Impact of high-intensity ultrasound waves on structural, functional, thermal and rheological properties of rice flour and its biopolymers structural features

1 Impact of High-Intensity Ultrasound Waves on structural, functional, thermal and rheological properties of rice flour and its biopolymers structural features Antonio J. Vela; Marina Villanueva; Ángela García Solaesa; Felicidad Ronda* Department of Agriculture and Forestry Engineering, Food Technology, College of Agricultural and Forestry Engineering, University of Valladolid, Spain *Corresponding author. E-mail: [email protected] Abstract Physical modifications of flours are an environment-friendly technology receiving increasing attention for widening the range of utilization of these raw materials. Rice flour was modified with ultrasound treatments at a frequency of 24 kHz and varying treatment time (2 to 60 min) and flour concentration (5 to 30%) in the dispersion. The effect of the modification was measured in the flours’ physical, functional, pasting and rheological properties. Particle size of treated samples was reduced, and particle’s disruption was observed by SEM; this had an impact on the water absorption ability, as shown by a sharp increase of swelling power. The thermal properties showed a significant reduction of gelatinization enthalpy, as well as narrowing of the gelatinization temperature range, characteristic of better packed starch crystalline structures after sonication. Modified patterns in starch and proteins were obtained with XRD and FTIR, which indicated impact to their crystalline and amide I secondary structures as a consequence of ultrasonication. Pasting profiles were found to be reduced with increasing treatment time, while higher concentrations did not significantly change the modification achieved. The pasting temperature was found to be significantly increased in all treated samples. Ultrasound treatment led to gels with higher strength, obtaining lower values of tan δ with increasing sonication time and flour concentration. Ultrasound allowed the modulation of rice flour functionality, being the effect independent on the concentration of the treated flour dispersion, up to 30%, and increased by treatment time up to 10 minutes; for longer treatments not many differences were found. Keywords: Rice flour; High-Intensity Ultrasound Treatment; Thermal properties; Pasting properties; Biopolymers structure; Gel Rheological properties. 2 1. Introduction Flours are among the main sources of carbohydrates in human consumption, commonly used in the food industry as thickener and bulking agents (Iida, Tuziuti, Yasui, Towata, & Kozuka, 2008). Rice flour is considered an interesting cereal grain in the gluten-free market because it can be used to replace wheat in an easier way than other cereals or pseudocereals, mainly because of its functional properties such as having a bland taste, its white color, digestibility and low allergenicity (Villanueva, Harasym, Muñoz, & Ronda, 2018; Wu, Chen, Li, & Wang, 2010). However, native flours have a limited industrial range of use due to their natural characteristics that limit their functionality, then modifications are applied to alter their physicochemical properties. The modification methods can be genetic, mechanical, chemical, enzymatic or physical (Zheng et al., 2013; Zhu, 2015). Physical modifications stand out for being an environment-friendly technology, and involving a reduced use of chemicals and processing time (Amini, Razavi, & Mortazavi, 2015). Ultrasounds (US) have been used in recent years in the food industry as a resource to generate physical modifications (Chemat, Zill-E-Huma, & Khan, 2011). Ultrasounds are acoustic waves above the human audible threshold, that can be classified in low intensity-high frequency US (>1MHz) and high intensity-low frequency US (20 - 100 kHz) (Zhu & Li, 2019). Sonication treatments are applied in a liquid-solid system, mostly using water as the medium (Zhu, 2015). Ultrasound treatments are capable of altering the structure of flour’s main biopolymers, starches and proteins, by acoustic cavitation and the generation of free radicals, but in the case of high intensity-low frequency US the amount of free radicals generated is low (Ashokkumar, 2015; O’Sullivan, Park, Beevers, Greenwood, & Norton, 2017). The acoustic energy of ultrasound is not absorbed by the treated molecules, so it is transformed to a chemically usable form through the cavitation phenomenon (Zhu, 2015). Acoustic cavitation is the cyclic generation and collapse of tiny bubbles in the aqueous medium as a result of pressure variation and US waves passing through. The cyclic collapsing of the bubbles results as water micro-jets shooting onto the particle’s surface, leading to local rises of temperatures and very high shearing forces, that cause mechanical damage to the sonicated matrix, detected as cracks, pitting and irregular surfaces, leading to a reduction of the particle size and molecular weight (Ashokkumar, 2015; Bai, Hébraud, Ashokkumar, & Hemar, 2017; Mir, Riar, & Singh, 2019; O’Sullivan et al., 2017; Zheng et al., 2013). The effect of the physical forces derived from cavitation is very homogeneous throughout the dispersions sonicated in water, since water has a high surface tension which makes