3. Results and discussion 3.1. Particle properties Table 1 presents the particle properties of the various nanoparticle suspensions, which were selected based on a preliminary work (Lima et al., 2021a). At both acetic acid concentrations, the addition of TPP at a low concentration (T1) resulted in smaller sized nanoparticles (≈25% smaller) than at a higher TPP concentration (T3). Hu et al. (2008) pointed out that more abundant TPP could result in larger particle size by inducing the linking of monoparticles. The T3 preparations showed lower (p <0.05) ζ potential values than T1. This effect could be associated with their larger particle size, since the higher neutralization of the exposed chitosan NH 3+ groups by the more abundant TPP phosphate groups reduces the surface charge of the resulting NPs (Hu et al., 2008). These findings are in agreement with those of Du et al. (2019), who reported an increase in the size of the nanoparticles and a decrease in the ζ potential with the decrease in the chitosan:TTP mass ratio. The higher acetic acid concentration (A3) slightly reduced (p <0.05) the particle size only when using a high TPP concentration (A3T3 vs A1T3), and did not induce significant differences in the ζ potential with respect to the A1 counterparts. The degree of chitosan protonation using both A1 and A3 was high enough and comparable, considering that the ζ potential values for the plain chitosan solutions (A1T0 vs A3T0) were highly electropositive and not significantly different (Table 1). However, the increase in the ionic strength of the system caused by the acetate anions and the NaOH consumed to neutralize the excess acid in the A3 preparations (pH adjusted to 5.0) could influence the molecular conformation of dissolved chitosan prior to TPP addition. To this respect, Fan et al. (2012) reported that higher ionic strength might cause chitosan molecules to adopt a more contracted conformation due to a decrease in intramolecular electrostatic repulsion through the screening of counter ions on the protonated amine groups. The lower ionic strength formulation (A1T1) was chosen to entrap the fish protein hydrolysate due to its small size and high ζ potential. The nanoparticles containing the hydrolysate (HA1T1) showed similar characteristics as their empty counterparts, with no significant differences (p >0.05) in size and ζ potential, and with lower (p <0.05) PdI. This lower PdI could be due to negatively charged amino acid residues in the hydrolysate, which may orient the crossover points between the chitosan molecules and the TPP, leading to a more uniform particle size. In contrast to the present work, an increase in particle size and decrease in ζ potential was previously reported in peptide-loaded chitosan nanoparticles (Piras et al., 2015; Hosseini et al., 2018). Table 1 Properties of chitosan nanoparticles and flow parameters of equation (1) for chitosan solutions and nanoparticle suspensions at 20 ◦C. Code z-average (nm) PdI ζ potential (mV) κ up (Pa⋅s -n ) n up κ down (Pa⋅s -n ) n down η 300 (Pa⋅s) A1T0 – – 74.4 ±0.7 0.00875 ±0.00017 e 1.000 ±0.002 e 0.00879 ±0.00018 d 1.000 ±0.002 f 0.00889 ±0.00018 d A3T0 – – 75.2 ±0.9 0.00879 ±0.00070 e 1.000 ±0.002 e 0.00867 ±0.00059 d 1.000 ±0.002 f 0.00956 ±0.00083 d A1T1 200.8 ±7.1 a 0.319 ±0.03 a 35.8 ±3.6 a 0.0140 ±0.0012 d 0.710 ±0.004 b 0.0298 ±0.0054 c* 0.591 ±0.007 a* 0.0026 ±0.0002 b A1T3 273.6 ±1.5 b 0.171 ±0.03 b 28.1 ±0.6 c 0.0147 ±0.0009 d 0.669 ±0.005 a 0.0168 ±0.0014 b* 0.634 ±0.007 b* 0.0022 ±0.0002 bc A3T1 199.8 ±7.4 a 0.371 ±0.02 a 37.1 ±1.7 a 0.0074 ±0.0015 c 0.752 ±0.007 c 0.0104 ±0.0022 a* 0.697 ±0.008 c* 0.0017 ±0.0004 ac A3T3 264.8 ±1.3 c 0.154 ±0.01 b 29.5 ±3.0 bc 0.0041 ±0.0005 b 0.850 ±0.004 d 0.0097 ±0.0009 a* 0.710 ±0.007 d* 0.0017 ±0.0002 a HA1T1 199.4 ±0.8 a 0.252 ±0.00 c 33.3 ±2.6 abc 0.0019 ±0.0002 a 0.996 ±0.002 e 0.0158 ±0.0021 b* 0.675 ±0.006 e* 0.0018 ±0.0002 ac Different letters (a,b,c…) in the same column indicate a significant difference among samples (p <0.05) for each parameter. * indicate