Effect of high pressure carbon dioxide processing on pectin methylesterase activity and other orange juice properties

2 Effect of high pressure carbon dioxide processing on pectin 1 methylesterase activity and other orange juice properties 2 H. Briongos, A.E. Illera, M. T. Sanz ∗, R. Melgosa, S. Beltrán, A.G. Solaesa 3 Department of Biotechnology and Food Science (Chemical Engineering Section), University of 4 Burgos, 09001 Burgos. Spain 5 Abstract: 6 Inactivation of pectinmethylesterase (PME) and quality parameters of orange juice have been 7 studied after high pressure carbon dioxide (HPCD) treatment. The HPCD treatment conditions 8 covered a wide range of temperature from 2 to 40 ºC, far below normal thermal treatment, while 9 operating pressure was varied from 10 to 30 MPa and exposure time from 3 to 60 min. A 10 decrease in PME activity was found, even at the lowest temperature studied in this work, 2 ºC. 11 Different inactivation kinetic models were used to correlate the PME residual activity: the two-12 fraction model, the fractional-conversion model and the Weibull model. The two-fraction model 13 presents the lowest mean relative deviation. Some quality parameters such as colour, pH, ºBrix, 14 turbidity, ascorbic acid, total acidity and particle size distribution (PSD) were also determined 15 right after HPCD treatment and along storage at 4ºC up to 12 days. PSD shows that HPCD 16 treatment results in a volume increase of small particles and a volume decrease of large particles 17 ∗ Corresponding author. Tel.: +34 947 258810. Fax: ++34947258831. E-mail address [email protected] 3 regarding the non-treated orange juice. Calcium content was also determined before and after 18 HPCD treatment to check for insoluble calcium carbonate formation but not significant changes 19 were observed in calcium content after HPCD treatment. 20 Keywords: Orange juice, HPCD, pectin methylesterase, cloud stability 21 1 Introduction 22 Fruit juice and nectars consumption amounted to 9.7 million litres in 2014 in the EU, of which 23 orange juice is one of the most consumed (European Fuit Juice Association, 2015). Cloud loss is 24 a quality defect in orange juice, since cloud particles are involved in the colour, flavour, texture 25 and aroma of orange juice (Klavons, Bennett, & Vannier, 1991). Additionally, consumers 26 associate the cloud loss with spoilage and quality loss. Citrus cloud is a complex mixture of 27 protein, pectin, lipid, hemicellulose, cellulose and other minor components. Cloud particles of 28 citrus juices range from 0.4 to 5 µm, being particles smaller than 2 µm the most stable clouds 29 (Ellerbee & Wicker, 2011). In the literature, one of the most accepted theories of cloud 30 destabilization is based on pectin demethylation by pectinmethylesterase (PME) (EC 3.1.1.11) in 31 a blockwise fashion. The negative charges generated by PME activity allow subsequent 32 formation of insoluble calcium pectate gels with calcium ions present in the juice. These gels can 33 precipitate pulling the cloud with them causing orange juice clarification due to the loss of 34 turbidity (Ellerbee & Wicker, 2011). Thermal treatment of orange juice at 90 ºC for 1 minute is 35 the method currently used to prevent microbial spoilage as well as the inactivation of the PME 36 (Oulé, Dickman, & Arul, 2013). However thermal treatment causes undesirable changes in 37 several quality parameters such as flavour, colour and texture and can also destroy heat-sensitive 38 nutritional components such as vitamins (Hu, Zhou, Xu, Zhang, & Liao, 2013). Non-thermal 39 4 technologies have gained interest and acceptance as food processing methods due to the 40 consumer increased demand for fresh-like products. Among them, high pressure carbon dioxide 41 (HPCD) has been proposed as an alternative non-thermal pasteurization technique for foods. 42 HPCD can also cause the inactivation of certain enzymes that affect the quality of some foods 43 such as PME in the orange juice under mild operation conditions (Damar & Balaban, 2006). In 44 HPCD treatments, operating temperatures can range between 5 – 60 ºC and pressures usually 45 below 50 MPa. Some other advantages of using HPCD as non-thermal treatment are that carbon 46 dioxide is nontoxic, nonflammable, inexpensive and readily available. It can also be easily 47 removed after treatment by depressurization. 