a very effective medium for 3 cavitation (Ashokkumar, 2015; Knorr, Zenker, Heinz, & Lee, 2004). The effect of US treatments on starch depends on variables such as the sonication frequency and power, time and temperature of the treatment, and botanical origin of the sample (Zhu, 2015). In the cereal science field, research has been mainly focused on the effect of US treatments on starches from diverse botanical origins such as corn (Amini et al., 2015; Flores-Silva et al., 2017; Jambrak et al., 2010), wheat (Sujka, 2017), potato (Bai et al., 2017; Degrois, Gallant, Baldo, & Guilbot, 1974), sweet potato (Zheng et al., 2013), maize (Bel Haaj, Magnin, Pétrier, & Boufi, 2013; Luo et al., 2008), barley (Kaur & Gill, 2019), tapioca (Iida et al., 2008), plantain (Carmona-García et al., 2016) and rice (Yu et al., 2013; Zuo, Knoerzer, Mawson, Kentish, & Ashokkumar, 2009), while less research has been centered on flours. It is also important to generate knowledge of the effect that US treatments have on modifying flours since they are more complex systems widely used as a food ingredient, having the advantage of being cheap and nutritious. According to our knowledge, the study of their effect on rice flour is unexplored. The aim of this research was to investigate the effect of US treatments of 24 kHz frequency on functional, pasting, rheological and physical properties of rice flour, as a function of treatment time and flour concentration in the dispersion. 2. Materials and methods 2.1 Rice Flour The rice flour used in this experiment was supplied by Emilio Esteban SA (Valladolid, Spain). The flour was stored at 4°C until it was used. The moisture content was 13.67%, protein 6.60%, fat 2.06% and ash 0.34% (data provided by the manufacturer). This native flour was used without any processing as the control in the study. Rice flour dispersions were prepared for the treatments by suspending the proper amount of flour in distilled water to achieve the desired concentration. All concentrations are given in g dry flour/100 g of dispersion. 2.2 Ultrasound Treatment The ultrasound generator consisted of Hielscher UP400St sonicator (Hielscher Ultrasonics, Germany) equipped with S24d22D titanium tip working at a constant frequency of 24 kHz with a maximum output power of 180W, at 80% on-off pulse. The rice dispersions (400g) were treated in a glass jacket containing circulating water from a water bath to set the target temperature at 20ºC and keep it constant during the treatment. When the effect of the 4 treatment time was studied, the concentration of rice flour in the dispersion was fixed at 10g/100g while the time varied from 2 to 60 min (see Table 1). For evaluating the effect of the flour concentration, the treatment time was maintained at 60 min and the concentration of the rice flour in the dispersion varied from 5 to 30g/100g (Table 1). All dispersions were stirred during the treatment to avoid sedimentation and to ensure a homogenous temperature. After sonication samples were freeze-dried with Genesis Pilot Lyophilizer (SP Industries Inc, Warminster, USA). All samples were sieved to <250µm and stored at 4°C. Table 1. Sonication conditions and ID of treated samples used in the study. Sample Treatment time (min) Concentration (g solids/100g) Control (native) 0 -- TIM-2 2 10 TIM-5 5 10 TIM-10 10 10 TIM-20 20 10 TIM-40 40 10 TIM-60 60 10 CON-5 60 5 CON-10 60 10 CON-20 60 20 CON-30 60 30 TIM-60 and CON-10 are samples with the exact same experimental conditions 2.3 Particle size distribution The particle size distribution of the studied samples was stablished using a Mastersizer 3000 laser diffraction particle size analyzer (Malvern Instruments Ltd, UK). Results are expressed in median diameter (D50) and ((D90-D10)/D50) as described in Abebe, Collar, & Ronda (2015). All samples were measured in triplicate. 2.4 Scanning Electron Microscopy (SEM) A Quanta 200FEG scanning electron microscope (FEI, Oregon, U.S.A.) was used to study the surface microstructure of the samples, equipped with an X-ray detector, which allowed 5 the analysis of the samples without prior metallization. The visualizations were performed with an accelerating voltage between 3 and 5 keV in low vacuum mode using a secondary electron detector at different magnifications. Representative micrographs were selected for illustrative purposes. 2.5 Starch Damage The percentage of damaged starch was measured following the AACC Official Method 76- 31.01 (AACC, 2014) using a Megazyme starch damage kit (K-SDAM). The results were expressed as g/100g of the sample on dry basis. All samples were evaluated in triplicate. 2.6 Hydration properties of flour Water absorption capacity (WAC) was measured following the centrifugation method described by Abebe, Collar & Ronda (2015). The results were expressed in g H2O/g flour dry matter. Water absorption index (WAI), water solubility index (WSI) and swelling power (SP) were determined with slight modifications of the method indicated by Abebe et al., (2015). A sample of 2g was dispersed in 40mL of distilled water in 50mL centrifuge tubes. Samples were boiled for 15 min and cooled down to room temperature before being centrifuged at 3000xg (Thermo Fisher Scientific, Waltham, U.S.A.) for 10 min. The supernatant was poured in a previously weighed evaporating capsule to determine its solid content, and the sediment was weighed. The weight of the soluble solids was determined by evaporating the water from the supernatant overnight at 110°C. WAI was expressed in g sediment/g flour dry matter (dm), WSI in g soluble solids/100 g flour dm and SP in g sediment/g of insoluble solids in flour. All results were referred to dry matter to avoid the effect of different water content in the samples. 