Effect of thermosonication batch treatment on enzyme inactivation kinetics and other quality parameters of cloudy apple juice

Accepted Manuscript Effect of thermosonication batch treatment on enzyme inactivation kinetics and other quality parameters of cloudy apple juice A.E. Illera, M.T. Sanz, O. Benito, S. Varona, S. Beltrán, R. Melgosa, A.G. Solaesa PII: S1466-8564(17)31225-0 DOI: https://doi.org/10.1016/j.ifset.2018.02.001 Reference: INNFOO 1921 To appear in: Innovative Food Science and Emerging Technologies Received date: 30 October 2017 Revised date: 2 February 2018 Accepted date: 2 February 2018 Please cite this article as: A.E. Illera, M.T. Sanz, O. Benito, S. Varona, S. Beltrán, R. Melgosa, A.G. Solaesa , Effect of thermosonication batch treatment on enzyme inactivation kinetics and other quality parameters of cloudy apple juice. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Innfoo(2017), https://doi.org/10.1016/j.ifset.2018.02.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT Effect of thermosonication batch treatment on enzyme inactivation kinetics and other quality parameters of cloudy apple juice. A.E. Illera, M. T. Sanz  , O. Benito, S. Varona, S. Beltrán, R. Melgosa. A. G. Solaesa Department of Biotechnology and Food Science (Chemical Engineering Section), University of Burgos, 09001 Burgos. Spain Abstract Cloudy apple juice has been treated by thermosonication in batch mode as an alternative processing to thermal treatment. Thermosonication was found to be effective to inactivate polyphenol oxidase; however, pectinmethylesterase was found to be more resistant. An increase of the working ultrasound amplitude and the amount of energy supplied to the juice led to lower enzyme residual activities. Enzyme inactivation kinetics were determined at different temperatures (from 44 to 67 ºC). Inactivation data were described by the first order kinetic model and the Weibull model, both models yielding good fitting. Thermosonication treatment caused a homogenization effect reflected in the shifting of the particle size distribution towards smaller diameters. The effect of the nature of dissolved gases in the juice on enzyme inactivation was studied by displacing the air dissolved in the juice by bubbling nitrogen or carbon dioxide, previous to the thermosonication experiments. Higher inactivation rates were obtained by displacing the air with nitrogen.  Corresponding author. Tel.: +34 947 258810. Fax: ++34947258831. E-mail address [email protected] ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Industrial relevance: Consumers demand of natural and fresh-like products has driven the food industry to investigate alternative technologies to replace conventional food heat treatments that may affect food quality. Among these technologies, thermosonication treatment is an attractive technology that can inactivate microorganisms and enzymes. This work shows that some enzymes that cause deleterious effect on cloudy apple juice can be more effectively inactivated by thermosonication than by thermal treatment, in the same temperature range, reducing the damages caused by heating Keywords: Cloudy apple juice, thermosonication, enzyme inactivation, dissolved gasses. 1. Introduction Cloudy apple juice is a good source of bioactive compounds such as ascorbic acid, polyphenols, and pectins. However, enzymatic browning causes serious product deterioration. One of the most important enzymes responsible for enzymatic browning is polyphenol oxidase (PPO). PPO catalyzes the oxidation of phenolic compounds to quinones, which will subsequently react with other compounds to form brown pigments that reduce juice quality. Other important quality parameter in cloudy apple juice is the cloud stability. One of the most accepted theories of cloud loss in juices, is based on the action of pectinmethylesterase (PME). This enzyme causes pectin demethylation that could expedite pectin precipitation with calcium ions present in the juice, causing clarification and turbidity loss (Briongos et al., 2016) Thermal treatments are widely used in the food industry to inactivate microorganisms and enzymes that cause deleterious effects on foods. However, consumers demand of natural and fresh-like products has driven the food industry to investigate alternative technologies to replace conventional food heat treatments, wich may affect food quality in terms of nutritional and physicochemical parameters. Among these technologies, ultrasonication is an attractive ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT technology that is considered to be simple, reliable, environmentally friendly and highly effective in achieving microbial decontamination (Dias et al., 2015). Inactivation of microorganisms and enzymes by sonication is attributed to physical (cavitation, mechanical effects) and chemical (formation of free radicals) effects (O’Donnell, Tiwari, Bourke, & Cullen, 2010). Inactivation of enzymes by ultrasonication has been attributed to the formation of localized hot spots upon collapse of bubbles, shear forces created by microstreaming and shock waves, as well as generation of free radicals through sonolysis of water (Kadkhodaee & Povey, 2008). Sonochemical activity (production of radicals) is influenced by operational parameters, being the effect of frequency and gas addition the most significant (Gielen et al., 2016). Furthermore, the