48 Some previous studies dealing with the effect of HPCD treatment on orange juice quality can be 49 found in the literature. The first work was carried out by Balaban, Arreola, Marshall, Peplow, 50 Wei, and Cornell (1991), who found 100 % PME inactivation when using a commercial Milton 51 Roy Supercritical X-10 System while only 86 % PME inactivation was achieved when a custom-52 made supercritical system was used. These authors also found that, when using the custom-made 53 system, cloud significantly increased. Kincal, Hill, Balaban, Portier, Sims, Wei, and Marshall 54 (2006) also reported a high increase in the cloud values (between 446 – 846%) in orange juice, 55 when using a continuous system but a maximum PME inactivation degree of only 46.3%. 56 Recently, Zhou, Bi, Xu, Yang, and Liao (2015) reviewed the effects of HPCD processing on 57 flavour, texture and colour of foods including orange juice. Combined technologies of high 58 power ultrasound assisted SC-CO2 (HPU-SCCO2) have been also reported to inactivate PME of 59 orange juice (Ortuño, Balaban, & Benedito, 2014). These authors found a lowest residual activity 60 of 10.65 %. Therefore, different inactivation degrees have been reported in the literature when 61 treated freshly squeezed orange juice to HPCD. This regard, in the literature it has been reported 62 5 an improvement of inactivation of different enzymes by increasing the CO2 concentration in the 63 enzyme solutions when CO2 was fed through a cylindrical filter nozzle (Ishikawa, Shimoda, 64 Kawano, & Osajime, 1995). Unfortunately, in most of the previous studies, no information about 65 the way CO2 is put in contact with the substrate can be found and comparison is difficult to 66 stablish. Additionally, differences in inactivation levels are related to cultivars, original pH of the 67 juice, isoenzyme forms, total solid content and other processing factors. 68 CO2 was used under supercritical conditions in previous reported HPCD treatments of orange 69 juice. The main objective of this work is to assess the effect of HPCD treatment under 70 supercritical and liquid conditions on PME activity. The effect of HPCD processing on other 71 physical and chemical parameters of orange juice will be also studied. 72 2 Materials and methods 73 2.1 HPCD equipment and processing 74 Valencia oranges were purchased from a local supplier. Oranges were squeezed in an orange 75 squeezer. The experimental apparatus used for the HPCD treatment has been designed in our 76 laboratory with a maximum operating pressure and temperature of 30 MPa and 80 ºC 77 respectively (Melgosa, Sanz, G. Solaesa, Bucio, & Beltrán, 2015). It consists of a CO2 reservoir, 78 a high pressure syringe pump with a pressure controller (ISCO 260 D) and 3 high pressure cells 79 immersed in a thermostatic water bath. In a typical HPCD experiment, orange juice was charged 80 into the high pressure cell, which was then placed in the thermostatic water bath at the preset 81 temperature. Afterwards, the system was pressurized and maintained at constant temperature and 82 pressure for a pre-established treatment time. CO2 was fed to the high pressure cell through a 83 sintered stainless steel micro-filter with a pore size of 10 µm to increase the concentration of 84 6 CO2 dissolved in the sample. The duration of the pressurization and depressurization was less 85 than 2-3 min and it was not included in the treatment holding time. The high pressure cells were 86 magnetically stirred. Experiments were carried out in a temperature (T) range from 2 to 40 ºC, 87 pressure (p) from 10 to 30 MPa and exposure time (t) from 3 to 60 min. Different pressure cells 88 were arranged in series to carry out experiments at different operating times. After HPCD 89 treatment, the high pressure cells were depressurized and the treated orange juice was analysed 90 (see section 2.2). During depressurization, a temperature decrease of the orange juice was 91 observed due to Joule-Thomson cooling effect depending on applied pressures (Zhou, Zhang, 92 Leng, Liao, & Hu, 2010). 