2.7 X-Ray diffraction (XRD) XRD patterns were obtained using a Bruker-D8-Discover-A25 diffractometer (Bruker AXS, Rheinfelden, Germany) equipped with a Cu-Kα radiation (λ = 0.154 nm) at a voltage of 40 kV and a current of 40 mA. Before measurement all samples were equilibrated to 15% humidity using a saturated humidity ICP260 incubator at 15°C (Memmert GmbH, Germany). The radiation intensities of the equilibrated samples were measured in the range of 5° to 40° of 2θ diffraction angle, with a scan step size of 0.02°, receiving slit width of 0.02nm, scatter slit width of 2.92°, divergence slit width of 1° and a rate of 1.2°/min. The crystallinity of the 6 samples was calculated as the ratio between the reduced peak area assigned to the crystalline part and the total area using peak-fitting software DifracEVA with PDF2-2004 and Crystallography Open Database. 2.8 Fourier Transform Infrared Spectroscopy (FTIR) FTIR spectra of the samples were recorded using a FT-IR Nicolet iS50 spectrophotometer (Thermo Fisher Scientific, U.S.A.) coupled with an attenuated total reflectance (ATR) device equipped with a diamond crystal. All samples were equilibrated at a humidity of 15% (as described in 2.7). All measurements were performed in duplicate in the wavenumber range of 600-4000cm-1 with a resolution of 4cm-1 and accumulation of 64 scans. All spectra were normalized using OMNIC software (Thermo Fisher Scientific, U.S.A.). Amide I bands (1700- 1600 cm-1) were analyzed using PeakFit 4.12 (SeaSolve Software Inc., USA). The individual components of deconvolved curves were determined by iterative fitting procedure that assumed Gaussian band shapes. Positions of the band centers were evaluated from second derivative analysis. The relative areas of these bands were used to estimate the percentage of secondary structural features: β-turns (1700-1660 cm-1), α-helix (1658-1650 cm-1), random coil (1650-1640 cm-1) and β-sheet (1640-1600 cm-1) (Byler & Susi, 1986; Zhang et al., 2018). 2.9 Differential scanning calorimetry (DSC) The samples’ thermal properties were measured using a DSC3 calorimeter (Mettler Toledo, Spain). The sample (~6 mg) was weighed in a 40 μL aluminum pan. The corresponding amount of distilled water was added to reach a solid concentration of 30% w/w. The pans were sealed and allowed to rest for 30 min at room temperature before performing the measurement. The scan was performed from 0 to 110°C at a rate of 5°C/min, using an empty sealed pan as reference. The values measured were onset (TO), peak (TP) and endset (TE) temperatures (°C) and the enthalpy of gelatinisation (ΔH) (J/g flour dm). The gelatinized samples were kept in the pans at 42°C and after 7 days they were measured again with the same method to study their retrogradation transition. All measurements were performed in duplicate. 2.10 Pasting analysis The pasting properties of the samples were studied using a Kinexus Pro+ rheometer (Malvern Instruments Ltd, UK) equipped with a starch pasting cell geometry following the AACC 7 International Method 76-21.02 Standard 2 (AACC, 2017). The samples (3.50g on a moisture basis of 14%) were transferred to the test canister and 25.0±0.1mL of distilled water were added. The paddle speed was set at 160 rpm. The pasting temperature (PT), peak viscosity (PV), trough viscosity (TV), breakdown viscosity (BV), final viscosity (FV) and setback viscosity (SV), were calculated using the rSpace for Kinexus software (Malvern Instruments Ltd, UK). Samples were measured in duplicate. 2.11 Rheological properties of gels The dynamic oscillatory tests were performed using a Kinexus Pro+ rheometer (Malvern Instruments Ltd, UK) with serrated parallel plate geometry (40 mm diameter) and a working gap of 1 mm. The gel samples were prepared following the protocol described in section 2.10 and left on the bottom plate to rest for 5 min to allow relaxation. Stress sweeps were performed from 0.1 to 100 Pa at a constant frequency of 1Hz. Frequency sweeps were performed from 10 to 1 Hz, at 1.0 Pa, within the linear viscoelastic region (LVR). The tests were carried out at constant temperature (25 °C) controlled with a Peltier KNX2002 C25P plate (Malvern Instruments Ltd, UK). The data obtained from frequency sweeps were adjusted to potential equations as described by Ronda, Villanueva, & Collar (2014). Tests were carried out in duplicate. 2.12 Statistical analysis Analysis of variance (ANOVA) was carried out using Statgraphics Centurion v.6 (Bitstream, Cambridge, MN, U.S.A.). Tukey’s honest significant difference (HSD) test was used to evaluate significant differences between samples at p ≤ 0.05. 