significant differences among up and down parameters for each sample. For up and down curves r 2 >0.9990 (equation (1)). Note: For both Newtonian controls (A1T0 and A3T0) κ is equal to dynamic viscosity (Pa⋅s). C.A. Tovar et al. Journal of Functional Foods 93 (2022) 105092 4 3.2. Rheological properties of chitosan solutions and nanoparticle suspensions Flow curves of protonated chitosan solutions and nanoparticle suspensions were analysed in terms of the relationship between the instantaneous values of apparent viscosity ( η ) with increasing shear rate (dγ/dt =γ ⋅) for both up and down curves. Thus, η was fitted to γ ⋅ using the power law (equation (1)). η =κ ˆ A⋅(dγ dt)n−1 (1) where κ is the consistency coefficient, i.e. the viscosity value at 1 s −1 , and n is the flow index for the flow characterization. Both chitosan solutions (A1T0 and A3T0) exhibited a Newtonian behaviour (n =1) in up and down curves, so that viscosities ( η up and η down) were high and similar (Fig. 1). In both controls, the κ coefficient is the dynamic viscosity ( η ) because in both cases the flow resistance is independent of the shear rate. The Newtonian response may be explained by the high ζ potential, associated with a greater potential barrier among the positively charged chitosan molecules. Repulsive electrostatic forces and attractive interactions by hydrogen bonding would produce some energy equilibrium supporting the hypothesis of intermolecular interactions and subsequent inter-frictional effect among chitosan chains. The combination of both energy and structural contributions produced an ideal flow, exhibiting similar flow parameters (p >0.05) for both A1T0 and A3T0 solutions (Table 1). Using a 2% acetic acid solution, Li & Huang (2012) also reported a Newtonian flow of plain chitosan solutions even at concentrations higher than those of the present work (8–15 mg/ mL). When TPP was added (T1 and T3), new flow rheological-units were formed in the nanoparticle suspensions, conferring an initial structure which was progressively reduced (shear-aligned) with increasing shear rate, thus viscosity values decreased explaining the shear-thinning (pseudoplastic) flow (n <1) in the up curve (Table 1). Samples exhibited a stationary shear-thinning response, thus, no thixotropy was found in any case. This result was evidenced by the significantly higher values (p <0.05) of κ down vs κ up (Table 1). This timeindependent response could be related to the high level of electrostatic interactions, which may be reversibly broken and recombined (Tanaka & Edwards, 1992). Moreover, during the down curve, new flow rheological-units may be formed by new inter-nanoparticle associations, as evidenced by the higher viscosities in the down curves close to the initial rate (Fig. 1b). To clarify the comparison between samples, a column including apparent viscosities ( η 300 ) at fixed shear rate (300 s −1 ) has been added in Table 1. This value was chosen because 300 s −1 is an intermediate value between 100 s −1 and 800 s −1 , which provided a homogenous response (more reproducible data). Noteworthy is the significant decrease (p <0.05) in η 300 of all NP suspensions as compared with the chitosan solutions. A possible explanation for this effect is that the hydrodynamic volumes of cross-linked chitosan nanoparticles are altogether smaller than the extended polymer chains of pure chitosan, hence reducing particle friction and improving fluidity in the nanoparticle suspensions (Li & Huang, 2012). At a low TPP concentration (T1), the higher concentration of acetic acid (A3) produced a significantly lower η 300 than that at a lower acid concentration (A1) (Table 1). The drop in viscosity with decreasing shear rate and the subsequent reduction in the consistency coefficient, for both up and down curves (A3T1 vs A1T1), could be attributed to the greater repulsive electrostatic interactions produced by the larger amount of Na + cations needed to raise the pH up to 5 in A3T1 vs A1T1. These repulsive forces naturally increased