93 PME activity, pH and calcium content were determined before and after HPCD treatment at 94 different operating conditions. To evaluate the effect of HPCD treatment on the self-life of 95 orange juice, a sample of orange juice treated at 30 MPa and 40 ºC for 40 min was stored in the 96 refrigerator (4ºC). Aliquots were taken after 5 and 12 days of storage, and different quality 97 parameters of orange juice were determined and compared with original freshly squeezed orange 98 juice. 99 2.2 Physico-chemical analysis 100 2.2.1 Determination of pectin methylesterase activity. PME activity was determined by using 101 an automatic titrator system (Metrohm Titrando). A 1% of pectin solution (Alfa Aesar Pectin 102 Citrus) prepared in NaCl 0.3 M was used as substrate. 50 mL of pectin solution mixed with 5 mL 103 of orange juice were adjusted to pH 7.5 with NaOH 0.02 N. During hydrolysis at room 104 temperature, pH was maintained at 7.5 by adding NaOH 0.02 N. The amount of NaOH added for 105 30 minutes was recorded. One PME activity unit (UPE) is defined as the micromoles of 106 7 carboxylic groups produced per minute and mL of juice at pH 7.5 and room temperature. PME 107 activity was calculated according to the following equation: 108 𝑈𝑈𝑈𝑈𝑈𝑈 𝑚𝑚𝑚𝑚 ⁄=(𝑚𝑚𝑚𝑚 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁)·(𝑁𝑁𝑁𝑁𝑟𝑟𝑚𝑚𝑁𝑁𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑁𝑁𝑜𝑜 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁)·(𝑜𝑜𝑁𝑁𝑓𝑓𝑚𝑚𝑁𝑁𝑟𝑟 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁)·(1000) (𝑚𝑚𝑚𝑚 𝑗𝑗𝑗𝑗𝑚𝑚𝑓𝑓𝑗𝑗)·(𝑚𝑚𝑚𝑚𝑚𝑚𝑗𝑗𝑚𝑚𝑗𝑗𝑚𝑚) [1] Results are presented as residual PME activity, defined as the relationship between PME activity 109 after and before HPCD treatment: 110 𝑅𝑅𝑗𝑗𝑚𝑚𝑚𝑚𝑅𝑅𝑗𝑗𝑁𝑁𝑚𝑚 𝑈𝑈𝑃𝑃𝑈𝑈 𝑁𝑁𝑓𝑓𝑚𝑚𝑚𝑚𝑎𝑎𝑚𝑚𝑚𝑚𝑚𝑚 = 𝑈𝑈𝑃𝑃𝑈𝑈 𝑁𝑁𝑓𝑓𝑚𝑚𝑚𝑚𝑎𝑎𝑚𝑚𝑚𝑚𝑚𝑚 𝑁𝑁𝑜𝑜𝑚𝑚𝑗𝑗𝑟𝑟 𝑁𝑁𝑈𝑈𝐻𝐻𝐻𝐻 𝑈𝑈𝑃𝑃𝑈𝑈 𝑁𝑁𝑓𝑓𝑚𝑚𝑚𝑚𝑎𝑎𝑚𝑚𝑚𝑚𝑚𝑚 𝑏𝑏𝑗𝑗𝑜𝑜𝑁𝑁𝑟𝑟𝑗𝑗 𝑁𝑁𝑈𝑈𝐻𝐻𝐻𝐻=𝐴𝐴 𝐴𝐴𝑜𝑜 [2] 2.2.2 Determination of pH, ºBrix, total acidity, Vitamin C and colour. pH of orange juice 111 was determined with a pH-meter (Crison pH & Ion-Meter GLP 22). ºBrix were measured with 112 a Milton Roy refractometer (Model 334610) at 25ºC. Temperature and acidity corrections were 113 made (Kimball, 1999). 114 Total acidity was determined by using an automatic titrator (Metrohm Titrando). A sample of 115 2 mL of orange juice was mixed with 50 mL of distilled water. The mixture was titrated with 116 0.02 N NaOH. Titrable acidity was expressed as citric acid percentage (g citric acid/100g). 117 Vitamin C was determined with 2,6-dichloroindophenol titrimetric method (Kimball, 1999). 118 Colour was evaluated by a Konica Minolta  CM-2600d colorimeter. The L*, a* and b* values 119 were obtained representing lightness, red to green colour and yellow to blue colour, respectively. 120 Other conditions are illuminant D65 (daylight source) and a 10º standard observer (perception of 121 a human observer) following the CIE recommendations. Changes in colour were expressed as: 122 8 ∆𝑈𝑈=��𝑚𝑚𝑏𝑏𝑏𝑏𝑏𝑏𝑜𝑜𝑏𝑏𝑏𝑏 ∗ −𝑚𝑚𝑎𝑎𝑏𝑏𝑎𝑎𝑏𝑏𝑏𝑏 ∗ �2+�𝑁𝑁𝑏𝑏𝑏𝑏𝑏𝑏𝑜𝑜𝑏𝑏𝑏𝑏 ∗ −𝑁𝑁𝑎𝑎𝑏𝑏𝑎𝑎𝑏𝑏𝑏𝑏 ∗ �2+�𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑜𝑜𝑏𝑏𝑏𝑏 ∗ −𝑏𝑏𝑎𝑎𝑏𝑏𝑎𝑎𝑏𝑏𝑏𝑏 ∗ �2 [3] Differences in perceivable colour can be classified analytically as not noticeable (0-0.5) slightly 123 noticeable (0.5-1.5), noticeable (1.5-3) well visible (3.0-6.0) and great (6.0-12.0) (Yuk, 124 Sampedro, Fan, & Geveke, 2014). 125 Another parameter that can be used to evaluate alterations in colour of a beverage is the chroma, 126 C, which measures colour intensity: 127 𝐻𝐻=�(𝑁𝑁∗)2+(𝑏𝑏∗)2 [4] 2.2.3 Determination of turbidity and particle size distribution. Cloud quality was determined 128 by using a spectrophotometric method. The orange juice sample was centrifuged at 9000 r.p.m. 129 for 30 min. The supernatant was poured into a quartz cuvette and absorbance at 660 nm was 130 measured in a Hitachi® spectrophotomter (Model U-2000). Distillated water was used as a blank. 131 Percent cloud change was calculated as: 132 𝑈𝑈𝑗𝑗𝑟𝑟𝑓𝑓𝑗𝑗𝑚𝑚𝑚𝑚 𝑓𝑓𝑚𝑚𝑁𝑁𝑗𝑗𝑅𝑅 𝑓𝑓ℎ𝑁𝑁𝑚𝑚𝑎𝑎𝑗𝑗=(𝐹𝐹𝑚𝑚𝑚𝑚𝑁𝑁𝑚𝑚 𝑓𝑓𝑚𝑚𝑁𝑁𝑗𝑗𝑅𝑅 𝑎𝑎𝑁𝑁𝑚𝑚𝑗𝑗𝑗𝑗−𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑁𝑁𝑚𝑚 𝑓𝑓𝑚𝑚𝑁𝑁𝑗𝑗𝑅𝑅 𝑎𝑎𝑁𝑁𝑚𝑚𝑗𝑗𝑗𝑗) 𝐼𝐼𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑁𝑁𝑚𝑚 𝑓𝑓𝑚𝑚𝑁𝑁𝑗𝑗𝑅𝑅 𝑎𝑎𝑁𝑁𝑚𝑚𝑗𝑗𝑗𝑗 ·100 [5] Particle size distribution (PSD) was determined by laser diffraction with a Mastersizer 2000 133 (Malvern® Inst., MA). The system uses a laser light at 750 nm wavelength to size particles from 134 0.4 to 2000 µm by light diffraction. Particle size distribution was calculated by the Fraunhofer 135 model. Size distributions (volume fractions against particle size) and the weight average size 136 expressed as the equivalent surface area mean diameter, D(3,2) and the equivalent volume mean 137 diameter, D(4,3) were calculated before and after HPCD treatment and along storage. 