presence of gases (air, oxygen or an inert gas) in the solution has been reported as necessary for enzyme inactivation (Grintsevich, Adzerikho, Mrochek, & Metelitza, 2001). In any case, enzyme inactivation mechanisms are specific to each enzyme and depend on their amino acid composition and their conformational structure (Anaya-Esparza et al., 2017). When ultrasound is combined with heating, the resulting technique, thermosonication (TS), is usually more effective as preservation technique even TS temperatures being lower than those used in thermal treatments. TS treatment has been found to be more efficient than thermal treatment for the inactivation of different types of enzymes responsible for deterioration of fruits and vegetable juices, among them, pectin methylesterase, PPO, lipoxygenases and peroxidases (Baltacioglu, Bayindirli, & Severcan, 2017). Recently, Anaya Esparza et al. (Anaya-Esparza et al., 2017) reviewed the advantages and limitations offered by the application of TS on different fruit and vegetable juices. In this work, the effect of TS on PPO and PME inactivation in Golden Delicious cloudy apple juice has been studied. The effect of different variables, such as amount of energy transmitted to the juice, expressed as power density, PD (W/mL), ultrasoung (US) amplitude, temperature and treatment time, on enzyme activity has been determined. In addition, ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT the effect of the nature of the dissolved gas (air, nitrogen and carbon dioxide) on enzyme inactivation has been evaluated. No previous studies have been found in the literature regarding the effect of this variable on fruit juices processed by TS. Finally, enzyme inactivation by TS was compared with previous results obtained by high pressure carbon dioxide (HPCD) treatment, as non-thermal technology (Illera et al., 2018), and with thermal treatment alone in the same temperature range. The effect of TS on other quality parameters of cloudy apple juice, such as particle size distribution (PSD), color and total polyphenols and hydroxymethylfurfural content has been also determined. 2. Materials and methods 2.1 Juice preparation Golden delicious apples were peeled, cut in cubes and submerged in a 0.3 % L-ascorbic acid solution to avoid enzymatic browning during processing. Apple cubes were wiped and immediately squeezed with a screw juice extractor. The liqueur was filtered with 2 layers of cheesecloth. The pH of the juice obtained was 3.89 ± 0.01 and the soluble solids content was 12.5 ± 0.1 °Brix. The juice was stored frozen at −18 °C until further use. Carbon dioxide (99.9 %) was supplied by Air Liquide S.A. (Spain). Nitrogen (98.5 %) was obtained by a Zefiro 25 LCMS nitrogen generator. 2.2 Thermosonication A 750 W Sonics MaterialTM with a 13 mm probe was used for thermosonication. Samples were processed at a constant ultrasound frequency of 20 kHz. Cloudy apple juice was introduced in a thermostated vessel (Φ = 4.8 cm, V = 199 cm3) and the probe was submerged in the juice at a constant depth of 2 cm from the bottom of the vessel. According to the manufacturer the US amplitude is 79 m when the amplitude control is set at 100 % and the 13 mm probe can handle volumes in the range from 50 to 150 mL. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Firstly, TS was performed in a continuous mode and fixing the amplitude at different levels (from 25 to 100 %), keeping constant the rest of operating variables: TS time (15 min), treated volume of apple juice (80 mL) and temperature of the jacketed water (40ºC). This way, the effect of amplitude on PSD and PPO inactivation was first analyzed. During each TS experiment, the temperature profile was registered. The energy input was recorded after the experiment. The ultrasonic power density, PD, was evaluated as: 𝑃𝐷=𝐸 𝑡·𝑉 =𝑃 𝑉 [1] where E, is the energy input, J, t is the ultrasonication time, s, P, the ultrasonic power (J/s =W) and V is the sample volume, mL. TS experiments were also performed by varying the treated sample volume from 60 to 120 mL, in continuous and in pulse (5 s on and 5 s off) modes, at maximum amplitude, during 15 min of TS and keeping constant the temperature of the jacketed water at 40ºC. By varying the sample volume at fixed amplitude, the power density was also varied (see equation 1) and its effect on PPO inactivation was analyzed. Enzyme inactivation kinetics for PME and PPO were determined at maximum amplitude for a treated sample volume of 80 mL at different TS temperatures, by varying the temperature of the jacketed water from 20 to 50ºC. Previous to the TS treatment, cloudy apple juice was preheated (90 s (Sulaiman, Soo, Yoon, Farid, & Silva, 2015)) to the initial temperature of TS (temperature of the jacketed water) in all the experiments, then temperature was continuously recorded during treatment. No enzyme inactivation was observed during the preheating time (90 s) at the initial TS temperatures essayed in this work, from 20 to 50ºC. The effect of the nature of the dissolved gas in the juice was also considered. Previous to the thermosonication experiments, air dissolved in the juice was displaced by bubbling nitrogen or ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT carbon dioxide in the sample at atmospheric pressure until oxygen concentration was below 0.4 mg/L (YSI ProODO optical dissolved oxygen meter). After that, TS experiments were performed as previously described. 