the inter-particle distances among the rheological flow-units improving the fluidity of the suspensions in A3T1 vs A1T1. The same qualitative trend was found at a higher TPP concentration (T3). So, the greater positive difference for κ up in A1T3 vs A3T3 indicates a greater level of inter-particle friction in the A1T3 suspension. This fact is consistent with the higher particle size (Table 1), which promote inter-particle contact increasing the flow resistance in A1T3 vs A3T3. The flow of the sample which contained the protein hydrolysate (HA1T1) was different from the empty nanoparticles (A1T1). Thus, whereas the up curve exhibited an ideal (Newtonian) flow (Fig. 1a), evidenced by the flow index (n =0.996), in the down curve HA1T1 exhibited a shear-thinning response (Fig. 1b), indicated by n down <1 (Table 1). Moreover, both κ up and κ down were significantly lower than in A1T1, as was η 300 (Table 1). This fact could be explained by the greater proportion of hydrophilic (62%) vs hydrophobic (35%) residues in the hydrolysate, which, according to the entrapment efficiency, remained ≈40% free in the NP suspension. Thus, ionic groups in HA1T1 increased the repulsive forces in the nanoparticle suspension (HA1T1), which would separate the different flow units (nanoparticles and hydrolysate peptide chains) enhancing the system fluidity. This result is consistent with the lower PdI in HA1T1 vs A1T1, as it decreases when anionic polymerization occurs (Sperling, 2001). 3.3. Vinaigrette properties The nanoparticle suspensions made with low acetic acid and TPP concentrations, without and with fish hydrolysate (NP =A1T1 and NPH Fig. 1. Flow curves for plain chitosan solutions and nanoparticle suspensions: up curves (a) and down curves (b). T =20 ◦C. C.A. Tovar et al. Journal of Functional Foods 93 (2022) 105092 5 =HA1T1, respectively) were incorporated to the formulation of vinaigrettes. The effects were compared to those of soluble chitosan added at the same concentration in its protonated polymer form (A1T0). The vinaigrette composition (50:25:25, oil:vinegar:water) was selected from a visual point of view to ensure a fluid emulsion-like appearance immediately after shaken and poured, allowing for an easy distribution in a typical Mediterranean salad. In the corresponding vinaigrettes, chitosan dissolved in acetic acid or cross-linked into nanoparticle suspensions, was added as a substitute for water, and despite its inclusion at a very low concentration (0.05%, w/v), it led to substantial changes with respect to the control sample. 3.3.1. Phase separation The visual appearance of newly shaken vinaigrettes (oil/vinegar mixtures) before and after complete gravitational phase separation (3 h) is shown in Fig. 2a. Fig. 2b represents the creaming index (CI) values of the different samples during phase separation. The time elapsed until complete vinegar/oil phase separation varied among the different samples, and proceeded initially faster (inset in Fig. 2b) in the nanoparticle-containing samples, especially in those with the entrapped hydrolysate (VNPH) (p <0.05). In contrast, from 15 min onwards, VNP showed the lowest (p <0.05) CI values. Phase separation was almost completed after 30 min in all vinaigrettes, excluding VCS, which still progressed until 180 min (p <0.05). At this time, CI was noticeably lower (p <0.05) in VNP, followed by VNPH, while no significant differences (p >0.05) were found between VC and VCS. The separated oil phase at the top showed a stable emulsion-like appearance in both nanoparticle-containing samples even after 24 h, unlike in VCS and VC samples. It should be noted that the emulsion appearance in the latter was completely lost with time, and a flocculated sediment appeared at the bottom, which was discarded for subsequent analysis. The high CI values indicated the poor stability of the emulsified vinaigrettes, which became prone to phase separation under the middle ratio of the components (oil/vinegar +water) and their intrinsic acidic conditions (pH 3.2–3.6), in