<%BANNER%>

Effect of Dissolved Oxygen and Deoxygenation on the Quality of Orange Juice

Permanent Link: http://ufdc.ufl.edu/UFE0042014/00001

Material Information

Title: Effect of Dissolved Oxygen and Deoxygenation on the Quality of Orange Juice
Physical Description: 1 online resource (172 p.)
Language: english
Creator: Garcia-Torres, Rosalia
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: acid, aroma, ascorbic, color, deaeration, dissolved, juice, orange, oxygen
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre: Food Science and Human Nutrition thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Orange juice was the most consumed fruit juice in the 2007/08 season, representing 50% (14.4 L, Single Strength Equivalent per capita) of the total per capita consumption of fruit juices. Because processing conditions influence the quality of the end product, it is important to understand the changes that orange juice experiences during processing. Deaeration is a processing step applied immediately before pasteurization to remove air, in particular oxygen, incorporated into orange juice during extraction or mixing. Vitamin C is believed to be the compound the most affected (oxidized) by dissolved oxygen (DO) in orange juice. However, there is a lack of information about changes in color and aroma under different levels of DO. The objective of this study was to understand the effect of initial DO, presence of air in the headspace and storage temperature on vitamin C, color and aroma of not from concentrated (NFC) orange juice in order to preserve fresh-like characteristics. The first storage study correlated changes in vitamin C (ascorbic and dehydroascorbic acid) and aroma compounds of commercial NFC orange juice (cultivar Valencia,) with selected initial DO levels during storage for 60 days at 5 degreeC in amber glass containers. The same study also evaluated the effect of having air in the headspace. The selected initial levels of DO in juice with no headspace were: deaerated (42.3 plus or minus 2.7 microM), oxygen saturated (494.7 plus or minus 1.8 microM), control (177.6 plus or minus 7.9 microM), and juice with air in the headspace (180.3 plus or minus 9.3 microM). The juice with air headspace had the highest loss of ascorbic acid (AA) of 42% after 60 days of storage at 5 degreeC. Pseudo-first order rate constant of DO consumption was 2.8x10-1 plus or minus 4.1x10-2 1/d. No changes in color were visually perceived among treatments or over time. Pseudo-first order rate constant of AA oxidation at 5 degreeC was 1.9x10-2 plus or minus 3.0x10-3. Aroma active and volatile compounds were analyzed in all juices by GC-O and GC-MS using solid phase micro extraction (SPME) headspace sampling. Among aroma active compounds, methional was perceived at higher intensity in orange juice samples with air headspace compared to control only. Nonanal had the main contribution to aroma intensity of control, oxygen saturated and air headspace orange juices. At day 1 of storage beta-elemene, alpha-selinene and delta 3-carene had the highest amounts in deaerated orange juice compared with the other juices. At day 60 of storage, the content of most of the compounds decreased with respect to day 0 except for E-2-hexenal and limonene oxide whose content increased in orange juice with air headspace. Principal component analysis (PCA) of volatile compounds allowed to differentiating orange juices deaerated and with air headspace between days 0 and 6 of storage. A second storage study with not-from concentrated (NFC) orange juice (cultivar Hamlin) described the impact of DO and storage temperature on volatile compounds, AA and color during storage at 5, 13, 21.5, 30.5 and 40 degreeC. The DO content in the juice samples was modified by gas sparging, followed by pasteurization and storage in amber glass containers at the specified temperatures. Similar to the previous study, the different initial levels of DO in juice with no headspace were: deaerated (58.7plus or minus 6.3 microM), oxygen saturated (209.7 plus or minus 76.3 microM), control (152.4 plus or minus 21.3 microM), and juice control with air headspace (340.9 plus or minus 64.5 microM). The juice with the highest loss of AA was the one air headspace with 100%, 73.5% and 54.4% of AA loss after 60 days of storage at 5 degreeC, 3 days of storage at 40 degreeC and 6 days of storage at 30.5 degreeC respectively. Color change was perceived by direct observation only in air headspace orange juice stored 3 days at 40 degreeC, Color change was also measured using a and b parameters than changed toward the red and blue respectively. No correlation with dissolved oxygen content was established. Volatile compounds were measured by SPME headspace and GC-MS after gas sparging and pasteurization and identified by matching with database and LRI values. Each replicate was obtained from a different batch of orange juice and relative standard deviations of more than 50 were observed in the raw peak areas of the replicates of the same sample. Although it was initially thought that differences in oil content could be the cause of the variability, differences between replicates were not detected by Scott oil measurement. Three different normalizations were applied to volatiles raw peaks: normalization with 4-heptadecanone added as internal standard to all the samples, normalization with total peak area and normalization with respect to day 0. Normalization with total peak area gave relative standard deviations smaller than the other two normalizations and was selected for further ANOVA and PCA analysis.alpha-pinene content was higher in juice with air headspace juice, beta-myrcene and beta-elemene were lower in control and oxygen saturated juices and octanal was lower in deaerated and oxygen saturated juices. Furfural, alpha-terpineol, beta-terpineol were only detected at day 6, and 1-hexanol content increased over the time. These observations were significantly different and confirmed with PCA. The results of this research suggest that since dissolved oxygen does not have an important impact on vitamin C, color and aroma of orange juice, deaeration of juice may not be necessary in citrus processing for short term storage. Rather, removing air headspace from storage containers or replacing air headspace with nitrogen will reduce the oxidation of vitamin C during storage. The results of this research may possibly be applied to other fruit juices, nectars or purees containing oxygen sensitive compounds and high concentration of vitamin C.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Rosalia Garcia-Torres.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Reyes De Corcuera, Jose Ignacio.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042014:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042014/00001