138 9 2.2.4 Determination of Calcium content. Calcium in orange juice before and after HPCD 139 treatment was determined by atomic absorption spectrometry (Perkin Elmer 3300). The orange 140 juice was centrifuged (Eppendorf Centrifugue 5804) at 9000 rpm for 30 minutes (Zhou et al., 141 2010). The precipitate was discharged and the calcium content of the supernatant was 142 determined. La2O3 (Merck) was added to samples to a final concentration of 0.5% of lanthanum 143 in the medium. The addition of lanthanum avoids the interference of phosphates in the calcium 144 determination. HCl was also added (5% in the sample) to promote dissolution of both calcium 145 and lanthanum in the medium. Calcium content was obtained by calibration with different 146 standard solutions of calcium (Merck Certipur®, 1 g/L) by following the same method as with 147 the original freshly squeeze orange juice. 148 Some experiments were also performed with a McIlvaine buffer solution containing 0.05 M 149 citric acid and 0.1 M disodium hydrogen phosphate, at pH close to the orange juice (pH ≈ 4), to 150 which calcium was added to achieve a content similar to that in orange juice (around 100 ppm -151 2.5·10-3 MCa2+) using two types of calcium salts (chloride, citrate). A McIlvaine buffer solution 152 was chosen since this solution had a buffer capacity similar to that of orange juice (Yoshimura, 153 Furutera, Shimoda, Ishikawa, Miyake, Matsumoto, Osajima, & Hayakawa, 2002). 154 2.3 Kinetic data analysis 155 Different kinetic models were tested to correlate the inactivation kinetics of PME (Hu et al., 156 2013) . 157 Two-fraction kinetic model. This model takes into account the existence of several isoenzymes 158 of PME in orange juice, grouped into two fractions, a labile and a stable fraction. Both enzymes 159 10 were considered to be inactivated according to first-order kinetics, but independently of each 160 other: 161 𝐴𝐴=𝐴𝐴𝐿𝐿𝑗𝑗𝑒𝑒𝑒𝑒(−𝑘𝑘𝐿𝐿𝑚𝑚)+𝐴𝐴𝑆𝑆𝑗𝑗𝑒𝑒𝑒𝑒(−𝑘𝑘𝑆𝑆𝑚𝑚) [6] where AL and AS (AS = 1 - AL) are the activity of the labile and stable fractions respectively and 162 kL and kS (min-1) the inactivation rate constants of both the labile and stable fractions 163 respectively. 164 Fractional-conversion model. A fraction-conversion model is a special case of a first order 165 kinetic model that takes into account the non-zero residual activity after prolonged heating 166 and/or pressure (A∞) treatment: 167 𝑚𝑚𝑚𝑚(1−𝑜𝑜)=𝑚𝑚𝑚𝑚�(𝐴𝐴−𝐴𝐴∞) (𝐴𝐴𝑜𝑜−𝐴𝐴∞)�=−𝑘𝑘𝑚𝑚 [7] 𝐴𝐴=𝐴𝐴∞+(𝐴𝐴𝑜𝑜−𝐴𝐴∞)𝑗𝑗𝑒𝑒𝑒𝑒(−𝑘𝑘𝑚𝑚) [8] Weibull model. This model can be written in the power-law form as (Ortuño et al., 2014): 168 𝑚𝑚𝑁𝑁𝑎𝑎 10 �𝐴𝐴 𝐴𝐴𝑜𝑜�=−𝑏𝑏𝑚𝑚𝑛𝑛 [9] where b is a non-linear rate parameter and n is the shape factor. 169 2.4 Statistical analysis 170 All analyses were conducted using software Statgraphics X64. The results are presented as a 171 mean ± standard deviation of at least three replicates. The significance of the differences was 172 17 slightly lower values after HPCD treatment than in buffer solutions, no significant differences 306 have been determined among sample means of buffer and orange juices when applying the 307 Tukey’s HSD method. 308 In this regard, the effect of different experimental variables on CaCO3 solubility has been 309 recently reported in the literature (Coto, Martos, Peña, Rodríguez, & Pastor, 2012). CaCO3 310 solubility increased with operating pressure (pressurized CO2) and decreased with temperature 311 and pH of the medium. For instance, at 40ºC CaCO3 solubility in water at 1 bar and 40 bar is 312 about 4.2·10-4 MCa2+ and 2.3·10-2 MCa2+, respectively. Although CaCO3 solubility at atmospheric 313 pressure is very small, depends strongly on pH increasing as pH decreases (for instance, at 25 ºC, 314 at pH = 7 MCa2+=2·10-3 and at pH = 6 MCa2+=9·10-3). Therefore, taking into account that at the 315 low pH of orange juice, only a small amount of dissolved CO2 is converted into bicarbonate 316 dissociating into free hydrogen ions (Zhou et al., 2015) and that calcium carbonate solubility 317 increases by decreasing pH, calcium content in orange juice was in fact not expected to change 318 much after HPCD processing. To our knowledge, the only measurement of calcium content 319 before and after HPCD treatment, was reported by Zhou et al. (2010), who found no significant 320 effects on the calcium content of peach juice, which is also an acidic juice (pH ~ 3.8), after 321 HPCD treatment. 