agreement with Zhang et al. (2021a). The mixture containing dissolved chitosan (VCS) barely reduced the CI, denoting the poor emulsifying properties of chitosan at such an acidic pH, given the protonation state of chitosan’s amine groups (Liu et al., 2012). In contrast, the nanoparticles contributed to decrease the extent of phase separation and to increase the emulsion stability in the oil phase. Franco-Ribeiro et al. (2020) reported the stabilizing effect of TPPcrosslinked chitosan nanoparticles in roasted coffee oil-in-water Pickering emulsions, based on their capacity to adsorb and concentrate at the oil droplets surface, which tend to separate from each other due to repulsive forces promoted by the high nanoparticles ζ potential. Likewise, chitosan nanoparticles similar to those of the present work have already shown a clear stabilizing effect in emulsions formed by sunflower oil and egg white (Lima et al., 2021a). The stabilizing effect of these nanoparticles, however, was not strong enough to avoid the vinegar/oil phase separation, probably because, in addition to the low pH, the nanoparticles concentration was not high enough (Zhang et al., 2021b). Fig. 2. Phase separation and creaming index (CI) variation of vinaigrettes during 180 min. VC: control with water; VNP: with empty nanoparticles; VNPH: with hydrolysate-loaded nanoparticles; VCS: with protonated chitosan polymer. C.A. Tovar et al. Journal of Functional Foods 93 (2022) 105092 6 3.3.2. Rheological properties (flow analysis) Vinaigrette Although flow curves were obtained by shearing up to 1000 s −1 , to simplify the manuscript only data up to 100 s −1 , where the shearthinning response was relevant, are shown. So, between 100 s −1 and 1000 s −1 the samples exhibited a Newtonian behaviour in up and down ramps without relevant differences (data not shown). From 0.1 s −1 to 100 s −1 the flow curves of vinaigrettes (control, with the chitosan solution and with nanoparticle suspensions) exhibited a shear-thinning response (n <1) for both up and down curves, thus the apparent viscosity was progressively reduced up to 100 s −1 (Fig. 3 a, b). This result indicates that diverse micro-structural rearrangements were shear-induced in the plane of the applied stress. The vinaigrette containing dissolved chitosan (VCS) had a tendency to present higher κ up and κ down values and lower n up and n down flow indices than the control (Table 2). Thus, the chitosan-enriched vinaigrette naturally increased the consistency enhancing the velocity at which the inner rearrangements allow molecules to align in the shear direction. This effect was intensified when nanoparticles were added to the vinaigrette, particularly in the case of the empty nanoparticles, where higher values for κ up and κ down and significant lower (p <0.05) exponents n up and n down were found as compared with the other samples (Table 2). So, these lower exponents, especially in the down curve (n down ), indicate a faster ability to re-organize the inner structure during the downward ramp. Combining both the higher consistency and lower exponents would explain the increased emulsion stability reflected in the lower CI value. Moreover, the great surface area per unit mass with nanoparticles may promote the fatty adsorption fostered by the greater inter-particle dynamism of the mixture evidenced by the lower n up and n down indices. In the vinaigrette with the hydrolysate-containing nanoparticles (VNPH) the structural role of the nanoparticles was screened and was found to be partially mitigated as compared with the empty ones. So, the greater ionic strength associated to the electrostatic charges of the free hydrolysate would increase the repulsive forces among nanoparticles reducing, consequently, the response velocity for both up and down ramps. This trend may be observed in the significantly higher (p <0.05) values for both n up and n down in VNPH vs VNP (Table 2). So, when flow indices increased between 0 and 0.8, the shear-thinning response became less intense, indicating