Material Information

Title: Effect of Dissolved Oxygen and Deoxygenation on the Quality of Orange Juice
Physical Description: 1 online resource (172 p.)
Language: english
Creator: Garcia-Torres, Rosalia
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: acid, aroma, ascorbic, color, deaeration, dissolved, juice, orange, oxygen
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre: Food Science and Human Nutrition thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Orange juice was the most consumed fruit juice in the 2007/08 season, representing 50% (14.4 L, Single Strength Equivalent per capita) of the total per capita consumption of fruit juices. Because processing conditions influence the quality of the end product, it is important to understand the changes that orange juice experiences during processing. Deaeration is a processing step applied immediately before pasteurization to remove air, in particular oxygen, incorporated into orange juice during extraction or mixing. Vitamin C is believed to be the compound the most affected (oxidized) by dissolved oxygen (DO) in orange juice. However, there is a lack of information about changes in color and aroma under different levels of DO. The objective of this study was to understand the effect of initial DO, presence of air in the headspace and storage temperature on vitamin C, color and aroma of not from concentrated (NFC) orange juice in order to preserve fresh-like characteristics. The first storage study correlated changes in vitamin C (ascorbic and dehydroascorbic acid) and aroma compounds of commercial NFC orange juice (cultivar Valencia,) with selected initial DO levels during storage for 60 days at 5 degreeC in amber glass containers. The same study also evaluated the effect of having air in the headspace. The selected initial levels of DO in juice with no headspace were: deaerated (42.3 plus or minus 2.7 microM), oxygen saturated (494.7 plus or minus 1.8 microM), control (177.6 plus or minus 7.9 microM), and juice with air in the headspace (180.3 plus or minus 9.3 microM). The juice with air headspace had the highest loss of ascorbic acid (AA) of 42% after 60 days of storage at 5 degreeC. Pseudo-first order rate constant of DO consumption was 2.8x10-1 plus or minus 4.1x10-2 1/d. No changes in color were visually perceived among treatments or over time. Pseudo-first order rate constant of AA oxidation at 5 degreeC was 1.9x10-2 plus or minus 3.0x10-3. Aroma active and volatile compounds were analyzed in all juices by GC-O and GC-MS using solid phase micro extraction (SPME) headspace sampling. Among aroma active compounds, methional was perceived at higher intensity in orange juice samples with air headspace compared to control only. Nonanal had the main contribution to aroma intensity of control, oxygen saturated and air headspace orange juices. At day 1 of storage beta-elemene, alpha-selinene and delta 3-carene had the highest amounts in deaerated orange juice compared with the other juices. At day 60 of storage, the content of most of the compounds decreased with respect to day 0 except for E-2-hexenal and limonene oxide whose content increased in orange juice with air headspace. Principal component analysis (PCA) of volatile compounds allowed to differentiating orange juices deaerated and with air headspace between days 0 and 6 of storage. A second storage study with not-from concentrated (NFC) orange juice (cultivar Hamlin) described the impact of DO and storage temperature on volatile compounds, AA and color during storage at 5, 13, 21.5, 30.5 and 40 degreeC. The DO content in the juice samples was modified by gas sparging, followed by pasteurization and storage in amber glass containers at the specified temperatures. Similar to the previous study, the different initial levels of DO in juice with no headspace were: deaerated (58.7plus or minus 6.3 microM), oxygen saturated (209.7 plus or minus 76.3 microM), control (152.4 plus or minus 21.3 microM), and juice control with air headspace (340.9 plus or minus 64.5 microM). The juice with the highest loss of AA was the one air headspace with 100%, 73.5% and 54.4% of AA loss after 60 days of storage at 5 degreeC, 3 days of storage at 40 degreeC and 6 days of storage at 30.5 degreeC respectively. Color change was perceived by direct observation only in air headspace orange juice stored 3 days at 40 degreeC, Color change was also measured using a and b parameters than changed toward the red and blue respectively. No correlation with dissolved oxygen content was established. Volatile compounds were measured by SPME headspace and GC-MS after gas sparging and pasteurization and identified by matching with database and LRI values. Each replicate was obtained from a different batch of orange juice and relative standard deviations of more than 50 were observed in the raw peak areas of the replicates of the same sample. Although it was initially thought that differences in oil content could be the cause of the variability, differences between replicates were not detected by Scott oil measurement. Three different normalizations were applied to volatiles raw peaks: normalization with 4-heptadecanone added as internal standard to all the samples, normalization with total peak area and normalization with respect to day 0. Normalization with total peak area gave relative standard deviations smaller than the other two normalizations and was selected for further ANOVA and PCA analysis.alpha-pinene content was higher in juice with air headspace juice, beta-myrcene and beta-elemene were lower in control and oxygen saturated juices and octanal was lower in deaerated and oxygen saturated juices. Furfural, alpha-terpineol, beta-terpineol were only detected at day 6, and 1-hexanol content increased over the time. These observations were significantly different and confirmed with PCA. The results of this research suggest that since dissolved oxygen does not have an important impact on vitamin C, color and aroma of orange juice, deaeration of juice may not be necessary in citrus processing for short term storage. Rather, removing air headspace from storage containers or replacing air headspace with nitrogen will reduce the oxidation of vitamin C during storage. The results of this research may possibly be applied to other fruit juices, nectars or purees containing oxygen sensitive compounds and high concentration of vitamin C.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Rosalia Garcia-Torres.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Reyes De Corcuera, Jose Ignacio.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042014:00001


This item has the following downloads:


Full Text

PAGE 1

1 EFFECT OF DISSO LVED 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 2010

PAGE 2

2 2010 Rosal a Garca Torres

PAGE 3

3 To my parents Zenn Garca Cruz and Rosalia Torres Prez

PAGE 4

4 ACKNOWLEDGMENTS I would like to express my gratitude to my advisor Dr. Reyes De Corc uera for his constant guidance, encouragement and support. He constantly challenged me to perform my work with excellence and I learn ed from him the enthusiasm for research. Dr. Reyes example of integrity and discipline in research has influenced my ethica l and scientific criteria. My deepe st gratitude also goes to CONACY T Mxico for providing financial support during the PhD program and to Dr. Jim Graham for allowing me to work in his lab. I would like to extend my gratitude to my committee members, Dr. R ouseff, Dr. Goodrich Schneider and Dr. Eh s ani. To Dr. Rouseff for allow me to work in his lab and for having really enjoyable scientific conversations. To Dr Goodrich Schneider for reviewing my manuscripts and to Dr. Eh s ani for taking the time to reading my dissertation. I am very thankful to my lab mates (Juan Manuel, Mike, Narsi and Brittany) for their help and support when needed, with Shelley for being such a good lab to Lis that was an important contributor to my research helping during the experiments and applying her sensitive sense of smelling. I thank to John Henderson and Rocky for their help with the technical issues I faced during my research. I want to thank my friends for becoming my family in the United States especially to Fatima, Raquel, Sunny, Juan Carlos, Juan Manuel, Milena and my group of gators from Central Florida with whom I learn ed the meaning of friendship and I do not have words to express my gr atitude to them. Being on the CREC was a great experience that I will never forget.

