EFFECT OF DISSOLVED OXYGEN AND DEOXYGENATION ON THE QUALITY OF ORANGE JUICE

Transcription

1 EFFECT OF DISSOLVED OXYGEN AND DEOXYGENATION ON THE QUALITY OF ORANGE JUICE By ROSALIA GARCIA TORRES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA

2 2010 Roslí Grcí-Torres 2

3 To my prents, Zenón Grcí-Cruz nd Rosli Torres-Pérez 3

4 ACKNOWLEDGMENTS I would like to express my grtitude to my dvisor Dr. Reyes-De-Corcuer for his constnt guidnce, encourgement nd support. He constntly chllenged me to perform my work with excellence nd I lerned from him the enthusism for reserch. Dr. Reyes exmple of integrity nd discipline in reserch hs influenced my ethicl nd scientific criteri. My deepest grtitude lso goes to CONACYT México for providing finncil support during the PhD progrm nd to Dr. Jim Grhm for llowing me to work in his lb. I would like to extend my grtitude to my committee members, Dr. Rouseff, Dr. Goodrich-Schneider nd Dr. Ehsni. To Dr. Rouseff for llow me to work in his lb nd for hving relly enjoyble scientific converstions. To Dr. Goodrich-Schneider for reviewing my mnuscripts nd to Dr. Ehsni for tking the time to reding my disserttion. I m very thnkful to my lb mtes (Jun Mnuel, Mike, Nrsi nd Brittny) for their help nd support when needed, with Shelley for being such good lb mnger nd friend, to Jck for his help during the dys I spent in Dr. Rouseff s lb, to Lis tht ws n importnt contributor to my reserch helping during the experiments nd pplying her sensitive sense of smelling. I thnk to John Henderson nd Rocky for their help with the technicl issues I fced during my reserch. I wnt to thnk my friends for becoming my fmily in the United Sttes especilly to Ftim, Rquel, Sunny, Jun Crlos, Jun Mnuel, Milen nd my group of gtors from Centrl Florid with whom I lerned the mening of friendship nd I do not hve words to express my grtitude to them. Being on the CREC ws gret experience tht I will never forget. 4

5 I m deeply grteful to my fmily for their support nd for believing in me. My prents hve lwys be n exmple to follow nd my strength when I hve been bout to give up. My mother is the best friend God gve me nd my role model s womn. My fther s scrifice nd unconditionl support hve been the motor of my professionl creer. My brother hs lwys been the one tht grounded me when I lose the perspective of life. My sister-in-lw hs enriched my knowledge bout the U.S. To my unts Cuc nd Cristin which time invested in my erly eduction hs been productive. To my grndprents tht in their prticulr wy tught me to live life courgeously nd with hppiness. Finlly, I thnk God for filling my life with thousnds of blessings. 5

6 TABLE OF CONTENTS pge ACKNOWLEDGMENTS... 4 LIST OF TABLES... 9 LIST OF FIGURES LIST OF ABBREVIATIONS ABSTRACT CHAPTER 1 INTRODUCTION AND LITERATURE Oxygen Solubility Interction of Dissolved Oxygen with Food Components Ascorbic Acid Degrdtion Effect of dissolved oxygen concentrtion Color Nonenzymtic browning Enzymtic browning Arom Methods of Mesurement of DO Methods of Removl of DO Vcuum-Deertion Gs Sprging Membrne Deertors Enzyme-Bsed Deertion Oxygen Scvengers Specific Objectives ASSESSMENT OF DISSOLVED OXYGEN EFFECT ON VITAMIN C AND AROMA COMPOUNDS IN PASTEURIZED NOT-FROM CONCENTRATE ORANGE JUICE Introduction Mterils nd Methods Ornge Juice Preprtion nd Chrcteriztion Smple Preprtion nd Storge Study Microbil Anlysis DO Mesurement Vitmin C Mesurement Color Mesurement

7 Arom Anlysis Hedspce smpling Gs Chromtogrphy-FID/ Olfctometry Gs Chromtogrphy - Mss Spectrometry (GC-MS) Rte of Rection Sttisticl Anlysis Results nd Discussion Microbil Anlysis Dissolved Oxygen (DO) Color Arom Active Compounds Voltile Compounds Conclusions IMPACT OF DISSOLVED OXYGEN LEVEL AND TEMPERATURE ON VOLATILE COMPOUNDS, VITAMIN C AND COLOR IN psteurized NOT- FROM-CONCENTRATE ORANGE JUICE Introduction Mterils nd Methods Ornge Juice Preprtion nd Chrcteriztion Smple Preprtion nd Storge Study Microbil Anlysis DO Mesurements Vitmin C Mesurement Color Mesurement Oil in Juice Assy Voltile Compounds Anlysis Hedspce smpling Gs Chromtogrphy - Mss Spectrometry (GC-MS) Rte of Rection Sttisticl Anlysis Results nd Discussion Microbil Anlysis DO Consumption AA Degrdtion Color Voltile Compounds Conclusions COMPREHENSIVE RESULTS AND FUTURE WORK APPENDIX A DISSOLVED OXYGEN CONTENT IN NFC OJ (HAMLIN CULTIVAR)

8 B COLOR CHANGES IN NFC OJ (HAMLIN CULTIVAR) STORED 3 DAYS AT 40 C C GAS SPARGING EXPERIMENTAL SETUP D E NORMALIZATION OF VOLATILE COMPOUNDS IN NFC OJ (HAMLIN CULTIVAR) AND PCA ANALYSIS POINTS TO CONTROL FOR REDUCING VARIABILITY DURING SAMPLE PREPARATION LIST OF REFERENCES BIOGRAPHICAL SKETCH

9 LIST OF TABLES Tble pge 1-1 Ascorbic cid content of different fruit juices (mg 100 g -1 ) Rection rte constnts for AA degrdtion obtined under reducing nd oxidizing conditions t 90 C Summry of selected studies of scorbic cid degrdtion in juices in which the effect of oxygen ws ssessed Comprison of different methods used to mesure DO Selected commercilly vilble O 2 scvengers used in liquids Microbil counts in NFC ornge juice during 60 dys of storge t 5 C Pseudo first nd second order rte of DO consumption in not-from concentrted ornge juice during storge t 5 C Pseudo first rte constnt of AA oxidtion nd DHA formtion in not-from concentrted ornge juice during storge t 5 C Color chnge in NFC ornge juice during storge t 40 C nd expressed s L, nd b with respect to control ornge juice Proportion of rom ctive nd voltile compounds in NFC ornge juice t dy 1 of storge t 5 C Arom ctive compounds identified in commercil NFC ornge juice with different levels of DO fter 1 dy of storge t 5 C Voltile compounds identified by GC-MS in NFC ornge juice with different levels of DO fter 1 dy of storge t 5 C Voltile compounds identified by GC-MS in NFC ornge juice with different levels of DO fter 60 dys of storge t 5 C Dissolved oxygen nd scorbic cid smpling schedule during storge study Pseudo- first order rte constnt (d -1 ) of dissolved oxygen consumption in NFC ornge juice AA loss due to psteuriztion nd storge t selected tempertures Pseudo- first order rte constnt (d -1 ) of AA degrdtion in NFC ornge juice with ir hedspce

10 3-5 Color chnge in NFC ornge juice during storge t 40 C nd expressed s L, nd b with respect to control ornge juice Oil content in NFC ornge juice t dy 0 of storge t 40 C clculted using Scott oil method Normliztion methods pplied to voltile compounds of NFC ornge juice Comprison of rw pek re nd two different normliztions pplied to selected voltile compounds in control not-from- concentrted ornge juice t dy 0 of storge t 40 C Voltile compounds identified in NFC ornge juice t different DO content t dy 0 of storge t 40 C Voltile compounds identified in NFC ornge juice t different DO content t dy 0.5 of storge t 40 C Voltile compounds identified in NFC ornge juice t different DO content t dy 6 of storge t 40 C D-1 Voltile compounds identified in NFC ornge juice t different DO content t dy 0 of storge t 40 C D-2 Voltile compounds identified in NFC ornge juice t different DO content t dy 0.5 of storge t 40 C D-3 Voltile compounds identified in NFC ornge juice t different DO content t dy 6 of storge t 40 C D-4 Voltile compounds identified in NFC ornge juice t different DO content t dy 0 of storge t 40 C D-5 Voltile compounds identified in NFC ornge juice t different DO content t dy 0.5 of storge t 40 C D-6 Voltile compounds identified in NFC ornge juice t different DO content t dy 6 of storge t 40 C

11 LIST OF FIGURES Figure pge 1-1 Ascorbic cid degrdtion mechnism modified from Dvey nd others (2000) Rections ctlyzed by ctechol oxidses Rections ctlyzed by lccses Vcuum deertor scheme Gs sprging method scheme Hollow-fiber membrne seprtor (modified from (Polypore 2007) Dissolved oxygen (DO) concentrtion in NFC ornge juice during 60 dys of storge t 5 C Kinetic models fitted to dissolved oxygen (DO) concentrtion in NFC ornge juice during 60 dys of storge t 5 C Chnges in (A) scorbic cid nd (B) dehydroscorbic cid in NFC ornge juice during 60 dys of storge t 5 C Arom ctive compounds orgnized by rom descriptor in NFC ornge juice t dy 1 of storge t 5 C. (A) deerted, (B) control, (C) oxygen sturted nd (D) ir hedspce ornge juices PCA score plot of voltile compounds in ornge juice t different DO concentrtions t dy 1 nd 60 of storge t 5 C Loding plot of voltile compounds in ornge juice t different DO concentrtions t dy 1 nd 60 of storge t 5 C Flow digrm of smple preprtion of not-from-concentrted ornge juice (Hmlin vr.) Different kinetic models pplied to DO dt of not-from-concentrte ornge juice (cultivr Hmlin) stored t 5 C First-order kinetic dt of not-from-concentrte ornge juice (Hmlin vr.) stored t 5 C Ascorbic cid nd dissolved oxygen content in not-from concentrted ornge juice (Hmlin vr) during storge t 5 C

12 3-5 Ascorbic cid nd dissolved oxygen content in not-from concentrted ornge juice (Hmlin vr) during storge t 40 C Proportion of voltile compounds identified in NFC ornge juice stored t 40 C. (A) 0 dys of storge nd fresh squeezed ornge juice, (B) 6 dys of storge Score plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with totl pek re of voltile compounds Loding plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with totl pek re of voltile compounds D-1 Score plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with internl stndrd D-2 Loding plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with internl stndrd D-3 Score plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with dy D-4 Loding plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with dy

13 LIST OF ABBREVIATIONS AA AcCys APDA CIP Cys DAD DHA DO GC-MS GC-O HMF MF NFC OSA PCA SPME Ascorbic cid N-cetyl-L-cysteine Acidified potto-dextrose gr Clen in plce Cysteine Diode rry detector Dehydroscorbic cid Dissolved oxygen Gs chromtogrphy mss spectrometry Gs chromtogrphy-olfctometry Hydroxymethyl-furfurl Methylfurfurl Not-from concentrted Ornge serum gr Plte count gr Solid phse micro extrction 13

14 Abstrct of Disserttion Presented to the Grdute School of the University of Florid in Prtil Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECT OF DISSOLVED OXYGEN AND DEOXYGENATION IN QUALITY OF ORANGE JUICE Chir: José I. Reyes De Corcuer Mjor: Food Science nd Humn Nutrition By Roslí Grcí Torres August 2010 Ornge juice ws the most consumed fruit juice in the 2007/08 seson, representing 50% (14.4 L, Single Strength Equivlent per cpit) of the totl per cpit consumption of fruit juices. Becuse processing conditions influence the qulity of the end product, it is importnt to understnd the chnges tht ornge juice experiences during processing. Deertion is processing step pplied immeditely before psteuriztion to remove ir, in prticulr oxygen, incorported into ornge juice during extrction or mixing. Vitmin C is believed to be the compound the most ffected (oxidized) by dissolved oxygen (DO) in ornge juice. However, there is lck of informtion bout chnges in color nd rom under different levels of DO. The objective of this study ws to understnd the effect of initil DO, presence of ir in the hedspce nd storge temperture on vitmin C, color nd rom of not from concentrted (NFC) ornge juice in order to preserve fresh-like chrcteristics. The first storge study correlted chnges in vitmin C (scorbic nd dehydroscorbic cid) nd rom compounds of commercil NFC ornge juice (cultivr Vlenci,) with selected initil DO levels during storge for 60 dys t 5 C in mber 14

15 glss continers. The sme study lso evluted the effect of hving ir in the hedspce. The selected initil levels of DO in juice with no hedspce were: deerted ( µm), oxygen sturted ( µm), control ( µm), nd juice with ir in the hedspce ( µm). The juice with ir hedspce hd the highest loss of scorbic cid (AA) of 42% fter 60 dys of storge t 5 C. Pseudo-first order rte constnt of DO consumption ws 2.8x10-1 ±4.1x10-2 d -1. No chnges in color were visully perceived mong tretments or over time. Pseudo-first order rte constnt of AA oxidtion t 5 C ws 1.9x10-2 ± 3.0x10-3. Arom ctive nd voltile compounds were nlyzed in ll juices by GC-O nd GC-MS using solid phse micro extrction (SPME) hedspce smpling. Among rom ctive compounds, methionl ws perceived t higher intensity in ornge juice smples with ir hedspce compred to control only. Nonnl hd the min contribution to rom intensity of control, oxygen sturted nd ir hedspce ornge juices. At dy 1 of storge -elemene, -selinene nd 3-crene hd the highest mounts in deerted ornge juice compred with the other juices. At dy 60 of storge, the content of most of the compounds decresed with respect to dy 0 except for E-2-hexenl nd limonene oxide whose content incresed in ornge juice with ir hedspce. Principl component nlysis (PCA) of voltile compounds llowed to differentiting ornge juices deerted nd with ir hedspce between dys 0 nd 6 of storge. A second storge study with not-from concentrted (NFC) ornge juice (cultivr Hmlin) described the impct of DO nd storge temperture on voltile compounds, AA nd color during storge t 5, 13, 21.5, 30.5 nd 40 C. 15

16 The DO content in the juice smples ws modified by gs sprging, followed by psteuriztion nd storge in mber glss continers t the specified tempertures. Similr to the previous study, the different initil levels of DO in juice with no hedspce were: deerted ( µm), oxygen sturted ( µm), control ( µm), nd juice control with ir hedspce ( µm). The juice with the highest loss of AA ws the one ir hedspce with 100%, 73.5% nd 54.4% of AA loss fter 60 dys of storge t 5 C, 3 dys of storge t 40 C nd 6 dys of storge t 30.5 C respectively. Color chnge ws perceived by direct observtion only in ir hedspce ornge juice stored 3 dys t 40 C, Color chnge ws lso mesured using nd b prmeters thn chnged towrd the red nd blue respectively. No correltion with dissolved oxygen content ws estblished. Voltile compounds were mesured by SPME hedspce nd GC-MS fter gs sprging nd psteuriztion nd identified by mtching with dtbse nd LRI vlues. Ech replicte ws obtined from different btch of ornge juice nd reltive stndrd devitions of more thn 50 were observed in the rw pek res of the replictes of the sme smple. Although it ws initilly thought tht differences in oil content could be the cuse of the vribility, differences between replictes were not detected by Scott oil mesurement. Three different normliztions were pplied to voltiles rw peks: normliztion with 4-heptdecnone dded s internl stndrd to ll the smples, normliztion with totl pek re nd normliztion with respect to dy 0. Normliztion with totl pek re gve reltive stndrd devitions smller thn the other two normliztions nd ws selected for further ANOVA nd PCA nlysis. pinene content ws higher in juice with ir hedspce juice, β-myrcene nd β-elemene 16

17 were lower in control nd oxygen sturted juices nd octnl ws lower in deerted nd oxygen sturted juices. Furfurl, -terpineol, β-terpineol were only detected t dy 6, nd 1-hexnol content incresed over the time. These observtions were significntly different nd confirmed with PCA. The results of this reserch suggest tht since dissolved oxygen does not hve n importnt impct on vitmin C, color nd rom of ornge juice, deertion of juice my not be necessry in citrus processing for short term storge. Rther, removing ir hedspce from storge continers or replcing ir hedspce with nitrogen will reduce the oxidtion of vitmin C during storge. The results of this reserch my possibly be pplied to other fruit juices, nectrs or purees contining oxygen sensitive compounds nd high concentrtion of vitmin C. 17

18 CHAPTER 1 INTRODUCTION AND LITERATURE In 2007, the U.S. totl fruit production ws 26.6 million metric tons. Citrus nd noncitrus production were 9.4 nd 15.3 million metric tons, respectively. The vlue of fruit production ws $14.5 billion, with citrus ccounting for $3.1 billion; this is similr to the $14.4 billion vlue of vegetbles nd nuts (USDA 2007; 2008b). Those most commonly processed into juice were ornge, grpe, pple, grpefruit, nd pech. In 2007, 7% of the fruit production ws consumed s juice in the United Sttes. The U.S. per cpit consumption of fruit juices for the sme yer ws 28.9 L of single-strength equivlent (SSE). SSE is the volume of juice t the soluble solid content (SSC) t which it is consumed (for reconstituted ornge juice, 11 Brix). In fruit juices, qulity is determined in terms of SSC, cidity, nutritionl content, color, flvor, microbil sfety, cloud stbility, fiber content, sugr content, rheologicl properties, defects, nd dultertion. Air is nturlly present in the intercellulr spces of fruits. During fruit mcertion, homogeniztion, nd juice extrction cells re crushed, the cell wll is disrupted nd the ir is mixed into the juice. Air cn be present s dissolved gs in solution or ssocited with the pulp prticles, for exmple, in ornge juice (Joslyn 1961). During cell disruption, metbolites nd enzymes tht re normlly comprtmentlized re mixed, producing chemicl nd biochemicl rections. Oxygen in ir, present in the spces between the juice vesicles nd from the surroundings, sturtes the juice producing oxidtion rections tht often result in browning, chnges in rom, nd loss of nutritionl vlue. These rections re ccelerted by the increse in temperture during psteuriztion nd reduce the overll qulity of the product during storge. Interctions of oxygen with the different food components result in severl sequentil nd prllel rections. 18

19 Product interctions mke it difficult to identify nd to differentite the compounds produced by direct rection with oxygen from the ones produced indirectly in subsequent rections. The common pproch hs been to follow the chnges of 1 or 2 compounds t time in the juices or in model solutions. Oxygen Solubility For the purpose of this review, DO is considered to be the moleculr or dirdicl triplet oxygen, which is the most bundnt nd stble form of oxygen. Singlet oxygen my lso be present in food nd requires less energy ctivtion to produce oxidtion thn triplet oxygen nd hs greter rte of oxidtion of food substrtes. Min nd Boff ( 2002) reviewed the chemistry nd mechnisms of formtion; nlyticl techniques for detection nd quntifiction; rection mechnisms with food components such s lipids, proteins, nd vitmins; quenching mechnisms to reduce oxidtion; nd kinetics of singlet oxygen. The min differences between triplet nd singlet oxygen re in the electron rrngement nd in the energy level. Electrons locted in the externl orbitl re pired in ntiprllel spins in singlet oxygen nd in prllel spins in triplet oxygen. Energy level for singlet oxygen is 22.5 kcl mol 1 reltive higher to triplet oxygen. Moreover, the detection of singlet oxygen is difficult becuse of its short life (50 to 700 μs). Detection requires the use of electron spin resonnce spectroscopy (ESR), lser deflection clorimetry, or time-resolved singlet oxygen detection (Min nd Boff 2002). The concentrtion of oxygen in liquids cn be described by Henry s lw tht estblishes tht gs dissolved in liquid is in equilibrium with the gs bove the liquid surfce. The concentrtion of the gs dissolved in the liquid is proportionl to the prtil pressure of the sme gs in the vpor phse (Eq. 1-1): 19

20 X A PA H (T) (1-1) where XA, PA, nd H(T) re concentrtion of gs A in the liquid phse, prtil pressure of gs A in the vpor phse, nd Henry s constnt, respectively (Gill nd Menneer 1997; Ringblom 2004). Oxygen solubility increses with incresing tmospheric pressure nd decreses with incresing temperture nd slinity (Lewis 2006b). There re severl methods used to quntify DO, but most hve been developed for wter nlysis. A thorough compendium of these methods is the U.S. Geologicl Survey (USGS) mnul tht includes tbles of oxygen solubility in wter t tempertures from 0 to 40 C, pressures from 600 to 795 mmhg, correction fctors for slinity t tempertures of 0 to 35 C, nd conductivity of 0 to μs cm 1 (t 25 C). Oxygen solubility cn be clculted using Eq. 1-2 (Weiss 1970; Lewis 2006b) ln C* / T ln( T /100) ( T 2 S ( T /100) ( T /100) /100) (1-2) where C, T, nd S re the oxygen solubility in milliliter per liter or milliliter per kilogrm from wter-sturted ir t one tmosphere, the bsolute temperture in K, nd the slinity in grm per kilogrm, respectively. The estimted precision from this eqution is ml L 1 of oxygen. The stndrds used for clibrtion re ir-sturted wter nd sodium sorbte-sturted solutions s 100% nd 0% of DO concentrtion, respectively (Singh nd others 1976; Kennedy nd others 1992) nd the oxygen solubility tbles for wter (Lewis 2006b) re used s references for clculting concentrtions in milliliter per liter t specified temperture nd pressure conditions. Sdler nd others (1988) found tht the presence of solutes such s sugrs, NCl, nd citric nd scorbic cids t concentrtions similr to fruit juices reduce the oxygen solubility by less thn 10% t 20

21 concentrtions similr to tht of fruit juices. They developed method to predict oxygen solubility in juices s function of sugr concentrtion t tempertures between 4 nd 40 C t tmospheric pressure. The method uses multiple regression eqution (Eq. 1-3) ble to predict DO concentrtion within 5% of literture vlues nd the coefficient of determintion of R 2 = The limittion of this method is tht it only pplies to ir-sturted sugr/juice solutions nd does not incorporte pressure effects. O B T ln 2 (1-3) where [O2] is the oxygen concentrtion in milliliter per kilogrm, B is soluble solids concentrtion in Brix, nd T is temperture in C. Dissolved oxygen concentrtion vlues determine the degrdtion rte of severl compounds of interest in fruit juices. DO mesurement is chllenging s oxygen is present in the ir nd cn esily incorporte into the liquid food during smple preprtion nd mesurement. This cn ffect DO quntifiction nd mke it necessry to mnge the smples under n oxygen-free tmosphere. Interction of Dissolved Oxygen with Food Components Ascorbic Acid One of the most importnt food components in juice tht is ffected by DO is L scorbic cid (L-AA). L-AA is nturl ntioxidnt present in most fruits. Tble 1-1 summrizes the scorbic cid content in some fresh fruits, juices, nd purees. Ornge juice, guv pulp, nd grpefruit juice hve the highest scorbic cid concentrtions in fruit juices. 21

22 Degrdtion Ascorbic cid degrdtion in ornge juice is result of erobic nd nerobic mechnisms (Hughes 1985; Johnson nd others 1995) tht occur simultneously (Kennedy nd others 1992; Bino nd others 2004). Figure 1-1 presents the oxidtion nd hydrolysis rections involved in the scorbic cid degrdtion. Ascorbic nd dehydroscorbic cid re the ctive forms of vitmin C. Ascorbic cid in queous solution is oxidized to dehydroscorbic cid (DHA) in 2-step process tht produces monodehydroscorbte (MDHA), lso clled scorbte-free rdicl. As intermedite MDHA cn be reduced bck to scorbic cid or form DHA, which hs pro-oxidnt effect. The oxidtion of AA to DHA is reversible process; the direction is ffected by the presence of oxidizing or reducing gents (Serpen nd Gökmen 2007). Dehydroscorbic cid lso undergoes irreversible hydrolytic ring clevge to produce 2, 3-diketogulonic cid (2,3-DKG) (Figure 1-1). Equtions 1-4 nd 1-5 show the simplified mechnism dopted to describe the kinetic rte of the L-AA degrdtion under erobic conditions (Mck nd others 1976; Singh nd others 1976; Lin nd Aglloco 1979; Trmmell nd others 1986; Ptki nd others 2002; Serpen nd Gökmen 2007). 2 k1 L AA O2 2 DHA 2 k 2 H O 2 (1-4) 2 3 k DHA 2 H 2O 2 DKG (1-5) where k 1, k 2, nd k 3 re the rection rtes of the scorbic cid oxidtion, dehydroscorbic cid reduction, nd dehydroscorbic cid hydrolysis, respectively. In scorbic cid queous solutions nd in the presence of ir, the reconversion of DHA to L-AA (k 1 ) is very slow compred to the L-AA oxidtion to DHA (k 2 ) nd the subsequent 22