322 To explain the cloud enhancement after HPCD treatment, the possible effect of homogenization 323 induced by gas expansion during the depressurization step has been studied by determining the 324 particle size distribution, PSD. 325 3.2.3 Particle size distribution. PSD of orange juice before and after HPCD treatment has been 326 represented in Figure 5. Two maximums around 0.8 µm and 850 µm can be observed. The size 327 of stable cloud particle has been reported to be in the range of 0.4-5 μm, with the most stable 328 18 cloud having particle sizes of 2 μm and smaller (Ellerbee & Wicker, 2011). The larger particle 329 size in Figure 5 is due to the presence of some settling pulp. In this regard, it must be emphasized 330 that laser diffraction methods generates a volume distribution. The total volume of all particles 331 with diameters less than 5 µm represents 24.5 % of the total volume of particles but represents 332 100 % in number distribution. That is, although the number of bigger particles is very small 333 represents a high volume distribution when compared to cloud particles. Clarification of juice 334 takes place when stable cloud showed aggregation by shifting the PSD distribution to larger 335 diameters (Corredig, Kerr, & Wicker, 2001). However, HPCD treatment results in an increase of 336 the volume peak of the smaller particles and a decrease of large particles (Figure 5). This 337 behaviour of orange juice PDS helps to understand the cloud enhancement after HPCD 338 treatment. This fact has been explained in terms of the effect of the homogenization caused by 339 HPCD treatment due to several reasons, such as high internal stress surpassing the tensile 340 strength of the particles when CO2 is removed from the vessel (Niu et al., 2010). Figure 5 also 341 presents PSD of treated orange juice after 5 and 12 days storage. It can be observed that the 342 volume peak of the small particles increased during the storage, while the volume peak of the 343 large particles decreased. This behaviour could be attributed to the remained active PME that 344 could decompose the high molecular weight compound and then reduce the size of particles. 345 Values of D[3,2] and D[4,3] of freshly squeezed juice, after treatment and after 5 and 12 days 346 storage are presented in Table 4. According to Figure 5, the values after HPCD treatment were 347 lower than those of freshly squeezed orange juice and a continuous decrease was observed with 348 increasing storage time, but this decrease was not significant different along storage (Table 4). In 349 any, case, no shift of PSD to larger diameters can be observed. This tendency can be also 350 19 observed in the values of d(0.1), d(0.5) and d(0.9) which correspond to the size of particle below 351 which 10%, 50% and 90% of the sample lies, respectively. 352 3.2.4 pH, total acidity, ºBrix, ascorbic acid and colour. 353 pH, ºBrix and total acidity did not change significantly in orange juice after HPCD treatment and 354 remained essentially constant during storage (Table 5). The low pH of the original orange juice 355 made difficult the dissociation of the carbonic acid, formed by dissolved CO2 in the juice, into 356 H+ ions. However, in some studies, a decrease in the pH of orange juice has been reported during 357 HPCD treatment (Oulé et al., 2013; Balaban et al., 1991). The pH decrease has been thus 358 associated with the state and density of CO2 in the juice during the treatment (Oulé et al., 2013). 359 Further investigations should be performed to analyse the pH of the sample during treatment, due 360 to different results found in the literature (see section 3.1). 