PAGE 5

5 I am deeply grateful to my family for their support and for believing in me. My parents have always be an example to follow and my strength when I have been about to give u p. My mother is the best friend God gave me and my role model as a woman. career. My brother has always been the one that ground ed me when I lose the perspective of life My sist er in law has enriched my knowledge about the U S. To my aunts Cuca and Cristina which time invested in my early education has been productive. To my grandparents that in their particular way taught me to live life courageously and with happiness. Finally I thank God for filling my life with thousand s of blessings.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 11 LIST OF ABBREVIATIONS ................................ ................................ ........................... 13 ABSTRACT ................................ ................................ ................................ ................... 14 C H A P T E R 1 INTRODUCTION AND LITERATURE ................................ ................................ ..... 18 Oxygen Solubility ................................ ................................ ................................ .... 19 Interaction of Dissolved Oxygen with Food Components ................................ ........ 21 Ascorbic Acid ................................ ................................ ................................ .... 21 Degradation ................................ ................................ ............................... 22 Effect of dissolved oxygen concentration ................................ ................... 24 Color ................................ ................................ ................................ ................. 31 Nonenzymatic browning ................................ ................................ ............. 31 Enzymatic browning ................................ ................................ ................... 33 Aroma ................................ ................................ ................................ ............... 34 Met hods of Measurement of DO ................................ ................................ ............. 36 Methods of Removal of DO ................................ ................................ ..................... 38 Vacuum Deaeration ................................ ................................ ......................... 38 Gas Sparging ................................ ................................ ................................ ... 39 Membrane Deaerators ................................ ................................ ..................... 40 Enzyme Based Deaeration ................................ ................................ ............... 41 Oxygen Scavengers ................................ ................................ ......................... 43 Specific Objectives ................................ ................................ ................................ 47 2 ASSESSMENT OF DISSOLVED OXYGEN EFFECT ON VITAMIN C AND AROM A COMPOUNDS IN PASTEURIZED NOT FROM CONCENTRATE ORANGE JUICE ................................ ................................ ................................ ..... 61 Introduction ................................ ................................ ................................ ............. 61 Materials and Methods ................................ ................................ ............................ 62 Ora nge Juice Preparation and Characterization ................................ ............... 62 Sample Preparation and Storage Study ................................ ........................... 63 Microbial Analysis ................................ ................................ ............................. 63 DO Measurement ................................ ................................ ............................. 64 Vitamin C Measurement ................................ ................................ ................... 65 Color Measurement ................................ ................................ .......................... 66

PAGE 7

7 Aroma Analysis ................................ ................................ ................................ 66 Headspace sampling ................................ ................................ ................. 66 Gas Chromatography FID/ Olfactometry ................................ ................... 66 Gas Chromatograph y Mass Spectrometry (GC MS). .............................. 67 Rate of Reaction ................................ ................................ ........................ 68 Statistical Analysis ................................ ................................ ..................... 68 Results and Discussion ................................ ................................ ........................... 68 Microbial Analysis ................................ ................................ ............................. 68 Dissolved Oxygen (DO) ................................ ................................ .................... 69 Color ................................ ................................ ................................ ................. 71 Aroma Active Compounds ................................ ................................ ................ 72 Volatile Compounds ................................ ................................ ......................... 73 Conclusions ................................ ................................ ................................ ............ 75 3 IMPACT OF DISSOLVED OXYGEN LEVEL AND TEMPERATURE ON VOLATILE COMPOUNDS, VITAMIN C AND COLOR IN pasteurized NOT FROM CONCENTRATE ORANGE JUICE ................................ ............................. 92 Introduction ................................ ................................ ................................ ............. 92 Materials and Methods ................................ ................................ ............................ 93 Orange Juice Preparation and Ch aracterization ................................ ............... 93 Sample Preparation and Storage Study ................................ ........................... 94 Microbial Analysis ................................ ................................ ............................. 95 DO Measurements ................................ ................................ ........................... 96 Vitamin C Measurement ................................ ................................ ................... 97 Color Measurement ................................ ................................ .......................... 98 Oil in Juice Assay ................................ ................................ ............................. 98 Volatile Compounds Analysis ................................ ................................ ........... 98 Headspace sampling ................................ ................................ ................. 98 Gas Chromatography Mass Spectrometry (GC MS). .............................. 99 Rate of Reaction ................................ ................................ ........................ 99 Statistical Analysis ................................ ................................ ............................ 99 Results and Discussion ................................ ................................ ......................... 100 Microbial Analysis ................................ ................................ ........................... 100 DO Consumption ................................ ................................ ............................ 100 AA Degradation ................................ ................................ .............................. 102 Color ................................ ................................ ................................ ............... 104 Volatile Compounds ................................ ................................ ....................... 105 Conclusions ................................ ................................ ................................ .......... 108 4 COMPREHENSIVE RESULTS AND FUTURE WORK ................................ ......... 134 APP ENDI X A DISSOLVED OXYGEN CONTENT IN NFC OJ (HAMLIN CULTIVAR) ................. 138

PAGE 8

8 B COLOR CHANGES IN NFC OJ (HAMLIN CULTIVAR) STORED 3 DAYS AT 40 C ................................ ................................ ................................ .......................... 140 C GAS SPARGING EXPERIMENTAL SETUP ................................ ......................... 141 D NORMALIZATION OF VOLATILE COMPOUNDS IN NFC OJ (HAMLIN C ULTIVAR) AND PCA ANALYSIS ................................ ................................ ....... 142 E POINTS TO CONTROL FOR REDUCING VARIABILITY DURING SAMPLE PREPARATION ................................ ................................ ................................ .... 160 LIST OF REFERENCES ................................ ................................ ............................. 162 BIOGRAP HICAL SKETCH ................................ ................................ .......................... 172

PAGE 9

9 LIST OF TABLES Table page 1 1 Ascorbic acid 1 ) ................................ .. 48 1 2 Reaction rate constants for AA degradation obtained under reducing and oxidizing conditions at 90 C. ................................ ................................ ............. 49 1 3 Summary of selected studies of ascorbic acid degradation in juices in which the effect of oxygen was assessed. ................................ ................................ .... 50 1 4 Comparison of different methods used to measu re DO ................................ ...... 53 1 5 Selected commercially available O 2 scavengers used in liquids ......................... 54 2 1 Microbial counts in NFC orange juice during 60 day s of storage at 5 C ............ 76 2 2 Pseudo first and second order rate of DO consumption in not from concentrated orange juice during storage at 5 C ................................ ............... 77 2 3 Pseudo first rate constant of AA oxidation and DHA formation in not from concentrated orange juice during storage at 5 C ................................ ............... 77 2 4 Color change in NFC orange juice during storage at 40 C and expressed as L, a and b with respect to control orange juice. ................................ ............ 77 2 5 Proportion of aroma active and volatile compounds in NFC orange juice at day 1 of storage at 5 C ................................ ................................ ...................... 78 2 6 Aroma active compounds identified in commercial NFC orange juice with different levels of DO after 1 day of storage at 5 C. ................................ .......... 79 2 7 Vol atile compounds identified by GC MS in NFC orange juice with different levels of DO after 1 day of storage at 5 C. ................................ ........................ 81 2 8 Volatile compounds identified by GC MS in NFC orange juice with differe nt levels of DO after 60 days of storage at 5 C. ................................ .................... 84 3 1 Dissolved oxygen and ascorbic acid sampling schedule during storage study. 109 3 2 Pseudo first order rate constant (d 1 ) of dissolved oxygen consumption in NFC orange juice. ................................ ................................ ............................. 109 3 3 AA loss due to pasteurization and storage at selected temperatures. .............. 110 3 4 Pseudo first order rate constant (d 1 ) of AA degradation in NFC orange juice with air headspace. ................................ ................................ ........................... 110