23 hydrolysis of DHA (k 3 ). This is similr to rections occurring in the presence of oxygen nd Fe 3+ s oxidizing gents. But in the presence of reducing gent, the conversion of DHA to L-AA is fster thn the other 2 rections, s Serpen nd Gökmen (2007) concluded by pplying Lplce trnsformtion to clculte the kinetic rtes of the rections tht pper in Tble 1-2. Ptki nd others (2002) suggested tht scorbic cid degrdtion to 2,3-diketogulonic cid under erobic conditions t ph lower thn 5 is 1-step rection, but t ph 8, oxidtion proceeds s 2-step rection (Eq. 1-4 nd 1-5). Perhps t low ph, the scorbic cid degrdtion process lso consists of 2 rections, but the dehydroscorbic cid degrdtion (k3) is fster thn the DHA formtion (k1). L-AA erobic degrdtion during storge is promoted by temperture increse, presence of copper, iron, nd lkli (Lin nd Aglloco 1979; Dvey nd others 2000), incresed light exposure (Mck nd others 1976; Singh nd others 1976; Lin nd Aglloco 1979; Lee nd Kder 2000), differences in ph (Lin nd Aglloco 1979; Ptki nd others 2002) nd increse of DO content (Mck nd others 1976; Singh nd others 1976; Lin nd Aglloco 1979; Trmmell nd others 1986; Kennedy nd others 1992; Lee nd Kder 2000). The combined effect of ph nd temperture on L-AA degrdtion under erobic nd nerobic conditions for L-AA queous model solutions ws described with Eq. 1-6 nd 1-7, respectively. Although is not cler how AA concentrtion is considered in these equtions. Eqution 1-6 pplies for ph 3 to 7 nd Eq. 1-7 for ph >3 (Lin nd Aglloco 1979). E ( ph) (1-6) vlid for erobic conditions nd ph 3 to 7 23

24 E ( ph) (1-7) vlid for nerobic conditions nd ph > 3 Under nerobic conditions, L-AA is degrded to furfurl t slower rte thn erobic degrdtion nd 1st-order kinetics hs been suggested (Ptki nd others 2002). However, the difference in rtes under erobic nd nerobic conditions hs not been quntified (Bino nd others 2004). Effect of dissolved oxygen concentrtion Although the scope of this disserttion is the degrdtion of scorbic cid under erobic conditions, it is importnt to mention tht nerobic degrdtion of scorbic cid lso ffects the qulity of fruit juices. For exmple, t ph below 4 nd in the bsence of oxygen, the multistep decomposition of L-AA to furfurl hs been ssocited with browning in stored citrus products nd model systems (Rouseff nd others 1992). Kinetic models hve been pplied to describe the rte of scorbic cid degrdtion (Eq. 1-4 nd 1-5) but they re not conclusive bout the kinetic order. One reson my be becuse they re bsed only on L-AA concentrtion, while there re other substrtes tht my be limiting the rection becuse they re present in smller mounts. For exmple, DO concentrtion in ornge juice is 10 times smller thn L-AA concentrtion. The following section discusses the kinetic models tht hve been pplied to scorbic cid oxidtion. The reltionship between the rte of L-AA degrdtion under erobic conditions or oxidtion nd the concentrtion of DO in liquid foods hs been studied since the 1930s. Eddy (1936) studied the role of oxygen in the destruction of reducing fctor nmed scorbic cid in ornge juice. Further studies showed tht when light effect is bsent, 24

25 the removl of DO is more effective in protecting L-AA oxidtion during processing thn during storge. Kefford nd others (1959) followed the L-AA degrdtion of cnned ornge juice during psteuriztion, erly storge (0 to 100 h), nd lter storge (200 to 9000 h) t 30 C on oxygen sturted, ir-sturted, nd oxygen-removed conditions. During psteuriztion, loss of AA of 46%, 7%, nd not significnt loss of L-AA degrdtion in oxygen-sturted, ir-sturted, nd oxygen-removed, respectively, ws observed. The storge study ws divided into erly storge (when DO is present) nd lte storge (DO is bsent). During the erly storge, the L-AA degrdtion followed liner trend nd rtes were higher thn in lter storge tht ws described with polynomil of 2nd order. The conclusion of the study ws tht in the cnned juice system there is competition for oxygen between corrosion rections nd scorbic cid nd other compound oxidtions tht ffect color nd flvor. This ide pplies to different juices nd pckging mterils on which creful nlysis of ll the oxygen-relted rections will give comprehensive chrcteriztion of the system. Over time, L-AA degrdtion due to DO occurs t the beginning of the storge period rnging from hours to weeks depending on storge conditions such s temperture, light intensity, nd permebility of pckging. For exmple, in mngo juice (cultivr Ogbomoso) stored t 6 C for 8 wk, glss protected L-AA better thn polyethylene film nd PET hving loss of scorbic cid of 11%, 68%, nd 50%, respectively (Alk nd others 2003). Kennedy nd others (1992)found tht fter 64 d of storge of single-strength ornge juice (SSOJ) in Tetr Brik crtons t 4, 20, 37, 76, nd 105 C, DO concentrtions dropped to 2, 1.5, 0.5, nd less thn 0.5 mg L 1, respectively. Increse of initil DO concentrtion increses the rte of L-AA 25

26 degrdtion. In SSOJ t different DO levels, relted to processing conditions, μm (10.1 mg L 1 ; oxygen sturted juice), μm (6.5 mg L 1 ; norml plnt opertion conditions), 56.2 μm (1.8 mg L 1 ; commercilly deerted ornge juice), nd μm (0.6 mg L 1 ; lowest level reched using gs equilibrtion technique), nd stored for 22 wk t 22 C in n inert tmosphere (nitrogen flushed), the percent of L-AA lost ws linerly relted to the initil DO concentrtion over the time, suggesting 1st-order rection with respect to DO. The mximum loss ws 32% t μm (10 mg L 1 ) of initil DO fter 22 wk (Trmmell nd others 1986). The effect of temperture nd DO in L-AA degrdtion under regulr storge conditions ws evluted in SSOJ pcked in Tetr Brik crtons t 4.5 mg L 1 initil DO concentrtion. L-AA loss fter 64 d ws 39.6%, 51.4%, nd 88.1% t 4, 20, nd 37 C, respectively. In contrst, t 76 nd 105 C, L-AA degrdtion ws 98% fter 6 d nd 96.4% fter 3 d, respectively (Kennedy nd others 1992). Ascorbic cid concentrtion in ornge juice decresed by 86% when stored exposed to fluorescent light for 84 d t 4 C in n oxygen impermeble mteril, compred to 50% loss when protected from light (Nelson 2005). Tble 3-3 summrizes different studies bout scorbic cid degrdtion in model solutions nd citrus juices in which oxygen ws removed or concentrtion ws vried t different levels. A totl of 8 of the 17 studies shown in Tble 1-3 gree in describing L- AA oxidtion s 1st-order rection under different DO conditions rnging from nerobic conditions (Johnson nd others 1995; Rojs nd Gerschenson 2001) to grdul decrese of oxygen concentrtions (Lewis nd McKenzie 1947; Khn nd Mrtell 1967; Robertson nd Smniego 1986; Bino nd others 2004; Dhuique- 26

27 Myer nd others 2007). However, in some cses zero-order (Kennedy nd others 1992) or 2nd-order (Singh nd others 1976; Hsieh nd Hrris 1993) seemed to describe the kinetics s well. These pprent contrdictions re due to the fct tht when DO is present in limited mounts, prticulrly t concentrtions below sturtion, it becomes limiting rectnt of the scorbic cid oxidtion (Kennedy nd others 1992). Under these conditions, 1st-order models do not fit properly becuse they only consider one rectnt concentrtion, either scorbic cid or DO. Under limited oxygen concentrtions, 2ndorder kinetic models my better describe the scorbic cid oxidtion thn 1 st order models (Mck nd others 1976; Singh nd others 1976). A model tht ccounts for DO nd tht cn potentilly be pplied to fruit juices ws reported in study on n infnt formul contining 55 mg L 1 of L-AA. A 2nd-order kinetic model (Eq. 1-8) described the decrese in both the scorbic cid nd DO concentrtion over time (Singh nd others 1976): AA ( ) [ ] ln o X AA ln DO 0 ( X ) [ DO] AA DO 0 0 AA 0 kt ln DO 0 (1-8) where [AA]0 nd [DO]0 re initil concentrtions of L-AA nd DO, respectively; [AA] nd [DO] re scorbic cid nd DO concentrtions t time t, respectively; nd X is the mount of L-AA or DO tht hs rected t time t. This eqution ws solved using the computer progrm KINFIT. The 2nd-order k vlues t different light intensities nd initil DO combintions hd vlues between nd L mg 1 h 1, nd ctivtion energies were not clculted becuse experiments were performed t 7.2 C only. 27

28 Although 2nd-order kinetics described correctly the scorbic cid oxidtion, reversible 2nd kinetic order would be even better since it is lso considering the reversible conversion of DHA to L-AA. The previously mentioned kinetic studies did not consider the reversible nture of dehydroscorbic cid nd ssumed only 1 rection direction. Kinetic studies tht consider the scorbic cid oxidtion s reversible rection re scrce. Serpen nd Gökmen (2007) considered the oxidtion of AA to DHA s reversible rection nd clculted the equilibrium between the rtes of the reversible oxidtion rection nd the subsequent DHA hydrolysis in L-AA solutions (Tble 1-2). Another pproch to clculte the rte of scorbic cid oxidtion is to nlyze the rection using different kinetic orders over time. For exmple, Ski nd others (1987) pplied kinetic model tht consists of 2 consecutive rections (Eq. 1-9) of different kinetic order tht pplies only under constnt DO concentrtion. k A A kb B C (1-9) Prmeters A, B, C, ka, nd kb re scorbic cid, intermedite product, oxidized product, zero-order rte, nd 1st-order rte constnts, respectively, B nd C products re not defined. But bsed on the chemistry (Figure 1-1), we cn ssume tht B is the monodehydroscorbte (MDHA) nd C is the dehydroscorbic cid. Under those definitions, the 1st rection is zero-order with respect to scorbic cid concentrtion nd the 2nd rection is 1st-order with respect to MDHA. This model clcultes the concentrtion of scorbic cid (A) over time bsed on the mesurement of the initil concentrtion of scorbic cid nd the rtio of the rte constnts s shown in Eq nd

29 When A 0 t k A k A k B A A B A k t 1 e t ' 0 A A k 0 When t nd A hs disspered k A B (1-10) A' B k k A B 1 e kb k A e t [ A 0 ] k 0 B A k A (1-11) [A]0 is the initil concentrtion of scorbic cid nd ka nd kb were defined in Eq This method ws vlidted by compring the scorbic cid concentrtion vlues obtined using Eq nd 1-11 with those obtined from queous scorbic cid solutions t 30 C. To describe scorbic cid oxidtion in ornge juice under ir sturtion Mnso nd others (2001) proposed kinetic model tht considers the degrdtion of AA to DHA s 1st-order rection, the reconversion of DHA to AA s 2nd-order rection, nd DHA degrdtion to 2, 3-DKG s 1st-order rection. However, they pointed out tht there is no evidence tht DHA degrdtion follows 2nd-order kinetics. Mnso nd others (2001) lso proposed Weibull model (Eq. 1-12) tht fitted the experimentl dt well (R 2 dj > 0.995). i C C * e t (1-12) where C,Ci, α, nd β re the L-AA concentrtion t time t, the initil L-AA concentrtion, scle constnt (which inversely corresponds to the rection time constnt), nd the shpe constnt or behvior index, respectively. A prticulr cse of the Weibull function when β = 1, is equivlent to the 1st-order model. However, when 29

30 one or more sequentil rections tke plce, this model clusters them into the shpe constnt without providing ny mechnistic informtion. The temperture effect on ny chemicl rection cn be quntified with the ctivtion energy (E ) clculted using the Arrhenius eqution. From the kinetic studies in L-AA degrdtion presented in Tble 1-3, only 6 studies reported E. The temperture rnges between 20 nd 124 C, nd t lest 3 different tempertures were used in the clcultions. E rnged from 35.9 to kj mol 1 but similr E vlues (38.6 ginst 35.9 kj mol 1 ) were reported for citrus juices (SSOJ ginst ornge tngerine juice) even when different models (Weibull ginst 1st kinetic order) nd different temperture rnges were used (below 45 C ginst higher thn 50 C) (Mnso nd others 2001; Dhuique-Myer nd others 2007). Brddock nd Sdler (1989) used the Arrhenius pproch (Eq. 1-13) to estimte L-AA loss t ll stges, t different tempertures during ornge juice concentrtion in 7-stge thermlly ccelerted short-time evportor: 1 ln k * T (1-13) where k is the rection constnt in percent per second nd T is the bsolute temperture in K. The experimentlly determined DO level ws below 0.1 mg L 1 fter stge 3; estimted L-AA loss fter the 7 stges ws 3.7 mg L 1 ; rection constnt vlues rnged between to s 1 for 76 to 86 C, respectively. In ddition to DO concentrtion, intrinsic vribility in the fruit such s culitvr, mturity, ntive soil constituents, nd processing conditions such s equipment nd continer mteril ffect fruit juice composition including ph nd trce metls. Differences in ph ffect the number of scorbic cid oxidtion steps (1 or 2 steps) s 30

31 mentioned in the section Degrdtion (Ptki nd others 2002). Fe 3+ nd Cu 2+ ctlyze scorbic cid oxidtion (Khn nd Mrtell 1967). For exmple, 2 mg L 1 Fe 3+ in n scorbic cid solution incresed L-AA loss by 12.5% compred to the control fter incubting for 6 h t 90 C (Serpen nd Gökmen 2007). Furthermore, ttempts to correlte the rtes of DO consumption nd L-AA degrdtion in pure solutions with those in fruit juices re most likely erroneous becuse severl other rections occur simultneously in fruit juices. In view of the different experimentl conditions reported in the literture nd the order-of-mgnitude differences in reported pprent kinetic prmeters nd ctivtion energy (see Tble 1-3), it ppers difficult, if not impossible to extrpolte reported dt to industril processing. However, vluble informtion cn be drwn from kinetic studies tht ccount for the effect of 1 single vrible on the system such s DO (Robertson nd Smniego 1986), light (Singh nd others 1976), sucrose, nd ph (Hsieh nd Hrris 1993). Juice degrdtion in the presence of oxygen cn be described s severl rections competing for the DO. The following sections describe the chnges in color nd rom due to degrdtion of L-AA nd other compounds in the presence of oxygen in the juices. Color Nonenzymtic browning Commercilly produced fruit juices receive therml tretment during processing to hve long shelf life. The heting step reduces the microbil lod (5 log reduction or more) nd inctivtes the enzymes present in the rw fruit. However, heting promotes chnges in color nd flvor. Nonenzymtic browning, including Millrd type rections (Cámr nd others 2003; Reineccius 2006), scorbic cid (Johnson nd others 1995), 31

32 nd sugr degrdtions (Rouseff nd others 1992) re importnt rections in psteurized fruit juices. Oxygen cn promote browning, but is not directly responsible for it (Shw nd others 1993). Oxygen s indirect effect on browning is by wy of scorbic cid oxidtion to dehydroscorbic cid with further formtion of decomposition products such s furfurl tht hve been ssocited to browning. For exmple, in study on SSOJ t initil DO levels of 0.6 to 10.1 mg L 1, browning nd scorbic cid loss were linerly correlted with the initil DO concentrtion during 22 d of storge t 22 C (Trmmell nd others 1986). In not from concentrte (NFC) ornge juice pckged in pouches of oxygen scvenger (EVOH/oxygen scvenger/cst polypropylene) nd oxygen brrier (EVOH/cst polypropylene) film stored t 25 C, the browning index (bsorbnce t 420 nm) nd the percentge of loss of scorbic cid were lower in pckges with oxygen scvenger (0.35% nd 24%, respectively) thn in pckges with only oxygen brrier (0.44% nd 24%, respectively) fter 365 d of storge. But t 4 C, the differences due to pckging mteril were less evident (Zerdin nd others 2003). Adms nd Brown (2007) reviewed different discolortion processes in rw nd processed fruits nd vegetbles, including enzymtic nd nonenzymtic browning, nd they lso concluded tht in fruit juices such s lemon, ornge, nd pinepple, nonenzymtic browning is relted to scorbic cid nd sugr degrdtions. Sulfur-contining mino cids hve ntioxidtive nd ntitoxic effects due to their bility to ct s reducing gents, oxygen scvengers, nd good nucleophiles (Friedmn nd Molnr-Perl 1990). These mino cids cn be used to inhibit enzymtic nd nonenzymtic browning (Friedmn nd Molnr-Perl 1990; Molnr-Perl nd Friedmn 32

33 1990; Hlev-Toledo nd others 1999). Formtion of (hydroxymethyl)-furfurl (HMF) nd methylfurfurl (MF), both relted to browning in stored citrus juices, ws studied in 1 to 10 mm L-cysteine (Cys) nd N-cetyl-L-cysteine (AcCys) buffered solutions (ph 3, 5, or 7) nd incubted t 70 C for 48 h. HMF nd MF formtion ws inhibited significntly (P < 0.01) t ph 3 nd 5 in the presence of Cys nd AcCys; but t the sme ph, browning (mesured s bsorbnce t 420 nm) ws promoted (Hlev-Toledo nd others 1999). In nother study (Nim nd others 1994), Cys nd AcCys (0.5 to 2.5 mm) reduced browning (opticl density t 420 nm) significntly (P < 0.05) in SSOJ stored for 14 d t 45 C. The increse in browning t higher concentrtions of Cys nd AcCys ws ttributed to products formed from glucose or rhmnose Cys interction t low ph nd ws not relted to the presence of oxygen, since the sme results were obtined under nerobic conditions (Hlev-Toledo nd others 1999). Enzymtic browning Enzymtic browning consists of the oxidtion of phenolic compounds to o- quinones tht combine with mines nd sulfur groups from proteins nd reducing sugrs to produce brown-colored polymers (melnines). Enzymtic browning occurs when the cell structure is dmged nd the phenolic compounds locted in the vcuole re brought in contct with the enzymes responsible for the oxidtion of phenols, which re locted in plstids (Artés nd others 1998; Burton 2003; Acree nd Arn 2004; Frnck nd others 2007). In fruits nd vegetbles, the enzymes responsible for enzymtic browning re polyphenol oxidses (PPO), lthough synergistic effect between PPO nd peroxidses (POD) hs been suggested. There re 2 types of PPO: ctechol oxidse (EC ) nd lccse (EC ). Both contin copper in their structure nd use 33

34 moleculr oxygen s secondry substrte (Mrtinez nd Whitker 1995; Artés nd others 1998; Acree nd Arn 2004; Gómez nd others 2006; Dhuique-Myer nd others 2007; Frnck nd others 2007). Ctechol oxidse ctlyses 2 rections referred to s hving cresolse nd ctecholse ctivity. They differ in tht cresolse converts monophenols into orthophenols nd ctecholse oxidizes o-phenols to o-quinones. (Figure 1-2), while lccse ctivity oxidizes p-phenols nd o-phenols (Figure 1-3) (Artés nd others 1998; Dhuique-Myer nd others 2007). Enzymtic browning is not concern in most fruit juices becuse the psteuriztion process, when pplied t the proper temperture time conditions, inctivtes the enzymes present in the rw fruit. Since this review is focused on psteurized fruit juices, no further discussion of this is included here. Arom The mixture of voltile compounds with rom ctivity is responsible for the chrcteristic rom of fruit juices. Dissolved oxygen my ffect the flvor of fruit juices by formtion of precursors such s dehydroscorbic cid, precursor of ldehydes formed during the Strecker degrdtion. It is difficult to quntify the effect of DO in rom becuse there re biochemicl pthwys occurring during the ripening of fruit tht do not involve oxygen nd lso produce rom-ctive voltile compounds. Perez nd Snz (2008) described these biochemicl pthwys in detil nd no further discussion is included in this review. Rections in which oxygen is the rectnt or tht re consecutive to those ffected by the presence of DO nd produce rom-ctive compounds re described subsequently. The oxidtion of L-AA to DHA s previously mentioned produces color nd flvor chnges. DHA contins n α-dicrbonyl structure tht prticiptes in the Strecker 34

35 degrdtion. Strecker degrdtion is the rection between the α-dicrbonyl group from DHA nd primry mino cid to form n ldehyde with crbon chin tht is 1 crbon shorter thn the precursor mino cid chin nd tht is rom-ctive. Well-known exmples of Strecker ldehydes re methionl nd cetldehyde. Methionl is derived from methionine, hs cooked potto rom, nd is n importnt contributor to the cooked sensory ttribute of cnned ornge juices (Brdley nd Min 1992; Reineccius 2006; Perez-Ccho nd others 2007). Acetldehyde is formed from the mino cid lnine nd hs fruity rom (Cremer nd Eichner 2000; Reg nd others 2003). Depending on fruit cultivr nd extrction method, the mount of peel oil present in ornge juice cn be s high s 0.06%. But becuse the excess of peel oil produces mouth burning senstion, the desired concentrtion of peel oil in ornge juice is between 0.02% nd 0.03% (Brddock 1999). About 90% of citrus oil is d-limonene. Becuse d-limonene hs high rom threshold, it does not contribute to the overll rom perception, but serves s crrier for other rom ctive compounds (Perez- Ccho nd Rouseff 2008b). Even d-limonene oxidtion in the presence of moleculr oxygen produces crveol nd crvone (Kutty nd others 1994). Of these 2 compounds, only the presence of crvone in ornge juice hs been reported nd contribution to the rom hs not been ttributed to it, even if the mount of crvone (110 ppb) ws higher thn its sensory threshold in wter (6.7 ppb) (Moshons nd Shw 1994). In citrl buffer contining d-limonene nd in the presence of oxygen, decrese of 20% in d-limonene concentrtion ccounted for n increse of crvone nd crveol of 200 nd 100 ppm, respectively, during the first 2 wk of storge (Kutty nd others 1994). 35

36 Trmmell nd others (1986) studied the effect of oxygen on SSOJ tste. The juice ws psteurized, septiclly filled in sterile glss bottles in nitrogen tmosphere nd stored t 22 C for 22 wk protected from light. Sensory evlution ws performed in juices with DO concentrtions between 0.6 nd 10.1mg L 1 during storge. Juices with the lower DO concentrtion (0.6 mg L 1 ) were sttisticlly preferred over the high DO concentrtion (10.1 mg L 1 ) until week 4 but not lter. Reserchers concluded tht bsed on tste, deertion did not extend the shelf life of the SSOJ. Methods of Mesurement of DO The U.S. geologicl survey (USGS) stndrd field methods to mesure dissolved oxygen re mperometric, luminescence, Winkler titrtion, nd colorimetric methods (Tble 1-4). Severl limittions were ssocited with some of these methods. The industry stndrd technique for DO mesurement uses conventionl Clrk-type electrode (mperometric method). Limittions ssocited with electrochemicl probes include drifts in clibrtion due to oxygen consumption by the electrode, need for turbulent flow to prevent the formtion of concentrtion grdients, nd electricl interference. The most recently developed methods re solid-stte luminescence methods bsed on fluorescence pek quenching or multi frequency phse shift. Typiclly, fluorophore cotings re deposited on the tip of fiber optic probes. The most common fluorophores re ruthenium or pltinum complexes (Sng- Kyung nd Ichro 1997) becuse they re excited nd emit t wvelengths in the visible light rnge nd they hve high sensitivity to oxygen nd lrge Stokes shifts. When oxygen molecules collide with fluorophore in its excited stte, the energy is trnsferred to the oxygen molecule in nonrditive wy, decresing the fluorescence intensity. Ruthenium complex-bsed sensors re the most common nd re used for queous 36