361 The content of ascorbic acid decreased after HPCD treatment (around 14%) and continued 362 decreasing during storage. However the decrease of the ascorbic content after HPCD treatment is 363 lower than after pasteurization treatments. Oule et al. (2013) reported a decrease of the vitamin C 364 content of 13% after HPCD treatment (25 MPa, 40ºC) while after pasteurization vitamin C 365 decreased 43% (90ºC and 60 s). This difference is attributed not only to the lower temperatures 366 employed in HPCD treatments but also to the O2-free environment and the low solubility of 367 vitamin C in SC-CO2. In this work, higher loss of vitamin C has been observed during storage 368 (10% and 24% after 5 and 12 days respectively) than other values found in the literature during 369 storage. For instance, Oulé et al. (2013) found around 5% vitamin C loss after 56 days storage at 370 4 ºC. This high percentage loss could be due to the presence of O2 during storage, but also to the 371 pH of the orange juice used in this work (4.11-4.12), since it is known that more acidic 372 conditions stabilized ascorbic acid (Bull et al., 2004). 373 20 Table 6 lists the L*, a*, b* parameters of freshly squeezed orange juice, after HPCD treatment 374 and during storage. Lightness (L*) and yellowness (b*) significantly decreased indicating the 375 darkening of the orange juice and less yellow and more blue colour after HPCD processing. On 376 the contrary, redness (a*) was not significant different in the untreated and HPCD processed 377 orange juice. In the literature, differences have been reported for the lightness, redness and 378 yellowness in HPCD-treated orange juice (Zhou et al., 2015). During storage, lightness (L*), 379 redness (a*) and yellow (b*) did not change significantly after 5 days storage; however redness 380 and yellowness decreased significantly after 12 days storage. According to Zhou et al. (2015) the 381 colour of foods can be influenced by biochemical or chemical reaction as well as physical effects 382 induced by HPCD. Among other mechanisms, oxidation of ascorbic acid could also lead to the 383 colour change (Zhou et al., 2015). This would agree with the results reported in Table 5 of 384 ascorbic acid content. The change in colour, ∆E (Eq 3) is also presented in Table 6 and visible 385 differences in colour after HPCD treatment have been determined (∆E ≈ 5). Kincal et al. (2006) 386 also reported ∆E values as high as 13.83, at 72 MPa and a ratio of 0.64 CO2/juice (w/w). Chroma 387 values listed in Table 6 show that HPCD treatment results in a significant lower colour intensity 388 juice after processing but no significance differences have been observed along storage. 389 4. Conclusions 390 Freshly squeezed orange juice has been treated by HPCD under different operating conditions. 391 PME in orange juice was effectively inactivated by HPCD showing a fast initial decrease that 392 remained nearly constant after prolonged HPCD treatment. The inactivation degree increased 393 with pressure and temperature. Different inactivation kinetic models were used to correlate the 394 residual PME activity, being the two-fraction model the best with the lowest mean relative 395 deviation. Inactivation of PME in orange juice after HPCD treatment seems to be reversible 396 21 since its activity is slightly recovered along storage at 4 ºC. PSD shows an increase of the 397 volume peak of the smaller particles (0.3-5 μm) and a decrease of large particles after HPCD 398 treatment, supporting the cloud enhancement observed. Calcium content does not change 399 significantly after HPCD treatment, proving that insoluble calcium content was not formed. 400 Further investigation should be done to analyse the effect of cloud enhancement after HPCD 401 treatment. 402 ACKNOWLEDGMENTS 403 To Hyperbaric. To the Spanish Government through MINECO (CTQ2015-64396-R). R. 404 Melgosa acknowledges MINECO for a grant (BES-2013-063937). A. G. Solaesa acknowledges 405 the Burgos University for a pre-doctoral fellowship. 