PAGE 10

10 3 5 Color change in NFC orange juice du ring storage at 40 C and expressed as L, a and b with respect to control orange juice. ................................ .......... 111 3 6 Oil content in NFC orange juice at day 0 of storage at 40 C calculated using Scott oil method. ................................ ................................ ............................... 111 3 7 Normalization methods applied to volatile compounds of NFC orange juice .... 112 3 8 Comparison of raw peak area and two differe nt normalizations applied to selected volatile compounds in control not from concentrated orange juice at day 0 of storage at 40 C. ................................ ................................ ................. 113 3 9 Volatile compounds identified in NFC orange juice at different DO content at day 0 of storage at 40 C.. ................................ ................................ ................ 114 3 10 Volatile compounds identified in NFC orange juice at different DO content at day 0.5 of storage at 40 C. ................................ ................................ .............. 118 3 11 Volatile compounds identified in NFC orange juice at different DO content at day 6 of storage at 40 C. ................................ ................................ ................. 122 D 1 Volatile compounds identified in NFC orange juice at different DO content at day 0 of storage at 40 C.. ................................ ................................ ................ 142 D 2 Volatile compounds identified in NFC orange juice at different DO content at day 0.5 of storage at 40 C.. ................................ ................................ ............. 146 D 3 Volatile compounds identified in NFC orange juice at different DO content at day 6 of storage at 40 C. ................................ ................................ ................. 149 D 4 Volatile compounds id entified in NFC orange juice at different DO content at day 0 of storage at 40 C. ................................ ................................ ................. 155 D 5 Volatile compounds identified in NFC orange juice at different DO content at day 0.5 of storage at 40 C. ................................ ................................ .............. 156 D 6 Volatile compounds identified in NFC orange juice at different DO content at day 6 of storage at 40 C. ................................ ................................ ................. 157

PAGE 11

11 LIST OF FIGURES Figure page 1 1 Ascorbic acid degradation mechanism modified from Davey and others (2000). ................................ ................................ ................................ ................ 56 1 2 Reactions catalyzed by catechol oxidases. ................................ ........................ 56 1 3 Reactions catalyzed by laccases. ................................ ................................ ....... 57 1 4 Vacuum deaerator scheme. ................................ ................................ ................ 58 1 5 Gas sparging method scheme. ................................ ................................ ........... 59 1 6 Hollow fiber membrane separator (modified from (Polypore 2007). ................... 60 2 1 Dissolved oxygen (D O) concentration in NFC orange juice during 60 days of storage at 5C. ................................ ................................ ................................ .... 86 2 2 Kinetic models fitted to dissolved oxygen (DO) concentration in NFC orange juice during 60 days of storage at 5C ................................ ................................ 87 2 3 Changes in (A) ascorbic acid and (B) dehydroascorbic acid in NFC orange juice during 60 days of storage at 5C. ................................ ............................... 88 2 4 Aroma a ctive compounds organized by aroma descriptor in NFC orange juice at day 1 of storage at 5C. (A) deaerated, (B) control, (C) oxygen saturated and (D) air headspace orange juices. ................................ ................................ 89 2 5 PCA sco re plot of volatile compounds in orange juice at different DO concentrations at day 1 and 60 of storage at 5 C ................................ .............. 90 2 6 Loading plot of volatile compounds in orange juice at different DO concentr ations at day 1 and 60 of storage at 5 C.. ................................ ............ 91 3 1 Flow diagram of sample preparation of not from concentrated orange juice (Hamlin var.) ................................ ................................ ................................ ..... 126 3 2 Different kinetic models applied to DO data of not from concentrate orange juice (cultivar Hamlin) stored at 5 C. ................................ ............................... 127 3 3 First order kinetic data of not from concentrate orange j uice (Hamlin var.) stored at 5 C ................................ ................................ ................................ ... 128 3 4 Ascorbic acid and dissolved oxygen content in not from concentrated orange juice (Hamlin var) during storage at 5 C.. ................................ ........................ 129

PAGE 12

12 3 5 Ascorbic acid and dissolved oxygen content in not from concentrated orange juice (Hamlin var) during storage at 40 C ................................ ........................ 130 3 6 Proportion of volatile compou nds identified in NFC orange juice stored at 40 C. (A) 0 days of storage and fresh squeezed orange juice, (B) 6 days of storage.. ................................ ................................ ................................ ........... 131 3 7 Score plot of PCA applied to volatile compounds from NF C orange juice stored at 40 C normalized with total peak area of volatile compounds ............ 132 3 8 Loading plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C normalized wi th total peak area of volatile compounds. ........... 133 D 1 Score plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C normalized with internal standard. ................................ ........... 153 D 2 Loading plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C normalized with internal standard. ................................ ........... 1 54 D 3 Score plot of PC A applied to volatile compounds from NFC orange juice stored at 40 C normalized with day 0. ................................ ............................. 158 D 4 Loading plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C norma lized with day 0. ................................ ............................. 159

PAGE 13

13 LIST OF ABBREVIATION S AA Ascorbic acid AcCys N acetyl L cysteine APDA Acidified potato dextrose agar CIP Clean in place Cys Cysteine DAD Diode array detector DHA Dehydroascorbic acid DO Dissolved oxy gen GC MS Gas chromatography mass spectrometry GC O Gas chromatography olfactometry HMF Hydroxymethyl furfural MF Methylfurfural NFC Not from concentrated OSA Orange serum agar PCA Plate count agar SPME Solid phase micro extraction

PAGE 14

14 Abstract of Disser tation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECT OF DISSOLVED OXYGEN AND DEOXYGENATION IN QUALITY OF ORANGE JUICE By Rosala Garc a Torres August 2010 Chair: Jos I. Reyes De Corcuera Major: Food Science and Human Nutrition Orange juice was the most consumed fruit juice in the 2007/08 season representing 50% (14.4 L, Single Strength Equivalent per capita ) of the total per capit a consumption of fruit juices. Because processing conditions influence the quality of the end product, it is important to understand the changes that orange juice experiences during processing. Deaeration is a processing step applied immediately before pas teurization to remove air in particular oxygen, incorporated into orange juice during extraction or mixing. Vitamin C is believed to be the compound the most affected ( oxidized ) by dissolved oxygen (DO) in orange juice. However, there is a lack of informa tion about changes in color and aroma under different levels of DO The objective of this stu dy was to understand the effect of initial DO presence of air in the headspace and storage temperature on vitamin C, color and aroma of not from concentrated (NFC ) orange juice in order to preserve fresh like characteristics. The first storage study correlated changes in vitamin C (ascorbic and dehydroascorbic acid) and aroma compounds of commercial NFC orange juice ( cultivar Valencia,) with selected initial DO lev els during storage for 60 days at 5 C in amber