37 liquids nd vpors. Sensors hving pltinum complex s fluorophore re the most sensitive nd, therefore, re used for low levels of DO. Tng nd others (2003) reported tht Tris (4,7-diphenyl-1,10-phennthroline) ruthenium(ii) (Ru[dpp]32+) immobilized in Octyl-triEOS/TEOS sol-gel hs nitrogen-to-oxygen intensity rtio (IN2/IO2) of Wheres, Yeh nd others (2006) reported tht oxygen sensitive dyes Pltinum tetrkis pentfluorophenylporphine (PtTFPP) nd pltinum octethylporphine (PtOEP) immobilized in sme mtrix Octyl-triEOS/TEOS sol-gel hs IN2/IO2 of 22 nd 47, respectively. There re some commercilly vilble ruthenium-bsed sensors tht cn be used for the determintion of DO in orgnic liquids nd vpors. Chu nd Lo (2007) reported tht PtTFPP nd PtOEP immobilized in support mtrix n-propyl-trimos/tfp- TriMOS hs higher sensitivities compred to Octyl-triEOS/TEOS sol-gel. When clibrting fiber optic probe, stndrds should hve the sme refrctive index s the smple. Limittions of fiber optic sensors re sensitivity to surrounding light, photo bleching of the fluorophore, nd signl intensity dependence of fiber bending. Attempts to isolte the fluorophore coting from surrounding light using drk overcot resulted in slow response time, poor reproducibility nd fouling. Another wy of mesuring oxygen using luminescent-bsed sensor system is using phse modultion techniques. The phse shift between the exciting nd the emitted light source is correlted with DO concentrtion (Brnes nd others 1990; Jing nd others 2002). Phse mesurements re more stble nd essentilly independent of the luminescence dye concentrtion nd photobleching which ffect signl intensity (Ogurtsov nd Ppkovsky 1998). 37

38 Methods of Removl of DO Methods of oxygen removl such s vcuum-deertion, nitrogen-sprging, membrne deertors, enzyme-bsed deertors, nd oxygen scvengers re discussed in the subsequent sections. Vcuum-Deertion Vcuum-deertion is bsed on the reduction of the pressure of the gs bove the juice. As pressure in the juice hedspce decreses, dissolved ir is relesed from the liquid phse following Henry s lw. In the citrus industry, this is the most common deertion method nd it performs the dul role of removing oxygen nd removing excess peel oil from ornge juice before psteuriztion (Brddock 1999). This opertion is clled flsh deertion becuse of the sudden decrese in pressure tht preheted juice undergoes s it enters the deertion tnk, producing prcticlly instntneous seprtion. Figure 1-3 shows digrm of vcuum-deertion system tht consists of vcuum system, vcuum chmber, juice feeding with spry inlet, dischrge of juice in the bottom of the tnk nd ir nd voltile compounds outlet t the top of the tnk. In the citrus industry, to ensure high elimintion of ir, juice is preheted to 50 to 60 C before deertion. The pressure in the vcuum chmber nd the inlet temperture re djusted so tht the inlet temperture is 2 to 5 C bove the boiling point of the OJ t tht pressure (Ringblom 2004). The 1st type of deertor in the juice industry ws the btch type tht ws lter substituted by continuous centrifugl, spry, or film deertors. In centrifugl deertors, the juice is fed to vcuum chmber contining revolving pn, which lets the juice flow through smll orifices in the rim. Spry deertors lso feed chmber, but in this cse juice is pssed t high velocity through smll nozzle to disperse it into 38

39 smll droplets, s Figure 1-4 shows. In the film-type deertor, juice is dispersed by spreding jet into number of thin lyers (Joslyn 1961). U.S. ptent estblished deertion method for fruit juices in which wter nd concentrted source of fruit juice re deerted seprtely nd subsequently mixed. Wter is deerted t temperture rnge of 50 to 60 C, with vcuum residence times between 1 nd 3 min to rech finl DO concentrtion between 0 nd 0.4 mg L 1. Fruit juice concentrte is deerted under milder conditions to void dmging voltile components nd vitmins, t tempertures in the rnge of 0 to 5 C, vcuum residence time of 3 to 6 min, nd finl DO concentrtion lower thn 3 mg L 1. The vcuum pressure suggested is 0.1 br (Rhrooh nd Engel ). Gs Sprging Gs sprging consists of displcing the DO with nother gs such s nitrogen or crbon dioxide. The gs is bubbled into the liquid or meets countercurrently with the liquid in chmber. Figure 1-5 depicts gs being bubbled into the juice. The size of the bubbles of gs, the height of the chmber, nd the flow rtes of gs nd liquid determine the rte nd extent of oxygen removl. For exmple in OJ, round 99% of O 2 ws removed using N 2 sprging compred with 77% of O 2 removed with vcuumdeertion (Joslyn 1961). Sprging with helium t flow rte of 40 ml s 1 for 30 s in one liter of queous buffer ws more effective t removing DO thn vcuum-deertion t 90 kp (Degenhrdt nd others 2004). The disdvntge of this deertion process is tht flvor voltile components re removed nd lost to the tmosphere (Jordn nd others 2003). 39

40 Membrne Deertors Membrne gs seprtion is bsed on the use of membrne s brrier between the liquid nd the gs phse; for exmple, between contminted wter nd sweeping ir phse (Mhmud nd others 1998). A membrne seprtor is device tht performs mss trnsfer between gs nd liquid or 2 liquids without dispersion of the phses on ech other by pssing the fluids on opposite sides of microporous membrne. The membrne configurtion cn be flt sheet or hollow fiber (Gbelmn nd Hwng 1999). Hollow fiber configurtion hs high surfce-to-volume rtio mking the membrne more efficient by providing lrge contct surfce re compred with the flt sheet configurtion (Mhmud nd others 1998). In wter deertion, hollow fiber configurtion membrne hs surfce-to volume rtio higher thn conventionl pcked tower ir-stripping (Mhmud nd others 1998). Since the 1970s, the cpcity of selective hollow fibers to remove oxygen hs been studied. It ws demonstrted tht polypropylene hollow fiber seprtor of 7.7 ft 3 volume removed 96% of DO in flux of L d 1 of wter (Cole nd Genetelli 1970). Applictions in the food industry of hollow fiber membrne contctors re crbontion of beverges, nitrogention of beer to give dense fom hed, deoxygention of beer to preserve flvor, s low-cost lterntive to supercriticl extrction, extrction of ethnol nd other metbolites during fermenttion, seprtion of products during enzymtic trnsformtion, protein extrction, osmotic distilltion to concentrte fruit nd vegetble juices, nd in the production of low-lcohol wine (Gbelmn nd Hwng 1999). This technology is commercilly vilble; for exmple, the Liqui-Cel TM membrne contctors schemtized in Figure 1-6, consist of membrne seprtor with severl hollow fiber membrnes knitted into fbric nd wrpped round center tube clled the 40

41 distribution/collection tube. Liquid flows into the distribution tube nd is forced to pss rdilly through the membrne, where vcuum or swept gs is pplied to remove the oxygen from the liquid. The membrne is mde of porous polypropylene nd polyolefin. These hydrophobic membrnes llow only gses to pss through membrne pores. As in ny type of membrne, fouling cn be problem in hollow-fiber seprtors. Fouling in wter tretment pplictions hs been reported due to suspended solids ccumultion, growth of bcteri nd slime development, clcium crbonte deposition, nd iron oxide fouling (Mhmud nd others 1998). Of these mentioned, suspended solid ccumultion nd growth of bcteri cn be problem when processing fruit juices, mking this technology unsuitble for nonclrified fruit juices. Enzyme-Bsed Deertion Glucose oxidse (GOx)-ctlse complex hs the bility to remove oxygen from solutions nd hedspce bsed on the rections shown in Eq nd 1-15: 2C H O glucose oxidse 6H12O6 2O2 2H2O 2C6 H12O7 2 2H H O O ctlse 2O (1-14) (1-15) GOx-ctlse ctlyzes the oxidtion of D-glucose to gluconic cid in the presence of ditomic oxygen. The by-product hydrogen peroxide is converted by ctlse to wter nd oxygen. Thus, the overll effect of the enzymtic rection is removl of 0.5 mol of oxygen per mole of oxidized D-glucose (Sgi nd Mnnheim 1990). At ph 4 to 7, GOx from Aspergillus niger hs t lest 90% of ctivity compred to its optiml ctivity t ph 5.5 to 6 (Klisz nd others 1990). Ctlse, from Micrococcus sp., hs working ph of 3 to 9 nd the ph optimum is 5.5 (Bioctlysts 2004; Schomburg 2007), nd ctlse from E. coli hs working ph of 6 to 8 (Vrndo nd 41

42 others 2004). GOx-ctlse complex showed mximum ctivity t ph 6 nd 30 C in model sld dressing (Mistry nd Min 1992). GOx-ctlse hs been studied in ornge, grpe, pple, nd per (Sgi nd Mnnheim 1990; Pickering nd others 1998; Prpinello nd others 2002). In ornge juice, 17 units L 1 of GOx-ctlse enzymtic preprtion were enough to reduce the DO concentrtion to less thn 1 ppm in 30 min t 20 C, but lso n importnt loss of vitmin C ws reported. This loss ws ttributed to the rection of L-AA with hydrogen peroxide (Sgi nd Mnnheim 1990). In Riesling grpe (ph 3.08, free SO mg L 1, nd bound SO mg L 1 ) treted with GOxctlse, the low ph inhibited the enzyme rection. At the sme GOx-ctlse concentrtion (2 g L 1 ), the gluconic cid produced t ph 6 (80 g L 1 ) ws lmost 3 times the gluconic cid produced t ph 3.1 (30 g L 1 ). In the previously described study, the enzyme ws used to reduce the mount of glucose vilble for fermenttion to produce reduced-lcohol wine (Pickering nd others 1998). A mixture of GOx-ctlse clled GOX100 (Gist-Brocdes, Lille, Frnce) ws compred with scorbic cid nd peroxidse in terms of oxygen removl nd browning control in pple nd per purees. The unit of ctivity of GOx ws defined s 1 μmol of gluconic cid formed per min, t 35 C, ph 5.1, nd for ctlse s 1 μmol of hydrogen peroxide consumed per min, t 25 C, ph 7. In the ctivity ssy t different ph vlues, GOx nd ctlse showed its mximum ctivity t 50 nd 12 units mg 1, t ph 5.5 nd 6.5, respectively (Prpinello nd others 2002). In this study, the reduction to ph 4 decresed GOx nd ctlse ctivity to 15 nd 5 units mg 1, respectively. Ascorbic cid ws the most effective in preventing nonenzymtic browning. The min disdvntge of GOxctlse is tht its ctivity is reduced t low ph vlues limiting its ppliction to fruit 42

43 juices. Therefore, studies hve been done on incorporting GOx into the pckge to interct with the juice but preventing migrtion into the food. The ctlytic efficiency, defined s kct Km 1, is mesure of the substrte specificity nd ctlytic bility of the enzyme nd ws compred between immobilized GOx nd GOx in solution. Immobilized GOx embedded on UV polymerizble resin hd similr (no significnt difference) kct Km 1 vlues to GOx in solution t 30, 25, nd 10 C, but ws 100 times lower (significnt difference) t 5 C (Kothplli nd others 2007). More studies need to be done to improve the ctlytic ctivity of immobilized GOx. Oxygen Scvengers Kinetic studies of DO in liquid foods describing oxygen consumption hve been done in pckged nd unpckged products. Ahrne nd others (1997) developed mthemticl model for pckged liquid products tht consider both oxygen consumption nd oxygen trnsfer through the pckge. The model considered oxygen consumption s 1st-order kinetic rection nd oxygen trnsfer s 3-step process: (1) oxygen from the tmosphere being bsorbed in the pckging mteril; (2) diffusion of oxygen through the pckge; (3) desorption of oxygen from pckge nd solubiliztion into the food. Eqution 1-16 describes both processes: C O eq initil eq Kt CO C O CO e (1-16) eq where C O, C 2 O 2, nd C 2 initil O re the DO concentrtion in the food inside the pckge t time t, the DO concentrtion t equilibrium, nd the initil DO concentrtion, respectively. K is the rte constnt estimted for this mthemticl model. C eq O2 K K R D C k out O2 ' pk D (1-17) 43

44 K K R k ' p K D (1-18) where K R, K D, nd K re rection rte constnts for oxygen consumption, DO mss trnsfer coefficient, nd rte constnt of the model, respectively, ll of them in (d 1 ' ). k is rtio of prtition coefficient nd is considered equl to 1. For ornge juice, p K vlues were between 2.69 nd d 1 nd E ws 46 kj mol 1. The model djusted well to the reported dt for ornge nd pple juices (Ahrne nd others 1997). Oxygen scvengers re commonly used s the bsis for ctive pckging to remove residul oxygen from food nd extend shelf life. Commonly used oxygen scvengers re bsed on iron powder oxidtion (Kerry nd others 2006), scorbic cid oxidtion (Brody 2005), photosensitive dye oxidtion (Vermeiren nd others 1999), enzyme-medited oxidtion (glucose oxidse ctlse nd lcohol oxidse) (Andersson nd others 2002; López-Rubio nd others 2004; Ozdemir nd Floros 2004), unsturted ftty cids oxidtion, nd polymers oxidtion (Solis nd Rodgers 2001). Oxygen scvenging mterils re dded to the pckge s schets, dhesive lbels, cp liners, incorported into the mtrix of plstic continers during the continer fbriction, or s multilyer ctive film (Vermeiren nd others 1999). A list of oxygen scvenging systems ptented from 1994 to 2002 cn be found in Ozdemir nd Floros (2004). Oxygen scvenger performnces cn be mesured in terms of their bsorption cpcity nd bsorption rte. Absorption cpcity refers to the mount of oxygen tht cn be removed. Absorption rte of iron oxidtion nd enzyme-medited oxidtion oxygen scvengers hve been described s 1st-order rection. Reported vlues for iron oxidtion-bsed oxygen scvengers t tempertures between 5 nd 35 C re nd h 1 g 1 of oxygen scvenger (Chrles nd others 2006) nd between 44

45 0.26 nd 2.46 h 1 for tempertures of 1.5 to 25 C (Tewri nd others 2002). E, for different brnds of iron oxidtion-bsed oxygen scvengers hd vlues between nd kj mol 1 nd for n enzyme ctlyzed oxidtion-bsed oxygen scvenger, vlue of kj mol 1 (Tewri nd others 2002; Chrles nd others 2006). Migrtion of the oxygen scvenger mteril to the pckged food is concern, especilly when using nonedible substnces (iron slts) in liquid foods. Lopez- Cervntes nd others (2003) compred migrtion in plstic cup nd schet ironbsed oxygen scvenger. The overll migrtion vlues were 5.35 (plstic cup) nd 44.9 (schet) mg L 1 in wter nd nd mg L 1 in 3% cetic cid solution. The study concluded tht the use of schets in cidic liquid foods is not desirble becuse the overll migrtion of iron (63.15 mg L 1 ) is higher thn 60 mg L 1, the mximum llowed vlue estblished by the Europen Union (EU). Incorportion of the oxygen scvenger during continer fbriction into the mtrix of plstic continers or s multilyer ctive film is desirble for liquid foods becuse there re some mterils tht do not hve migrtion problems. Oxygen removl vi polymer oxidtion is bsed on the use of n oxidizble plstic nd coblt s ctlyst. The typicl oxidizing plstics re polyethylene terephthlte (PET) nd polymides such s MXD-6 nylon, nd the rection is triggered by humidity. The Chevron Phillips OSP TM or oxygen scvenger polymer is system tht consists of 90% ethylene methylcrylte cyclohexenylmethyl crylte (EMCM) nd 10% coblt slt with proprietry photoinititor. The dvntges of this system re tht it is inctive until exposed to UV light, it hs FDA pprovl, cn be used in rigid nd flexible pckging, nd it is recommended for mets, cheese, sncks, bred, coffee, te, nuts, cndies, fruit juices, 45

46 beers, nd wines. The OSP system hs scvenging cpcity of 45 to 70 ml O 2 g 1 of scvenger. Compring ornge juice pckged in regulr crtons nd crtons with OSP, the ltter retined higher concentrtions of vitmin C nd showed lower browning index (Solis nd Rodgers 2001). Tble 1-5 shows list of oxygen scvengers commercilly vilble nd recommended to use on liquid foods. Most of them re bsed in polymer oxidtion with moisture or UV light s the ctivting gent, nd in lower proportion they re bsed on scorbic cid nd GOx. There re lso O 2 indictors tht re useful to detect O 2 inside the pckge due to low ctivity of the O 2 scvenger or to O 2 entering the pckge vi leks. Commercil brnds re, for exmple, Ageless Eye TM (Mitsubishi Gs Chemicl Co. Inc.,Tokyo, Jpn) used for the mster pckge of O2 scvenging schets (Ageless Eye) with color chnges due to the presence (blue) or bsence (pink) of O 2. O 2 indictors nd bsorbers re used together for chilled nd shelf stble redy mde pckged foods (Rodrigues nd Hn 2003). Current reserch on oxygen scvengers focuses on the development of pckges with oxidizing polymers nd biologicl entities such s enzymes nd microorgnisms; for exmple, the incorportion of GOx in pckge, using UV polymeriztion (Kothplli nd others 2007); nd the entrpment of erobic microorgnisms (Kocuri vrins nd Pichi subpelliculos) into polymeric mtrix (Altieri nd others 2004). Oxygen is incorported into fruit juices during their preprtion nd processing nd remins dissolved during storge, ffecting prmeters correlted with the qulity of the juice such s L-AA concentrtion, color, nd rom. The direct erobic oxidtion of L- AA indirectly ffects color nd rom profile. L-AA oxidtion occurs t the beginning of 46

47 storge. Most of the kinetic studies of the effect of DO in fruit juices re on L-AA degrdtion nd browning development nd only few of them describe the kinetics of DO consumption. It is necessry to develop models tht ccurtely describe the scorbic cid nd oxygen interction during oxidtion due to the importnce of these rections. Determintion of kinetic prmeters in rel systems insted of models or pure solutions provides pprent kinetic prmeters tht, lthough highly vrible due to intrinsic rw mteril vribility, re more likely to be of prcticl vlue to juice processors. Mechnisms nd compounds directly relted to DO nd rom compounds lso need to be studied. Oxygen removl methods suitble for pulpy or cloudy juices re vcuum-deertion nd gs sprging, but these methods hve the min disdvntge of removing importnt voltile rom compounds. The methods tht cn remove dissolved oxygen but preserve the rom compounds re membrne deertors nd GOx ctlse but only the 1st one is currently used in the fruit juice industry. Oxygen cn lso be removed during storge by dding enzymes or n oxygen scvenger mteril inside the pckge or incorporting them into the pckging mteril. Specific Objectives The centrl hypothesis of this reserch is tht current processing conditions deteriorte ornge juice qulity nd cn be improved. The overll objective of this reserch is to understnd the effect of initil dissolved oxygen (DO) nd temperture in vitmin C, color nd rom of not from concentrted ornge juice to increse fresh-like qulity retention for industril processing nd storge. The first specific objective is to understnd the effect of selected initil DO levels in vitmin C, rom, voltile compounds nd color of psteurized commercil ornge juice during storge t 5 C. 47

48 The second specific objective is to understnd the effect of selected initil DO on the rte of DO consumption, vitmin C, voltile compounds, nd color of fresh nd psteurized ornge juice during storge t selected tempertures, to increse fresh-like qulity retention nd identify opportunities for processing improvement. The working hypothesis for both specific objectives is tht DO decreses vitmin C, deteriortes rom nd color of ornge juice during psteuriztion nd storge. Tble 1-1. Ascorbic cid content of different fruit juices (mg 100 g -1 ) Ornge juice, frozen concentrte, unsweetened, undiluted White guv, frozen pulp Red guv, frozen pulp 56.0 Grpefruit juice, white, frozen concentrte, unsweetened, undiluted Ppy rw 61.8 Ornge juice, rw 50.0 Pinepple nd grpefruit juice drink, cnned 46.0 Lemon juice, rw 46.0 Crnberry juice cocktil, bottled 42.3 Grpefruit juice, pink nd white, rw 38.0 Ornge juice, cnned, unsweetened 34.4 Ornge juice, chilled, includes from concentrte 32.9 Grpe drink, cnned 31.4 Grpefruit juice, white, cnned, unsweetened 29.2 Mngo rw 27.7 Lemon juice, cnned or bottled 24.8 Pinepple nd ornge juice drink, cnned 22.5 Tngerine juice, cnned, sweetened 22.0 Apricot rw 10.0 Pinepple juice, cnned, unsweetened, without dded scorbic cid 10.0 Kiwi rw 9.3 Lime juice, cnned or bottled, unsweetened 6.4 Pssion fruit frozen pulp 4.3 Prune juice, cnned 4.1 Apple juice, cnned or bottled, unsweetened, without dded scorbic cid 0.9 Modified from (Aymoto-Hssimotto nd others 2005; Holden 2006; Genovese nd others 2008) 48

49 Tble 1-2. Rection rte constnts for AA degrdtion obtined under reducing nd oxidizing conditions t 90 C. K1 (h -1 ) K2 (h -1 ) K3 (h -1 ) Control: L-AA solution (200 mg L -1 ) x x x x 10-3 L-AA solution under reducing conditions: cysteine dded 20 mg L -1 cysteine x x x 10-3 L-AA solution under oxidizing conditions: Fe 3+ dded 20 mg L -1 Fe x x x x 10-3 Dt from (Serpen nd Gökmen 2007) 49

50 Tble 1-3. Summry of selected studies of scorbic cid degrdtion in juices in which the effect of oxygen ws ssessed. Product DO concentrtion (mg L -1 ) Initil Finl Kinetic order Rte constnt k Ornge juice ND < 0.4 ND ND ND Concentrted ornge juice Model solution First Anerobic First E vlue (kj mol-1) T ( C) 4 nd x10-4 to 4.6 x10-2 s -1 ND 25 to x10-5 to 3.98 x10-8 s to to 90 Ornge juice 2.7 < 0.04 ND ND ND 4 nd 25 ph ND ND 0 Storge time (d) 160 to ND ND 373 Other vribles Pckging mteril, light bsence Processed with TASTE C evportor Nonelectrolytes nd electrolytes Pckging mteril FCOJ < 1 ND ND ND ND Pckging Reference (Ros- Chumills nd others 2007) (Brddock nd Sdler 1989) (Rojs nd Gerschenson 2001) (Zerdin nd others 2003) (Berlinet nd others 2006) Citrus-like beverge ND First 2.64 x x10-8 s -1 ND 5 nd Pckging mteril permebility (Bino nd others 2004) FCOJ 1 ~0 ND ND ND 8 ND 52 Clrified ornge juice Anerobic First Ornge juice Anerobic First 7.17 x 10-7 to x 10-4 s (serum) to 7.5 x 10-7 to x s Presence of light, pckging mteril Brix content (Solomon nd others 1995) (Johnson nd others 1995) Psteurized lemon juice First 1.93 x x10-7 ND light bsence 2.08 x10-7 s -1 (Robertson nd Smniego 1986) 50