3. Results and Discussion 3.1 Morphology and particle size of samples Scanning electron micrographs of selected flours are illustrated in Fig. 1. The images of native flour showed the characteristic morphology of rice flour, which are polygonal compact structure of starch granules packed very tightly in the cell, being entwined with globular protein bodies and lipids (Villanueva, Harasym, et al., 2018). In all treated samples it was observed that sonication caused a general disruption of these structures (see Fig. 1 – B2- C2-D2-E2). In rice starch, it has been reported that ultrasound leads to the formation of cracks and pores (Yang et al., 2019), which were not observed in rice flour, probably as 8 consequence of the protective effect that the remaining components of flour (mainly proteins and lipids) have on starch granules. Subsequent magnifications revealed structural fragmentation (higher amount of small size particles and loosen starch granules), and changes in the surface shape of the macrostructures, being more rugged and uneven after treatment, with starch granules having a wider exposed area in the surface, caused by the influence that shear forces of cavitation have on the granules’ surface (Luo et al., 2008). Granulation and uniformity of particle size have long been assumed to be important factors affecting the processing performance of flours. The particle size distribution of the treated samples showed significant changes compared to the native flour (see Supplementary Fig. 1). While the control sample showed a small percentage of particles in the range of 1 to 10μm, all ultrasonicated samples presented a marked percentage increase in this range, and showed a clear bimodal particle size distribution. The results of the median diameter (D50) and size distribution ((D90-D10)/D50) (Table 2) also confirmed that sonication led to a reduction of particle size. Two minutes of sonication reduced D50 from 123 to 65 m; five minutes decreased it up to 49 m. Longer treatment times did not result in additional reduction. The effect of the flour concentration on the resulting particle size distribution was also significant (p<0.05). The greatest effect was obtained for the lowest flour concentration (CON-5) that led to a reduction of 72% in D50 with respect to the control flour. The effect was slightly diminished at higher concentrations with decreases in D50 of 52-58% for CON-10 to CON-30 samples. This reduction observed in every sonicated sample can be explained as a fragmentation caused by the cavitation phenomenon, given that collapsing bubbles generate material fatigue, followed by a gradual tearing off of microscopic particles (Degrois et al., 1974). It is believed that the fragmentation of the starch chains induced by sonication does not happen in a random manner, and there is a limit to that degradation and once that limit is reached, no further chain scission happens (Czechowska-Biskup, Rokita, Lotfy, Ulanski, & Rosiak, 2005). The decreased fragmentation in flours treated at higher concentration can be the result of a reduction of acoustic energy, due to a change of medium impedance (Amini et al., 2015; Degrois et al., 1974). 9 Figure 1. SEM images of (A) Control flour, (B) TIM-10, (C) TIM-60 (CON-10), (D) CON-5 and (E) CON-30 at a magnification of (1) 100x, (2) 500x, (3) 1500x and (4) 3000x. E1 E2 E3 E4 D1 D2 D3 D4 C4 C3 C2 C1 B1 B2 B3 B4 A1 A2 A3 A4 16 to an increase of the disordered structure, and indicates sensitivity of α-helix and β-sheet to ultrasonication. The differences in the secondary structure after ultrasonication are due to the shear forces of US mechanical action, disrupting the interactions between the protein molecules and influencing the protein molecule internal structure (X. Yang et al., 2017). A decrease of the intensity peaks (~1600cm-1) has also been indicated after sonication of wheat and sweet potato flour (Cui & Zhu, 2020) and quinoa proteins (Vera, Valenzuela, Yazdani- Pedram, Tapia, & Abugoch, 2019) at a frequency of 20kHz. The amide I zone is altered by oscillations of the polypeptide structure, which could be caused by flexural vibration frequencies of the intra- and inter-molecular hydrogen bonds, as a consequence of the cavitation phenomenon (Vera et al., 2019). Table 3. Protein secondary structure analysis of the studied samples. Protein secondary structure analysis (%) Sample α-Helix β-Sheet β-Turn Random coil Effect of treatment time Control 21.6e 36.7e 20.8bc 21.0a TIM-2 18.5a 36.1e 17.6a 27.8c TIM-5 21.2de 33.4d 22.6ef 22.9b TIM-10 19.2ab 31.5c 22.1de 27.4c TIM-20 20.1bc 25.0a 23.1f 31.9d TIM-40 19.0a 33.3d 20.1b 27.7c TIM-60 20.5cd 27.8b 21.3cd 32.4d SE 0.2 0.2 0.2 0.2 Analysis of variance and significance (p-values) *** *** *** *** Effect of flour concentration Control 21.6c 36.7d 20.8d 21.0a CON-5 23.5d 32.6c 14.1a 29.9c CON-10 20.5b 27.8b 21.3d 32.4d CON-20 19.5a 26.0a 19.2c 35.3e CON-30 21.9c 33.0c 17.3b 27.9b SE 0.2 0.2 0.2 0.2 Analysis of variance and significance (p-values) *** *** *** *** SE: Pooled standard error from ANOVA. The different letters in the corresponding column within each studied factor indicate statistically significant differences between means at p < 0.05. Analysis of variance and significance: *** p < 0.001. ** p < 0.01. * p < 0.05. ns: not significant. 17 Figure 4. Deconvolved amide I bands of control and selected sonicated samples. A) Control; B) TIM-10; C) TIM-40; D) TIM-60 = CON-10; E) CON-5; F) CON-30. Deconvolved FTIR spectra are represented by a continuous line and the fitted curves by a discontinuous line. Bands: β-turns (1700-1660 cm-1), α-helix (1658-1650 cm-1), random coil (1650-1640 cm-1) and β-sheet (1640-1600 cm-1). 