406 407 22 References: Agcam, E., Akyıldız, A., & Evrendilek, G. A. (2014). Effects of PEF and heat pasteurization on PME activity in orange juice with regard to a new inactivation kinetic model. FOOD CHEMISTRY, 165, 70–76. http://doi.org/10.1016/j.foodchem.2014.05.097 Arreola, A. G., BalaBan, M. O., Marshall, M. R., Peplow, A. J., Wei, C., & Cornell, J. A. (1991). Supercritical Carbon Dioxide Effects on Some Quality Attributes of Single Strength Orange Juice. Journal of Food Science, 56(4), 1030–1033. Balaban, M. O., Arreola, A. G., Marshall, M., Peplow, A., Wei, C. I., & J., C. (1991). Inactivation of Pectinesterase in Orange Juice by Supercritical Carbon Dioxide. Journal of Food Science, 56(3), 743–750. Bull, M. K., Zerdin, K., Howe, E., Goicoechea, D., Paramanandhan, P., Stockman, R., … Stewart, C. M. (2004). The effect of high pressure processing on the microbial , physical and chemical properties of Valencia and Navel orange juice. Innovative Food Science and Emerging Technologies, 5, 135–149. http://doi.org/10.1016/j.ifset.2003.11.005 Chen, J. S., Balaban, M. O., Wei, C. I., Marshall, M. R., & Hsu, W. Y. (1992). Inactivation of polyphenol oxidase by high-pressure carbon dioxide. Journal of Agricultural and Food Chemistry, 40, 2345–2349. http://doi.org/10.1021/jf00024a005 Corredig, M., Kerr, W., & Wicker, L. (2001). Particle size distribution of orange juice cloud after addition of sensitized pectin. Journal of Agricultural and Food Chemistry, 49(5), 2523–2526. http://doi.org/10.1021/jf001087a Coto, B., MArtos, C., Peña, J. L., Rodríguez, R., & Pastor, G. (2012). Fluid Phase Equilibria Effects in the solubility of CaCO 3 : Experimental study and model description. Fluid Phase Equilibria, 324, 1–7. http://doi.org/10.1016/j.fluid.2012.03.020 Damar, S., & Balaban, M. O. (2006). Review of Dense Phase CO 2 Technology : Microbial and Enzyme Inactivation , and Effects of Food Quality. Journal of Food Science, 71(1), 1–11. Ellerbee, L., & Wicker, L. (2011). Calcium and pH influence on orange juice cloud stability. Journal of the Science of Food and Agriculture, 91(May 2010), 171–177. http://doi.org/10.1002/jsfa.4169 European Fruit Association, (2015). Liquid Fruit. Market report. Retrieved from http://www.aijn.org/files/default/aijn2015-report.pdf Gui, F., Chen, F., Wu, J., Wang, Z., Liao, X., & Hu, X. (2006). Inactivation and structural change of horseradish peroxidase treated with supercritical carbon dioxide. Food Chemistry, 97, 480–489. http://doi.org/10.1016/j.foodchem.2005.05.028 Hu, W., Zhou, L., Xu, Z., Zhang, Y., & Liao, X. (2013). Enzyme Inactivation in Food Processing using High Pressure Carbon Dioxide Technology. Critical Reviews in Food Science and Nutrition, 53(2), 145–161. http://doi.org/10.1080/10408398.2010.526258 Ishikawa, H., Shimoda, M., Kawano, T., & Osajime, Y. (1995). Inactivation of Enzymes in an Aqueous Solution by Micro-bubbles of Supercritical Carbon Dioxide. Bioscience, Biotechnology, and Biochemistry, 59(4), 628–632. Kimball, D. A. (1999). Citrus Processing. A Complete Guide, 2a ed. New York: Kluwer 23 Academic/Plenum Publishers. http://doi.org/10.1017/CBO9781107415324.004 Kincal, D., Hill, W. S., Balaban, M., Portier, K. M., Sims, C. A., Wei, C. I., & Marshall, M. R. (2006). A continuous high-pressure carbon dioxide system for cloud and quality retention in orange juice. Journal of Food Science, 71(6), 338–344. http://doi.org/10.1111/j.17503841.2006.00087.x Klavons, J. A., Bennett, R. D., & Vannier, S. H. (1991). Nature of the Protein Constituent of Commercial Orange Juice Cloud. J. Agric. Food Chemistry, 39(9), 1545–1548. Melgosa, R., Sanz, M. T., G. Solaesa, Á., Bucio, S. L., & Beltrán, S. (2015). Enzymatic activity and conformational and morphological studies of four commercial lipases treated with supercritical carbon dioxide. The Journal of Supercritical Fluids, 97, 51–62. http://doi.org/10.1016/j.supflu.2014.11.003 Meyssami, B., Balaban, M. O., & Teixeira, A. A. (1992). Prediction of pH in Model Systems Pressurized with Carbon Dioxide7. Biotechnology Progress, 8, 149–154. Niu, L., Hu, X., Wu, J., Liao, X., Chen, F., Zhao, G., & Wang, Z. (2010). Effect of Dense Phase Carbon Dioxide Process on Physicochemical Properties and Flavor Compounds of Orange Juice. Journal of Food Processing and Preservation, 34(2010), 530–548. http://doi.org/10.1111/j.1745-4549.2009.00369.x Ortuño, C., Balaban, M., & Benedito, J. (2014). Modelling of the inactivation kinetics of Escherichia coli , Saccharomyces cerevisiae and pectin methylesterase in orange juice treated with ultrasonic-assisted supercritical carbon dioxide. The Journal of Supercritical Fluids, 90, 18–26. http://doi.org/10.1016/j.supflu.2014.03.004 Oulé, K. M., Dickman, M., & Arul, J. (2013). Properties of Orange Juice with Supercritical Carbon Dioxide Treatment. International Journal of Food Properties, 16(February 2014), 1693–1710. http://doi.org/10.1080/10942912.2011.604893 