51 Tble 1-3. Continued. Product DO concentrtion (mg L -1 ) Kinetic order Rte constnt k E vlue (kj mol-1) T ( C) ph Storge time (d) Other vribles Reference Cnned ornge juice Initil Finl 3.5% < 0.1% Zero SSOJ D 4.7 to 5.67 Ascorbic cid solutions Ornge nd tngerine juice mixture Cnned ornge juice 0.5 to 1 tm Weibull model 0.4 tm Pseudo first First 3.5% < 0.1% Anerobic Zero Polynomil First SSOJ 7.0 ~0 First mg L -1 h -1 ND 30 to x 10-5 s to x to 2.31 x 10-4 s -1 ND nd x 10-5 to 6.2 x 10-5 s to to 4 Different methods to remove DO, cn mteril Light bsence 3 to 3.85 ND ph, bsence nd presence of Cu 2+ nd Fe ND ND mg L -1 h -1 0 to x10-2 to 4.7 x10-2 mg 3.5 L -1 h -1 to (Liner) ND x to 375 to 2.1 x10-3 mg L-1 h -1 (Qudrtic) 1.0 x10-8 to 1.3 x to 375 s 4.89 x 10-8 s -1 Different methods to remove DO, cn mteril (Kefford nd others 1959) (Mnso nd others 2001) (Khn nd Mrtell 1967) (Dhuique- Myer nd others 2007) (Kefford nd others 1959) Concentrted ornge juice First 1.09 x 10-7 s (Rssis nd ND 32 ND 35 Presence of Cu Sguy 1995) 51

52 Tble 1-3. Continued. Product DO concentrtion (mg L -1 ) Kinetic order Rte constnt k E vlue (kj mol-1) T ( C) ph Storge time (d) Other vribles Reference Initil Finl Ascorbic cid queous solutions 7.5 (t 760 mm Hg) Constnt bubbling of ir Model of consecutive zero nd first Zero: 81.3 mg L -1 h -1 First:4.69 x ND 30 ND 0.46 ND 10-5 s -1 (Ski nd others 1987) FCOJ 4.5 < 0.5 to 2 Zero First 2.29 x 10-5 to 4.34 x 10-5 mg L -1 s x 10-8 to 3.36 x 10-7 s -1 ND 4 to 105 ND 64 Pckged in TetrBrik crtons (Kennedy nd others 1992) Acette/ceti c buffer solution SSOJ 9.2 to to to < Pseudo - First Second 9.0 x 10-6 to to 1.06 x 10-4 s x 10-4 to 3.27 x 10-3 M -1 s -1 ND 26.5 to 33 3 to ND ND ND 22 ND 150 Sucrose concentrtion, ph Psteuriztion method (Hsieh nd Hrris 1993) (Trmmell nd others 1986) 5.56 x 10-6 to to 7.2 to No quntifiction (Mck nd <1.0 First Infnt formul 6.67 x 10-5 s -1 ND of scorbic cid others 1976) 1.27 x 10-4 to 1 to Infnt formul 0.4 Second (Singh nd ND 7.2 ND 1 Light intensities 8.71 L mg -1 h -1 others 1976) FCOJ= from concentrted ornge juice; b SSOJ =single-strength ornge juice; c TASTE=thermlly ccelerted short-time evportor; ND= not determined. 52

53 Tble 1-4. Comprison of different methods used to mesure DO Method Advntges Disdvntges Electrochemicl probe Vriety of shpes, sizes nd consumbles Some probes hve build on thermistor for temperture compenstion (Lewis 2006) O2 consumption.electrolyte solution should be free of bubbles to void interferences Membrne fouling, stgntion or rupture Opticl probe Ese of minituriztion Do not consume oxygen Sensitive to low levels of oxygen Fst response time: s (Choi nd Xio 2000) Continuous mesurements. Sensitive to mbient light Photo bleching of the fluorophore Position dependence of optic fiber Winkler titrtion Officil reference method (ASTM D ) Highly dependent on nlyst experience. Contmintion with tmospheric O2 during titrtion. Chemicl nd color interferences Colorimetric methods Officil methods (ASTM D ). Kit vilble. Able to mesure ppb of O2. Not for continuous mesurements 53

54 Tble 1-5. Selected commercilly vilble O 2 scvengers used in liquids Trde nme Mnufcturer Country Scvenger mechnism Incorportion method Shelf-plus Cib Specilty United PET copolyester Into the plstic Chemicls Sttes oxidtion film Activting gent Moisture Used in Soups, beverges PureSel W.R. Grce Co. Ltd United Sttes Ascorbte/ metllic slts On bottle crowns NS Liquid food ZerO2 CSIRO- Food Science Austrli/ Visy Industries Austrli Polymer oxidtion & photosensitive dye compound Into the plstic film UV light Beverges O2-displcer closure system Alco CSI United Sttes NS Shell molded liner NS Beer Amosorb 3000 BP Amoco P.L.C United Sttes Orgnic oxygen bsorber for polyester Into the PET bottles Metl slt Beverges Strshield TM Crown Cork & Sel Co., Inc. United Sttes Polymer oxidtion Into the PET bottles Moisture Wines Cryovc OS1000 & 2000 Cryovc Seled Air Corportion United Sttes Polymer oxidtion Into the plstic film UV-light Not specified Oxygurd Toyo Seikn Kish, Inc Jpn Iron oxidtion Cp liner, into the plstic cup NS Soft drinks, beer, wine, soup 54

55 Tble 1-5. Continued Trde nme Mnufcturer Country Tri-SO2RB Tri-Sel United EVA Interntionl, Inc Sttes Scvenger Incorportion Activting mechnism method gent Used in Polymer oxidtion Closure liner NS Beverges OSPTM Chevron Phillips Chemicl Co. United Sttes Polymer oxidtion Blended or s lyer in pckge UV light Fruit juices, beers, wines Oxbr, Strshield bottles Constr United Sttes, Frnce, UK Coblt ctlyzed oxidtion of MXD-6 Into the plstic bottle NS For wines, beer, suces, fruit juices nd flvored lcoholic beverges Oxycp Stnd Industrie Frnce NS Cp liner NS Wine, beer, fruit juice, suces, soft drinks Drex technology NS= not specified Grce Drex United Sttes Ascorbic cid oxidtion Into the plstic film NS Juices, te, beer, sport drinks 55

56 Figure 1-1. Ascorbic cid degrdtion mechnism modified from Dvey nd others (2000). Figure 1-2. Rections ctlyzed by ctechol oxidses. 56

57 Figure 1-3. Rections ctlyzed by lccses. 57

58 Figure 1-4. Vcuum deertor scheme. 58

59 Figure 1-5. Gs sprging method scheme. 59

60 Figure 1-6. Hollow-fiber membrne seprtor (modified from (Polypore 2007). 60

61 CHAPTER 2 ASSESSMENT OF DISSOLVED OXYGEN EFFECT ON VITAMIN C AND AROMA COMPOUNDS IN PASTEURIZED NOT-FROM CONCENTRATE ORANGE JUICE Introduction Ornge juice ws the most consumed fruit juice in the 2007/08 seson, representing 50% (14.4 L, single strength equivlent per cpit) of the totl nnul consumption of fruit juices (USDA 2008). Although psteurized ornge juice lcks of the freshness of non psteurized ornge juice nd hs cooked-like roms, it is commercilly preferred becuse psteuriztion inctivtes enzymes nd destroys microorgnisms incresing the shelf life of the juice (Perez-Ccho nd Rouseff 2008). Another fctor tht is thought to ffect qulity of ornge juice is the dissolved oxygen tht is incorported during extrction which my promote oxidtion of scorbic cid (Ringblom 2004). Ascorbic cid cn be oxidized to dehydroscorbic cid, which my prticipte in Strecker degrdtion producing rom ctive ldehydes such s methionl (cooked potto rom) nd cetldehyde (fruity rom) (Perez-Ccho nd Rouseff 2008). Ornge juice is considered n importnt source of scorbic cid. Therefore, preventing scorbic cid degrdtion during processing is necessry in order to protect nutritionl vlue nd rom of the juice. Ascorbic cid degrdtion is two step process: 1) scorbic cid is oxidized to dehydroscorbic cid in reversible rection nd 2) the dehydroscorbic cid is irreversibly hydrolyzed to 2,3 - diketogulonic cid. In the citrus industry, deertion before psteuriztion hs been implemented to remove the ir incorported into the juice. However, importnt voltile compounds such s lcohols, ldehydes, ketones, esters nd terpene hydrocrbons re lso removed during these procedures reducing the rom intensity of the ornge juice (Brddock 61

62 1999; Jordn nd others 2003; Ringblom 2004; Perez-Ccho nd Rouseff 2008; Grcí-Torres nd others 2009). Studies on the effect of oxygen during storge of citrus juices hve focused on chnges in vitmin C content (Kennedy nd others 1992; Dhuique-Myer nd others 2007), browning (Trmmell nd others 1986; Rssis nd Sguy 1995; Solomon nd others 1995), voltile sulfur compounds (Jblpurwl nd others 2010) nd pckging (Zerdin nd others 2003). Most of these prmeters hve been studied seprtely. Comprehensive studies of chnges in ornge juice chrcteristics need to be performed to hve full ssessment of the effect of oxygen in fruit juices. The objective of this study ws to correlte the chnges in vitmin C content nd rom profile of ornge juice with different dissolved oxygen (DO) levels during storge t refrigertion conditions (5 C). Mterils nd Methods Ornge Juice Preprtion nd Chrcteriztion Not-from concentrte (NFC) psteurized erly-mid Vlenci ornge juice ws obtined from Citrosuco S.A. (Lke Wles, FL, U.S.A.). Soluble solids, ph, trittble cidity nd suspended pulp content were mesured in the ornge juice before tretments. Soluble solids were mesured using digitl refrctometer (Abbe Mrk II, Reichert, Depew, NY), ph using ph meter (Orion 5 Str,Thermo Fisher Scientific, Beverly, MA), titrtble cidity (TA) by titrtion with 0.1 N NOH using n utomtic titrtor (Schott, Elmsford, NY). Suspended pulp ws mesured by centrifugtion of 50 ml of ornge juice during 10 min t 365 x g nd clcultions were done s percentge of volume of pulp precipitted per totl volume of centrifuged smple. Screened pulp ws mesured by pssing 500 ml of ornge juice through 20 mesh screen using 62

63 FMC FoodTech quick fiber device (JBT Food Tech, Lkelnd, FL) tht shkes the screen during 2 min, until the pulp is free of excess of juice. Results re expressed s weight of pulp recovered in the screen by volume of juice used for the test (FMC 2002). Smple Preprtion nd Storge Study Sixteen liters of ornge juice were divided in four portions of 4 L ech: A nitrogen tmosphere ws creted by filling vcuum-emptied hnds in-bg tmospheric chmber (Spilfyter, Green By, WI) with nitrogen. The first smple portion ws bubbled with industril grde nitrogen (Prxir, Inc. Dnbury, CT) in the nitrogen tmosphere; the second portion ws bubbled with ultr high purity oxygen (Prxir, Inc. Dnbury, CT); the third nd fourth portions were not bubbled. Gs-bubbling consisted of pssing the gs t flow of 1.5 L min -1 for 10 min through 700 ml of ornge juice in 1-L gs wshing bottle (Chemglss, Vinelnd, NJ) s ppendix C shows. All ornge juice portions were poured into 30 ml glss mber vils. No hedspce ws left in vils filled with smples from first, second, nd third portions. Air hedspce equivlent to one third of continer totl volume (10 ml) ws left in vils filled with smples from the fourth portion. For smples bubbled with nitrogen, vils were filled in the nitrogen tmosphere. Storge of ll smples ws t 5 C during 60 dys. Single nlysis of three seprte vils from the sme btch of juice ws done. Dissolved oxygen (DO) nd vitmin C were mesured during the storge. Voltile nlysis ws performed t dy 1 of storge for ll the juice portions nd t dy 60 only nitrogen-bubbled nd hedspce smples were vilble for nlysis. Microbil Anlysis Ornge juice ws received in septic pckges tht were kept t 4 C during 3 dys until smple preprtion, ll the continers nd utensils used to mnipulte nd 63

64 store the juice were sterilized using n utoclve (40 min t 121 C), ornge juice ws poured into finl continers for storge within lminr flow hood to minimize product contmintion becuse in preliminry tests we observed some contmintion. Microbil popultion ws quntified in ll juices during storge to insure tht observed chnges in DO nd voltile compounds were not the result of microbiologicl ctivity. Microbil nlysis ws performed t dys 1, 6, 15, 30 nd 60. Aerobic plte counts were conducted using plte count gr (PCA), cidophilic bcteri ssocited with spoilge of citrus products nd molds, yests were quntified using ornge serum gr (OSA) nd cidified potto dextrose gr (PDA), respectively. Smples were serilly diluted with sterile 0.1% peptone s needed nd poured in the pltes by duplicte. PCA, OSA, PDA pltes were incubted 24 h t 35 C, 24 h t 30 C nd 2 dys t 25 C, respectively. DO Mesurement DO concentrtion ws mesured under nitrogen tmosphere using n electrochemicl probe model Orion D nd ph meter model Orion 5 Str (Thermo Fisher Scientific, Beverly, MA) tht ws inserted in the vils contining ornge juice then the cp ws tightly ttched to void entry of ir nd no hedspce ws left. Constnt stirring ws mintined during the mesurements nd DO concentrtion vlues were collected digitlly using computer progrm written in LbVIEW 7.1 (Ntionl Instruments, Austin, TX). Clibrtion of the probe nd mesurement of DO ws done t the sme s the storge temperture (5 C) for ech smple. A stndrd curve ws generted by correltion of the DO probe current response expressed in nnomperes vs. DO concentrtion expressed in µm from n oxygen-sturted nd n oxygen- free stndrds t the sme temperture. Oxygen-sturted stndrd ws ir-sturted wter 64

65 for which DO concentrtion hs been reported by Lewis (2006b) t different pressure nd temperture conditions. Oxygen-free stndrd ws 13% glucose solution with 0.05% of the enzyme glucose oxidse tht ws llowed to consume oxygen for 5 min before mesuring oxygen nd DO concentrtion ws ssumed to be 0 µm. The use of smll mount of enzyme ws enough to decrese DO content. Use of glucose oxidse insted of sodium sulfite to prepre the 0% DO stndrd hd the dvntge of not fouling the probe membrne. Vitmin C Mesurement Vitmin C ws quntified s scorbic (AA) nd dehydroscorbic (DHA) cid. The cpillry electrophoresis (CE) system used ws model P/ACE TM MDQ with DAD with the softwre Krt 32 version 5.0 from Beckmn Coulter (Fullerton, CA, U.S.A.). CE system ws run t 21 kv, 23 C, detection t 263 nm nd 35 mm sodium borte with 5% (v/v) cetonitrile s running buffer ccording to the method described by Cnclon (2001). The cpillry ws bre fused-silic from Polymicro Technologies (Phoenix, AZ, U.S.A) 50 mm id, 56 cm totl length. Clibrtion curve ws done with stndrds contining scorbic cid concentrtions of 284, 852, 1419, 1987, 2555 nd 3123 µm. Smples were diluted (1:4) with 1000 mg L -1 Ethylenedimine tetr cetic cid dipotssium slt (EDTA) solution nd filtered through 0.45 µm filter. To determine dehydroscorbic cid (DHA), L-dithiothreitol (DTT) ws dded t 13 mm (0.2 %) finl concentrtion in smples in order to reduce DHA to AA. The DHA content of the smple ws clculted by subtrcting the AA content before dding DTT from the totl AA content fter dding DTT. Tryptophn ws used s internl stndrd t 200 mg L -1 finl concentrtion nd ws detected t 214 nm. 65

66 Color Mesurement A Minolt colorimeter model CR-400 (Konic Minolt, Jpn) ws used to mesure Hunter prmeters nd L,, b were clculted with respect to control ornge juice. Arom Anlysis Hedspce smpling Under nitrogen tmosphere, 10 ml of smple were plced into 40 ml glss vil contining stirring br nd tightly closed with screw cp with Teflon-coted septum. The smple ws equilibrted t 40 C for 20 min in wter bth t slow stirring, nd 1 cm 50/30 µm (DVB/Crboxen/PDMS) SPME stble flex fiber (Supelco, Bellefonte, PA, U.S.A.) ws inserted into the vil nd exposed for 30 min. The fiber ws then inserted into the injector port of the GC t 220 C for 5 min to llow the voltiles to be desorbed. Gs Chromtogrphy-FID/ Olfctometry Arom intensity of voltile compounds ws nlyzed using n HP-6890A GC (Hewlett-Pckrd Inc., Plo Alto, CA, U.SA.) equipped with sniffing port. Smples were run on polr column (DB-wx, 30 m x 0.32 mm i.d. x 0.5 m; J&W Scientific, Folsom, CA, U.S.A.). The oven temperture ws progrmmed from 40 to 240 C t 7 C min -1, with holding time of 5 min t the mximum temperture. Helium ws used s crrier gs t flow rte of 2 ml min -1. Injector nd detector tempertures were 200 nd 290 C respectively. The column effluent ws split with one-third flow directed to FID nd the other two-thirds to the humidified olfctory port for sniffing. Two trined ssessors evluted ech smple in duplicte. A time- intensity method using liner potentiometer s previously described (Bzemore nd others 1999) ws used to record 66

67 their sensory impressions (retention times, rom descriptors nd intensities). Olfctory ssessor nd FID responses were seprtely recorded nd integrted using two chnnels with ChromPerfect (Justice Lbs, Melbourne, FL, U.S.A.) softwre. Intensity of ech odor-ctive compound ws expressed s normlized pek intensity. Pek normliztion, consisted in doing n verge of ll intensities of ll odor ctive compounds for ech smple nd then normlize to vlue of 10 given for the highest pek mong ll smples (pek of octdien-3-one, 1,5(z) in deerted juice fter 60 dys of storge). A pek ws considered to be rom-ctive if detected 2 times or more out of totl four responses for ech smple. Arom ctive compounds were identified by mtching odor descriptors with liner retention index (LRI) vlues on GC-O nd GC-MS. Confirmtion of identity ws done by comprison with those reported on the literture. Gs Chromtogrphy - Mss Spectrometry (GC-MS). Arom ctive compounds were identified using GC-MS. A Perkin-Elemer Clrus 500 qudrupole mss spectrometer with Turbo Mss softwre version (Perkin- Elmer, Shelton, CT, U.S.A.) polr column (Stbilwx; 60 m x 0.25 mm i.d. x 0.5 m film thickness, Restek, Bellefonte, PA, U.S.A.) ws used. Helium ws used s crrier gs in constnt flow mode of 2 ml min. Source nd injection port tempertures were mintined t 200 C nd the trnsfer line temperture ws 260 C. The oven ws progrmmed from 40 to 240 C t 7 C min. Electron impct ioniztion (70 ev) ws operted in the totl ion current mode. Helium ws used s crrier gs t flow rte of 1.5 ml min -1. Mss spectr mtches were mde by comprison of NIST 2005 version 2.0 stndrd spectr (NIST, Githersburg, MD, U.S.A.). Only compounds with spectrl fit vlues > 900 were considered positive identifiction. Smples were run in triplicte, 67

68 ech replicte ws collected from different SPME hedspce smple. Identity of compounds ws verified by the liner retention index (LRI) vlues, dtbse mtching nd by comprison with those reported on the literture (Acree nd Arn 2004; Rouseff 2008). Rte of Rection Zero, 0.5, Pseudo first nd second order rte model were pplied to DO dt. Sttisticl Anlysis One-wy ANOVA nlysis ws employed to determine significnt differences in dissolved oxygen nd vitmin C (AA nd DHA). A two-wy ANOVA ws used to find significnt differences in voltile compounds between the different tretments (α =0.05). PCA nlysis ws performed for voltile compounds using MATLAB. Results nd Discussion Control not-from concentrte ornge juice hd ph of , Brix, pulp nd cidity percentge. First, second, third nd fourth portions will be referred s deerted, oxygen sturted, control nd ir hedspce ornge juice respectively. Microbil Anlysis No microbil growth ws detected in the NFC ornge juice during the 60 dys of storge for ll the tretments except deerted ornge juice tht hd finl microbil count of1.5, 1.8 nd 1.8 log CFU ml -1 for erobic plte counts (PCA), cidophilic bcteris (OSA) nd molds nd yest (PDA) respectively s Tble 2-1 shows. The microbil popultion in deerted ornge juice ws below the spoilge level of 5 log CFU ml -1 nd it ws not considered to ffect rom, voltiles, vitmin C or color of the ornge juice. 68

69 Dissolved Oxygen (DO) Since smple preprtion (gs sprging nd pouring of vils) required one dy, DO mesurement strted on dy one of storge. At dy one DO concentrtion in deerted ornge juice ws reduced by 76% with respect to control. In oxygen sturted ornge juice, DO concentrtion incresed 278% compred with control. As shown in Figure 2-1, deerted, control nd oxygen sturted juices reched the lowest constnt DO concentrtion of round M from dy 17 until the end of the storge period. A similr fst reduction of DO concentrtion of seven dys ws observed in lemon juice stored t 36 C in glss continers tightly seled nd without hedspce (Robertson nd Smniego 1986). In this study, for smples with ir hedspce DO concentrtion in the ornge juice remined between M until dy 38 nd therefter until reching 35 M, except for dy 51 when DO content ws higher. There is no resonble explntion other thn mnipultion error nd thus cn be considered n outlier (Figure 2-1). These results suggest tht the presence of ir in the hedspce hs more effect in the DO concentrtion thn bubbling oxygen, possibly due to the equilibrium of oxygen reched in the system ccording to Henry s lw (Ringblom 2004). Thus, s oxygen is consumed in the ornge juice it is replenished with oxygen from ir in the hedspce to rech equilibrium. This equilibrium cn be kept s long s the oxygen supply of the hedspce lsts. Following the idel gs lw, the mount of oxygen present in the hedspce t 5 C ws estimted s 9.21x10-5 moles tht re pproximtely 10 times higher thn 7.98x10-6 moles in oxygen sturted ornge juice. From the kinetic models pplied to DO dt, pseudo first nd second order hd the best fit (Figure 2-2). Dt ws fitted to first nd second order models Second-order model 69

70 didn t djust well to oxygen sturted ornge juice. A second order model tht ccounts for dissolved oxygen nd scorbic cid described better AA oxidtion (Singh nd others 1976). However in our experiments AA ws in excess. The pseudo first-order model pplies when one of the rectnts is present in excess over the other nd the rte of rection is proportionl to the concentrtion of the rectnt not in excess (Dissolved oxygen). In second order rection, the rte of rection is proportionl to the squre of the concentrtion of one of the rectnts (Dissolved oxygen). First order rte constnt vlues of 1.3 x 10-1 ± 2.3 x 10-2, 2.8 x 10-1 ± 3.8 x 10-2 nd 2.8 x 10-1 ±4.1 x 10-2 d -1 were obtined for deerted, control nd oxygen sturted juices. It ws expected to hve the sme rection rte for ll the juices since initil DO should not ffect the rte rection nd the lower rte rection for deerted juice is ttributed to hving insufficient dt points vilble for the rte rection clcultions. These vlues were bout double the vlue of rte constnts of DO consumption clculted in Chpter 3 for NFC ornge juice from Hmlin ornges stored t 5 C (Tble 3-2). This difference in rte constnt vlues ws ttributed to different processing conditions nd pulp content, becuse the screened pulp content of the juice used on this study ws 0% compred to 5 % screened pulp in the juice used for Chpter 3. Vitmin C On dy 1, AA concentrtion in oxygen sturted ornge juice ws significntly ( =0.05) lower thn tht of the other smples (Figure 2-3A) tht corresponded to the mximum initil DO concentrtion. This result suggests AA oxidtion. On dy 17, ir hedspce ornge juice reched AA concentrtion s low s oxygen sturted juice (Figure 2-3A). Finlly, on dy 60, smples with ir hedspce showed the lowest AA 70