a slower velocity in the internal rearrangements (Mezger, 2014). Separated oil phase The separated oil phase, which became demulsified after 24 h in the control sample (EC), exhibited a Newtonian response, evidenced by the practically constant and low values of viscosity with increasing shear rate (Fig. 3c,d). When the chitosan solution was added (ECS) the flow response became shear-thinning. The ECS sample exhibited significantly greater (p <0.05) κ up and κ down values, significantly decreasing (p < 0.05) the corresponding flow indices n up and n down as compared with those in EC (Table 2). Thus, chitosan molecules increased the fluid consistency enhancing the directed rearrangements in the shearing plane. This fact was reinforced in the ENPH sample, which exhibited the highest consistencies (κ up and κ down ) and the lowest n up and n down values (Table 2). In this case, the charged peptides energize the shear-thinning flow promoting faster inter-particle rearrangements during up and down ramps. Thus, the untrapped protein hydrolysate fraction might play an additional role as a surfactant by adsorbing onto the surfaces of the oil droplets and stabilising them through electrostatic repulsive and steric forces. The greater dynamism (lower flow indices) of the shear-thinning Fig. 3. Flow curves for the different vinaigrettes and separated oil phases: up curves (a,c) and down curves (b,d). T =20 ◦C. VC, EC: control with water; VNP, ENP: with empty nanoparticles; VNPH, ENPH: with hydrolysate-loaded nanoparticles; VCS, ECS: with protonated chitosan polymer. C.A. Tovar et al. Journal of Functional Foods 93 (2022) 105092 7 Table 2 Flow parameters (from Eq. (1)) for vinaigrettes and emulsion-gels (separated oil phase), and power law parameters (from Eqs. (2) and (3)) of mechanical spectra for emulsion-gels: VC, EC (control with water), VNP, ENP (with empty nanoparticles), VNPH, ENPH (with hydrolysate-loaded nanoparticles) and VCS, ECS (with chitosan in solution) at 20 ◦C. Samples κ up (Pa⋅s -n ) n up κ down (Pa⋅s - n ) n down G 0 ′ (Pa) n’ r 2 G 0 ′’ (Pa) n’’ r 2 tanδ VC 0.40 ±0.15 b 0.379 ± 0.015 d 0.211 ± 0.026 b 0.471 ±0.026 g VCS 1.11 ± 0.38 bc 0.162 ± 0.003 c 0.76 ± 0.42 bcd 0.177 ± 0.004 d VNP 1.88 ±0.35 c 0.090 ± 0.002 a 1.56 ± 0.72 bd 0.0074 ± 0.0001 b VNPH 1.22 ± 0.32 bc 0.167 ± 0.002 c 1.41 ±0.17 d 0.0406 ± 0.0006 a EC 0.062 ± 0.002 a 0.97 ±0.13 g 0.060 ± 0.002 a 0.98 ±0.10 h ECS 0.76 ±0.11 b 0.445 ± 0.017 e 1.15 ± 0.47 cbd 0.287 ±0.010 e 14.7 ± 3.8 a 0.087 ± 0.003 c 0.974 2.62 ± 0.66 a* 0.306 ± 0.015 c 0.957 0.178 ± 0.004 c ENP 0.53 ±0.27 b 0.562 ± 0.013 f 0.636 ± 0.070 c 0.395 ±0.019 f 27.5 ± 6.6 a 0.078 ± 0.003 b 0.975 5.42 ± 0.70 b* 0.243 ± 0.012 b 0.954 0.201 ± 0.024 b ENPH 1.97 ± 0.79 bc 0.122 ± 0.004 b 1.45 ± 0.70 abcd 0.126 ±0.004 c 57.9 ± 6.9 b 0.062 ± 0.001 a 0.989 8.72 ± 0.70 c* 0.212 ± 0.011 a 0.954 0.151 ± 0.005 a Values are given as mean ±standard deviation. Different letters (a,b,c…) in the same column indicate a significant difference between samples (p <0.05) for each parameter. * indicate significant differences among up and down parameters for each sample. For up and down curves r 2 >0.90 (Eq. (1)). Fig. 4. Mechanical spectra for the different vinaigrettes and separated oil phases: Viscoelastic moduli (G’ and G’’) (a) and loss factor (tanδ) (b) of vinaigrettes; Viscoelastic moduli (G’ and G’’) (c) and loss factor (tanδ) (d) of separated oil phases. T =20 ◦C. VC, EC: control with water; VNP, ENP: with empty nanoparticles; VNPH, ENPH: with hydrolysate-loaded nanoparticles; VCS, ECS: with protonated chitosan polymer. C.A. Tovar et al. Journal of Functional Foods 93 (2022) 105092 8 flow (ENPH) is consistent with the faster (oil/vinegar) phase separation already observed after 2.5–5 min (inset Fig. 2). In addition, the higher consistency of ENPH vs ENP is coherent with the lower volume occupied by the oil separated phase in VNPH vs VNP (Fig. 2). 