2.3 Physical-chemical analyses 2.3.1. Determination of enzyme activity Polyphenol oxidases. PPO activity was determined by adding 100 µL of apple juice into 2.9 mL of substrate solution consisting of a 0.05 M catechol (Sigma Aldrich) solution prepared in a 0.1 M phosphate buffer (pH 6.5). Oxidation of catechol was determined immediately by the absorbance increase at 420 nm by using a Jasco V-750 spectrophotometer equipped with a Peltier thermostated cell holder at 30 ºC. The PPO activity was determined by using the very first linear part of the reaction curve (Illera et al., 2018). Pectinmethylesterase. Samples were analyzed by using an automatic titrator system (Metrohm  Titrando). A pectin solution (1 %) (Alfa Aesar  pectin citrus) was prepared in NaCl (0.3 M). This solution was used as substrate. 50 mL of pectin solution were mixed with 1 mL of cloudy apple juice and pH was adjusted to 7.5 with NaOH 0.005 N. During pectin hydrolysis at 30 ºC, pH was maintained at 7.5 by adding NaOH 0.005 N. The amount of NaOH added for 15 min was recorded. One PME activity unit (UPE) is defined as the micromoles of carboxylic groups produced per minute and mL of juice, at pH 7.5 and 30 ºC (Briongos et al., 2016). Relative residual activities of PPO and PME were evaluated as: Residual activity = Enzyme specific activity after TS treatment Enzyme specific activity before TS treatment ·100% [2] As previously explained, during the preheating time (90s) no inactivation was observed, therefore the enzyme activity right before TS treatment was considered to be the enzyme activity of the untreated juice. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 2.3.2. Determination of hydrogen peroxide, nitrate and nitrite formation One of the mechanisms proposed on enzyme inactivation by US is the formation of free radicals. In the literature, hydrogen peroxide (H2O2) formation has been determined as an estimation of cavitation intensity (Raviyan, Zhang, & Feng, 2005). However, H2O2 generation during an ultrasound treatment in a food system is complicated due to the presence of ions and other colloidal components (Raviyan et al., 2005). Therefore, in this work, H2O2 generation during TS has been determined in a citrate buffer solution at the same pH as the one of the apple juice (pH = 3.9). Formation of nitrite and nitrate in citrate buffer solutions was also determined, since its sonochemical formation in water has been known for a long time (Supeno & Kruus, 2000). Different citrate buffer solutions (pH = 3.9) in the presence of dissolved air, nitrogen or carbon dioxide, prepared as indicated in section 2.2, were thermosonicated (80 mL of solution at 60ºC) for 20 minutes. Samples were collected at different time intervals, to determine the H2O2, nitrate and nitrite formation during the sonication process. H2O2 was determined spectrophotometrically by a colorimetric method based on the one described by Mead et al. (Mead, Sutherland, & Verrall, 1976). 1 mL of solution A [1g NaOH, 33g KI, and 0.1g (NH4)6Mo7O24H2O in 500mL H2O] was mixed with 1 mL of solution B [10g C8H5O4K in 500 mL H2O] and subsequently mixed with 2 mL of the thermosonicated buffer solution taken at different time intervals. Absorbance at 350 nm was registered with time and it was found that 45 min was time enough not to observe any further change in the reaction medium color. Citrate buffer solution, not subjected to thermosonication, was used as a blank. The concentration of H2O2 was calculated based on a standard curve at different H2O2 concentrations. Nitrate and nitrite analysis were performed by using Dr. Lange Küvetten-Test analysis kits and a Hach Lange DR2800 spectrophotometer. LCK 339 kit was used for measuring nitrates. This ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT measurement is based on the reaction of nitrate ion with 2,6-dimethylphenol that causes the formation of 2,6-dimethyl-4-nitrophenol. For measuring nitrites, LCK 341 kit, which is based on the reaction of nitrite with aromatic primary amines, producing the formation of an azo dye, was used. 2.3.3. Particle size distribution Particle size distribution of cloudy apple juice was determined by laser diffraction at 750 nm with a Mastersizer 2000 (Malvern® Inst., MA). Particle size distribution was calculated by the Fraunhofer model. Size distributions (volume fractions against particle size) before and after thermosonication treatment were calculated and the weight-average sizes expressed as:  The equivalent surface area mean diameter: D(3,2)=∑ncdlc 3∑nc ⁄dlc 2 [3]  The equivalent volume mean diameter: D(4,3)=∑ncdlc 4∑nc ⁄dlc 3 [4] where dlc is the diameter of the particle and nc is the percentage of particles. Other useful parameters Dv,0.9, Dv,0.1 and Dv0.5 correspond to the particle size bellow which, 90%, 10% and 50% of the particles lie respectively. 