71 concentrtion (Figure 2-3A). DHA concentrtion ws higher in the smples with more oxidized AA (oxygen sturted nd ir hedspce) during the first 30 d of storge (Figure 2-3B). Although for oxygen-sturted ornge juice DHA concentrtion hd constnt decrese over the time. After 60 d of storge, ornge juice with ir hedspce lost the most AA (42%). Our results greed with Beltrán-González nd others (2008) tht observed tht higher oxygen content in hedspce correlted with higher reduction in AA concentrtion in mndrin juice. However, in the bsence of ir in the hedspce, DO cn oxidize only smll mount of AA since the DO-to-AA proportion ws 0.25 (495 µm DO/1940 µm AA). In this experiment the initil DO concentrtion rnged from 495 to 42 µm (Figure 2-1) wheres scorbic cid rnged from 1940 to 1282 µm (Figure 2-3A) for ll the juices t dy one of storge nd pproximtely 13% of AA ws degrded by dy 60 for smples without hedspce independent of the initil DO content. In the cse of ir hedspce smples, the DO/AA rtio (0.1) ws bout the sme during most of the storge. However, t dy 60 it dropped to less thn hlf (0.04) becuse DO reched its lowest level (35 µm). Pseudo first rte constnt of AA oxidtion ws 1.9x10-2 ± 3.0x10-3 d -1. Finlly besides deertion, voiding ir in hedspce of ornge juice continers is criticl to prevent AA degrdtion. Color No chnge in color ws visully detected mong tretments nd over time. L, nd b vlues pper in Tble 2-4. L nd prmeters did not chnge over the time, b prmeter chnged slightly from yellow to blue. 71

72 Arom Active Compounds The proportion of rom ctive compounds in NFC ornge juice t dy one of storge is shown in Tble 2-5. Aldehydes hd the highest impct in rom profile (29 51%) for ll the tretments nd in deerted ornge juice, terpenes were similrly importnt (24%). About 12 to 23% of rom ctive compounds could not be identified becuse only polr column ws used (DBWx). The use of non polr column (DB5) in ddition to polr column will llow to identify polr nd non polr compounds in the juices. Our results gree with other reserchers tht observed tht ldehydes hve n importnt contribution to ornge juice rom (Perez-Ccho nd Rouseff 2008). Limonene ws present in ornge juice t the highest mount but the pek could not be integrted becuse it ws out of the chromtogrm scle, this compound is not rom ctive nd its function is only s crrier of rom ctive compounds (Brddock 1999). Twenty-four rom ctive compounds were detected using GC-O nlysis (Tbles 2-6 nd 2-7). Tble 2-6 shows the rom ctive compounds identified in the NFC ornge juice t dy one of storge t 5 C. Nonnl described s piney nd herbl hd the min contribution to rom intensity of control, oxygen sturted nd ir hedspce juices presenting normlized pek height of 8.2, 9.5 nd 9.1 for ech juice respectively. Figure 2-4 shows the rom ctive compounds orgnized by rom descriptor. By comprison with control ornge juice, deerted ornge juice cooked descriptor ws the 34% of rom profile compred to 17% in control ornge juice (Figure 2-4A nd B) this corresponded to pek intensity of 8.8 nd 3.5 in deerted nd control ornge juice for n unknown compound with cooked rice descriptor. Oxygen sturted juice rom profile ws 44% s herbl piney compred with 33% for control (Figure 2-4C nd B) due to the higher intensity of nonnl nd -terpinolene in 72

73 the oxygen sturted juice of 9.5 nd 9.2 compred with 7.2 nd 7.3 for control respectively. Cooked note in ornge juice with ir in the hedspce ws 26% nd in control 17%. This difference is due to methionl being perceived s 7.5 of pek intensity in juice stored with ir hedspce but not perceived in control juice. (Tble 2-6). Methionl ws lso detected in deerted nd oxygen sturted ornge juice with pek intensity of 5.7 nd 7.2. The presence of methionl in oxygen sturted nd ir hedspce ornge juice is explined by the ccumultion of DHA vi AA oxidtion tht cn promote the formtion of methionl from methionine (Perez-Ccho nd Rouseff 2008). But the presence of methionl in deerted ornge juice ws not expected since deertion removed the oxygen to prevent oxidtion of AA. Voltile Compounds Alcohols, ldehydes nd terpenes were the voltile compounds detected t the highest percentge of 11-17, 9-13 nd 6-10% respectively, for ll the totl voltile compounds in ll the smples. Esters nd ketones constituted less thn 5% of the totl voltile compounds (Tble 2-5). The proportion of voltile compounds in this study is under 17, 39, 14, 19 nd 11% of lcohols, ldehydes, terpenes, esters nd ketones respectively, reported from consensus of 36 rom ctive compounds in fresh squeezed ornge juice (Perez-Ccho nd Rouseff 2008b). Deertion, either by gs sprging or vcuum heting, is known to remove the most voltile lcohols, ldehydes nd terpene hydrocrbons (Ringblom 2004). Jordn nd others (2003) observed significnt losses of 50% or more in voltile compounds of fresh squeezed ornge juice fter deertion. On this study, none of the detected compounds decresed significntly in content fter 10 min of either nitrogen or oxygen 73

74 bubbling (Tble 2-7) probbly becuse n importnt portion of voltiles ws lost during the previous storge of the juice t the juice processing plnt. Crvone ws significntly lower in control ornge juice but similr in deerted nd oxygen sturted juices, since crvone is product of limonene oxidtion it ws expected tht this compound content be higher in oxygen sturted juice nd lower in deerted juice with respect to control. From the compounds with significnt differences t dy one of storge t 5 C, -terpinene nd cis -terpineol were detected in ll juices except in control juice. -elemene, -selinene nd 3-crene hd the highest mount in deerted ornge juice compred with the other juices. Acetldehyde, sbinene, - cymene, -terpinolene nd limonene oxide were detected in deerted nd control juices but not in oxygen sturted nd ir hedspce juices. At dy 60 of storge the content of most of the compounds decresed with respect to dy 0 (Tble 2-8) except for E-2-hexenl nd limonene oxide whose content incresed in ir hedspce ornge juice nd 1-hexnol tht incresed t dy 60 of storge in deerted ornge juice. Principl Component Anlysis (PCA) pplied to the voltile compounds, explined the 63% of vrition (Figures 2-5 nd 2-6) mong the juices. It ws possible to differentite between deerted nd hedspce ornge juice t dy one nd 60 of storge s indicted by seprted well defined clusters. Control nd deerted juices t dy one of storge grouped together. According to the PCA loding plot (Figure 2-6) ir hedspce ornge juice t dy one hd significnt higher mounts of nonnl, octnl nd decnl from ir hedspce juice t dy 60 of storge. Crvone ws present in significnt higher mounts in oxygen sturted nd ir hedspce ornge juices compred with deerted nd control juices t dy one of storge. Acetldehyde ws 74

75 present only in deerted nd control ornge juice nd ws the only significntly different s indicted with n sterisk in Figure 2-6. These trends could be used s reference for future studies, lthough it should be kept in mind tht other rections such s cid-ctlyzed hydrtion my hve greter impct thn oxygen in chnges of ornge juice rom (Shw nd others 1993). Exmples of compounds formed by cidctlyzed hydrtion rections re -terpineol, cis nd trns-1,8-p methnediol (Mrshll nd others 1986). Conclusions At low storge temperture (5 C) nd 60 dys of storge DO did not produce chnges in color tht could be visully perceived. The mount of DO present in ornge juice, even t sturted levels, only ccounted for smll consumption of AA t the beginning of the storge. And the presence of ir in hedspce of continers cused detectble AA degrdtion. From the rom ctive compounds, methionl ws perceived t higher intensity in ir hedspce ornge juice with respect to control only nd nonnl hd the min contribution to rom intensity of control, oxygen sturted nd ir hespce ornge juices. At dy one of storge -elemene, -selinene nd 3- crene hd the highest mounts in deerted ornge juice compred with the other juices. At dy 60 of storge, the content of most of the compounds decresed with respect to dy 0 except for E-2-hexenl nd limonene oxide whose content incresed in ir hedspce ornge juice. 75

76 Tble 2-1. Microbil counts in NFC ornge juice during 60 dys of storge t 5 C Deerted Control Oxygen sturted Air hedspce Dy PCA OSA b PDA c PCA OSA PDA PCA OSA PDA PCA OSA PD A (log CFU ml -1 ) ±0.7 ±0.7 ± ±0.8 ±0.5 ± erobic plte counts, b cidophilic bcteris, c molds nd yest, -not detected 76

77 Tble 2-2. Pseudo first nd second order rte of DO consumption in not-from concentrted ornge juice during storge t 5 C dt points used Pseudo first order k (d -1 ) Second order k (µm d) -1 Deerted 3 1.3x10-1 ± 5.2x10-2 r 2 = x10-1 ± 2.3x x10-3 ± 9.4x10-4 r 2 =0.94 r 2 =0.95 Control 4 2.8x10-1 ± 3.8x x10-3 ± 7.9x10-4 Oxygen Sturted r 2 =0.96 r 2 = x10-1 ±4.1x x10-3 ± 4.5x10-4 r 2 =0.96 r 2 =0.92 Tble 2-3. Pseudo first rte constnt of AA oxidtion nd DHA formtion in not-from concentrted ornge juice during storge t 5 C Pseudo first order Air hedspce dt points used k (d -1 ) AA 5 1.7x10-2 ± 4.7x10-3 r 2 = x10-2 ± 3.0x10-3 r 2 =0.91 DHA 3 2.3x10-2 ± 7.7x10-3 r 2 = x10-2 ± 3.4x10-3 r 2 =0.94 Tble 2-4. Color chnge in NFC ornge juice during storge t 40 C nd expressed s L, nd b with respect to control ornge juice. Deerted Oxygen sturted Air hedspce Time (d) L b L b L b

78 Tble 2-5. Proportion of rom ctive nd voltile compounds in NFC ornge juice t dy 1 of storge t 5 C Arom ctive compounds % pek height Deerted Control Oxygen sturted Air hedspce lcohols ldehydes ketones terpenes* esters unknowns Voltile compounds % pek re Deerted Control Oxygen sturted Air hedspce lcohols ldehydes ketones terpenes esters Without limonene 78

79 Tble 2-6. Arom ctive compounds identified in commercil NFC ornge juice with different levels of DO fter 1 dy of storge t 5 C. Vlues re verge of normlized pek height ± stndrd error. Different letters indictes significnt difference (Duncn s test =0.05) Normlized pek height LRI Compound Descriptor seinene grins, musty 1706 unknown 1659 unknown grins,musty, cooked pottoe 1192 sbinene minty,musty 1284 unknown cooked egg rosted pecn, medicine, egg yolk, cooked rice 1443 methionl cooked pottoe 1805 decdienl (E,E) 2,4 cooked pottoe, cooked 1527 decnl 1228 ethyl hexnote fruity cooked,mushroom, cooked, melon 1540 linlool fruity, bubble gum 1033 ethylbutnote fruity, bubble gum, tutti fruity Deerted 7.5 ± ±1.8b 8.8 ± ± ± ±2.1 Control Oxygen sturted 0.0b 3.7 ±0.3b Air hedspce 0.0b 0.0b 0.0b 6.3 ± ±1.4b 5.2 ± b 4.7 ± ± ± ± ± ±0.4b 0.0b 4.0 ± b 0.0b 0.0b 7.2 ± b 7.2 ± ± ± ± ± ± b 6.6 ± b 0.0b 7.0 ± ± ± ± b 7.8 ±0.6 79

80 Tble 2-6. Continued LRI Compound Descriptor 1332 ethyl heptnote fruity, sweet, 1051 unknown fruity, florl, citrus 998 cetldehyde 1817 nerol ornge peel, fruity, cooked ornge citrus lemonde, citrus, green 1366 nonnl herbl, piney, 1489 nonenl (z)-2 herbl, ftty, medicine, piney, terpineol herbl, green 1076 hexnl herbl, green, grpe octnol herbl, musty, florl,sopy terpinolene herbl, piney 1848 Perill ldehyde herbl,lemon grss, herbl 1383 unknown plstic, herbl DBWx Normlized pek height Deerted Oxygen Control sturted b 5.6 ± ± ± ±1.0b 3.8 ±1.4 ± ±1.7b 4.3 ± b 0.0b 5.8 ± ±1.8 ±1.5 ± ± b 8.5 ± b 0.0b 4.5 ± ± b 5.0 ± ± ±1.8 ± ± b 6.1 ± ± b 6.7 ±0.7 Air hedspce 5.7 ± b 6.0 ± b 9.1 ± b 0.0b 6.7 ± b 0.0b 9.2 ± b ±0.7 ± b 80

81 Tble 2-7. Voltile compounds identified by GC-MS in NFC ornge juice with different levels of DO fter 1 dy of storge t 5 C. Vlues re verge of normlized pek re ± stndrd error. Different letters indictes significnt differences (Duncn s test =0.05) LRI Compound Deerted b Control b sturted b Oxygen Aldehydes (8) 719 Acetldehyde 6.50E-01 ±2.90E Hexnl 8.66E-02 ±1.94E E-2-hexenl 2.76E-02 ±8.49E Octnl 3.01E-01 ±8.16E Nonnl 5.87E-02 ±1.23E E-2-octenl 2.74E-03 ±3.87E Decnl 1.47E-01 ±2.56E Perillldehyde 2.29E-02 ±2.95E-03 Esters (5) 1048 Ethylbutnote 1.12E-01 ±4.08E E-01 ±6.09E E-02 ±6.94E-03 Air Hedspce b - c E-02 ±1.06E E-01 ±1.05E E-01 ±1.42E E-02 ±1.09E E-01 ±7.18E E-02 ±8.80E E-01 ±1.06E methyl hexnote E-03 ±7.50E ethyl hexnote 3.05E E-02 ±5.13E-03 ±7.85E Ethyloctnote 2.96E-03 ±2.64E ethyl 3- hydroxyhexnote 5.74E-02 ±7.70E-03 Terpenes (17) pinene 2.51E-02 ±7.37E β-myrcene 2.85E-01 ±1.09E E-01 ±1.76E E-02 ±6.68E E-03 ±4.90E E-02 ±1.50E E-02 ±9.26E E-02 ±8.23E E-01 ±2.74E E-02 ±4.20E E-01 ±1.41E E-02 ±2.06E E-02 ±6.13E E-02 ±4.79E E-02 ±5.16E E-02 ±5.86E E-02 ±3.73E E-01 ±2.37E E-02 ±1.57E E-03 ±5.23E E-01 ±3.31E E-02 ±2.42E E-02 ±6.97E E-01 ±4.86E-02 81

82 Tble 2-7. Continued LRI compound Deerted Control sturted b Air Hedspce b b b Oxygen crene 8.86E ±5.48E terpinene 1.04E-02 ±4.58E E-03 ±4.38E E-03 ±1.99E humulene 7.90E Limonene 6.11± Sbinene 3.54E-02 ±1.20E cymene 9.67E-03 ±5.52E terpinolene 6.09E-03 ±2.46E elemene 3.27E-03 ±9.18E limonene oxide 3.46E-02 ±7.19E Vlencene 5.09E-01 ±1.32E selinene 3.74E-02 ±7.41E Crvone 8.04E-02 ±1.28E pnsinsen 3.27E-02 ±8.04E-03 Alcohol terpenes(6) 1555 Linlool 4.79E-01 ±1.04E cis terpineol 3.64E-03 ±7.12E terpineol 2.54E-01 ±3.98E-02 ±1.82E ± E-02 ±4.49E E-03 ±1.19E E-03 ±3.19E ± ± E-02 ±2.97E E-01 ±1.55E E-02 ±1.50E-04 b 3.94E-02 ±3.93E-03 b 2.20E-02 ±4.64E E-01 ±3.69E E-01 ±1.64E E-01 ±3.84E E-02 ±3.56E-03 b 6.46E-02 ±5.09E-03 b 1.41E-02 ±1.60E E-01 ±5.14E E-03 ±1.76E E-01 ±2.86E E-01 ±2.70E E-02 ±2.72E-04 b 7.33E-02 ±4.71E E-02 ±2.37E E-01 ±4.34E E-03 ±2.65E E-01 ±1.30E-02 82

83 Tble 2-7. Continued LRI compound Deerted b Control b sturted b Oxygen terpineol 8.19E E E-02 ±1.30E-02 ±4.65E-03 ±7.92E cis/trns 1-crveol 2.06E-02 ±4.13E citronellol 3.38E-03 ±3.38E-03 b Alcohols(6) 949 Ethnol 2.09E-01 ±6.61E heptnol 3.11E-03 ±1.46E Octnol 3.95E-02 ±7.75E nonnol 6.17E-03 ±8.15E Nerol 9.00E-03 ±1.95E E-02 ±1.54E E-03 ±4.76E-04 b 1.81E-01 ±2.64E E-03 ±1.30E E-02 ±3.77E E-03 ±5.02E E-03 ±2.06E E-02 ±1.96E E-03 ±1.13E-03 b 1.11E-01 ±7.91E E-03 ±2.03E E-02 ±4.86E E-03 ±1.62E E-03 ±4.55E-05 DBWx, b verge ± std error, c Not detected, tenttively identified Air Hedspce b 1.05E-01 ±5.57E E-02 ±4.57E E-03 ±3.89E E-01 ±3.45E E-03 ±1.07E E-02 ±4.42E E-03 ±4.90E E-02 ±7.70E-04 83

84 Tble 2-8. Voltile compounds identified by GC-MS in NFC ornge juice with different levels of DO fter 60 dys of storge t 5 C. Vlues re verge of normlized pek re ± stndrd error. Different letters indictes significnt difference (Duncn test =0.05) LRI Compound Deerted b Air hedspce b Aldehydes (7) nd Ketones (1) 719 cetldehyde - c hexnl E-02 ±1.23E E-2-hexenl E-02 ±1.06E octnl 2.43E-02 ±6.35E E-02 ±4.80E nonnl 1.11E-02 ±2.79E decnl 4.48E-02 ±1.28E E-02 ±2.23E Perill ldehyde E-02 ±4.18E E E-03 ±3.14E-03 ±2.06E methyl-5-hepten-2-one Esters (3) 1048 ethylbutnote 1239 ethyl hexnote 3.86E-02 ±7.10E E-03 ±5.32E E-02 ±2.98E E-02 ±2.74E E-02 ±1.00E ethyl 3-hydroxyhexnote Terpenes (11) 8.15E E β-myrcene ±3.94E-02 ±8.86E E crene ±1.10E E limonene ±1.07 ±1.39E E E sbinene ±6.59E-03 ±5.69E E E vlencene ±5.04E-02 ±1.54E-03b 1.02E selinene ±4.54E E crvone - ±2.14E-03 84

85 Tble 2-8. Continued LRI Compound Deerted b Air hedspce b pnsinsen 8.34E-03 ±2.93E limonene oxide E-02 ±9.05E-04 Alcohol terpenes(6) 1555 linlool 3.31E-01 ±2.10E E-01 ±1.56E terpineol 1.78E-01 ±2.22E E-01 ±9.94E terpineol 5.89E-02 ±2.91E E-02 ±2.18E cis/trns 1-crveol E-02 ±1.49E-03 Alcohols(2) 949 ethnol 3.45E-01 ±2.45E E-01 ±1.19E hexnol 2.36E-02 ±3.25E nerol E-03 ±3.96E octnol 2.42E-02 ±2.98E E-02 ±1.79E-04 DBWx, b verge + std error, c Not detected 85

86 Figure 2-1. Dissolved oxygen (DO) concentrtion in NFC ornge juice during 60 dys of storge t 5 C. ( ) deerted, ( ) control, ( ) oxygen sturted, ( ) ir hedspce 86

87 Figure 2-2. Kinetic models fitted to dissolved oxygen (DO) concentrtion in NFC ornge juice during 60 dys of storge t 5 C 87

88 Figure 2-3. Chnges in (A) scorbic cid nd (B) dehydroscorbic cid in NFC ornge juice during 60 dys of storge t 5 C. ( ) deerted, ( ) control, ( ) oxygen sturted, ( ) ir hedspce 88

89 Figure 2-4. Arom ctive compounds orgnized by rom descriptor in NFC ornge juice t dy 1 of storge t 5 C. (A) deerted, (B) control, (C) oxygen sturted nd (D) ir hedspce ornge juices. 89

90 Figure 2-5. PCA score plot of voltile compounds in ornge juice t different DO concentrtions t dy 1 nd 60 of storge t 5 C ( deerted, control, oxygen sturted nd ir hedspce) 90

91 Figure 2-6. Loding plot of voltile compounds in ornge juice t different DO concentrtions t dy 1 nd 60 of storge t 5 C. *significntly different. 91

92 CHAPTER 3 IMPACT OF DISSOLVED OXYGEN LEVEL AND TEMPERATURE ON VOLATILE COMPOUNDS, VITAMIN C AND COLOR IN PASTEURIZED NOT-FROM- CONCENTRATE ORANGE JUICE Introduction Deertion of ornge juice in citrus industry is pplied in the production of notfrom-concentrte ornge juice before psteuriztion with the purpose of removing oxygen nd essentil oil (Brddock 1999; Ringblom 2004). The importnce of the deertion step to the citrus industry hs relied on the ssumption tht if oxygen (21% in ir) is present during psteuriztion nd storge, it cuses scorbic cid degrdtion, s well s chnges in rom nd color (Ringblom 2004). The min disdvntge of deertion is tht voltile compounds tht give the fresh-like rom to the juice re lso removed (Jordn nd others 2003) during this step. Ascorbic cid (AA) degrdtion of citrus juices in the presence of oxygen nd during storge hs been described under different initil dissolved oxygen (DO) concentrtions (Kefford nd others 1959; Robertson nd Smniego 1986), storge tempertures (Brddock nd Sdler 1989) nd pckging mterils (Kennedy nd others 1992; Zerdin nd others 2003; Ros- Chumills nd others 2007). AA degrdtion in ornge juice hs been described s first-order for concentrted ornge juice (Burdurlu nd others 2006); for membrneclrified concentrted ornge juice; for fresh squeezed ornge juice (Dhuique-Myer nd others 2007); s well s zero- nd first-order, for deerted nd non-deerted single strength from-concentrted ornge juice respectively, (Sores nd Hotchkiss 1999). Voltile compounds in ornge juice thought to be product of oxidtion include crvone nd crveol from limonene oxidtion (Kutty nd others 1994) nd methionl 92

93 from dehydroscorbic cid prticiption in Strecker degrdtion of methionine (Perez- Ccho nd Rouseff 2008). Time temperture effects on rom of ornge juice during storge cn be ssessed following -terpineol production due to linlool nd limonene degrdtion by cid-ctlyzed hydrtion-dehydrtion pthwy (ACHD) nd not oxidtion (Mrcotte nd others 1998). The influence of DO in non-enzymtic browning hs been suggested s consequence of AA oxidtion but no significnt effects on color of ornge juice hs been observed (Rssis nd Sguy 1995). Oxidtion rections cn be ccelerted in the juice depending on storge nd pckging conditions. Previous studies hve focused on individul compounds (such s vitmin C) tht re ffected by time, pckging mteril, or storge conditions insted of doing comprehensive studies. There is lck of informtion on the single effect of dissolved oxygen in vitmin C degrdtion, excluding ny other vribles such s pckging mteril. Therefore, to improve processing conditions, it is importnt to better understnd the mechnisms of how oxygen deteriortes ornge juice qulity. The objective of this study ws to elucidte the chnges in vitmin C, rom nd color of psteurized not-from-concentrte (NFC) ornge juice with selected initil levels of DO during storge t selected tempertures without hedspce. We lso compred the chnges in AA, color nd rom in the presence of ir hedspce under the sme storge conditions. Mterils nd Methods Ornge Juice Preprtion nd Chrcteriztion Ornges (Hmlin vr.) were obtined from CREC groves (Lke Alfred, FL) nd extrcted using Fresh n Squeeze citrus juicer (JBT Food Tech, Lkelnd, FL). Soluble solids content (SSC), ph, titrtble cidity, suspended nd screened pulp 93