PAGE 15

15 glass containers. The same study also evaluated the effect of having air in the headspace. The selected initial levels of DO in juice with no headspace were: deaerated (42.3 + 2.7 M ), oxygen saturated (494.7 + 1.8 M ), control (177.6 + 7.9 M ) and juice with air in the headspace (180.3 + 9.3 M ) The juice with air headspace had the highest l oss of ascorbic acid (AA) of 42 % after 60 days of storage at 5 C. Pseudo f irst order r ate constant of DO consumption was 2.8x10 1 4.1x10 2 d 1 No chang es in color were visually perceived among treatments or over time. Pseudo first order rate constant of AA oxidation at 5 C was 1 .9x10 2 3.0x10 3 Aroma active and volatile compounds were analyz ed in all juices by GC O and GC MS using solid phase micro extraction (SPME) headspace sampling. Among aroma active compounds, methional was perceived at higher intensity in orange juice samples with air headspace compared to control only N onanal had the main contribution to aroma intensity of control, oxygen saturated and air hea d space orange juices. At day 1 of storage elemene, selinene and 3 carene had the highest amounts in deaerated orange juice compared with the other juices. At day 60 of storage, the content of most of the compounds decreased with respect to day 0 except for E 2 hexenal and limonene oxide whose content increased in orange juice with air headspace P rinc ipal component analysis ( PCA ) of volatile compounds allowed to differentiating orange juices deaerated a nd with air headspace between days 0 and 6 of storage. A second storage study with not from concentrated ( NFC ) orange juice ( cultivar Hamlin) described the impact of DO and storage temperature on volatile compounds, AA and color during storage at 5, 13, 21.5, 30.5 and 40 C.

PAGE 16

16 The DO content in the juice samples was modified by gas sparging, followed by pasteurization and storage in amber glass containers at the specified temperatures. Similar to the previous study, the different initial levels of DO in juice with no headspace were: deaerated ( 58.7 + 6.3 M ), oxygen saturated (209.7 + 76.3 M ), control ( 152.4 + 21.3 M ), and juice control with air headspace ( 340.9 + 64.5 M ). The juice with the highest loss of AA was the one air headspace with 100%, 73.5% and 54.4% of AA loss af ter 60 days of storage at 5 C, 3 days of storage at 40 C and 6 days of storage at 30.5 C respectively. Color change was perceived by direct observation only in air headspace orange juice stored 3 da ys at 40 C, C olor change was also measured using a and b parameters than changed toward the red and blue respectively. N o correlation with dissolved oxygen content was established. Volat ile compounds were measured by SPME headspace and GC MS after gas sparging and pasteurization and identified by matching with database and LRI values. Each replicate was obtained from a different batch of orange juice and relative standard deviations of mo re than 50 were observed in the raw peak areas of the replicates of the same sample. Although it was initially thought that differences in oil content could be the cause of the variability, differences between replicates were not detected by Scott oil meas urement. Three d ifferent normalization s were applied to volatiles raw peaks: normalization with 4 heptadecanone added as internal standard to all the samples, normalization with total peak area and normalization with respect to day 0. Normalization with t otal peak area gave relative standard deviations smaller than the other two normalizations and was selected for further ANOVA and PCA analysis. pinene content was higher in juice with air headspace juice, myrcene and elemene

PAGE 17

17 were lower in control and oxygen saturated juices and octanal was lower in deaerated and oxygen saturated juices. Furfural, terpineol, terpineol were only detected at day 6, and 1 hexanol content increased over the time. These o bservations were significantly different and confirmed with PCA. The results of this research suggest that since dissolved oxygen does not have an important impact on vitamin C, color and aroma of orange juice, deaeration of juice may not be necessary in c itrus processing for short term storage Rather, removing air headspace from storage containers or replacing air headspace with nitrogen will reduce the oxidation of vitamin C during storage. The results of this research may possibly be applied to other fr uit juices nectars or purees containing oxygen sensitive compounds and high concentration of vitamin C

PAGE 18

18 CHAPTER 1 INTRODUCTION AND LIT ERATURE In 2007, the U.S. total fruit production was 26.6 million metric tons. Citrus and non citrus production were 9. 4 and 15.3 million metric tons, respectively. The value of fruit production was $14.5 billion, with citrus accounting for $3.1 billion; this is similar to the $14.4 billion value of vegetables and nuts (USDA 2007; 2008b) Those most commonly processed into juice were orange, grape, apple, grapefruit, and peach. In 2007, 7% of the fruit production was consumed as juice in the United States. The U.S. per capita consumption of fruit juices for the same year was 28.9 L of single strength equivalent (SSE). SSE is the volume of juice at the soluble solid content (SSC) at which it is consumed (for reconstituted orange juice, 11 Brix). In fruit juices, quality i s determined in terms of SSC, acidity, nutritional content, color, flavor, microbial safety, cloud stability, fiber content, sugar content, rheological properties, defects, and adulteration. Air is naturally present in the intercellular spaces of fruits. During fruit maceration, homogenization, and juice extraction cells are crushed, the cell wall is disrupted and the air is mixed into the juice. Air can be present as dissolved gas in solution or associated with the pulp particles, for example, in orange j uice (Joslyn 1961) During cell disruption, metabolites and enzymes that are normally compartmentalized are mixed, producing chemical and biochemical reactions. Oxygen in air, present in the spaces between the juice vesicles and from the surroundings, satur ates the juice producing oxidation reactions that often result in browning, changes in aroma, and loss of nutritional value. These reactions are accelerated by the increase in temperature during pasteurization and reduce the overall quality of the product during storage. Interactions of oxygen with the different food components result in several sequential and parallel reactions.