2.3.4. Color, hydroxymethyl furfural content and total polyphenol content The cloudy apple juice color was determined by using a Beckman DU-650 spectrophotometer with diode-array of UV–vis (Beckman Instruments). Following the CIE recommendations, illuminant D65 (daylight source) and a 10º standard observer (perception of a human observer) were used. L*, a* and b* values were obtained representing brightness, red to green color and yellow to blue color. Color difference (E) and chroma (C) were evaluated as: ∆E=√(Lbefore ∗−Lafter ∗)2+(abefore ∗−aafter ∗)2+(bbefore ∗−bafter ∗)2 [5] 𝐶 =√(𝑎∗)2+(𝑏∗)2 [6] ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT -166·PD +278, R2 = 0.987; at 67ºC: RA = -332·PD + 396 R2 = 1 (only two points); for PME at 67ºC RA = -43·PD + 110, R2= 0.920). The values of these slopes indicate a greater effect of the power density in enzyme inactivation when working at higher temperatures. This suggests that the synergic effect between temperature and ultrasound is more effective when working at high temperatures. Figure 4a also shows that when working in a pulse mode, the effect of power density on PPO enzyme inactivation is lower than when working in a continuous mode (RA= -30·AED + 100, R2= 0.867). Although pulse ultrasound can be considered as an energy saving operational mode (Al-Juboori, Yusaf, & Bowtell, 2015), in this study, it has been found to be less effective than continuous mode for PPO inactivation. Other values have been found in the literature for PPO and PME inactivation in apple juice. Values of PPO inactivation in apple juice in Golden delicious by TS treatment, with a 22 mm diameter probe (24 kHz) (200 mL of apple juice in a 250 mL vessel) in a pulse mode (50% and 100% of cycle), were reported by Baslar and Ertugay (Başlar & Ertugay, 2013). These authors observed that the inactivation degree increased sharply with amplitude and temperature, observing an increase in the inactivation degree from 5% to 30% when amplitude increased from 50 to 100 m at 50ºC, for 10 min of TS and pulse 1:1. At 60 ºC, these authors reported much higher inactivation degrees, reaching values of 30% and 80% at 50 and 100 m of amplitude respectively for 10 min of TS and pulse 1:1. Sulaimna et al. (Sulaiman et al., 2015) reported values of PPO residual activity, around 17%, for PPO of Royal Gala apple puree when treated by TS for 15 min at 58ºC and 210 m of amplitude in a continuous mode, with a 3 mm diameter probe, which corresponds to 1.3 W/g (25g in apple puree in a 34.4 mL vessel). These authors found that, under these conditions, nearly complete inactivation of PPO was reached at ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT temperatures of 72-73 ºC. Abid et al. (Abid et al., 2014) studied the TS of Fuji (Malus Domestica) variety apple juice (80 mL of apple juice in a 100 mL vessel), with a 0.5 inch probe at 20 kHz, and a value of the power density of 0.3 W/mL in a pulse mode (5s on and 5 s off). No value of the amplitude was reported. They reported residual activities of 97, 53 and 6 % for PPO at 20, 40 and 60ºC, respectively after 10 minutes of TS. For PME, these authors reported similar values of residual activities 97, 52 and 7 % at 20, 40 and 60ºC, respectively after 10 minutes of TS. In this work, lower inactivation degree, i.e. higher residual activity, was obtained for PME of Golden deliciuous, reaching the lowest residual activity of 50% at 67ºC after 15 min of TS at 79 m of amplitude and 1.44 W/mL. In this regard, it must be highlighted that PME from cloudy Golden deliciuous apple juice was found to be one of the apple cultivars presenting strongest thermostability (30% of residual activity after heating 5 min at 100 ºC) (Teleszko, Nowicka, & Wojdyło, 2016), although mechanism of TS and thermal treatment is probably different. Regarding PPO activity, higher residual activities have been also obtained compared to the results found in the literature. This could be attributed to the lower working amplitude, 79 m (maximum amplitude reported by the supplier for this type of probes). The variety of data on PPO inactivation from apple juice indicates that enzyme inactivation depends on many factors such as source, sub-type, environmental and physicochemical conditions (pH, temperature) (X. F. Cheng, Zhang, & Adhikari, 2013). Islam et al. (Islam, Zhang, & Adhikari, 2014) reviewed the inactivation mechanisms and factors affecting the enzyme inactivation by ultrasound; however, further investigation is needed to understand the specific mechanism by which the different enzymes in apple juice are inactivated. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT TS treatment was compared with thermal treatment by heating the cloudy apple juice during 15 min in the same temperature range as for the TS process, 52  1, 60  1 and 67  1ºC. Heating the juice at the lowest temperatures essayed in TS experiments 52ºC and 60ºC did not cause any loss of enzyme activity but even a significant increase (p  0.05), compared to the untreated juice (RA for PPO: 114 ± 12, 106 ± 16 , RA for PME: 135 ± 10, 157 ± 8 at 52ºC and 60ºC, respectively). At the highest temperature essayed, 67ºC a significant decrease (p  0.05), in enzyme activity by 24  4 % for PPO (RA = 76  4 %) and 9  2 % for PME (RA = 91  2 %) was observed. These results prove that TS significantly improved the inactivation of PPO and PME. 3.4 Inactivation kinetics of PPO at different operating temperatures. The TS inactivation kinetics of PPO was determined in the temperature range from to 52 to 67ºC final TS temperature. PPO of cloudy apple juice was inactivated faster by increasing operating temperature (Figure 5). First order kinetic model has been successfully used in the literature to correlate thermosonication inactivation kinetic data (Terefe, Buckow, & Versteeg, 2014). In this work, kinetic data were fitted to the first order kinetic model and the Weibull model. The corresponding kinetic parameters are listed in Table 2, together with the quality of the fitting. For the first order kinetic model, the inactivation rate constant, k, increased with temperature; therefore, D-values for PPO inactivation decreased with temperature from 130 min at 52ºC to 18 min at 67ºC. Baslar and Ertugay (Başlar & Ertugay, 2013) reported lower D values for PPO of Golden delicious apple juice, 146, 66.4 and 12.9 min at 40, 50 and 60ºC respectively when working at 100 m of amplitude by ultrasound. D values for PPO of Malus domestica cv Roya Gala apple juice ranged from 49 to 4 min at 33 and 72ºC, respectively, at 1.3 W/g, 210 m of amplitude (Sulaiman et al., 2015). ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT The temperature sensitive parameters, zT and Ea, were evaluated through the slopes of plotting equations 8 and 9, respectively, yielding 17.5 ± 0.8 ºC and 123 ± 4 kJ/mol, respectively. Both sensitivity parameters values show that the inactivation of PPO by TS is rather temperature sensitive. Process with high activation energy values are very temperature sensitive (Levenspiel, 1999) as well as processes with low zT values. In the literature, values of the same order were obtained for PPO inactivation of mushroom crude extract in the temperature range from 55 to 75ºC, with zT = 13.8 ºC and Ea=183 ± 32 kJ/mol (X. F. Cheng et al., 2013). Baslar and Ertugay (2013) reported a similar value of zT of 19 ºC for PPO inactivation by thermosonication of Golden delicious apple juice, same variety as in this work. These authors did not report a value for Ea; however, as previously mentioned, these authors reported D values in the temperature range from 40 to 60ºC and the activation energy could be easily evaluated through Arrhenius equation by using the inverse of the D-values (D=1/k), obtaining a value of 105 kJ/mol, similar to this work. On the contrary, other values reported for PPO of Malus domestica cv. Royal Gala apple puree inactivation by TS showed lower temperature sensitivity at 1.3 W/g, with values for activation energy and zT of, 52 kJ/mol and 39ºC respectively, in the temperature range from 33 to 72ºC (Sulaiman et al., 2015). The Weibull parameters are also listed in Table 2. Scale, α, and shape, , parameters were used to calculate the time required to inactivate 1 log (td=1) of PPO. Similar to the D-values for the first order kinetic model, td=1 decreased with temperature. The td=1 calculated by the Weibull model were lower than the corresponding D values obtained by the first order kinetic model. Therefore, in this case, the use of the first order kinetic model could lead to an overprocessing estimation for PPO inactivation. According to van Boekel (2002), the dependence on temperature of both parameters was analyzed. It was found that the scale parameter, α, was statistically significant dependent on temperature, when tested at the 95% significance level for a ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT linear relationship. The shape factor, , was higher than 1 indicating a downward concavity of the enzyme inactivation curve (Van Boekel, 2002), but it was not statistically significant dependent on temperature, when tested at the 95% significance level for a linear relationship. According to van Boekel (2002), the scale parameter, α, could be modelled in a similar way to the classical D value of the first order kinetic model, suggesting a linear dependence of the log α on temperature and considering that  did not depend on temperature: 𝑙𝑜𝑔𝛼=𝑎1−𝑏1𝑇 [13] A zT’ value can be also defined: 𝑧𝑇 ′=1 𝑏1 ⁄ [14] The value of the zT’ is given in Table 2, together with the quality of the fitting. Although the concept of zT and z’T is different since zT is obtained from the linear part, whereas z’T takes also into account the nonlinear part (Van Boekel, 2002), similar values were obtained for both parameters (zT = 17.5 ± 0.8 ºC and z’T = 20 ± 1 ºC). An Arrhenius type equation was also considered to relate the inverse of the scale parameter, 1/α, with temperature. Although 1/α cannot be considered a kinetic constant, the value obtained from the slope was found to be close to the Ea obtained by applying the first order kinetic model, being 107 ± 6 kJ/mol 3.4. Effect of the nature of dissolved gases The effect of the nature of dissolved gases on PPO inactivation kinetics is shown in Figure 6a at three different thermosonication temperatures, 44, 52 and 60 ºC (temperature of the plateau, see Figure 1) when treating 80 mL of apple juice at 100% of amplitude. Regarding the effect of nitrogen, faster inactivation kinetics were obtained compared to the inactivation kinetics in the presence of dissolved air in the apple juice at the same thermosonication temperature. The effect of dissolved carbon dioxide was studied only at 60ºC. In this case, initial inactivation reaction ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT rates (initial slope of the curve) in the presence of CO2 was of the same order as in the presence of air (Figure 5), however, faster reaction rates were obtained at longer TS times and therefore, lower residual activity after 20 min of TS was reached in presence of CO2, 18 ± 2 %, than in the presence of air, 36 ± 5 % at the same operating conditions. Nevertheless, the best results in terms of inactivation rate and final RA were obtained