160016201640166016801700 Absorbance (Arbitrary unit) Wavenumber (cm-1) A 160016201640166016801700 Absorbance (Arbitrary unit) Wavenumber (cm-1) B 160016201640166016801700 Absorbance (Arbitrary unit) Wavenumber (cm-1) C 160016201640166016801700 Absorbance (Arbitrary unit) Wavenumber (cm-1) D 160016201640166016801700 Absorbance (Arbitrary unit) Wavenumber (cm-1) E 160016201640166016801700 Absorbance (Arbitrary unit) Wavenumber (cm-1) F 18 Table 4. Thermal properties of native and sonicated rice flour First scan Second scan Sample ΔHgel (J/g) TO-gel (°C) TP-gel (°C) TE-gel (°C) ΔT (°C) ΔHam-lip (J/g) TP-am-lip (°C) ΔHret (J/g) TO-ret (°C) TP-ret (°C) TE-ret (°C) ΔHam-lip (J/g) TP-am-lip (°C) Effect of treatment time Control 9.58b 61.1a 74.6c 81.5a 20.4d 1.4a 96.6a 4.3b 38.6a 51.4a 63.0bc 3.1a 98.2b TIM-2 8.93a 61.1a 74.0a 79.7b 18.7c 1.8b 95.2a 3.4a 38.3a 51.1a 62.3ab 3.1a 97.0ab TIM-5 8.76a 61.5a 74.2abc 80.0b 18.5bc 1.6ab 95.5a 3.9ab 38.0a 50.8a 61.9a 3.3a 96.6ab TIM-10 9.02a 61.2a 74.18ab 79.3b 18.1abc 1.7ab 95.2a 3.5a 38.9a 51.5a 62.8abc 3.2a 97.0ab TIM-20 9.00a 61.8a 74.2ab 79.3b 17.5a 2.0b 95.3a 3.4a 38.9a 53.1b 62.6abc 3.2a 96.3a TIM-40 8.84a 61.6a 74.5bc 79.5b 17.9abc 2.0b 96.0a 3.7ab 38.1a 52.7b 61.8a 3.7b 96.7ab TIM-60 8.94a 61.7a 74.4abc 79.5b 17.8ab 1.9b 96.3a 3.9ab 38.4a 50.7a 63.5c 3.2a 97.4ab SE 0.13 0.2 0.1 0.3 0.2 0.1 0.4 0.1 0.5 0.2 0.2 0.1 0.4 Analysis of variance and significance (p-values) ** ns ** *** *** ** ns ** ns *** ** *** * Effect of flour concentration Control 9.58c 61.1ab 74.6bc 81.5b 20.4b 1.4a 96.6a 4.3b 38.6a 51.4ab 63.0ab 3.1ab 98.2a CON-5 9.23ab 61.3ab 74.3a 79.4a 18.1a 1.6ab 95.5a 4.1b 37.9a 50.9ab 62.7ab 2.7a 97.0a CON-10 8.94a 61.7b 74.4ab 79.5a 17.8a 1.9bc 96.3a 3.9ab 38.4a 50.7a 63.5b 3.1ab 97.4a CON-20 9.03a 60.9a 74.3a 79.5a 18.6a 1.7bc 96.4a 3.4a 37.8a 52.0b 62.7a 3.1ab 97.5a CON-30 9.49bc 61.5ab 74.7c 79.9a 18.4a 2.0c 96.2a 3.7ab 37.4a 51.4ab 62.4a 3.3b 97.4a SE 0.14 0.2 0.1 0.3 0.2 0.1 0.4 0.1 0.5 0.3 0.2 0.1 0.4 Analysis of variance and significance (p-values) ** * *** *** *** *** ns ** ns ns * * ns ΔHgel: Enthalpy of gelatinisation. To-gel, TP-gel, TE-gel: Onset, peak and endset temperatures of gelatinization. ΔT = (TE-gel - To-gel). ΔHam-lip = Enthalpy of the amylose-lipid complex dissociation. TP-am-lip = Peak temperature of the amylose-lipid complex dissociation. ΔHret = Enthalpy of melting of retrograded amylopectin. TO-ret, TP-ret, TE-ret: Onset, peak, and endset temperatures of melting of retrograded amylopectin. ΔHgel, ΔHret, ΔHam-lip are given in J/g dry matter. SE: Pooled standard error from ANOVA. The different letters in the corresponding column within each studied factor indicate statistically significant differences between means at p < 0.05. Analysis of variance and significance: *** p < 0.001. ** p < 0.01. * p < 0.05. ns: not significant. 19 3.6. Thermal properties The thermal properties observed from the phase transitions detected from the gelatinization (first) and retrogradation (second) scans are summarized in Table 4. DSC thermograms of all samples showed two endothermic peaks, the main one associated to the starch gelatinization transition and a smaller peak caused by the amylose-lipid complex dissociation, appearing at the temperature range of 90-105ºC (Villanueva, Ronda, Moschakis, Lazaridou, & Biliaderis, 2018). Gelatinization enthalpy ΔHgel was decreased after ultrasonication in all samples, from 9.58 J/g dm to 8.76-9.49 J/g dm. A similar behavior has been reported by Yang et al. (2019) when sonicating rice starch. Jambrak et al. (2010) stated that a decrease of the gelatinization enthalpy after ultrasonication possibly results from different alignments of hydrogen bonds within the starch molecules, as a consequence of the disruption of amorphous regions. The peak and endset gelatinization temperatures were decreased after sonication, while the onset gelatinization temperature was slightly increased. In all treated samples a narrowing of the gelatinization temperature range (ΔT) was determined, mainly due to a shift of TE-gel to significantly lower temperatures, going from 20.4°C, in the control, to up to 17.5°C at the more marked differences of time and concentration. Such reduction of ΔT, indicate a reinforcement of the starch structure after treatment probably because of a better packed crystalline leftovers structure after the breakage of amorphous regions and the melting of the weakest crystallites in the starch caused by cavitation (Amini et al., 2015; Chi et al., 2019; Luo et al., 2008). Even though an increase of TO-gel has been reported in several studies, the degree of said increase seems to vary depending mostly on the treatment conditions and treated matter, in this study the results obtained for sonicated samples were not found to be significantly different than the control (Carmona-García et al., 2016; Flores- Silva et al., 2017; Jambrak et al., 2010). Yang et al. (2019) did not report significant differences in the gelatinisation temperatures of rice starch after treatments at different sonication power levels, indicating that the range of modification achieved in TP-gel and TE-gel in this study