94 content were mesured in ornge juice. SSC were mesured using digitl refrctometer (Abbe Mrk II, Reichert, Depew, NY), ph using ph meter (Orion 5 Str,Thermo Fisher Scientific, Beverly, MA), titrtble cidity (TA) by titrtion with NOH 0.1 N until reching ph of 8.2 nd using n utomtic titrtor (Schott, Elmsford, NY). Suspended pulp ws mesured by centrifugtion of 50 ml of ornge juice during 10 min t 365 x g nd clcultions were done s percentge of volume of pulp precipitted per totl volume of centrifuged smple. Screened pulp ws mesured by pssing 500 ml of ornge juice through 20 mesh screen using FMC FoodTech quick fiber device (JBT Food Tech, Lkelnd, FL) tht shkes the screen during 2 min, until the pulp is free of excess of juice. Results re expressed s weight of pulp recovered in the screen by volume of juice used for the test (FMC 2002). Smple Preprtion nd Storge Study Ornge juice ws divided into four portions: A nitrogen tmosphere ws creted by filling vcuum-emptied glove bg ( hnds in-bg, Spilfyter,Green By, WI) with nitrogen. The first smple portion ws bubbled with industril grde nitrogen (Prxir, Inc. Dnbury, CT) in the nitrogen tmosphere; the second portion ws bubbled with ultr high purity oxygen (Prxir, Inc. Dnbury, CT); the third nd fourth portions were not bubbled. Gs-bubbling consisted of pssing the gs t flow of 1.5 L min -1 for 10 min through 700 ml of ornge juice in 1-L gs wshing bottle (Chemglss, Vinelnd, NJ). Gs flow ws regulted with vlve nd time monitored with timer. All ornge juice portions were psteurized in seprte btches t 90 C for 24 s using HTST psteurizer (Microthermics, Rleigh, NC). Ornge juice portions were poured into 30 ml mber vils, one per smpling time. No hedspce ws left in vils filled with juice from the first, second, nd third portions. An ir hedspce equivlent to one-third (10 ml) of 94

95 continer totl volume (30 ml) ws left in vils filled with juice from the fourth portion. Figure 3-1 shows the juices preprtion process. For juices bubbled with nitrogen, vils were filled in nitrogen gs tmosphere. Control, oxygen sturted nd ir hedspce juices were filled in lminr hood. All juice portions were poured to sterile glss beker from the glss bottle used to collect the juice from psteurizer, nd then mber vils were completely submerged into the beker nd cpped submerged to void ir bubbles. Glss mber vils were stored under regulr tmospheric conditions (ir) t 5, 13, 21.5, 30.5 nd 40 C during 60, 12, 12, 6 nd 6 dys, respectively. Only two replictes of ech tretment were prepred from the sme fruit btch becuse Hmlin vr. ornges were not vilble to prepre more juice lter in the seson. DO nd AA were mesured t the times indicted in Tble 3-1. Becuse color chnges were detected only in ir hedspce smple t 40 C nd voltile compounds t 40 C did not show significnt difference. Color nd voltile compounds t storge tempertures lower thn 40 C were not mesured. Microbil Anlysis In order to ssure the ornge juice shelf life, especilly t intermedite tempertures (13 to 30.5 C) fruit ws properly wshed, the extrctor ws snitized ccording to mnufcturer instructions, ll the continers nd utensils used to mnipulte nd store the juice were sterilized, nd the psteurizer ws clened using CIP protocol before nd fter processing the juice. Ornges were extrcted in clen room nd juice ws poured in finl continers for storge under lminr flow hood to minimize product contmintion becuse in preliminry tests we observed some contmintion. The clen in plce (CIP) protocol used ws performed t set flow of 1.4 L min -1 nd the following solutions were pssed through the psteurizer 95

96 1. 20 min wshing with 2% NOH solution t 80 C min wter rinsing t 80 C min 1% phosphoric cid t 30 C min wter rinsing t 30 C Microbil popultion ws quntified in ll juices during storge to insure tht observed chnges in DO nd voltile compounds were not the result of microbiologicl ctivity. Microbil nlysis ws performed t dys 1, 6, 15, 30 nd 60 t 5 C; t dy 1, 6, 12 t 13 nd 21.5 C, t dy 1 nd 6 t 30.5 C; t dy 1 nd 3 t 40 C. Aerobic plte counts were conducted using plte count gr (APCA), cidophilic bcteri ssocited with spoilge of citrus products nd molds, yests were quntified using ornge serum gr (OSA) nd cidified potto dextrose gr (APDA), respectively. Smples were serilly diluted with sterile 0.1% peptone s needed nd poured in the pltes by duplicte. PCA, OSA, PDA pltes were incubted 24 h t 35 C, 24 h t 30 C nd 2 dys t 25 C, respectively. DO Mesurements DO concentrtion ws mesured under nitrogen tmosphere using n electrochemicl probe model Orion D nd ph meter model Orion 5 Str (Thermo Fisher Scientific, Beverly, MA) tht ws inserted in the vils contining ornge juice then the cp ws tightly ttched to void entry of ir. Constnt stirring ws mintined during the mesurements nd DO concentrtion vlues were collected digitlly using computer progrm written in LbVIEW 7.1 (Ntionl Instruments, Austin, TX). Clibrtion of the probe nd mesurement of DO ws done t the sme s the storge temperture (5-40 C) for ech smple. A stndrd curve ws generted by correltion of the DO probe response expressed in na vs. DO concentrtion expressed in µm of n 96

97 oxygen sturted nd n oxygen free stndrds t the sme temperture. The oxygen sturted stndrd ws ir sturted wter whose DO concentrtion hve been reported by Lewis (2006b) for different pressure nd temperture conditions. Oxygen free stndrd ws 13% glucose solution with 0.05% of the enzyme glucose oxidse tht ws llowed to consume oxygen during 5 min before mesuring oxygen nd DO concentrtion ws ssumed to be 0 µm. The use of smll mount of enzyme ws enough to decrese DO content. Use of glucose oxidse insted of sodium sorbte to prepre the 0% DO stndrd hd the dvntge of not fouling the probe membrne. Vitmin C Mesurement Vitmin C ws quntified s scorbic (AA) nd dehydroscorbic (DHA) cid. A cpillry electrophoresis (CE) system used ws model P/ACE MDQ with DAD nd Krt 32 version 5.0 softwre from Beckmn Coulter (Fullerton, CA). The CE system ws run t 21 kv, 23 C, detection t 263 nm nd 35 mm sodium borte with 5% (v/v) cetonitrile s bckground electrolyte running buffer ccording to the method described by Cnclon (2001). The cpillry ws bre fused-silic from Polymicro Technologies (Phoenix, AZ) 50 µm id, 56 cm totl length. A clibrtion curve ws done with stndrds contining AA concentrtions of 284, 852, 1419, 1987, 2555 nd 3123 µm. Smples were diluted (1:4) with 1000 mg L -1 EDTA solution nd filtered through 0.45 µm syringe filter. To determine DHA, L-dithiothreitol (DTT) ws dded t 13 mm (0.2%) finl concentrtion in smples in order to reduce DHA to AA. The DHA content of the smple ws clculted by subtrcting the initil AA content from the totl AA content fter the conversion. Tryptophn ws used s n internl stndrd t 200 mg L -1 finl concentrtion nd ws detected t 214 nm. 97

98 Color Mesurement A Minolt colorimeter model CR-400 (Konic Minolt, Jpn) ws used to mesure Hunter prmeters nd L,, b were clculted with respect to control psteurized ornge juice. Oil in Juice Assy Scott Oil method ws used to determine oil content in SSOJ nd consists of distilltion nd titrtion. To 25 g of SSOJ 25 ml of wter nd the sme mount of isopropnol re dded nd mixed in 300 ml round distilltion flsk. An ntifom gent nd boiling chips re dded to the mixture. The mixture is distillted until ll isopropnol crrying the oil, is recovered. Then 10 ml of 4 M HCl nd drop of 1% methyl ornge indictor re dded to the distillted smple. Titrtion to colorless endpoint using N potssium bromide-bromte solution is performed. Percentge of oil in the juice is clculted s (3-1) Voltile Compounds Anlysis Hedspce smpling Under nitrogen tmosphere, 10 ml of smple were plced into 40 ml glss vil contining stirring br nd tightly closed with screw cp with Teflon-coted septum. Ech smple ws equilibrted t 40 C for 20 min in wter bth with slow stirring, nd 1 cm 50/30 µm (DVB/Crboxen/PDMS) StbleFlex SPME fiber (Supelco, Bellefonte, PA) ws inserted into the vil nd exposed for 30 min. The fiber ws then inserted into the injector port of the GC t 220 C for 5 min to llow the voltiles to be desorbed. 98

99 Gs Chromtogrphy - Mss Spectrometry (GC-MS). Voltile compounds were identified using GC-MS. A Perkin-Elemer Clrus 500 qudrupole mss spectrometer with Turbo Mss softwre version (Perkin-Elmer, Shelton, CT) with polr column (Stbilwx; 60 m x 0.25 mm i.d. x 0.5 m film thickness, Restek, Bellefonte, PA). Helium ws used s crrier gs in constnt flow mode of 2 ml min -1. Source nd injection port tempertures were mintined t 200 C nd the trnsfer line temperture ws 260 C. The oven ws progrmmed from 40 to 240 C t 7 C min. Electron impct ioniztion (70 ev) ws operted in the totl ion current mode. Helium ws used s crrier gs t flow rte of 1.5 ml min -1. Mss spectr mtches were mde by comprison of NIST 2005 version 2.0 stndrd spectr (NIST, Githersburg, MD). Only compounds with spectrl fit vlues > 900 were considered positive identifiction. Smples were run in triplicte, ech replicte ws collected from different SPME hedspce smple. Identifiction of compounds ws verified by the liner retention index (LRI) vlues, GC-MS identifiction nd by comprison with those reported on the literture (Acree nd Arn 2004; Rouseff 2008). Pek res were normlized using 4-heptdecnone s internl stndrd. Pek re vlues of ech voltile compound were divided between the pek re of the internl stndrd for the sme replicte. Then verge nd stndrd error of the three replictes ws clculted. Rte of Rection Zero, 0.5, first nd second order rte model were pplied to DO dt. Sttisticl Anlysis Anlysis of vrince (ANOVA) of two wys nd Duncn tests t =0.05 using SAS softwre (Cry, NC) were performed to determine significnt differences in voltile compounds between the different tretments. 99

100 Results nd Discussion The ornge juice (cultivr Hmlin) used for the storge study hd ph of 3.8± 0.08, soluble solids content of 11 ± 0.09 Bx, 16.0 ± 2.0% suspended pulp, 6.0 ± 0. 5% screened pulp, nd 0.7 ± 0.02% titrtble cidity expressed s citric cid. Microbil Anlysis Microbil popultion in not-psteurized fresh squeezed ornge juice ws 2.2±0.07, 2.3± 0.02 nd 1.9 ± 0.03 log CFU ml -1 in erobic plte counts (APCA), molds nd yests (APDA) nd cidophilic bcteri (OSA), respectively. These vlues re round hlf of 4.5 log CFU ml -1 reported by Sdler nd others (1992) in non-psteurized fresh squeezed ornge juice. During 60 dys of storge t 5 C, microbil counts in APCA were lower thn 1 log CFU ml -1 for ll the juices; in APDA nd OSA microbil counts were between nd log CFU ml -1 respectively for dy 60 of storge. No microbil growth ws observed in ornge juice stored t 40 C but juice stored t 13, 21.5 nd C presented microbil popultions of 1.7 to 3 log CFU ml -1 from the first dy of storge reching counts of 5 log CFU ml -1 t dys 12 (13 nd 21.5 C) nd 6 (30.5 C). Due to juices stored t 13, 21.5 nd 30.5 C reching 5 log CFU ml -1 microbil popultions, considered s microbil spoilge, no further nlyses were performed onto these juices. DO Consumption DO dt fitted pseudo-first nd second order kinetic model (Figure 3-2), but only pseudo first order rte constnt ws clculted for ll the tempertures becuse DO consumption hs been modeled s first order with respect to oxygen(ahrne nd others 1997) nd second order kinetic model with respect to oxygen nd scorbic cid (Singh nd others 1976). Pseudo-first-order djusted well for control nd oxygen 100

101 sturted juices. However for deerted juices, not enough dt ws vilble to hve good djustment for the model (Figure 3-3). For the juice with ir hedspce, DO concentrtion remined mostly unchnged during storge nd rte constnt could not be clculted. Rte constnt vlues re reported in Tble 3-2. Control nd oxygen sturted hve similr rte constnts t the sme temperture but deerted juice is not in greement probbly becuse more dt points re needed to do better model djustment. At 5 C k vlues were 0.05, 0.14 nd 0.13 d -1 for deerted, control nd oxygen sturted juices tht compred with 0.128, nd d -1 k vlues clculted for deerted, control, oxygen sturted juices in the Chpter 2 re bout hlf of the vlues. Ornge juices in both cses were stored t 5 C during 60 dys nd these differences re ttributed to the different processing conditions (deertion, psteuriztion) nd pulp content. Commercil NFC ornge juice used for chpter 2 hd 0% screened pulp nd Hmlin NFC ornge juice hd 5% screened pulp. Our results for control nd oxygen-sturted juices re in greement with those reported by Pongndl (2010), who used n optic fiber for the determintions insted of n polrogrphic probe s it ws done on this experiment. Pongndl (2010) reported pseudo-first-order rte constnts for pinepple juice, which DO content ws not modified, of 0.77, 1.12, 2.69 nd 4.1 d -1 t 13, 21.5, 30.5 nd 40 C, respectively. Our results for control nd oxygen-sturted juices re in greement with the previous results except t 13 C where the rte constnt for pinepple juice is more thn double (0.77 d -1 ) tht the vlue reported on this study for control ornge juice (0.29 d -1 ) mybe becuse t low tempertures the rte of DO consumption is very slow. DO consumption in AA solutions t 30 C, with or without sugr hd first-order rte of 1.4 nd 0.9 d

102 (Hsieh nd Hrris 1993) tht ws lower thn the rte (2.86 d -1 ) obtined in this study for oxygen-sturted ornge juice t similr temperture (30.5 C). The effect of temperture ssessed by ctivtion energy rnged from ± 0.47 to 81.2 ± 0.13 kj mol -1 nd ws in greement with 58.6 to 92.5 kj mol -1 reported for AA solutions with or without sugr t temperture rnge of 26.5 to 33 C (Hsieh nd Hrris 1993). However, it ws higher compred to pinepple juice (E =48.6 kj mol -1 ) t the sme temperture rnge, suggesting thn ornge juice is more sensitive to temperture effects thn pinepple, which is probbly due to the dsorption of O 2 onto the pulp. Glss vils were used in this study, to void gs exchnge with the surroundings. Sores nd Hotchkiss (1999) found tht deertion did not mke difference in DO concentrtion of ornge juice during 60 dys of storge t 7 C when stored in continers with oxygen permebility between ml O 2 dy -1 per continer of 229 cm 3 of volume. AA Degrdtion The stoichiometry of AA rection with oxygen is 1:1 (Hsieh nd Hrris 1993) nd from the dt obtined for control ornge juice t 5 C, 152 µm of DO nd 2060 µm of AA the rtio between DO/AA is Therefore, with respect to its initil concentrtion, only smll mount of AA is needed to consume the entire mount of DO present in the ornge juice. This is confirmed s no chnges in AA concentrtion were observed in control ornge juice during the period of time (1 dy t 40 C) when DO ws depleted (Figure 3-5B). During psteuriztion, the loss of AA incresed s initil DO incresed nd the vlues were 12, 14 nd 19% for deerted, control nd oxygen sturted ornge juice respectively (Tble 3-3). In some cses, t the end of the storge AA concentrtion ws higher compred to the beginning of the storge, this ws cuse for 102

103 n experimentl error during AA mesurement (cpillry nd equipment vribility) nd in tht cse the loss of AA could not be clculted. AA loss during 3, 6, 12, 12 nd 60 dys of storge t 40, 30.5, nd 5 C of ornge juices without hedspce ws between 3 nd 23%. AA loss in control nd oxygen sturted ornge juices ws similr for the sme temperture lthough it ws not the cse for 5 nd 40 C possibly due to experimentl errors during mesurement of AA. For ir hedspce stored 60 dys t 5 C, 6 dys t 30.5 C nd 3 dys t 40 C the loss ws 100, 54 nd 73%, respectively (Tble 3-3). As discussed in chpter 2, oxygen from the hedspce is constntly being supplied to the ornge juice promoting AA oxidtion. Loss of AA in NFC ornge juices studied in Chpters 2 nd 3 under sme storge conditions (60 dys of storge t 5 C) ws 42 nd 100 % respectively. The AA decrese in NFC ornge juice from Chpter 2 nd 3 ws 688 nd 2000 M respectively. The difference in AA loss could be due to the different therml tretment nd previous storge time-temperture conditions used for ech juice. Figures 3-4 nd 3-5 show tht AA decresed the most in ir hedspce juices (Figure 3-4 D nd 3-5 D) stored t 5 (Figure 3-4D) nd 40 C (Figure 3-5D) perhps due to the presence of considerble mount of DO during storge (350 to 150 µm DO). Therefore, to protect AA during storge, hedspce should be filled with n inert gs such s nitrogen which is common prctice in industry. Figures 3-4 nd 3-5 contrst concentrtion chnges in DO nd AA during storge t 5 or 40 C respectively. The overll trend for AA in deerted, control nd oxygen sturted ornge juices is to remin bout the initil concentrtion while DO content reches the lowest concentrtion t the beginning of the storge time. In juice with ir hedspce AA nd 103

104 DO hve similr trend for the decrese in concentrtion during storge (Figure 3-4D nd 3-5D). The pseudo first-order rte constnts of AA degrdtion were clculted only for ir hedspce smples (Tble 3-4) nd the vlues rnged from d -1 t tempertures between 5 40 C. These vlues re higher thn d -1 reported for single strength from concentrte ornge juice (SSOJ) stored t 4 37 C (Kennedy nd others 1992); thn d -1 reported for concentrted ornge juice stored t C (Burdurlu nd others 2006); nd 6.84x d -1 membrne-clrified concentrted ornge juice stored t 4 24 C (Lee nd Chen 1998).All the mentioned studies used first model. Activtion energy for AA degrdtion clculted in study (12.12 kcl mol -1 ) ws lower thn concentrted ornge juices: kcl mol -1 (Burdurlu nd others 2006), 34.3 kcl mol -1 (Lee nd Chen 1998), nd SSOJ 36 kj mol -1 (Dhuique-Myer nd others 2007). Color Hunter prmeters, b nd L were mesured in ornge juice stored t 40 C (Tble 3.5) nd only ir hedspce ornge juice stored t 40 C showed drker color fter 3 dys. The rest of the juices stored t less thn 40 C were not mesured becuse no color chnge ws visully observed. Although in future experiments visul observtions should be confirmed with instrumentl mesurements. For the ir hedspce ornge juice t 40 C, incresed towrd more positive direction corresponding to more redness nd b becme more negtive corresponding to less yellowness. These vlues re exctly opposite of color chnges in membrne clrified concentrted ornge juice (Lee nd Chen 1998). Furfurl, produced from AA 104

105 degrdtion hs been correlted to browning in ornge juice model solutions (Shinod nd others 2005). In this study furfurl ws identified in ll ornge juice smples t dy 6 of storge t 40 C s explined in the voltile compounds section. Voltile Compounds Thirty-five voltile compounds were identified in NFC ornge juice nd since ech replicte ws obtined from different btch of juice importnt vribility between replictes of the sme juice ws indicted s rw pek re vlues in Tble 3-7. Oil content is shown in Tble 3-6, similr vlues of 0.04% were observed for ll three btches of control ornge juice. Although oil content in deerted nd oxygen sturted juices were lower thn the control for the sme btch of juice rnging between 0.02 to 0.03% probbly becuse limonene ws removed during the nitrogen/oxygen bubbling. Mesurement of limonene in control ornge juice from the different btches by SPME nd GC-MS ws explored in order to explin smples vribility, the SPME conditions of 5 min equilibrtion t 40 C nd 10 s of fiber exposition nd sme GC-MS conditions used for smples nlysis were pplied to obtin limonene pek tht could be completely integrted, limonene rw pek res obtined for control ornge juice were 2.88 x 10 9, 2.35 x 10 9, 2.91 x 10 9 for btch 1, 2 nd 3 respectively but limonene pek ws still not well defined in the chromtogrm. The SPME nd GC-MS pproch still needs to be improved in SPME conditions for n ccurte mesurement of limonene. In order to reduce vribility between replictes, three normliztions were pplied to voltiles dt: with internl stndrd, with totl pek re nd with dy 0. Clcultions used for the different normliztions re presented in Tble 3-7. Normliztion with internl stndrd compenstes for experimentl error, normliztion with totl pek re compenstes for differences between the replictes nd normliztion with dy 0 105

106 compenstes for initil differences in replictes. Reltive stndrd devitions (RSD) of selected compounds in control ornge juice t dy 0 of storge t 40 C were compred between normliztions (Tble 3-8), normliztion with internl stndrd hd RSD vlues between 31 nd 36, normliztion with totl pek hd RSD vlues of 15 to 25 nd RSD could not be clculted for normliztion t dy 0 becuse the ssumption of ll peks t dy 0 hve vlue of 100. Normliztion with totl pek re ws selected for PCA nlysis nd further discussion becuse the lower RSD vlues. Normlized pek vlues nd PCA nlysis of the other two normliztions methods re included in Appendix D. Tbles 3-9 to 3-11 shows verge, stndrd error nd Duncn s groups ( =0.05) for the normliztion with totl pek re of the three replictes of juices t dy 0, 0.5 nd 6 respectively. Figure 3-6 shows the proportion of terpenes, lcohols, ldehydes nd esters present in ll the juices t dy 0 nd 6.Terpenes were the voltiles present in highest proportion t both storge times, even though limonene ws excluded from the clcultions. At dy 0 (Figure 3-6A) not psteurized fresh squeezed ornge juice hd the highest proportion of 0.8 for terpenes nd the lowest of 0.07 for ldehydes compred with the psteurized juices. Among the psteurized juices, deerted juice showed the sme trend thn fresh squeezed with respect to the rest of the juices. A low proportion of ldehydes in fresh squeezed ornge juice ws unexpected since ldehydes hve been reported s importnt contributors to ornge juice flvor (Perez-Ccho nd Rouseff 2008b). At dy 6 of storge (Figure 3-6B), deerted ornge juice hd bout the sme 0.25 proportion of terpenes, lcohols, ldehydes nd lcohol terpenes. Oxygen sturted hd bout the 106

107 similr proportion of voltile compounds thn deerted juice except tht terpenes represented proportion of 0.4 of the totl composition. Principl component nlysis (PCA) is tool tht helps to exmine lrge dt set reducing them by ssociting vribles (individul voltiles) to lower dimensions or principl components. PCA help to identify ptterns in the dt nd highlight similrities nd differences mong smples (different levels of DO) nd help to identify those voltiles tht re more different thn the entire dt set nd tht explin the differences between smples. 54% of vrition between juice smples ws explined by PCA nd distribution of smples in the loding plot (Figure 3-7) ws by regions depending of the storge time going in digonl from the right up corner to the left bottom corner from dy 0 to dy 6. Only replictes of deerted nd ir hedspce juices were clustered in well defined groups (Figure 3-7). If Figures 3-7 nd 3-8 re overlpped, the compounds in the loding plot (Figure 3-8) re in higher mount in smples locted in the sme section of the plot in Figure 3-7 nd tht compounds explin the differences between smples. In this cse, -pinene, limonene, β-myrcene, octnl, nonnl, decnl, β- elemene, gernil, ethyl octnote nd octyl cette re higher in juices t dy 0 compred to dy 6. The opposite trend ws observed for furfurl, -terpineol, β- terpineol tht were only detected t dy 6, nd 1-hexnol which content incresed over the time. The mentioned compounds were significnt different except limonene, nonnl, decnl, gernil, ethyl octnote nd octyl cette. From Duncn s test groups reported in Tbles 3-9 to 3-11, it ws observed tht mong juices -pinene content ws higher in ir hedspce juice, β-myrcene nd β-elemene were lower in 107