when the air was displaced by nitrogen, reaching a RA of 5 ± 2 % at 60ºC for 20 min of TS. Figure 6b shows the PME inactivation kinetics at 60ºC (temperature of the plateau) when dissolved air was displaced by nitrogen. After 20 min of TS, a minimal RA of 55 % was reached while no inactivation of PME was observed at the same temperature when TS was performed in presence of air (Figure 4b). However, at these conditions, results obtained with CO2 did not bring any improvement in the inactivation process of PME compared with the results obtained in the presence of air and inactivation was neither observed. PPO and PME inactivation kinetic data obtained displacing the air by N2 or CO2 were also fitted to the first order kinetic model and the Weibull model (Table 3). In general, a better fitting was obtained by the Weibull model. For the first order kinetic model, D-values were evaluated. It can be clearly observed that by displacing the dissolved air by previously bubbling nitrogen into the apple juice, Dvalues were lower than those reported in Table 2. At 60 ºC, in the presence of air, the D-value for PPO was 50 min, while it decreased down to 11 min in the presence of nitrogen and to 36 min in the presence of carbon dioxide. The temperature sensitive parameters, zT and Ea, in the presence of nitrogen were evaluated through the slopes of plotting equations 8 and 9, respectively, yielding 20 ± 4 ºC and 106 ± 18 kJ/mol respectively. Regarding the Weibull model, the time required to achieve one decimal reduction, td=1, was calculated by using the shape and scale parameters (Table 3). Displacing the air by nitrogen led ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT to lower values of td=1, compared to the values reported in the presence of air. It was found that the scale parameter, α, was statistically significant dependent on temperature, when tested at the 95% significance level for a linear relationship; but the shape factor,  was not. According to equations 13 and 14, zT’ for the Weibull was evaluated as 18 ± 4 ºC, being similar to zT for the first order kinetic model. An Arrhenius type equation was also considered to relate the inverse of the scale parameter, 1/α, with temperature. From the slope, a value close to the activation energy obtained by applying the first order kinetic model was obtained, 117 ± 24 kJ/mol. These results indicated that the nature of the dissolved gas in the apple juice played an important role in enzyme inactivation. The effect of the nature of dissolved gases on enzyme inactivation has not been frequently studied in the literature. Cheng et al. (L. H. Cheng et al., 2007) studied the sonication process on guava juice quality with previous carbonation of the sample. These authors proposed that dissolved carbon dioxide could reduce surface tension and could create more nuclei cavitation. However, they found that after sonication or a combination of carbonation + sonication, PPO showed higher activity than the untreated sample. Gielen et al. (Gielen et al., 2016) discussed the different effect of the dissolved gases on sonochemistry considering the solubility and thermal properties of the gases as well as the different radical production. In this work, the H2O2 rate production was determined in a citrate buffer solutions at the same pH as apple juice (pH = 3.9) at 60 ºC, 100 % of amplitude, 80 mL of treated volume during 20 min. H2O2 rate formation was determined in three types of citrate solutions with dissolved air and by displacing the air by bubbling nitrogen or carbon dioxide previous to TS experiments. No H2O2 production was observed during the first five minutes of TS in the presence of nitrogen or air, according to our analytical method. After that, it was found that the H2O2 production as a function of time followed a linear relationship, with a H2O2 rate production of (3.6 ± 0.2)·10-8 mol/(Lmin) (R2 = 0.994) in air, and (3.78 ± 0.05)·10-8 mol/(Lmin) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT (R2 = 0.999) for nitrogen. In the presence of CO2, no H2O2 formation was observed after 15 min of TS; subsequently, H2O2 formation rate was similar to that obtained with nitrogen and air (3.81·10-8 mol/(Lmin), from 15 min to 20 min). In any case, these results indicated that CO2 partially inhibited the formation of H2O2 during TS. These results agreed partially with the radical yield formation obtained by Gielen et al.(2016) in the presence of different gases in water at 24  2 ºC. These authors also found, that CO2, although being a highly soluble gas (39.2 mmol/L versus 0.85 and 0.71 mmol/L for air and nitrogen respectively, at 293 K and 1 atm), inhibited radical formation due to low bubble temperature attributed to its lower polytropic index and a higher thermal conductivity compared to air and nitrogen (Gielen et al., 2016). The similar H2O2 rate production for air and nitrogen obtained in this work, could be then attributed to their similar thermal properties. However, Gielen et al. (Gielen et al., 2016) found that nitrogen reduced radical yield by the production of radical scavengers and a lower bubble volume. In any case, no relation can be established between H2O2 production rate in presence of the different dissolved gases and enzyme inactivation, since higher inactivation rates were determined when nitrogen displaced the dissolved air of the apple juice. Nitrate and nitrite