could also be related to other components in the flour, not exclusively to starch. In the case of the amylose-lipid complex, the treatments significantly increased ΔHam-lip values obtained from the first scan going from 1.4 J/g dm in native flour to 1.6-2.0 J/g dm in the treated samples, probably due to higher availability of amylose leaked outside the starch granule after the attack to the amorphous regions caused by sonication (Amini et al., 2015). The temperature of this transition was also lower in sonicated-flour samples (Table 4). The fact that no significant differences were observed between treated samples could be 20 explained as a degradation process limited by a minimum chain length in polymers, given that sonication breaks polymeric chains in a non-random manner, which could have been possibly achieved in 10 min of treatment (Czechowska-Biskup et al., 2005). To study the retrogradation enthalpy of amylopectin, a second scan was performed after 7 days of sample storage at 4°C. Two peaks were also found during the second scan, the first one, very wide, at a peak temperature of 51°C, corresponding to the melting of recrystallized amylopectin, and the second one corresponding to the amylose-lipid complex dissociation (~97°C). The retrogradation enthalpy (3.9 J/g dm) and temperatures obtained in the second scan were slightly modified by the treatments (Table 4). The enthalpy of the amylose-lipid complex dissociation obtained in the second scan was similar for all (native and sonicated) samples, and higher than in the first (gelatinization) scan. Eliasson (1994) reported that the increased values usually found during a second scan are probably due to better conditions for complex formation after the first heating because the leaking of amylose from granules can occur at temperatures above the gelatinization temperature range. 3.7. Pasting properties The pasting parameters of studied flours are shown in Table 5. The pasting curves are shown in Figure 5. In flours, starch is the main component responsible for the development of pasting profiles. However, other components in rice flour also affect viscometric profiles, where proteins seem to play an important role (Meadows 2002). The results show that ultrasonication led to a decrease in the viscosity profile of rice flour during pasting. A similar effect was reported on ultrasonicated quinoa flour (Zhu & Li, 2019) and sweet potato starch (Zheng et al., 2013). Peak (PV), Breakdown (BV) and Setback viscosities (SV) were significantly decreased by ultrasound treatment. A decrease of PV could indicate that sonication weakens the granule structure (Chan, Bhat, & Karim, 2010). Lower PV values are usually related to an increased ratio of amylose to amylopectin in starch and to smaller granule size (Niu, Hou, & Zhao, 2017). Lower BV values after treatments indicate that the samples have a better capacity to withstand stress and heating than the control flour. The lower SV indicates a lower amylose retrogradation. BV was found to be significantly reduced with longer sonication time and higher concentration, up to -27% (TIM-60) and -32% (CON- 30), for the lowest values obtained for each studied variable (Table 5). SV also followed an increasing reduction with longer treatment time, (-13% in TIM-60), and lower concentrations, (-14% in CON-5). 21 Figure 5. Pasting profile of native and treated flours at different (A) time and (B) concentration. The grey line corresponds to temperature (°C). The pasting temperature (PT) was found to be significantly increased in all treated samples, which indicates that starch in rice flour was strengthened by the treatments as was observed from DSC assays. Opposite trend was reported by Yang et al (2019) in sonicated rice starch samples. They found increased viscometric profile and lower pasting temperatures in the treated samples with respect to native starch. The different effect of ultrasounds on rice starch and rice flour could be explained by the impact it has on rice proteins. As can be seen from FTIR spectra, proteins secondary structure was affected by US. Ultrasonication probably decreased proteins folding which would increase binding sites for binding water and other rice constituents. This would decrease the accessibility of water to granular components and reduce the rate and extent at which granular components leach out from the granule leading to lower pasting viscosities (Meadows, 2002). The increase in lipid-amylose complexation 0 10 20 30 40 50 60 70 80 90 100 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 0200 400 600 800 1000 1200 1400 Temperature (°C) Viscosity (Pa ·s) Time (s) Control TIM-2 TIM-5 TIM-10 TIM-20 TIM-40 TIM-60 0 10 20 30 40 50 60 70 80 90 100 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 0200 400 600 800 1000 1200 1400 Temperature (°C) Viscosity (Pa ·s) Time (s) Control CON-5 CON-10 CON-20 CON-30 B A 22 extent observed just after gelatinization in sonicated samples from DSC could explain the decrease in the amylose available to retrograde and SV values in treated samples. 