108 control nd oxygen sturted juices nd octnl ws lower in deerted nd oxygen sturted juices. Conclusions Ornge juice smples prepred with ir in the hedspce produced the highest AA degrdtion during storge s observed in the previous chpter. DO consumption in control nd oxygen-sturted ornge juice hd similr kinetic rte vlues under the sme storge conditions nd fit pseudo-first order model. For deerted juice, insufficient smpling in DO determintion did not llow ccurte determintion of the rte constnt. Experimentl error in AA mesurement hs to be ddress by running n AA stndrd the sme dy tht ornge juice smples, to compenste by differences mong dys of testing. AA nd DHA mesurements of ornge juice without hedspce should be repeted to hve enough dt to clculte percentge of AA loss nd rte constnt of AA oxidtion. Using the normliztion with totl pek, PCA nd ANOVA, -pinene, β-myrcene, octnl, β-elemene, re higher in juices t dy 0 compred to dy 6. Furfurl, - terpineol, β-terpineol were only detected t dy 6. 1-hexnol incresed over the time. Being ll this differences significnt over the time but not between the juices t different levels of DO. Additionl studies re needed to confirm if DO hs direct effect on voltile compounds. From Duncn s test groups reported in tbles 3-9 to 3-11, it ws observed tht mong juices -pinene content ws higher in ir hedspce juice, β- myrcene nd β-elemene were lower in control nd oxygen sturted juices nd octnl ws lower in deerted nd oxygen sturted juices. 108

109 Tble 3-1. Dissolved oxygen nd scorbic cid smpling schedule during storge study. T DO AA ( C) 5 Every 3 d for 2 wk nd every wk for 60 d At 1,9, 15, 30, 44, 60 d 13 At 6 h, 1 d, 3 d, 6 d, 12 d 21.5 Every 6 h for 1 d nd t dy 6 nd At 6 h, 12 h, 18 h, 1 d, 3 d, 6 d 40 2 h, 6 h, 12 h, 24 h, 3 d Tble 3-2. Pseudo- first order rte constnt (d -1 ) of dissolved oxygen consumption in NFC ornge juice. T ( C) Deerted Control Oxygensturted Control w/ir hedspce ±0.01 (r 2 =0.85) ±0.03 (r 2 =0.96) E (kj mol -1 ) ±0.02 (r 2 =0.99) 0.29±0.08 (r 2 =0.97) 0.96±0.03 (r 2 =0.94) ±0.8 (r 2 =0.98) 0.13±0.03 (r 2 =0.89) 0.24±0.02 (r 2 =0.95) 1.00±0.06 (r 2 =0.99) 2.86±0.1 (r 2 =0.91) 5.01±0.7 (r 2 =0.99) 81.24± ±0.47 Insufficient dt to determine rte constnt t low incubtion time, [DO] did not vry significntly during this experiment -b -b -b -b -b -b 109

110 Tble 3-3. AA loss due to psteuriztion nd storge t selected tempertures. AA loss due to psteuriztion with respect to not psteurized fresh squeezed (%) T ( C) storge time (d) Deerted Control Oxygensturted Control w/ir hedspce ±4 14 ±4 19 ±3 14 ±4 AA loss due to storge fter psteuriztion (%) T ( C) storge time (d) Deerted Control Oxygensturted Control w/ir hedspce ±5 5 ±3 23 ±2 100 ± ND ±5 ±1 ± ND ±3 ±3 ± ±1 19 ±3 20 ±4 54 ± ND ±2 - not clculted due to experimentl errors 18 ±4 73 ±11 Tble 3-4. Pseudo- first order rte constnt (d -1 ) of AA degrdtion in NFC ornge juice with ir hedspce. T ( C) k (d -1 ) r * * E kcl mol -1 *Insufficient dt to determine rte constnt t low incubtion time 110

111 Tble 3-5. Color chnge in NFC ornge juice during storge t 40 C nd expressed s L, nd b with respect to control ornge juice. Deerted Oxygen sturted Air hedspce Time (d) L b L b L b Tble 3-6. Oil content in NFC ornge juice t dy 0 of storge t 40 C clculted using Scott oil method. Oil content (%) Ornge juice btch control deerted oxygen sturted 1 4.5E-02 ±2.6E E-02 ±2.6E E-02± 1.4E E E E-02± 2 ±2.2E E-02 3 ±0 smple ws vilble for only one replicte 1.8E-02 ±1.2E E E-02± 2.0E

112 Tble 3-7. Normliztion methods pplied to voltile compounds of NFC ornge juice Normliztion Clcultion of normlized pek re Internl stndrd pek pek re compound re int ernl s tn drd(4heptdecn one) Totl pek re pek re compound * verge totl of replictes totl pek of replicte pek re re Dy 0 pek re compound normlized with int ernl s tn drd *100 pek re dy 0 normlized with int ernl s tn drd A vlue of 100 is ssigned to pek re t dy 0 of ll the compounds 112

113 Tble 3-8. Comprison of rw pek re nd two different normliztions pplied to selected voltile compounds in control not-from- concentrted ornge juice t dy 0 of storge t 40 C. Control ornge juice t dy 0 Rw pek re Compound Replicte RSD linlool 2.01E E E octnl 1.92E E E nonnl 5.36E E E heptdecnone (internl stndrd) 1.26E E E Normliztion with internl stndrd Compound Replicte RSD linlool octnl nonnl Normliztion with totl pek re Compound Replicte RSD linlool 1.25E E E octnl 1.19E E E nonnl 3.32E E E Normliztion with dy 0, All vlues were 100 nd RSD could not be clculted 113

114 Tble 3-9. Voltile compounds identified in NFC ornge juice t different DO content t dy 0 of storge t 40 C. Vlues re verge nd stndrd error of res normlized with totl pek re. Vlues followed by different letters re significntly different (Duncn s test t =0.05). LRI (ref)* LRI (DBWx) ALDEHYDES (9) compound Fresh squeezed Deerted Control Oxygen sturted Air hedspce 698 (f) 722 cetldehyde E+07 ±4.43E E+07 ±3.63E E+08 ±2.10E E+08 ±1.20E (c) 1217 octnl 5.28E+06 ±2.76E E+08 ±1.14E+07b 1.10E+08 ±1.58E E+07 ±2.21E+07b 1.44E+08 ±4.96E (b,c) 1317 nonnl 2.42E+06 ±7.20E E+07 ±5.53E E+07 ±2.97E E+07 ±1.15E E+07 ±1.44E (d) 1396 furfurl (b,c) 1415 decnl 5.67E+06 ±2.77E E+08 ±3.16E E+08 ±1.12E E+08 ±3.47E E+08 ±1.68E (g) 1585 nerl 1.27E+07 ±5.36E E+07 ±1.00E+06b 3.06E+06 ±9.76E+05b 1.63E+07 ±4.45E E+07 ±5.78E (f) 1619 dodecnl 1.10E+07 ±1.06E E+07 ±1.23E E+07 ±8.70E E+07 ±2.14E E+07 ±1.58E () 1651 gernil 1.32E+07 ±1.69E E+07 ±7.00E E+07 ±1.18E+06b 1.61E+07 ±2.16E+06b 2.62E+07 ±5.16E+06b 1835 (f) ESTERS (4) 1713 perillldehyde 2.27E+06 ±2.72E E+07 ±2.51E E+06 ±3.25E E+07 ±6.73E E+07 ±2.42E (b,c,e) 978 ethyl butnote 1.78E+07 ±4.32E E+07 ±1.53E E+07 ±8.51E

115 Tble 3-9. Continued LRI (ref)* 1448 (b,c,e) LRI (DBWx) 1344 compound ethyl octnote Fresh squeezed Deerted Control 1.17E+07 ±1.18E E+07 ±9.16E E+06 ±2.72E+05 Oxygen sturted 7.48E+06 ±6.77E+05 Air hedspce 6.73E+06 ±2.54E (c) 1383 octyl cette 7.45E+06 ±3.25E E+07 ±7.12E E+07 ±1.26E+06b 2.38E+07 ±3.49E+06b 2.08E+07 ±8.78E+05b 1706 (f) TERPENES (10) 1626 neryl cette E+07 ±1.21E E+06 ±8.51E+05b 8.91E+06 ±2.35E (,c,e) 964 -pinene 2.13E+07 ±2.32E E+08 ±1.37E+07b 1.22E+08 ±1.09E+07b 8.87E+07 ±2.91E+07b 1.66E+08 ±1.49E (,c,d,e) 1139 Limonene 4.90E+09 ±3.91E E+10 ±3.18E E+10 ±1.78E+08b 1.09E+10 ±1.21E+09b 1.56E+10 ±3.23E (,b,d,e) 1082 β myrcene 4.70E+07 ±4.55E E+09 ±7.73E E+08 ±1.00E+08b 6.72E+08 ±1.58E+08b (e) cymene 1.47E+06 ±7.85E E+06 ±3.74E (c,d,e) terpinolene 3.93E+06 ±4.57E E+07 ±1.22E E+07 ±2.47E E+07 ±3.68E E+06 ±6.98E () 1501 β elemene 8.13E+06 ±3.09E E+07 ±9.69E E+06 ±4.46E+05b 7.31E+06 ±4.54E+05b 2.91E+08 ±1.46E+07b 1731 (d,e) 1636 Vlencene 3.77E+08 ±6.69E E+08 ±2.20E E+06 ±1.22E+06b 2.71E+08 ±2.19E (,e) 1142 Sbinene 2.50E+07 ±4.94E E+08 ±1.24E E+08 ±1.96E

116 Tble 3-9. Continued LRI (ref)* LRI (DBWx) compound Fresh squeezed Deerted Control Oxygen sturted Air hedspce 1248 (,e) 1169 β -ocimene 3.07E+07 ±2.96E E+07 ±2.56E (e) selinene 3.19E+07 ±4.96E+06 ALCOHOL TERPENES (5) 1.99E+07 ±1.21E E+07 ±2.90E E+07 ±7.16E E+07 ±3.63E (,c,d) 1449 Linlool 1.27E+07 ±1.58E E+08 ±1.90E E+08 ±1.08E E+08 ±2.03E E+08 ±1.21E (e) terpineol E+07 ±3.61E E+07 ±4.98E E+07 ±2.24E E+07 ±3.57E (d) 1543 β terpineol (d) terpineol E+06 ±1.90E (,c) ALCOHOLS (7) 1655 β-citronellol 4.75E+06 ±4.27E E+07 ±1.04E E+06 ±1.84E E+07 ±6.62E E+07 ±1.92E (,e) 887 Ethnol 4.41E+07 ±8.50E E+08 ±3.51E E+08 ±1.03E E+08 ±1.14E E+08 ±9.12E () hexnol 2.77E+06 ±4.01E (d) hexen-1-ol 2.65E+06 ±3.55E E+06 ±5.82E (e) octnol 7.08E+06 ±8.14E E+07 ±1.32E E+07 ±3.11E E+08 ±6.51E E+07 ±5.54E

117 Tble 3-9. Continued LRI (ref)* LRI (DBWx) compound Fresh squeezed Deerted Control Oxygen sturted Air hedspce 1640 (e) nonnol E+07 ±1.61E E+06 ±2.80E E+07 ±8.68E E+07 ±7.52E (e) decnol E+07 ±1.82E E+07 ±4.66E E+07 ±8.78E E+07 ±2.87E (,d) 1736 nerol 4.35E+06 ±4.30E E+07 ±1.13E E+06 ±4.01E E+07 ±5.61E E+06 ±6.89E+05 *(Jblpurwl 2009); b (Aren nd others 2006); c (Berlinet nd others 2007); d(berlinet nd others 2006); e(brt nd others 2003); f(rouseff 2008); g (Mottrm 2010), - not detected. 117

118 Tble Voltile compounds identified in NFC ornge juice t different DO content t dy 0.5 of storge t 40 C. Vlues re verge nd stndrd error of res normlized with totl pek re. Vlues followed by different letters re significntly different (Duncn s test t =0.05) LRI (ref)* ALDEHYDES (9) LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 698 (f) 722 cetldehyde 1.06E+08 ±3.50E E+06 ±3.41E E+08 ±4.00E E+08 ±3.23E (c) 1217 octnl 4.68E+07 ±4.51E E+07 ±2.64E E+07 ±4.49E E+07 ±4.03E (b,c) 1317 nonnl 1.67E+07 ±3.30E E+07 ±4.40E E+07 ±2.04E E+07 ±1.92E (d) 1396 furfurl (b,c) 1415 decnl 1.21E+08 ±4.46E E+08 ±3.10E E+08 ±1.31E E+08 ±1.65E (g) 1585 nerl 4.67E+06 ±4.94E E+06 ±1.70E E+06 ±5.53E E+07 ±1.05E (f) 1619 dodecnl 1.65E+07 ±9.45E E+07 ±4.22E E+07 ±2.73E E+06 ±4.71E () 1651 gernil 1.62E+07 ±4.43E E+06 ±1.57E (f) 1713 perillldehyde 8.23E+06 ±1.66E E+07 ±3.56E E+06 ±1.17E E+07 ±2.84E

119 Tble Continued LRI (ref)* LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce ESTERS (4) 1049 (b,c,e) 978 ethyl butnote E+07 ±3.27E E+07 ±2.91E (b,c,e) 1344 ethyl octnote 6.24E+06 ±2.92E E+06 ±8.89E E+06 ±7.48E E+06 ±5.55E (c) 1383 octyl cette 1.77E+07 ±1.73E E+06 ±3.99E E+07 ±2.68E E+07 ±1.12E (f) TERPENES (10) 1626 neryl cette 5.22E+06 ±1.20E E+06 ±1.36E E+06 ±7.44E (,c,e) 964 -pinene 8.65E+07 ±1.50E E+07 ±9.46E E+07 ±1.58E E+08 ±1.05E (,c,d,e) 1139 Limonene 1.17E+10 ±1.07E E+09 ±3.99E E+10 ±4.72E E+09 ±3.48E (,b,d,e) 1082 β myrcene 2.42E+06 ±6.42E (e) cymene 2.72E+07 ±9.38E E+06 ±1.44E E+07 ±2.64E (c,d,e) terpinolene 4.94E+06 ±9.20E E+06 ±5.71E E+06 ±6.46E E+06 ±4.07E () 1501 β elemene 2.67E+08 ±1.67E E+08 ±5.41E E+08 ±1.45E E+08 ±1.57E (d,e) 1636 Vlencene E+07 ±5.66E E+07 ±3.07E

120 Tble Continued LRI (ref)* LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 1137 (,e) 1142 Sbinene 6.44E+08 ±1.16E E+07 ±6.38E E+08 ±2.78E E+08 ±3.79E (,e) 1169 β ocimene 2.85E+07 ±7.82E (e) ALCOHOL TERPENES (5) selinene 2.02E+07 ±2.16E E+06 ±1.79E E+07 ±1.44E E+07 ±3.87E (,c,d) 1449 Linlool 9.52E+07 ±7.28E E+08 ±6.66E E+08 ±7.66E E+08 ±1.91E (e) terpineol 3.52E+07 ±2.96E E+07 ±1.53E E+07 ±4.53E E+08 ±7.40E (d) 1543 β terpineol E+06 ±1.12E (d) terpineol (,c) ALCOHOLS (7) 1655 β-citronellol 9.74E+06 ±1.42E E+06 ±2.08E E+06 ±5.16E E+07 ±1.46E (,e) 887 Ethnol 1.07E+08 ±1.40E E+08 ±3.95E E+08 ±1.72E E+08 ±8.52E () hexnol 2.03E+06 ±5.27E E+06 ±1.25E E+06 ±3.91E (d) hexen-1-ol 4.80E+06 ±3.11E E+06 ±7.10E E+06 ±2.49E

121 Tble Continued LRI (ref)* LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 1534 (e) octnol 4.52E+07 ±5.85E E+07 ±2.08E E+07 ±2.51E E+08 ±8.47E (e) nonnol 1.32E+07 ±2.27E E+07 ±3.82E E+06 ±4.00E E+07 ±1.65E (e) decnol 1.23E+07 ±7.10E E+07 ±4.63E E+07 ±3.04E (,d) 1736 Nerol 5.97E+06 ±7.87E E+06 ±2.74E E+06 ±8.65E E+07 ±1.73E+07 *(Jblpurwl 2009); b (Aren nd others 2006); c (Berlinet nd others 2007); d(berlinet nd others 2006); e(brt nd others 2003); f(rouseff 2008); g (Mottrm 2010), - not detected. 121

122 Tble Voltile compounds identified in NFC ornge juice t different DO content t dy 6 of storge t 40 C. Vlues re verge nd stndrd error of res normlized with totl pek re. Vlues followed by different letters re significntly different (Duncn s test t =0.05) LRI (ref)* ALDEHYDES (9) LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 698 (f) 722 Acetldehyde 1.56E+08 ±3.87E E+08 ±4.63E E+08 ±1.35E E+08 ±1.78E (c) 1217 octnl 1.65E+07 ±7.59E E+07 ±4.11E E+07 ±4.71E E+07 ±3.04E (b,c) 1317 nonnl 8.51E+06 ±2.90E E+07 ±6.22E E+06 ±2.05E E+06 ±3.21E (d) 1396 furfurl 2.10E+06 ±8.25E E+06 ±2.03E E+06 ±1.22E E+06 ±9.84E (b,c) 1415 decnl 5.54E+07 ±1.13E E+07 ±3.49E E+07 ±8.80E E+07 ±1.04E (g) 1585 nerl 1.00E+06 ±2.81E E+06 ±2.75E (f) 1619 dodecnl 1.06E+07 ±2.20E E+07 ±2.12E E+06 ±4.87E () 1651 gernil 1.25E+06 ±3.20E (f) 1713 perillldehyde 8.52E+06 ±3.52E E+06 ±6.61E E+06 ±2.28E E+06 ±5.39E

123 Tble Continued LRI (ref)* LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce ESTERS (4) 1049 (b,c,e) 978 ethyl butnote 1.06E+07 ±1.12E E+07 ±2.36E E+07 ±4.61E (b,c,e) 1344 ethyl octnote 4.84E+06 ±5.98E E+06 ±3.85E E+06 ±5.04E E+06 ±6.07E (c) 1706 (f) TERPENES (10) 1383 octyl cette 1.55E+07 ±3.39E neryl cette 5.98E+06 ±1.73E E+07 ±1.82E E+06 ±1.25E E+07 ±3.89E E+06 ±4.54E E+07 ±1.50E E+06 ±2.20E (,c,e) 964 -pinene 2.24E+07 ±2.85E E+07 ±6.41E E+07 ±7.37E E+08 ±1.82E (,c,d,e) 1139 Limonene 5.35E+09 ±4.70E E+10 ±9.44E E+09 ±4.20E E+10 ±6.89E (,b,d,e) 1082 β myrcene 2.32E+06 ±8.57E (e) cymene 1.05E+07 ±2.05E E+07 ±2.74E E+06 ±1.50E E+07 ±5.83E (c,d,e) terpinolene 2.31E+06 ±5.52E E+06 ±5.63E E+06 ±1.08E () 1501 β elemene 1.84E+08 ±2.58E E+08 ±2.00E E+08 ±5.05E E+08 ±1.45E (d,e) 1636 Vlencene 2.68E+07 ±6.52E E+06 ±6.02E

124 Tble Continued LRI (ref)* LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 1137 (,e) 1142 Sbinene 4.95E+06 ±9.38E E+08 ±9.25E E+08 ±2.85E E+08 ±4.79E (,e) 1169 β ocimene 1.81E+06 ±1.71E (e) ALCOHOL TERPENES (5) selinene 1.31E+07 ±1.77E E+07 ±2.90E E+07 ±4.61E E+06 ±5.32E (,c,d) 1449 Linlool 1.68E+08 ±6.10E E+08 ±8.83E E+08 ±4.14E E+07 ±7.53E (e) terpineol 4.24E+07 ±1.36E E+07 ±3.06E E+07 ±1.15E E+07 ±1.57E (d) 1543 β terpineol 1.98E+07 ±8.89E E+06 ±1.65E E+07 ±1.10E E+06 ±2.77E (d) terpineol E+06 ±1.90E (,c) 1655 β-citronellol 1.14E+07 ±4.22E+06 - ALCOHOLS (7) 9.84E+06 ±3.26E E+07 ±3.88E (,e) 887 Ethnol 1.68E+08 ±4.87E E+08 ±1.67E E+08 ±2.96E E+08 ±1.08E () hexnol 7.28E+06 ±4.06E E+06 ±3.91E E+06 ±1.69E (d) hexen-1-ol 5.22E+06 ±2.08E E+06 ±4.29E E+06 ±1.54E

125 Tble Continued LRI (ref)* LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 1534 (e) octnol 1.07E+08 ±4.19E E+07 ±1.69E E+07 ±1.98E E+07 ±2.55E (e) nonnol 1.67E+07 ±7.04E E+06 ±4.29E E+06 ±2.35E E+06 ±3.00E (e) decnol 3.80E+07 ±1.57E E+07 ±2.44E E+07 ±4.76E E+06 ±1.08E (,d) 1736 Nerol 1.51E+07 ±4.88E E+06 ±1.73E E+07 ±5.24E E+06 ±1.69E+06 *(Jblpurwl 2009); b (Aren nd others 2006); c (Berlinet nd others 2007); d(berlinet nd others 2006); e(brt nd others 2003); f(rouseff 2008); g (Mottrm 2010), - not detected. 125

126 Figure 3-1. Flow digrm of smple preprtion of not-from-concentrted ornge juice (Hmlin vr.) 126

127 Figure 3-2. Different kinetic models pplied to DO dt of not-from-concentrte ornge juice (cultivr Hmlin) stored t 5 C. 127

128 Figure 3-3. First-order kinetic dt of not-from-concentrte ornge juice (Hmlin vr.) stored t 5 C. ( )deerted, ( ) control, ( )oxygen sturted, ( )ir hedspce 128

129 Figure 3-4. ( )scorbic cid nd ( )dissolved oxygen content in not-from concentrted ornge juice (Hmlin vr) during storge t 5 C. A) Deerted, B) control, C) Oxygen sturted, D) ir hedspce. 129

130 Figure 3-5. ( )scorbic cid nd ( ) dissolved oxygen content in not-from concentrted ornge juice (Hmlin vr) during storge t 40 C. A) deerted, B) control, C) oxygen sturted, D) ir hedspce. 130

131 Figure 3-6. Proportion of voltile compounds identified in NFC ornge juice stored t 40 C. (A) 0 dys of storge nd fresh squeezed ornge juice, (B) 6 dys of storge. Proportion of terpenes were clculted excluding limonene. 131

132 Figure 3-7. Score plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with totl pek re of voltile compounds 132

133 Figure 3-8. Loding plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with totl pek re of voltile compounds. *significnt differences 133

134 CHAPTER 4 COMPREHENSIVE RESULTS AND FUTURE WORK This reserch described the chnges in vitmin C content, color nd rom in notfrom concentrte ornge juice during processing nd storge t different tempertures. The first chpter presents the knowledge lredy vilble bout oxygen interctions with vitmin C, color nd rom in different fruit juices nd the methods used for DO removl. In most of these previously published studies the oxidtion of AA due to presence of oxygen is ssumed without monitoring chnges in DO concentrtion during the experiment. This reserch hypothesized tht DO decreses vitmin C nd deteriortes rom nd color of ornge juice during storge. The second nd third chpter confirmed tht the mount of oxygen present in ornge juice even t sturtion levels is bout 10 times smller thn vitmin C content nd just smll portion of vitmin C rected with DO oxygen when no dditionl oxygen ws supplied to the smple, then dissolved oxygen did not ccount for importnt loss of scorbic cid. The effect of hving ir in the hedspce in the loss of scorbic cid ws lso studied nd the loss of scorbic cid rnged from 42 to 100% fter two months of storge t 5 C. Under this experiment conditions, oxygen dissolved into the juice or present in the hedspce of continers did not seem to hve direct effect on voltile compounds since significnt differences could not be correlted to DO content. Although voltile compounds such s furfurl, produced due to scorbic cid degrdtion were identified. Color chnge ws visully detected only in ornge juice with ir hedspce fter 3 dys of storge t 40 C. For the rest of the juices no color chnge ws observed minly becuse storge times were short, less thn 15 dys t tempertures of C nd 60 dys t 5 C. 134