PAGE 19

19 Product interactions make it difficult to identify and to differentiate the compounds produced by direct reaction with oxygen fro m the ones produced indirectly in subsequent reactions. The common approach has been to follow the changes of 1 or 2 compounds at a time in the juices or in model solutions. Oxygen Solubility For the purpose of this review, DO is considered to be the molec ular or diradical triplet oxygen, which is the most abundant and stable form of oxygen. Singlet oxygen may also be present in food and requires less energy activation to produce oxidation than triplet oxygen and has a greater rate of oxidation of food subs trates. Min and Boff ( 2002) reviewed the chemistry and mechanisms of formation; analytical techniques for detection and quantification; reacti on mechanisms with food components such as lipids, proteins, and vitamins; quenching mechanisms to reduce oxidation; and kinetics of singlet oxygen. The main differences between triplet and singlet oxygen are in the electron arrangement and in the energy l evel. Electrons located in the external orbital are paired in antiparallel spins in singlet oxygen and in parallel spins in triplet oxygen. Energy level for singlet oxygen is 22.5 kcal mol r elative higher to triplet oxygen. Moreover, the detection of singlet oxygen is difficult because of its short life (50 to 700 s). Detection requires the use of electron spin resonance spectroscopy (ESR), laser deflection calorimetry, or time resolved singlet oxygen detection (Min and Boff 2002) that a gas dissolved in a liquid is in equilibrium with the gas above the liquid surface. The concentration of the gas dissolved in the liquid is proportional to the partial pressure of the same gas in the vapor phase (Eq. 1 1):

PAGE 20

20 (1 1) where X A P A and H ( T ) are concentration of gas A in the liquid phase, partial pressure of (Gill and Menneer 1997; Ringblom 2004) Oxygen solubility increases with increasing atmospheric pressure an d decreases with increasing temperature and salinity (Lewis 2006b) There are several metho ds used to quantify DO, but most have been developed for water analysis. A thorough compendium of these methods is the U.S. Geological Survey (USGS) manual that includes tables of oxygen solubility in water at temperatures from 0 to 40 C, pressures from 600 to 795 mmHg, correction factors for salinity at temperatures of 0 to 35 C, and conductivity of 0 to 67000 S cm (at 25 C). Oxygen solubility can be calculated using Eq. 1 2 (Weiss 1970; Lewis 2006b) (1 2) where C T and S are the oxygen solubility in milli liter per liter or milliliter per kilogram from water saturated air at one atmosphere, the absolute temperature in K and the salinity in gram per kilogram, respectively. The estimated precision from this equation is 0.015 mL L of oxygen. The standards u sed for calibration are air saturated water and sodium sorbate saturated solutions as 100% and 0% of DO concentration, respectively (Singh and others 1976; Kennedy and others 1992) and the oxygen solubility tables for wa ter (Lewis 2006b) are used as references for calculating concentrations in milliliter per l iter at specified temperature and pressure conditions Sadler and others (1988) found that the presence of solutes such as sugars, NaCl, and citric and ascorbic acids at concentrations similar to fruit juices reduce the oxygen solubility by less than 10 % at

PAGE 21

21 concentrations similar to that of fruit juices. They developed a method to predict oxygen solubility in juices as a function of sugar concentration at temperatures between 4 and 40 C at atmospheric pressure. The method uses a multiple regression equa tion (Eq. 1 3) able to predict DO concentration within 5% of literature values and the coefficient of determination of R 2 = 0.949. The limitation of this method is that it only applies to air saturated sugar/juice solutions and does not incorporate pressure effects. (1 3) where [O 2 ] is the oxygen concentration in milliliter per kilogram, B is soluble solids concentration in Brix, and T is temperature in C. Dissolved oxygen concentration values determine the degradation rate of severa l compounds of interest in fruit juices. DO measurem ent is challenging as oxygen is present in the air and can easily incorporate int o the liquid food during sample preparation and measurement. This can affect DO quantification and make it necessary to man age the samples under an oxygen free atmosphere. Interaction of Dissolved Oxygen with Food Components Ascorbic Acid One of the most important food components in juice that is affected by DO is L ascorbic acid (L AA). L AA is a natural antioxidant present i n most fruits. Table 1 1 summarizes the ascorbic acid content in some fresh fruits, juices, and purees. Orange juice, guava pulp, and grapefruit juice have the highest ascorbic acid concentrations in fruit juices

PAGE 22

22 Degradation Ascorbic acid degradation in o range juice is a result of aerobic and anaerobic mechanisms (Hughes 1985; Johnson and others 1995) that occur simultaneously (Kennedy and others 1992; Baiano and others 2004) Figure 1 1 presents the oxidation and hydrolysis reactions involved in the ascorbic acid degradation. Ascorbic and dehydroascorbic acid are the active f orms of vitamin C. Ascorbic acid in aqueous solution is oxidized to dehydroascorbic acid (DHA) in a 2 step process that produces monodehydroascorbate (MDHA), also called ascorbate free radical. As intermediate MDHA can be reduced back to ascorbic acid or f orm DHA, which has a pro oxidant effect. The oxidation of AA to DHA is a reversible process; the direction is affected by the presence of oxidizing or reducing agents (Serpen and Gkmen 2007) Dehydroascorbic acid also undergoes irreversible hydrolytic ring cleavage to produce 2, 3 diketogulonic acid (2,3 DKG) (Figure 1 1). Equ ations 1 4 and 1 5 show the simplified mechanism adopted to describe the kinetic rate of the L AA degradation under aerobic conditions (Mack and others 1976; Singh and others 1976; Lin and Agalloco 1979; Trammell an d others 1986; Patkai and others 2002; Serpen and Gkmen 2007) (1 4 ) (1 5) where k 1 k 2 and k 3 are the reaction rates of the ascorbic acid oxidation, dehydroascorbic acid reduction, and dehydroascorbic acid hydrolysis, respectively. In ascorbic acid aqueous solutions and in the presence of air, the reconversion of DHA to L AA ( k 1 ) is very slow compared to the L AA oxidation to DHA ( k 2 ) and th e subsequent

PAGE 23

23 hydrolysis of DHA ( k 3 ). This is similar to reactions o ccurring in the presence of oxygen and Fe 3+ as oxidizing agents. But in the presence of a reducing agent, the conversion of DHA to L AA is faster than the other 2 reactions, as Serpen and G kmen (2007) concluded by applying Laplace transformation to calcul ate the kinetic rates of the reactions that appear in Table 1 2. Patkai and others (2002) suggested that ascorbic acid degradation to 2,3 diketogulonic acid under aerobic conditions at pH lower than 5 is a 1 step reaction, but at pH 8, oxidation proceeds as a 2 step reaction (Eq. 1 4 and 1 5). Perhaps at low pH, the ascorbic acid degrada tion process also consists of 2 reactions, but the dehydroascorbic acid degradation ( k 3 ) is faster than the DHA formation ( k 1 ). L AA aerobic degradation during storage is promoted by temperature increase, presence of copper, iron, and alkali (Lin and Agalloco 1979; Davey and others 2000) increased light exposure (Mack and others 1976; Singh and other s 1976; Lin and Agalloco 1979; Lee and Kader 2000) differences in pH (Lin and Agalloco 1979; Patkai and others 2002) and increase of DO content (Mack and others 1976; Singh and others 1976; Lin and Agalloco 1979; Trammell and others 1986; Kennedy and others 1992; Lee and Kader 2000) The combined effect of pH and temperature on L AA degradation under aerobic and anaerobic conditions for L AA aqueous model solutions was described with Eq. 1 6 and 1 7, respectively Although is not clear how AA concentration is considered in these equations Equation 1 6 applies for pH 3 to 7 and Eq. 1 7 for pH > 3 (Lin and Agalloco 1979) (1 6) valid for aerobic conditions and pH 3 to 7