3.3.3. Viscoelastic properties Vinaigrette The mechanical spectrum of the control vinaigrette (VC) was the only one that showed the crossover point between storage (G’) and loss (G’’) moduli at ω =8 rad/s. So, for ω >8 rad/s sample C exhibited a fluid-like behaviour (G’’ >G’) while for ω <8 rad/s it showed a solidlike response (G’ >G’’) (Fig. 4a). This result indicates that some rearrangements were enhanced at lower frequencies (higher oscillation times), which promoted a shear-induced gelation (Pi˜ neiro-Lago et al., 2020). This result shows the natural instability of the control vinaigrette, evidenced by the noticeable higher slope of G’’ vs ω than that of G’ vs ω . Both trends converge in the continuous increase of the loss factor (tanδ) with increasing ω (Fig. 4b). This specific viscoelastic response is consistent with the higher degree of aqueous (vinegar) and oil phase separation observed in this sample. The vinaigrette containing dissolved chitosan (VCS) stabilised the mixture somewhat, as determined by the practically constant values of G’ with increasing ω , while G’’ increased slightly at higher frequencies (Fig. 4a). Thus, chitosan molecules increased the firmness of the modified vinaigrette resulting in a gel-like structure (G’ >G’’) in all the frequency-range. This result was also evidenced by the lower values of the loss factor (tanδ) in VCS vs VC samples indicating a greater energy content of bonds (energy cohesiveness) which is equivalent to a more solid-like property (Moreno et al., 2020) in the chitosan-modified vinaigrette (Fig. 4b). The nanoparticles with the hydrolysate (VNPH) led to a similar mechanical spectrum than that of the VCS sample, as was evidenced by the similar values of G’, G’’ and tanδ with frequency (Fig. 4a,b). For the empty nanoparticles, the mechanical spectrum of VNP exhibited a greater frequency dependence of G’, tanδ and particularly G’’ compared with the VNPH and VCS samples, indicating a greater weakness of the gel-like response for the vinaigrette containing empty nanoparticles (Rao, 2007), but a noticeably structuring effect when compared with the control vinaigrette. Separated oil phase The demulsified oil phase in the control sample (EC) exhibited a fluid-like response (G’’ >G’) for all frequency intervals (Fig. 4c). The slope of the linear fit (G’’ vs ω ) was 0.059 ±0.004 Pa⋅s, which was similar to the dynamic viscosity obtained from the flow analysis (Table 2). This similitude between both parameters corroborates the Newtonian behaviour of the EC fluid. The ideal flow entails high values of the loss factor (tanδ ≫ 1) for EC, irrespective of the frequency, therefore, those data are not comparable with those of the other (gellike) samples, and thus they were not included in Fig. 4d. When the chitosan solution (ECS), empty nanoparticles (ENP) or hydrolysate-loaded nanoparticles (ENPH) were added, the mechanical spectra of the oil-phase were essentially different than the control (EC) showing a gel-like behaviour with a higher frequency dependence of G’’ as compared with G’. This effect may be analysed through the power-law parameters (Eq. (2) and (3)): G’=G’ 0 ˆ A⋅ ω n’(2) G’’ =G’’ 0 ˆ A⋅ ω n’’ (3) where G 0 ′and G 0 ′’ are the storage and loss moduli at 1 rad/s, and n’ and n’’ denote the rate of increase of G’ and G’’, respectively, with increasing ω . In general, the nanoparticles increased both G 0 ′and G 0 ′’ compared with those in ECS (Table 2), showing the strengthening produced by the nanoparticles on the viscoelasticity of these “emulsion gels”. So, the greater surface area per unit mass with the smaller size of the nanoparticles improved the lipid adsorption, increasing consequently the emulsion-gel strength; specifically for the hydrolysate-containing nanoparticles ENPH, whose G 0 ’ and G 0 ’’ increased considerably (290% and 230%, respectively) in comparison with those in ECS (Table 2). The higher increase proportion in G 0 ’ vs G 0 ’’ shows the gain of energy cohesiveness of the emulsion-gel ENPH (lower tanδ =G 0 ’’/G 0 ’). Therefore, when the hydrolysate was added, a structural benefit, shown in the formation of a more solid-like