Yoshimura, T., Furutera, M., Shimoda, M., Ishikawa, H., Miyake, M., Matsumoto, K., … Hayakawa, I. (2002). Inactivation Efficiency of Enzymes in Buffered System by Continuous Method with Microbubbles of Supercritical Carbon Dioxide. Journal of Food Science, 67(9), 3227–3231. Yuk, H. G., Sampedro, F., Fan, X., & Geveke, D. J. (2014). Nonthermal Processing of Orange Juice Using a Pilot-Plant Scale Supercritical Carbon Dioxide System with a Gas-Liquid Metal Contactor. Journal of Food Processing and Preservation, 38, 630–638. http://doi.org/10.1111/jfpp.12013 Zhi, X., Zhang, Y., Hu, X., Wu, J., & Iao, X. (2008). Inactivation of Apple Pectin Methylesterase Induced by Dense Phase Carbon Dioxide. Journal of Agricultural and Food Chemistry, 56, 5394–5400. Zhou, L., Bi, X., Xu, Z., Yang, Y., & Liao, X. (2015). Effects of high-pressure CO2 processing on flavor, texture, and color of foods. Critical Reviews in Food Science and Nutrition, 55(6), 750–68. http://doi.org/10.1080/10408398.2012.677871 Zhou, L., Wu, J., Hu, X., Zhi, X., & LIao, X. (2009). Alterations in the Activity and Structure of Pectin Methylesterase Treated by High Pressure Carbon Dioxide. Journal of Agricultural and Food Chemistry, 57, 1890–1895. Zhou, L., Zhang, Y., Hu, X., Liao, X., & He, J. (2009). Comparison of the inactivation kinetics 24 of pectin methylesterases from carrot and peach by high-pressure carbon dioxide. Food Chemistry, 115, 449–455. http://doi.org/10.1016/j.foodchem.2008.12.028 Zhou, L., Zhang, Y., Leng, X., Liao, X., & Hu, X. (2010). Acceleration of precipitation formation in peach juice induced by high-pressure carbon dioxide. Journal of Agricultural and Food Chemistry, 58, 9605–9610. http://doi.org/10.1021/jf101430j 25 Table 1. PME inactivation values in orange juice after HPCD and other non-thermal treatments System pH p, MPa T, ºC t, min Inactivation Reference Milton Roy System 3.8 29 50 240 100% (Balaban et al., 1991) Custom made system 3.8 29 45 120 84% (Balaban et al., 1991) Static 3.9 40 55 10 36% (Niu et al., 2010) Static 3.9 40 55 60 95% (Niu et al., 2010) Continuous 3.7 72 24 10* 56% (Kincal et al., 2006) Static 4.12 30 40 20-60 90-92 % This work HPU-SCCO 2 3.8 23 41 10 89 % (Ortuño et al., 2014) Pasteurization 3.5 0.1 90 10 - 20 s 93 – 96 % (Agcam et al., 2014) (*) residence time; HPU-SCO2 = high power ultrasound assisted supercritical CO2. 26 Table 2. Estimated kinetic parameters of orange juice PME inactivation at different operating conditions, for different kinetic models Kinetic model T, ºC p, MPa Model parameters D value, min MRD r2 (p<0.05) Two-fraction 40 30 k L = 0.74 ± 0.07 A L = 0.85 ± 0.02 D L = 3.3 ± 0.3 5.8 0.999 k S = 0.010 ± 0.003 A S = 0.15 ± 0.02 D S = 230 ± 69 21 20 k L = 0.24 ± 0.01 A L = 0.652 ± 0.008 D L = 9.6 ± 0.4 1.4 0.999 k S = 0.0048 ± 0.0005 A S = 0.348 ± 0.007 D S = 480 ± 69 21 10 k L = 0.123 ± 0.003 A L = 0.628 ± 0.009 D L = 18.7 ± 0.5 1.2 0.999 k S = 0.0020 ± 0.0005 A S = 0.372 ± 0.009 D S = 1152 ± 288 Fractionalconversion 40 30 k = 0.63 ± 0.06 A ∞ = 0.11 ± 0.01 D = 3.7 ± 0.3 11.4 0.996 21 20 k = 0.18 ± 0.02 A ∞ = 0.28 ± 0.01 D = 13 ± 1 4.8 0.997 21 10 k = 0.112 ± 0.003 A ∞ = 0.336 ± 0.004 D = 20.6 ± 0.6 1.4 0.999 Weibull 40 30 b = 0.58 ± 0.03 n = 0.15 ± 0.02 D = 1.7 ± 0.1 5.8 0.990 21 20 b = 0.27 ± 0.02 n = 0.19 ± 0.02 D = 3.7 ± 0.3 2.8 0.998 21 10 b = 0.14 ± 0.03 n = 0.31 ± 0.06 D = 7 ± 2 5.3 0.985 MRD: mean relative deviation (Eq. 10) 33 Figure 3. Effect of operating time on PME inactivation at different operating conditions (△ 40ºC and 30 MPa,  21ºC, 20 MPa; ○, 21ºC, 10 MPa). Data: mean ± SD (n=3). Continuous lines correspond to the two-fraction model. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 010 20 30 40 50 60 70 Residual activity, A/Ao time, min 34 Figure 4. Evolution of PME residual activity and cloud percentage (Eq. 5) in HPDC treated (30 MPa, 40 ºC and 40 min) orange juice during storage at 4ºC. (◇ PME residual activity,  cloud percentage). Data: mean ± SD (n=3). 0 5 10 15 20 25 30 35 40 0 5 10 Residual activity, %; Cloud change % Storage, days 35 Figure 5. Particle Size Distribution (PSD) of orange juice freshly squeezed ( ─ ), immediately after treatment by HPCD at 30 MPa, 40ºC for 40 min (····); after 5 days storage at 4ºC (-·-·) ; after 12 days storage at 4ºC (− − −). 0 1 2 3 4 5 6 7 0.01 0.1 1 10 100 1000 10000 Volume (%) Particle size, (µm)