3.8. Rheological properties of gels made with sonicated rice flour The rheological properties of gels made from sonicated rice flour were determined and compared to those of the gel made from the control/untreated flour by dynamic oscillatory tests. Table 5 presents the parameters obtained from fitting the power law model to the frequency sweeps data, and max values and the stress at the crossing point (G’=G’’ and tan =1) obtained from stress sweeps. The results showed that sonication increased max in samples sonicated longer than 10 min and at concentrations above 20% reaching values up to 47% (TIM-10) and 56% (CON-30) higher than the native sample. This denotes sonicated samples led to stronger gel that resisted higher stress before its structure disruption. The stress at which gels go from a solid-like behavior to a viscous-like one (cross over, G’=G’’) did not varied as result of sonication. The oscillatory tests showed flour sonication had a significant impact on viscoelastic properties of gels. As can be seen in Table 5, although the elastic modulus, G1’, remained unaltered after sonication, the viscous one, G1’’, decreased up to -35% in both TIM-60 and CON-5 samples, with the concomitant decrease in (tan )1, that went from 0.17 to 0.11. These changes mean an increase in the relative predominance of the solid/elastic behavior of the gels (lower G''/G' ratio), while their consistency is decreased (equal G' and lower G''). This can be translated into softer gels that are able to maintain better their shape. Kaur & Gill (2019) when treating rice starch at high intensity ultrasound concluded that (tan )1 remains unchanged for treatment times below 30 min, while in rice flour significant differences were found in all studied sonication times, indicating that the other components found in flour play a decisive role in the samples’ rheological properties. The decrease in the viscosity of the gels could be explained by the severe damage suffered by starch granules due to shear forces, causing the straightening out of amylose molecules resulting in a decrease of viscosity (Kaur & Gill, 2019). On the other side, the possible chains released from broken ramifications of amylopectin are unable to form a consolidated compact network during the gelatinization process (Carmona-García et al., 2016). Not many differences were found within treated samples, in agreement with the idea that the effect of ultrasonication has a disruption limit, as stated by Czechowska-Biskup et al. (2005), which in this study seems to happen in 10 minutes. 23 Table 5. Pasting parameters of studied samples. Sample PV (Pa · s) FV (Pa · s) TV (Pa · s) PT (°C) BV (Pa · s) SV (Pa · s) G1' (Pa) a G1'' (Pa) b tan ()1 c max (Pa) Cross over (Pa) Effect of treatment time Control 4.06d 4.15d 1.77a 80.2a 2.37d 2.37c 187a 0.092c 31c 0.342a 0.165d 0.250a 121a 175a TIM-2 3.81c 4.06bcd 1.76a 82.4b 2.05c 2.30bc 169a 0.087bc 24ab 0.354b 0.138c 0.267ab 124a 166a TIM-5 3.77c 4.08cd 1.74a 83.4cd 1.97c 2.36c 182a 0.084abc 25b 0.356bc 0.135bc 0.272abc 115a 166a TIM-10 3.66b 4.13cd 1.75a 83.0bc 1.92bc 2.37c 191a 0.069ab 23ab 0.360bc 0.124abc 0.291bcd 179b 194a TIM-20 3.61b 3.97bc 1.75a 83.5d 1.83ab 2.19b 201a 0.071ab 22ab 0.366c 0.117ab 0.295cd 174b 189a TIM-40 3.48a 3.75a 1.71a 83.2cd 1.79a 2.06a 190a 0.064a 21ab 0.366bc 0.111a 0.302d 174b 192a TIM-60 3.53a 3.87ab 1.80a 83.1c 1.73a 2.06a 174a 0.078abc 20a 0.367c 0.120abc 0.289bcd 147ab 174a SE 0.02 0.05 0.04 0.2 0.04 0.04 9 0.005 1 0.003 0.005 0.006 12 7 Analysis of variance and significance (p-values) *** *** ns *** *** *** ns ** *** *** *** *** *** * Effect of flour concentration Control 4.06d 4.15c 1.77a 80.2a 2.37c 2.37c 187ab 0.092b 31c 0.342a 0.165b 0.250a 121a 175a CON-5 3.53c 3.72a 1.72a 82.6b 1.83b 2.03a 167a 0.075ab 20a 0.363b 0.146b 0.288b 141ab 171a CON-10 3.53c 3.87b 1.80a 83.1b 1.73b 2.06ab 174a 0.078b 20a 0.367b 0.120a 0.289b 147ab 174a CON-20 3.45b 3.94b 1.80a 82.9b 1.71ab 2.10ab 191ab 0.073ab 23b 0.362b 0.119a 0.289b 178bc 188a CON-30 3.37a 3.94b 1.79a 82.9b 1.60a 2.13b 205b 0.055a 22ab 0.365b 0.105a 0.310b 189c 204a SE 0.02 0.04 0.03 0.2 0.04 0.02 8 0.006 1 0.004 0.005 0.009 12 9 Analysis of variance and significance (p-values) *** *** ns *** *** *** * ** *** *** *** *** *** ns PV = Peak Viscosity. FV = Final Viscosity. TV = Trough Viscosity. PT = Pasting Temperature. BV = Breakdown Viscosity. SV = Setback Viscosity. G1’ (elastic modulus), G1’’ (viscous modulus) and tan ()1 (loss tangent) are the coefficients obtained from fitting the frequency sweeps data to power law model and represent the moduli and loss tangent values at a frequency of 1 Hz. The a, b and c exponents quantify the dependence degree of dynamic moduli and the loss tangent with the oscillation frequency. max represents the maximum stress tolerated by the sample in the LVR. SE: Pooled standard error from ANOVA. The different letters in the corresponding column within each studied factor indicate statistically significant differences between means at p < 0.05. Analysis of variance and significance: *** p < 0.001. ** p < 0.01. * p < 0.05. ns: not significant 24 4. Conclusion Rice flour treated by ultrasounds at 24 kHz showed significant differences in functional, thermal, pasting and rheological properties. Physical differences could be observed at surface microstructure and particle size of treated samples, which define their interaction with water and, consequently, their pasting profile and rheological behavior. The thermal properties showed an important narrowing of gelatinization