rate production was also determined during TS for the three buffer solutions at the same operating conditions. The highest nitrate rate production was obtained when nitrogen was previously dissolved in the buffer, (2.6  0.2·10-7) mol/(Lmin) (R2 = 0.982) and the same value was obtained for CO2 and air, 1.8  0.2 10-7 mol/(Lmin), R2 = 0.910 and 0.960 respectively. Nitrite rate formation was lower than nitrate for all the gases being 3.2  0.2·10-8, 3.3  0.4·10-8 and 1.5  0.2·10-8 mol/(Lmin), for nitrogen, air and CO2 (R2 = 0.993, 0.960, 0.967) respectively. These results could partially explain the higher inactivation rate for PPO and PME ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT found by displacing the air with nitrogen. However, further studies are needed to understand the effect of the nature of the dissolved gas on enzyme inactivation by TS. 3.5. Comparison with HPCD Among the different non-thermal technologies, the use of high pressure carbon dioxide (HPCD) treatment is having also growing attention to inactivate certain microorganisms and enzymes. Typically, operating pressure does not exceed 50 MPa and temperature ranges between 20 and 50 ºC (Briongos et al., 2016), being lower than the temperature employed in this work in TS. Figure 7 compares the PPO inactivation kinetics by TS at 44 ºC, temperature of the plateau, (V=80 mL and 100 % of amplitude) in the presence of air and by displacing the air with nitrogen, with results obtained in a previous study on inactivation kinetic of PPO from apple juice at 45ºC by HPCD at 20 MPa (Illera et al., 2018). At 44 ºC, in the presence of air, short treatment times of TS led to an increase of PPO activity. As indicated in section 3.2, this increase in enzyme activity has been attributed to the release of enzymes bond to cell walls due to the acoustic shock waves (Başlar & Ertugay, 2013) or to the activation of PPO latent forms at low intensities and short treatment times. In any case, at longer treatment times, PPO was inactivated. Figure 7 shows that, at the same operating temperature, HPCD treatment at 20 MPa led to lower PPO residual activity values. This indicates a different enzyme inactivation mechanism for these two non-thermal technologies. 3.6. Effect of thermosonication on different quality parameters of cloudy apple juice Some quality parameters have been determined right after TS treatment at 60 ºC (final TS temperature), 100 % of amplitude, 15 min of treatment time and 80 mL treated volume of apple juice. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Color parameters. Table 4 lists the L*, a*, b* parameters of cloudy apple juice before and after different TS treatments. As general trend, there were significant differences in all color attributes after TS. The lightness (L*) slightly increased, while the red and yellow components decreased compared to the untreated juice. The increase in lightness after TS has been previously reported in the literature and it has been attributed to a homogenization effect of sonication (Tiwari, Muthukumarappan, O’Donnell, & Cullen, 2008). A decrease in a* and b* parameters has been also observed in different juices, such as orange juice (Tiwari et al., 2008) or soursop juice (Dias et al., 2015). Color difference, ΔE, (Equation 4) is also listed in Table 4. In all TS experiments, a color difference was observed compared to the untreated juice. According to Yuk et al. (Yuk, Sampedro, Fan, & Geveke, 2014), the color difference could be estimated to be not noticeable (0–0.5), slightly noticeable (0.5–1.5), noticeable (1.5–3), well visible (3–6) and great (6–12). Based on this classification, at 60ºC, the change in color was noticeable (1.5-3) while at 67ºC was well visible (3-6). According to color parameters, the chroma value presented the lowest value at 67ºC and the lowest treated volume, 60 mL (higher power density). Therefore, although higher TS temperatures, led to higher inactivation rates, change in color was more visible. Color changes during TS have been attributed to Maillard reactions that may occur at long treatment time and high temperature, as well as to cavitation, that involves various physical, chemical and biological reactions (Anaya-Esparza et al., 2017). Total polyphenolic compounds. Total polyphenolic compounds (TPCs) slightly increased after TS compared to the untreated apple juice. Significant differences of the total polyphenolic compounds were found at 60 and 67ºC and in the apple juice treated by previously displacing the air with nitrogen (Table 4). This TPCs increase after TS has been attributed to the enhanced disruption of cell walls that might lead to the release of some chemically bound phenolic ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure captions: Figure 1. Temperature profile of cloudy apple juice during thermosonication of different volumes of cloudy apple juice in a continuous mode ( 60 mL-1.57 W/mL, □ 80 mL1.36 W/mL, ◇ 100 mL-1.21 W/mL, ○ 120 mL-1.09 W/mL) and pulse mode (x 80 mL1.18 W/mL). 