emulsion-gel network evidenced by the lower values of tanδ (Table 2), was produced. So, energy cohesiveness expresses the bond strength between the possible interactions: waternanoparticle-oil and nanoparticle-nanoparticle-water, specifically considering the effects derived from the presence of peptide chains, which could remain free or adsorbed to the nanoparticle surface. This peculiar cohesiveness is in line with the volume contraction in the oil-phase, in accordance to the higher creaming index in VNPH compared with VNP (Fig. 2b). However, with the empty nanoparticles (ENP), the relative increase in G 0 ′vs G 0 ′’ compared with that in ECS was inverse, so G 0 ′increased (87%) less than G 0 ′’ (105%). This result led to a significant increase (p < 0.05) in the loss factor (tanδ) (Table 2) resulting in a loss in the ideal network fraction (elastic gel character) of the emulsion-gel network (ENP) as compared with that in ENPH. The n’ and n’’ exponents quantify the change velocity of G’ and G’’ with ω respectively. They are useful tools to compare the time-stability of diverse gel networks (Borderías et al., 2020). In general, the frequency dependence was different in G’ and G’’, being n’’>n’ in the modified emulsion-gels irrespective of the additive (Table 2). This difference in both exponents justifies the natural instability of these emulsion-gels. Thus, n’’ >n’ expresses a great inner dynamism in the network which tends to flow at a higher frequency, while at lower frequencies it tends to present a more solid-like behaviour. Both kinds of nanoparticles introduced certain time stability, as was evidenced by the lower relative difference between n’ and n’’ as opposed to that in ECS (Table 2). 3.3.4. Antioxidant properties All studied vinaigrettes showed slight antioxidant properties, both in terms of radical scavenging capacity (ABTS) and ferric ion reducing power (FRAP) (Fig. 5a,b), which could be attributed in part to antioxidant compounds present in the sherry vinegar and in the sunflower oil. The addition of chitosan, both in the dissolved protonated state and in the cross-linked nanoparticle form, contributed noticeably (46–64%) to increase the radical scavenging capacity of vinaigrettes (Fig. 5a), despite the low concentration used. In contrast, no significant changes in FRAP values were found (Fig. 5b). The radical scavenging capacity of chitosan has been well documented, and it is known to be much higher than the reducing power (Younes & Rinaudo, 2015). The empty and hydrolysateloaded nanoparticles did not confer significant (p >0.05) radical scavenging capacity to the corresponding vinaigrettes (VNP and VNPH, respectively) beyond that provided by chitosan itself (VCS), despite the previously reported antioxidant capacity of this hydrolysate (Lima et al., 2019) and the corresponding hydrolysate-loaded nanoparticles (Lima et al., 2021a). The low hydrolysate concentration in the vinaigrettes and poor diffusion of peptides from the nanoparticles would explain the lack of apparent activity. Besides the potential bioactive interest of the nanoparticle-enriched vinaigrettes, the radical scavenging mechanism could play a role to prevent lipid oxidative instability. 3.3.5. Lipid digestion In order to explore the possible functional role of chitosan, empty nanoparticles and hydrolysate-loaded nanoparticles in decreasing the lipid digestibility, the vinaigrettes were subjected to an in vitro simulated gastrointestinal digestion. The amount of free fatty acids produced as a consequence of digestion is shown in Fig. 5c. As expected, the amount of free fatty acids after digestion increased in all vinaigrettes, indicating that lipids were hydrolysed by digestive enzymes and bile salts. However, the lipid digestion in all chitosan-containing vinaigrettes was C.A. Tovar et al. Journal of Functional Foods 93 (2022) 105092 9 lower (p <0.05) compared to the control sample (VC), being the reduction more pronounced when chitosan was in nanoparticle form (VNP and VNPH). The concentration of