range, indicating a reinforced starch structure, probably because of a better packed crystalline leftovers structure. The information showed by XRD and the intensity ratios calculated using the FTIR spectra indicate a disruption of the outer branches of amylopectin chains due to sonication. FTIR also confirmed the modification of proteins secondary structure with the decrease in α-helix and β-sheet structures in benefit of the random coil structure. These molecular changes would explain the lower pasting profiles and increased PT obtained in treated samples. Rheological properties revealed that ultrasonication leads to softer gels but with higher elastic behavior. The results showed that the effect of ultrasonication does not highly depend on the concentration of the flour during the treatment within the studied range, which would make it advisable to use the highest concentration studied, 30%. On the other hand, the effect of time was detectable only up to 10 minutes; beyond this point not many differences were found. It can be concluded that ultrasound is a feasible technique for physically modify rice flour that would allow to tailor-make and improve its functional properties. Acknowledgement The authors thank the financial support of the Ministerio de Economía y Competitividad and the European Regional Development Fund (FEDER) (AGL2015-63849-C2-2-R), the Ministerio de Ciencia e Innovación (PID2019-110809RB-I00) and the Junta de Castilla y León/FEDER VA072P17. A. Vela thanks the Junta de Castilla y León for the doctorate grant and M. Villanueva thanks the Alfonso Martín Escudero Foundation for the post-doctoral grant. References AACC. (2014). AACC International methods, 76-31.01. Determination of Damaged Starch- Spectrophotometric Method. AACC International Approved Methods. https://doi.org/10.1094/aaccintmethod-76-31.01 AACC. (2017). AACC International methods, 76-21.02. General Pasting Method for Wheat or Rye Flour or Starch Using the Rapid Visco Analyser. AACC International Approved Methods, (11th Edition). https://doi.org/https://doi.org/10.1094/AACCIntMethod-76- 21.01 25 Abebe, W., Collar, C., & Ronda, F. (2015). Impact of variety type and particle size distribution on starch enzymatic hydrolysis and functional properties of tef flours. Carbohydrate Polymers, 115, 260–268. https://doi.org/10.1016/j.carbpol.2014.08.080 Amini, A. M., Razavi, S. M. A., & Mortazavi, S. A. (2015). Morphological, physicochemical, and viscoelastic properties of sonicated corn starch. Carbohydrate Polymers, 122, 282– 292. https://doi.org/10.1016/j.carbpol.2015.01.020 Ashokkumar, M. (2015). Applications of ultrasound in food and bioprocessing. Ultrasonics Sonochemistry, 25(1), 17–23. https://doi.org/10.1016/j.ultsonch.2014.08.012 Bai, W., Hébraud, P., Ashokkumar, M., & Hemar, Y. (2017). Investigation on the pitting of potato starch granules during high frequency ultrasound treatment. Ultrasonics Sonochemistry, 35, 547–555. https://doi.org/10.1016/j.ultsonch.2016.05.022 Bajer, D., Kaczmarek, H., & Bajer, K. (2013). The structure and properties of different types of starch exposed to UV radiation : A comparative study. Carbohydrate Polymers, 98, 477–482. Barth, A. (2007). Infrared spectroscopy of proteins. Biochimica et Biophysica Acta - Bioenergetics, 1767(9), 1073–1101. https://doi.org/10.1016/j.bbabio.2007.06.004 Bel Haaj, S., Magnin, A., Pétrier, C., & Boufi, S. (2013). Starch nanoparticles formation via high power ultrasonication. Carbohydrate Polymers, 92(2), 1625–1632. https://doi.org/10.1016/j.carbpol.2012.11.022 Bernazzani, P., Peyyavula, V. K., Agarwal, S., & Tatikonda, R. K. (2008). Evaluation of the phase composition of amylose by FTIR and isothermal immersion heats. Polymer, 49(19), 4150–4158. https://doi.org/10.1016/j.polymer.2008.07.022 Byler, D. M., & Susi, H. (1986). Examination of the secondary structure of proteins by deconvolved FTIR spectra. Biopolymers, 25(3), 469–487. https://doi.org/10.1002/bip.360250307 Carmona-García, R., Bello-Pérez, L. A., Aguirre-Cruz, A., Aparicio-Saguilán, A., Hernández- Torres, J., & Alvarez-Ramirez, J. (2016). Effect of ultrasonic treatment on the morphological, physicochemical, functional, and rheological properties of starches with different granule size. Starch/Staerke, 68(9–10), 972–979. https://doi.org/10.1002/star.201600019 Chan, H. T., Bhat, R., & Karim, A. A. (2010). Effects of sodium dodecyl sulphate and sonication treatment on physicochemical properties of starch. Food Chemistry, 120(3), 703–709. https://doi.org/10.1016/j.foodchem.2009.10.066 Chemat, F., Zill-E-Huma, & Khan, M. K. (2011). Applications of ultrasound in food technology: Processing, preservation and extraction. Ultrasonics Sonochemistry, 18(4), 813–835. https://doi.org/10.1016/j.ultsonch.2010.11.023 Chi, C., Li, X., Lu, P., Miao, S., Zhang, Y., & Chen, L. (2019). Dry heating and annealing treatment synergistically modulate starch structure and digestibility International Journal of Biological Macromolecules Dry heating and annealing treatment synergistically modulate starch structure and digestibility. International Journal of Biological Macromolecules, 137(February 2020), 554–561. https://doi.org/10.1016/j.ijbiomac.2019.06.137 Cui, R., & Zhu, F. (2020). Effect of ultrasound on structural and physicochemical properties