135 The second chpter showed tht DO content reched its lowest concentrtion fter 7 dys of storge t 5 C. Considering the shelf-life of not-from-concentrte ornge juice is two months t 5 C, oxidtion cn hppen only t the beginning of the storge. Presence of ir in the hedspce of continers produced lrger AA degrdtion (42%) thn oxygen sturted juices without ir in hedspce during 60 dys of storge t 5 C. From the rom ctive compounds, methionl ws perceived t higher intensity in ir hedspce ornge juice with respect to control only nd nonnl hd the min contribution to rom intensity of control, oxygen sturted nd ir hespce ornge juices. At dy 1 of storge -elemene, -selinene nd 3-crene hd the highest mounts in deerted ornge juice compred with the other juices. At dy 60 of storge, the content of most of the compounds decresed with respect to dy 0 except for E-2-hexenl nd limonene oxide whose content incresed in ir hedspce ornge juice. In the third chpter, pseudo-first order ws used to clculte rte constnt nd ctivtion energy of DO consumption in NFC ornge juice. Loss of AA due to psteuriztion nd due to storge ws quntified seprtely. Loss of AA due to psteuriztion ws less thn 19% for ll the juices nd during storge losses of AA between 17 to 100% hppened in juice contining ir hedspce. Chnge in color ws visully detected only in ornge juice with hedspce t the highest temperture (40 C). Different normliztions were pplied to voltiles dt in order to compenste for differences between the replictes. Differences between replictes were not detected by Scott oil mesurement. Normliztion with totl pek re gve reltive stndrd devitions smller thn normliztion with internl stndrd nd with dy 0 nd it ws 135

136 selected for further ANOVA nd PCA nlysis. From Duncn s test groups it ws observed tht mong juices -pinene content ws higher in ir hedspce juice, β- myrcene nd β-elemene were lower in control nd oxygen sturted juices nd octnl ws lower in deerted nd oxygen sturted juices. Furfurl, -terpineol, β-terpineol were only detected t dy 6, nd 1-hexnol content incresed over the time. To ddress the remining questions from this reserch, the effect of oxygen on ech qulity prmeter (vitmin C, rom nd color) should be studied seprtely before performing storge study monitoring ll the prmeters. To clculte n ccurte pseudo first-order rte constnt for deerted ornge juice storge study with frequent smpling times t the beginning of storge nd longer smpling times towr the end of storge is required. Moreover mechnism for filling the vils fter psteuriztion tht voids reentry of ir needs to be implemented. Storge times of t lest 6 months re required to detect chnges in AA, DHA nd color in juices without hedspce, but since microbil growth is concern t tempertures between 13 nd 30.5 C filling method tht increses shelf life of ornge juice t these tempertures such s hot filling insted of cold filling should be utilized. Confirmtion of identity of rom ctive compounds should be done using second column (DB5) in order to verify the results obtined in the second chpter. Identifiction of voltile compounds should be repeted using replictes from the sme btch of ornge juice, pouring the juice immeditely fter gs sprging (nitrogen or oxygen) directly into the vils to use for the nlysis nd flushing the hedspce with nitrogen to hve only the effect of dissolved oxygen. 136

137 The effect of hedspce composition in ornge juice cn be studied by hving glss vils with hedspce volume of 3% of the totl continer volume filled with different gses such s nitrogen, ir nd oxygen. Future studies on the effect of DO on rom could be conducted relting to pckging of ornge juice, since the oxygen permebility of pckging mteril, volume nd composition of hedspce of pckges ffect the mount of dissolved oxygen in the juice. 137

138 APPENDIX A DISSOLVED OXYGEN CONTENT IN NFC OJ (HAMLIN CULTIVAR) 138

139 139

140 APPENDIX B COLOR CHANGES IN NFC OJ (HAMLIN CULTIVAR) STORED 3 DAYS AT 40 C 140

141 APPENDIX C GAS SPARGING EXPERIMENTAL SETUP Sprging gs (O 2 or N 2 ) Ornge juice 141

142 APPENDIX D NORMALIZATION OF VOLATILE COMPOUNDS IN NFC OJ (HAMLIN CULTIVAR) AND PCA ANALYSIS NORMALIZED WITH INTERNAL STANDARD Tble D-1. Voltile compounds identified in NFC ornge juice t different DO content t dy 0 of storge t 40 C. Vlues re verge nd stndrd error of normlized pek re using 4 heptdecnone s internl stndrd. Vlues followed by letters re significntly different (Duncn s test t =0.05). LRI (ref) LRI (DBWx) Compound Fresh squeezed Deerted Control Oxygen sturted Air hedspce ALDEHYDES (10) 698 (f) 722 cetldehyde E-02 ±7.06E E-02 ±7.45E E-01 ±6.61E E-01 ±2.16E (,d) 1297 (c) 1024 hexnl 5.08E-03 ±5.15E octnl 5.53E-03 ±4.20E E-01±2.95E E-01 ±5.69E E-02 ±8.29E E-01 ±1.02E (b,c) 1317 nonnl 6.33E-03 ±6.34E E-02±1.50E E-02 ±5.69E E-02 ±4.48E E-02 ±8.50E (d) 1396 furfurl (b,c) 1415 decnl 1.19E-02 ±1.00E E-01 ±1.00E E-01 ±4.74E E-01 ±2.80E E-01 ±1.19E (g) 1585 nerl 2.46E-02 ±1.62E E-02 ±7.11E E-03 ±3.19E-03b 1.73E-02 ±4.12E-04b 2.81E-02 ±1.15E (f) 1619 dodecnl 2.20E-02 ±2.87E E-02 ±1.21E E-02 ±6.46E E-02 ±7.67E E-02 ±1.67E

143 Tble D-1. Continued LRI (ref) LRI (DBWx) Compound Fresh squeezed Deerted Control Oxygen sturted Air hedspce 1766 () 1835 (f) 1651 gernil 2.33E-02 ±8.30E-04 perillldehyd 1713 e 4.35E-03 ±1.00E E-02 ±1.05E E-02 ±1.13E E-02 ±3.04E E-02 ±8.26E E-02 ±8.32E E-02 ±1.78E E-02 ±1.09E E-02 ±2.51E-03 ESTERS (5) 1049 (b,c,e) 1246 (b,c,d,e) 1448 (b,c,e) ethyl butnote ethyl hexnote ethyl octnote 4.09E-02 ±1.44E E-02 ±5.65E E-02 ±5.78E E-02 ±1.04E E-02 ±2.19E E-02 ±1.98E E-03 ±1.80E-03b 1.01E-02 ±1.03E-03b 7.97E-03 ±2.45E-04b 1485 (c) 1383 octyl cette 2.16E-02 ±1.17E E-02 ±1.31E E-02 ±3.92E E-02 ±4.11E E-02 ±9.49E (f) 1626 neryl cette E-02 ±3.55E E-03 ±1.98E-03b 9.66E-03 ±1.30E-03b - TERPENES (11) 1025 (,c,e) 964 -pinene 4.31E-02 ±9.74E E-01 ±2.63E E-01 ±8.29E E-01 ±1.22E E-01 ±2.02E (,c,d,e) 1139 Limonene 1.04E+01 ±1.45E E+01 ±7.40E E+01 ±3.65E E+01 ±7.23E E+01 ±2.99E (e) 1195 p-cymene 3.54E-03 ±2.34E E-03 ±1.92E

144 Tble D-1. Continued LRI (ref) LRI (DBWx) compound Fresh squeezed Deerted Control Oxygen sturted Air hedspce 1290 (c,d,e) terpinolene 8.38E-03 ±1.03E E-02 ±1.49E E-02 ±5.37E E-02 ±2.44E () 1501 β-elemene 2.24E-02 ±1.16E E-02 ±6.88E E-03 ±9.24E E-02 ±4.60E E-03 ±7.46E (e) 1517 cryophyllene 3.56E-02 ±4.58E (d,e) 1636 vlencene 7.59E-01 ±5.70E E-01 ±2.11E E-03 ±3.43E-03b 4.27E-01 ±3.14E-03b 3.41E-01 ±1.69E (,e) 1142 sbinene 5.03E-02 ±3.63E E-01 ±9.69E (,b,d,e) 1082 β-myrcene 1.36E-01 ±1.32E E+00 ±4.78E E+00 ±8.83E E+00 ±7.50E E+00 ±1.01E (,e) 1169 β-ocimene 5.66E-02 ±6.33E E-02 ±2.27E (e) selinene 5.44E-02 ±1.51E E-02 ±8.84E E-02 ±8.74E E-02 ±7.29E E-02 ±7.07E-03 ALCOHOL TERPENES (5) 1560 (,c,d) 1449 linlool 2.87E-02 ±3.48E E-01 ±7.79E E-01 ±1.47E E-01 ±9.03E E-01 ±2.06E (e) terpineol E-01 ±2.28E E-02 ±4.28E E-02 ±7.64E E-02 ±7.24E

145 Tble D-1. Continued LRI (ref) LRI (DBWx) compound Fresh squeezed Deerted Control Oxygen sturted Air hedspce 1616 (d) 1543 b-terpineol (d) terpineol E-02 ±5.45E () 1655 b-citronellol 9.35E-03 ±1.76E-03 ALCOHOLS (7) 1.73E-02 ±4.95E E-02 ±5.62E E-02 ±1.09E E-02 ±3.52E (,e) 887 ethnol 9.05E-02 ±3.31E E-01 ±2.62E E-01 ±2.28E E-01 ±1.28E E-01 ±3.70E () hexnol 6.28E-03 ±1.01E (d) hexen-1-ol 5.46E-03 ±1.44E E-03 ±2.80E (e) octnol 1.48E-02 ±3.30E E-01 ±3.29E E-02 ±7.12E E-02 ±4.72E E-02 ±7.56E (e) nonnol E-02 ±4.58E E-02 ±1.21E E-02 ±6.13E E-02 ±3.02E (e) decnol E-02 ±1.18E E-02 ±7.53E-03b 3.62E-02 ±8.67E-04b 3.57E-02 ±6.08E-03b 1818 (,d) 1736 nerol 7.93E-03 ±5.28E E-02 ±3.97E E-03 ±5.63E E-02 ±4.78E E-03 ±7.74E-04 (Jblpurwl 2009); b (Aren nd others 2006); c (Berlinet nd others 2007); d (Berlinet nd others 2006); e (Brt nd others 2003); f (Rouseff 2008); g (Mottrm 2010). 1 Pek not detected 145

146 Tble D-2. Voltile compounds identified in NFC ornge juice t different DO content t dy 0.5 of storge t 40 C. Vlues re verge nd stndrd error of normlized pek re using 4 heptdecnone s internl stndrd. Vlues followed by letters re significntly different (Duncn s test t =0.05). LRI (ref) ALDEHYDES (9) LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 698 (f) 722 Acetldehyde 1.33E-01 ±4.25E E-02 ±6.40E E-01 ±8.92E E-01 ±2.91E (c) 1217 Octnl 6.13E-02 ±6.93E E-01 ±7.25E E-02 ±2.25E E-02 ±5.22E (b,c) 1317 Nonnl 2.19E-02 ±2.45E E-02 ±1.23E E-02 ±4.78E E-02 ±1.00E (d) 1396 Furfurl (b,c) 1415 Decnl 1.59E-01 ±1.84E E-01 ±8.44E E-01 ±2.98E E-01 ±6.03E (g) 1585 Nerl 6.10E-03 ±9.32E E-02 ±4.61E E-03 ±1.02E E-03 ±1.05E (f) 1619 Dodecnl 2.14E-02 ±2.14E E-02 ±1.24E E-02 ±4.24E E-02 ±9.22E () 1651 Gernil 2.03E-02 ±4.53E E-02 ±3.68E (f) 1713 Perillldehyde 1.10E-02 ±3.00E E-02 ±8.07E E-02 ±3.35E E-02 ±1.50E-03 ESTERS (6) 1049 (b,c,e) 978 ethyl butnote E-02 ±8.11E E-02 ±9.19E

147 Tble D-2. Continued LRI (ref) 1448 (b,c,e) LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 1344 ethyl 8.12E E E E-03 octnote ±7.24E-04 ±3.17E-03 ±1.22E-03 ±2.44E (c) 1383 octyl cette 2.34E-02 ±4.27E E-02 ±7.19E E-02 ±6.23E E-02 ±5.22E (f) TERPENES (12) 1626 neryl cette 6.50E-03 ±6.93E E-03 ±1.21E E-03 ±1.21E (,c,e) 964 -pinene 1.17E-01 ±2.96E E-02 ±2.38E E-01 ±3.86E E-01 ±6.93E (e) cymene 2.99E-03 ±4.25E (c,d,e) terpinolene 3.31E-02 ±7.34E E-03 ±3.80E E-02 ±4.75E () elemene 6.59E-03 ±1.71E E-03 ±1.70E E-03 ±2.03E E-03 ±2.36E (d,e) 1636 Vlencene 3.46E-01 ±2.37E E-01 ±1.75E E-01 ±1.20E E-01 ±9.29E (,e) 1142 Sbinene E-01 ±2.73E E-01 ±2.17E (,b,d,e) myrcene 8.66E-01 ±2.24E E-01 ±8.75E E-01 ±2.27E E-01 ±3.87E (,e) ocimene 3.95E-02 ±1.32E (e) selinene 2.60E-02 ±2.39E E-02 ±7.36E E-02 ±8.78E E-02 ±7.76E

148 Tble D-2. Continued LRI (ref) ALCOHOL TERPENES (5) LRI (DBWx) Compound Deerted Control Oxygen sturted Air hedspce 1560 (,c,d) 1449 Linlool 1.25E-01 ±1.89E E-01 ±1.66E E-01 ±4.54E E-01 ±3.65E (e) terpineol 4.64E-02 ±7.50E E-01 ±4.10E E-02 ±1.44E E-02 ±1.25E (d) terpineol E-03 ±2.05E (d) terpineol (,c) ALCOHOLS (8) citronellol 1.24E-02 ±9.14E E-02 ±4.96E E-03 ±2.11E E-02 ±3.83E (,e) 887 Ethnol 1.44E-01 ±3.12E E-01 ±9.59E E-01 ±5.98E E-01 ±5.42E () hexnol 2.69E-03 ±8.68E E-03 ±2.51E E-03 ±1.08E E E (d) hexen-1-ol 5.44E-03 ±2.99E E-03 ±1.84E E-03 ±7.58E (e) octnol 6.01E-02 ±1.26E E-01 ±5.37E E-02 ±1.85E E-02 ±1.61E (e) nonnol 1.67E-02 ±1.76E E-02 ±9.24E E-02 ±2.20E E-02 ±1.83E-03 (Jblpurwl 2009); b (Aren nd others 2006); c (Berlinet nd others 2007); d (Berlinet nd others 2006); e (Brt nd others 2003); f (Rouseff 2008); g (Mottrm 2010). 1 Pek not detected 148

149 Tble D-3. Voltile compounds identified in NFC ornge juice t different DO content t dy 6 of storge t 40 C. Vlues re verge nd stndrd error of normlized pek re using 4 heptdecnone s internl stndrd. Vlues followed by letters re significntly different (Duncn s test t =0.05) LRI ref LRI (DBWx) compound ALDEHYDES (10) 698 f 722 cetldehyde 9.49E-02 ±2.74E c 1217 octnl 1.32E-02 Deerted Control Oxygen sturted Air hedspce ±1.14E-02b 1411 b, c 1317 nonnl 7.46E-03 ±4.60E d 1396 furfurl 2.12E-03 ±1.01E b, c 1415 decnl 4.78E-02 ±1.86E-02b 1630 g 1585 nerl 8.96E f 1619 dodecnl 9.34E-03 ±5.21E gernil 9.33E f 1713 perillldehyde 8.79E-03 ±4.08E E-01 ±1.02E E-02 ±8.37E E-02 ±1.37E E-03 ±5.08E E-01 ±2.40E-02 ±5.59E-05b E-02 ±6.29E E-01 ±2.71E E-02 ±1.16E-03b 1.22E-02 ±1.03E E-03 ±1.28E E-01 ±3.77E E-02 ±2.28E-03b 1.33E-02 ±6.40E E-03 ±1.73E E E-01 ±3.85E-03b ±5.22E E-02 ±2.71E E-02 ±1.03E-02 - ±5.28E E E-02 ±2.56E-03 ±5.91E E-03 ±9.78E-05 ESTERS (6) 1049 b,c,e 1448 b,c,e 978 ethyl butnote 1344 ethyl octnote 8.70E-03 ±1.31E E-03 ±6.83E-04b 2.10E-02 ±3.84E E-03 ±1.84E E-02 ±8.29E E-03 ±1.47E-04b 7.96E-03 ±2.16E

150 Tble D-3. Continued LRI LRI (ref) (DBWx) 1485 c f 1626 compound octyl cette neryl cette Deerted Control Oxygen sturted Air hedspce 1.40E-02 ±2.01E E-03 ±5.00E-04 TERPENES (12) 1025,c,e 964 -pinene 1.45E-02 ±1.22E-03b 1188,c,d,e 1139 limonene 3.67E+00 ±7.07E-01b 1148,b,d,e 1082 β-myrcene 5.39E-03 ±2.44E-03b 1244 e 1195 p-cymene 2.10E c,d,e terpinolene 2.39E-02 ±3.22E E-03 ±4.21E E-01 ±3.30E E+01 ±2.29E E+00 ±1.52E E-02 ±5.24E E-03 ±6.58E E-02 ±1.76E-03b 5.72E+00 ±7.66E-02b 8.22E-01 ±2.98E-01b 2.14E-02 ±3.05E E-03 ±3.96E E-01 ±2.38E E+01 ±1.31E E+00 ±6.55E-01 ±4.02E E-03 ±3.37E-03b β-elemene 1.92E-03 ±2.31E E-02 ±9.48E E-02 ±2.92E-05b 3.78E-03 ±1.68E E-02 ±1.17E E-03 ±2.43E d,e 1636 vlencene 1.53E-01 ±1.08E-02b 3.92E-01 ±8.44E ,e 1142 sbinene 2.09E-02 ±1.12E E-01 ±2.29E-02b 4.11E-01 ±1.20E E-03 ±7.47E

151 Tble D-3. Continued LRI (ref) 1248,e 1689 e Deerted Control Oxygen sturted Air hedspce LRI (DBWx) compound 1169 b-ocimene 1.31E-03 ±3.74E selinene 1.10E-02 ±2.62E-03 ALCOHOL TERPENES (5) 1560,c,d 1572 e 1616 d 1661 d 1777 ALCOHOLS (8) 1449 linlool 1.68E-01 ±1.22E terpineol 4.20E-02 ±9.25E b-terpineol 2.08E-02 ±1.05E terpineol 1.09E-01 ±4.94E E-02 ±1.13E E-01 ±4.33E E-02 ±1.43E E-02 ±6.47E E-01 ±4.51E b-citronellol 1.13E-02 ±7.10E ,e 887 ethnol 1.59E-01 ±1.22E hexnol 7.23E d hexen-1- ol ±2.27E E-03 ±3.04E e octnol 1.07E-01 ±7.17E e nonnol 1.64E-02 ±1.83E E-02 ±3.66E E-01 ±6.91E E-02 ±2.36E E-02 ±1.06E E-01 ±1.35E E-02 ±2.51E E E-01 ±2.12E-02 ±4.18E E-03 ±1.00E E-03 ±1.65E-04b E E-01 ±9.80E-03 ±4.85E E E-02 ±1.18E-03 ±4.01E E-02 ±1.24E E-01 ±9.71E E-02 ±4.93E E-02 ±3.55E E-02 ±2.54E E-02 ±7.97E E-01 ±1.65E E-03 ±3.08E E-03 ±3.47E-05b 6.38E-02 ±3.55E E-03 ±6.36E

152 Tble D-3. Continued LRI (ref) LRI (DBWx) compound 1745 e decnol 3.61E-02 ±6.37E ,d 1736 nerol 1.50E-02 ±2.90E-03 Deerted Control Oxygen sturted Air hedspce 3.24E-02 ±4.24E E-02 ±3.15E E-03 ±3.24E E-02 ±5.12E E-02 ±2.72E E-02 ±3.21E-03 (Jblpurwl 2009); b (Aren nd others 2006); c (Berlinet nd others 2007); d (Berlinet nd others 2006); e (Brt nd others 2003); f (Rouseff 2008); g (Mottrm 2010). 1 Pek not detected 152

153 Figure D-1. Score plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with internl stndrd. 153

154 Figure D-2. Loding plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with internl stndrd. 154

155 NORMALIZED WITH DAY O OF STORAGE Tble D-4. Voltile compounds identified in NFC ornge juice t different DO content t dy 0 of storge t 40 C. Vlues re verge nd stndrd error of normlized pek re with dy 0. Vlues followed by letters re significntly different (Duncn s test t =0.05) Oxygen LRI Compound Deerted Control sturted Air hedspce 716 cetldehyde 100±0 100±0 100±0 100±0 891 ethnol 100±0 100±0 100±0 100± pinene 100±0 100±0 100±0 100±0 984 ethylbutnote 100±0 100±0 100±0 100± b-myrcene 100±0 100±0 100±0 100± limonene 100±0 100±0 67±33 100± sbinene 100±0 100±0 100±0 100± b-ocimene 100±0 100±0 100±0 100± o-cymene 100±0 100±0 100±0 100± terpinolene 100±0 100±0 100±0 100± octnl 100±0 100±0 100±0 100± hexnol 100±0 100±0 100±0 100± hexen-1-ol, (z) 100±0 100±0 100±0 100± nonnl 100±0 100±0 100±0 100± ethyl octnote 100±0 100±0 100±0 100± octyl cette 100±0 100±0 100±0 100±0 -Not detected 155

156 Tble D-5. Voltile compounds identified in NFC ornge juice t different DO content t dy 0.5 of storge t 40 C. Vlues re verge nd stndrd error of normlized pek re with dy 0. Vlues followed by letters re significntly different (Duncn s test t =0.05) Air LRI compound Deerted Control Oxyge sturted hedspce 716 cetldehyde 202±63 50±40 136±48 91± ethnol 66±18 35±12 211±107 64± pinene - 86±14 82± ethylbutnote 60±14 13±8 129±129 66± b-myrcene 96±27 52±18 75± limonene 63±29 69±23 66±35 43± sbinene 79±31 246± ± b-ocimene 148±54 174±77 100± o-cymene 52±18 93±6 75±6 45± terpinolene 68±21 124±14 119±20 71± octnl 53±16 98±7 75±7 49± hexnol 72±29 228±36 87±33 65± hexen-1-ol, (z) 75±35 273±22 83±27 66± nonnl 63±21 153±25 88±17 60± ethyl octnote 32±9 307±129 46±2 24± octyl cette 91±41 54± Not detected 156

157 Tble D-6. Voltile compounds identified in NFC ornge juice t different DO content t dy 6 of storge t 40 C. Vlues re verge nd stndrd error of normlized pek re with dy 0. Vlues followed by letters re significntly different (Duncn s test t =0.05) Air LRI Compound Deerted Control Oxygen sturted hedspce 716 cetldehyde 241±92 103±2 97±36 155± ethnol 14±8 93±13 70±38 64± pinene 100±0 89±9 72± ethylbutnote - 128±18 74±34 84± b-myrcene 34±20 370±178 36±8 100± limonene 18±11 47±1 11±7 19± sbinene 100±0 100±0-100± b-ocimene 108±2 100±0-100± o-cymene 18±8 39±5 36±6 20± terpinolene 35±10 102±16 90±21 80± octnl 16±6 45±9 33±9 26± hexnol 62±7 91±7 82±22 43± hexen-1-ol, (z) 79±10 106±13 106±19 55± nonnl 41±3 91±12 77±15 51± ethyl octnote 4±1-64± octyl cette 5± Not detected 157

158 Figure D-3. Score plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with dy

159 Figure D-4. Loding plot of PCA pplied to voltile compounds from NFC ornge juice stored t 40 C normlized with dy

160 APPENDIX E POINTS TO CONTROL FOR REDUCING VARIABILITY DURING SAMPLE PREPARATION Points list 1. To hrvest fruit t the sme mturity stge, size nd to extrct the juice t the sme time for ll the replictes. 2. Set the gs flow using wter nd before pouring the ornge juice into the bottles used for bubbling. 3. Set flow nd constnt temperture in microthermics before strting psteuriztion. 4. To void reentry of ir to ornge juice fter psteuriztion by filling the vils under nitrogen tmosphere or using mechnism tht void the contct of 160