free fatty acids in VCS was 32.8% lower with respect to VC, while in VNP and VNPH, free fatty acids were reduced by 45.2% and 46.2%, respectively. The decrease in lipid digestibility was attributed to the fat binding capacity of chitosan, which increased significantly when included in the nanoparticle form. The great surface area per unit mass of nanoparticles facilitated the adsorption of lipids (Zhang et al., 2012). The protective chitosan layer around the lipid droplets promoting their flocculation could have inhibited the lipase-fat interaction inside the micelles (Mun, Decker, Park, Weiss & McClements, 2006; Hur, Kim, Choi & Lee, 2013). Fig. 6 shows the microscopic images (40x) of micelles after vinaigrettes digestion. These images confirmed that chitosan and chitosan nanoparticles had the capacity to aggregate the micelle droplets, thereby reducing lipid digestion (Hur et al., 2013). However, no great differences between VCS, VNP and VNPH were observed. Hur et al. (2013) also found that the micellar size after in vitro digestion of egg yolk was much larger when egg yolk was encapsulated with chitosan. 4. Conclusions Simple emulsion-like oil/vinegar mixtures (vinaigrettes) are an example of a food application in which dilute acid nanoparticle suspensions could be directly added without introducing any further modification. Despite the different TPP and acetic acid concentrations, particle properties varied within a narrow range (200 – 274 nm; ζ potential: 28 – 36 mV), including those with the entrapped hydrolysate. Vinaigrettes exhibited a stationary shear-thinning flow, which was reinforced after adding dilute nanoparticle suspensions. Chitosan and nanoparticles, especially those loaded with the hydrolysate, increased the energy cohesiveness of vinaigrettes. The addition of chitosan and chitosan nanoparticles contributed to increase>45% the antioxidant properties of vinaigrettes. Furthermore, the nanoparticles reduced more effectively the lipid digestibility (around 45%) than the dissolved chitosan (33%). The present work demonstrated that chitosan nanoparticles improved the stability and rheological properties of vinaigrettes, which could also be useful to supplement weight loss diets. Ethical statement Hereby, I, M. Carmen G´ omez-Guill´ en, consciously assure that for the manuscript “The effect of chitosan nanoparticles on the rheo-viscoelastic properties and lipid digestibility of oil/vinegar mixtures (vinaigrettes)” the following is fulfilled: 1) This material is the authors’ own original work, which has not been previously published elsewhere. 2) The paper is not currently being considered for publication elsewhere. 3) The paper reflects the authors’ own research and analysis in a truthful and complete manner. 4) The paper properly credits the meaningful contributions of coauthors and co-researchers. 5) The results are appropriately placed in the context of prior and existing research. 6) All sources used are properly disclosed (correct citation). 7) All authors have been personally and actively involved in substantial work leading to the paper, and will take public responsibility for its content. CRediT authorship contribution statement Clara A. Tovar: Methodology, Investigation, Formal analysis, Data curation. Karina Oliveira Lima: Investigation, Formal analysis, Data curation. Ail´ en Alem´ an: Investigation, Formal analysis, Data curation. Fig. 5. Antioxidant properties and lipid digestibility of vinaigrettes: (a) ABTS radical scavenging capacity, (b) ferric ion reducing power; (c) free fatty acids content before and after digestion. VC: control with water; VNP: with empty nanoparticles; VNPH: with hydrolysate-loaded nanoparticles; VCS: with protonated chitosan polymer. Different letters (a,b,c…) indicate significant differences (p <0.05) among samples. Fig. 6. Photomicrographs of the digested vinaigrettes (40x). VC: control with water; VNP: with empty nanoparticles; VNPH: with hydrolysate-loaded nanoparticles; VCS: with protonated chitosan polymer. C.A. Tovar et al.