PAGE 24

24 (1 7) valid for anaerobic conditions and pH > 3 Under anaerobic conditions, L AA is degraded to furfural at a slower rate than aerobic degradation and a 1st order kinetics has been suggested (Patkai and others 2 002) However, the difference in rates under aerobic and anaerobic conditions has not been quantified (Baiano and oth ers 2004) Effect of dissolved oxygen concentration Although the scope of this dissertation is the degradation of ascorbic acid under aerobic conditions, it is important to mention that anaerobic degradation of ascorbic acid also affects the quality of fr uit juices. For example, at pH below 4 and in the absence of oxygen, the multistep decomposition of L AA to furfural has been associated with browning in stored citrus products and model systems (Rouse ff and others 1992) Kinetic models have been applied to describe the rate of ascorbic acid degradation (Eq. 1 4 and 1 5) but they are not conclusive about the kinetic order. One reason may be because they are based only on L AA concentration, while there are other substrates that may be limiting the reaction because they are present in smaller amounts. For example, DO concentration in orange juice is 10 times smaller than L AA concentration. The following section discusses the kinetic models that have bee n applied to ascorbic acid oxidation. The relationship between the rate of L AA degradation under aerobic conditions or oxidation and the concentration of DO in liquid foods has been studied since the 1930s. Eddy (1936) studied the role of oxygen in the destruction of a reducing factor named

PAGE 25

25 the remov al of DO is more effective in protecting L AA oxidation during processing than during storage. Kefford and others (1959) followed the L AA degradation of canned orange juice during pasteurization, early storage (0 to 100 h), and later storage (200 to 9000 h) at 30 C on oxygen saturated, air saturated, and oxygen removed conditions. During pasteuriza tion, a loss of AA of 46%, 7%, and not significant loss of L AA degradation in oxygen saturated, air saturated, and oxygen removed, respectively, was observed. The storage study was divided into early storage (when DO is present) and late storage (DO is ab sent). During the early storage, the L AA degradation followed a linear trend and rates were higher than in later storage that was described with a polynomial of 2nd order. The conclusion of the study was that in the canned juice system there is competitio n for oxygen between corrosion reactions and ascorbic acid and other compound oxidations that affect color and flavor. This idea applies to different juices and packaging materials on which a careful analysis of all the oxygen related reactions will give a comprehensive characterization of the system. Over time, L AA degradation due to DO occurs at the beginning of the storage period ranging from hours to weeks depending on storage conditions such as temperature, light intensity, and permeability of packag ing. For example, in mango juice ( cultivar Ogbomoso ) stored at 6 C for 8 wk, glass protected L AA better than polyethylene film and PET having a loss of ascorbic acid of 11%, 68%, and 50%, respectively (Alaka and others 2003) Kennedy and others (1992) found that after 64 d of storage of single strength orange juice (SSOJ) in Tetra Brik cartons at 4, 20, 37, 76, and 105 C, DO concentrations dropped to 2, 1.5, 0.5, and less than 0.5 mg L respectively. Increase of initial DO concentration increases the rate of L AA

PAGE 26

26 degradation. In SSOJ at different DO levels, related to processing conditions, 315.6 M (10.1 mg L ; oxygen saturated juice), 203.1 M (6.5 mg L ; normal plant operation co nditions), 56.2 M (1.8 mg L ; commercially deaerated orange juice), and 18.75 M (0.6 mg L ; lowest level reached using a gas equilibration technique), and stored for 22 wk at 22 C in an inert atmosphere (nitrogen flushed), the percent of L AA lost was linearly related to the initial DO concentration over the time, suggesting a 1st order reaction with respect to DO. The maximum loss was 32% at 312.5 M (10 mg L ) of initial DO after 22 wk (Trammell and others 1986) The effect of te mperature and DO in L AA degradation under regular storage conditions was evaluated in SSOJ packed in Tetra Brik cartons at a 4.5 mg L 1 initial DO concentration. L AA loss after 64 d was 39.6%, 51.4%, and 88.1% at 4, 20, and 37 C, respectively. In contrast, at 76 and 105 C, L AA degradation was 98% after 6 d and 96.4% after 3 d, respectively (Kennedy and others 1992). Ascorbic acid co ncentration in orange juice decreased by 86% when stored exposed to fluorescent light for 84 d at 4 C in an oxygen impermeable material, compared to 50% loss when protected from light (Nelson 2005) Table 3 3 summarizes different studies about ascorbic acid degradation in model solutions and citrus juices in which oxygen was removed or concentration was varied at different levels. A total of 8 of the 17 studies shown in Table 1 3 agree in describing L AA oxidation as a 1st order reaction under different DO conditions ranging from anaerobic conditions (Johnson and others 1995; Rojas and Gerschenson 2001) to gradual decrease of oxygen concentrations (Lewis and McKenzie 1947; Khan an d Martell 1967; Robertson and Samaniego 1986; Baiano and others 2004; Dhuique

PAGE 27

27 Mayer and others 2007) However, in some cases a zero order (Kennedy and others 1992) or 2nd order (Singh and others 1976; Hsieh and Harris 1993) seemed to describe the kinetics as well. These apparent contradictions are due to the fact th at when DO is present in limited amounts, particularly at concentrations below saturation, it becomes a limiting reactant of the ascorbic acid oxidation (Kennedy and others 1992). Under these conditions, 1st order models do not fit properly because they on ly consider one reactant concentration, either ascorbic acid or DO. Under limited oxygen concentrations, 2nd order kinetic models may better describe the ascorbic acid oxidation than 1 st order models (Mack and others 1976; Singh and others 1976). A model t hat accounts for DO and that can potentially be applied to fruit juices was reported in a study on an infant formula containing 55 mg L 1 of L AA. A 2nd order kinetic model (Eq. 1 8) described the decrease in both the ascorbic acid and DO concentration ove r time (Singh and others 1976): (1 8) where [ AA ] 0 and [ DO ] 0 are initial concentrations of L AA and DO, respectively; [AA] and [DO] are ascorbic acid and DO concentrations at time t respectively; and X is the amount of L AA or DO th at has reacted at time t This equation was solved using the computer program KINFIT. The 2nd order k values at different light intensities and initial DO combinations had values between 1.17 10 and 18.16 10 L mg h 1 and activation energies were not calculated because experiments were performed at 7.2 C only.