100 % of amplitude (76 m), 40ºC of the jacketed water. Figure 2. Effect of the amplitude on PPO residual activity (□), final temperature of the juice () and power density, W/mL (○) for a treated volume of 80 mL and 40ºC of the jacketed water. Lines represent the linear regression. Figure 3. Particle size distribution of fresh cloudy apple juice (a ) and treated by TS at different amplitudes 19 m (25% of maximum amplitude, b − − − −), 38 m (50% of maximum amplitude, c ∙−∙−∙−∙), 57 m (75% of maximum amplitude, d ▬ ▬ ▬ ), 76 m (100% of maximum amplitude, e ∙∙∙∙∙∙∙). Treated volume, 80 mL and 40ºC of the jacketed water. Figure 4. Enzyme residual activity after 15 min of thermosonication at100 % of amplitude (79 m) and different temperatures ( 52ºC, □ 60ºC, Δ 67ºC, T of the plateau) in a continuous mode and in pulse mode at 54ºC (T of the plateau) () as a function of AED, W/mL (a) PPO (b) PME. Lines represent the linear relationship. Figure 5. PPO inactivation kinetics during TS treatment at 100 % of amplitude (79 m) at different operating temperatures ( 52ºC, □ 60ºC, ◇ 64ºC, ○ 67ºC, T of the plateau). Treated volume = 80 mL. Continuous lines represent the Weibull model. Figure 6. PPO (6a) and PME (6b) inactivation kinetics during TS treatment at 100 % of amplitude (79 m) at different operating temperatures (◇ 44ºC;  52ºC, □, ■ 60ºC, T of the plateau) in presence of nitrogen (hollow symbols) and carbon dioxide (full symbols). Treated volume = 80 mL. Lines represent the Weibull model (continuous: N2, dashed line: CO2). ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 7. Comparison of PPO inactivation kinetics at 44 - 45 ºC for different non-thermal treatments: Δ HPCD at 20 MPa, ◇ TS in the presence of N2 (V=80 mL, 100 % of amplitude), □ TS in the presence of air (V=80 mL, 100 % of maximum amplitude). ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 1. Temperature profile of cloudy apple juice during thermosonication of different volumes of cloudy apple juice in a continuous mode ( 60 mL-1.57 W/mL, □ 80 mL1.36 W/mL, ◇ 100 mL-1.21 W/mL, ○ 120 mL-1.09 W/mL) and pulse mode (x 80 mL-1.18 W/mL). 100 % US amplitude (76 m), 40ºC of the jacketed water. 35 40 45 50 55 60 65 0 2 4 6 8 10 12 14 16 Temperature (ºC) Time, min ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 2. Effect of the amplitude on PPO residual activity (□), final temperature of the juice () and power density, W/mL (○), for a treated volume of 80 mL and 40ºC of the jacketed water. Lines represent the linear regression. -0.2 0.2 0.6 1.0 1.4 1.8 0 20 40 60 80 100 120 140 0 20 40 60 80 Power density, W/mL PPO Residual activity (%), T (ºC) Amplitude, m ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 3. Particle size distribution of fresh cloudy apple juice (a ) and treated by TS at different amplitudes 19 m (25% of maximum amplitude, b − − − −), 38 m (50% of maximum amplitude, c ∙−∙−∙−∙), 57 m (75% of maximum amplitude, d ▬ ▬ ▬ ), 76 m (100% of maximum amplitude, e ∙∙∙∙∙∙∙). Treated volume, 80 mL and 40ºC of the jacketed water. 0 1 2 3 4 5 6 7 8 9 10 0.01 0.1 1 10 100 1000 10000 Volume (%) Particle size, m a b c d e ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 4. Enzyme residual activity after 15 min of thermosonication at100 % of amplitude (79 m) and different temperatures ( 52ºC, □ 60ºC, Δ 67ºC, T of the plateau) in a continuous mode and in pulse mode at 54ºC (T of the plateau) () as a function of power density, W/mL (a) PPO (b) PME. Lines represent the linear relationship (continuous mode , pulse mode - - - ). 0 20 40 60 80 100 0.8 1.0 1.2 1.4 1.6 1.8 PPO Residual activity (%) Power density, W/mL 0 20 40 60 80 100 120 0.8 1.0 1.2 1.4 1.6 1.8 PME Residual activity (%) Power density, W/mL (b) (a) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 5. PPO inactivation kinetics during TS treatment at 100 % of amplitude (79 m) at different operating temperatures ( 52ºC, □ 60ºC, ◇ 64ºC, ○ 67ºC, T of the plateau). Treated volume = 80 mL. Continuous lines represent the Weibull model. 0 20 40 60 80 100 0 5 10 15 20 25 PPO Residual activity (%) Time, min ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 6. PPO (6a) and PME (6b) inactivation kinetics during TS treatment at 100 % of amplitude (79 m) at different operating temperatures (◇ 44ºC;  52ºC, □, ■ 60ºC (T of the plateau) in presence of nitrogen (◇, , □) and carbon dioxide (■). Treated volume = 80 mL. Lines represent the Weibull model (continuous: N2, dashed line: CO2). 0 20 40 60 80 100 0 5 10 15 20 25 PPO Residual activity (%) Time, min 0 20 40 60 80 100 0 5 10 15 20 25 PME Residual activity (%) Time, min ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 7. Comparison of PPO inactivation kinetics at 44 - 45 ºC for different non-thermal treatments: Δ HPCD at 20 MPa, ◇ TS in the presence of N2 (V=80 mL, 100 % of amplitude), □ TS in the presence of air (V=80 mL, 100 % of amplitude) (Lines are to guide the eye). 0 20 40 60 80 100 120 010 20 30 40 PPO residual acitivy (%) Time, min ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Table 1. Effect of amplitude in the thermosonication treatment on the particle size distribution (m) of cloudy apple juice Time D[3,2] D[4,3] d(0.1) d(0.5) d(0.9) Cloudy apple juice 2.24 ± 0.04d 70 ± 2d 0.690 ± 0.003d 20.9 ± 0.8c 249 ± 4c 19 m (25 %) 0.82 ± 0.02c 44 ± 2c 0.36 ± 0.01c 0.88 ± 0.01b 158 ± 4b 38 m (50 %) 0.168 ± 0.003b 4.66 ± 0.06b 0.086 ± 0.003b 0.202 ± 0.003a 0.70 ± 0.06a 57 m (75 %) 0.140 ± 0.001a 0.238 ± 0.001a 0.077 ± 0.000a 0.164 ± 0.001a 0.513 ± 0.001a 76 m (100 %) 0.138 ± 0.001a 0.217 ± 0.001a 0.078 ± 0.001a 0.160 ± 0.001a 0.434 ± 0.001a Data: mean ± SD (n=3). Different letters in a column indicate significant differences by the Tukey’s honestly significant difference (HSD) method at p-value ≤ 0.05. ACCEPTED MANUSCRIPT