PAGE 28

28 Although a 2nd order kinetics described correctly the ascorbic acid oxidation, a reversible 2nd kinetic order would be even bette r since it is also considering the reversible conversion of DHA to L AA. The previously mentioned kinetic studies did not consider the reversible nature of dehydroascorbic acid and assumed only 1 reaction direction. Kinetic studies that consider the ascorb ic acid oxidation as a reversible reaction are scarce. Serpen and Gkmen (2007) considered the oxidation of AA to DHA as a reversible reaction and calculated the equilibrium between the rates of the reversible oxidation reaction and the subsequent DHA hydr olysis in L AA solutions (Table 1 2). Another approach to calculate the rate of ascorbic acid oxidation is to analyze the reaction using different kinetic orders over time. For example, Sakai and others (1987) applied a kinetic model that consists of 2 con secutive reactions (Eq. 1 9) of different kinetic order that applies only under constant DO concentration. (1 9) Parameters A B C k A and k B are ascorbic acid, intermediate product, oxidized product, zero order rate, and 1st order rate constants, respectively, B and C products are not defined. B ut based on the chemistry (Figure 1 1), we can assume that B is the monodehydroascorbate (MDHA) and C is the dehydroascorbic acid. Under those definitions, the 1st reaction is zero order wit h respect to ascorbic acid concentration and the 2nd reaction is 1st order with respect to MDHA. This model calculates the concentration of ascorbic acid (A) over time based on the measurement of the initial concentration of ascorbic acid and the ratio of the rate constants as shown in Eq. 1 10 and 1 11.

PAGE 29

29 (1 10) (1 11) [ A ] 0 is the initial concentration of ascorbic acid and k A and k B were defined in Eq. 1 9. This metho d was validated by comparing the ascorbic acid concentration values obtained using Eq. 1 10 and 1 11 with those obtained from aqueous ascorbic acid solutions at 30 C. To describe ascorbic acid oxidation in orange juice under air saturation Manso and other s (2001) proposed a kinetic model that considers the degradation of AA to DHA as a 1st order reaction, the reconversion of DHA to AA as a 2nd order reaction, and DHA degradation to 2, 3 DKG as a 1st order reaction. However, they pointed out that there is n o evidence that DHA degradation follows 2nd order kinetics. Manso and others (2001) also proposed a Weibull model (Eq. 1 12) that fitted the experimental data well ( R 2 adj > 0.995). (1 12) where C a C ia and are the L AA concentration at time t the initial L AA concentration, a scale constant (which inversely corresponds to the reaction time constant), and the shape constant or behavior index, respectively. A particular case of the Weibull function w hen = 1, is equivalent to the 1st order model. However, when

PAGE 30

30 one or more sequential reactions take place, this model clusters them into the shape constant without providing any mechanistic information. The temperature effect on any chemical reaction can be quantified with the activation energy ( E a ) calculated using the Arrhenius equation. From the kinetic studies in L AA degradation presented in Table 1 3, only 6 studies reported E a The temperature ranges between 20 and 124 C, and at least 3 different t emperatures were used in the calculations. E a ranged from 35.9 to 143.5 kJ mol but similar E a values (38.6 against 35.9 kJ mol ) were reported for citrus juices (SSOJ against orange tangerine juice) even when different models (Weibull against 1st kinetic order) and different temperature ranges were used (below 45 C against highe r than 50 C) (Manso and others 2001; Dhuique Mayer and others 2007). Braddock and Sadler (1989) used the Arrhenius approach (Eq. 1 13) to estimate L AA loss at all stages, at different temperatures during orange juice concentration in a 7 stage thermally accelerated short time evaporator: (1 13) where k is the reaction constant in percent per second and T is the absolute temperature in K The experimentally determined DO level was below 0.1 mg L after stage 3; estimated L AA loss after the 7 stages was 3.7 mg L ; reaction constant values ranged between 5.1 10 4 to 4.6 10 s for 76 to 86 C, respectively. In addition to DO concentration, intrinsic variability in the fruit such as culitvar maturity, native soil constituents, and processing conditions such as equipment and container material affect fruit juice composition including pH and trace metals. Differences in pH affect the number of ascorbic acid oxidation steps (1 or 2 steps) as

PAGE 31

31 me ntioned in the section Degradation (Patkai and others 2002). Fe 3+ and Cu 2 + catalyze ascorbic acid oxidation (Khan and Martell 1967) For example, 2 mg L Fe 3+ in an ascorbic acid solution increased L AA loss by 12.5% compared to the control after incubating for 6 h at 90 C (Serpen and Gkmen 2007). Furthermore, attempts to correlate the rates of DO consumption and L AA degradation in pure solutions with those in fruit juices are most likely erroneous because several other reactions occur simultaneously in fruit juices. In view of the different experimental conditions reported in the literature and the order of magnitude differences in reported apparent k inetic parameters and activation energy (see Table 1 3), it appears difficult, if not impossible to extrapolate reported data to industrial processing. However, valuable information can be drawn from kinetic studies that account for the effect of 1 single variable on the system such as DO (Robertson and Samaniego 1986), light (Singh and others 1976), sucrose, and pH (Hsieh and Harris 1993). Juice degradation in the presence of oxygen can be described as several reactions competing for the DO. The following sections describe the changes in color and aroma due to degradation of L AA and other compounds in the presence of oxygen in the juices. Color Nonenzymatic browning Commercially produced fruit juices receive a ther mal treatment during processing to have a long shelf life. The heating step reduces the microbial load (5 log reduction or more ) and inactivates the enzymes present in the raw fruit. However heating promotes changes in color and flavor. Nonenzymatic browning, including Maillard type reactions (Cmara and others 2003; Reineccius 2006) ascorbic acid (Johnson and others 1995),

PAGE 32

32 and sugar degradations (Rouseff and others 199 2) are important reactions in pasteurized fruit juices. Oxygen can promote browning, but is not directly responsible for it (Shaw and others 1993) is by way of ascorbic acid oxidation to dehydroascorbic acid with further formation of decomposition products such as furfural that have been associated to browning For example, in a stud y on SSOJ at initial DO levels of 0.6 to 10.1 mgL browning and ascorbic acid loss were linearly correlated with the initial DO concentration during 22 d of storage at 22 C (Trammell and others 1986). In not from concentrate (NFC) orange juice packaged in pouches of oxygen scavenger (EVOH/oxygen scave nger/cast polypropylene) and oxygen barrier (EVOH/cast polypropylene) film stored at 25 C, the browning index (absorbance at 420 nm) and the percentage of loss of ascorbic acid were lower in packages with oxygen scavenger (0.35% and 24%, respectively) tha n in packages with only oxygen barrier (0.44% and 24%, respectively) after 365 d of storage. But at 4 C, the differences due to packaging material were less evident (Zerdin and others 2003) Adams and Brown (2007) reviewed different discoloration processes in raw and processed fruits and vegetables, including enzymatic and nonenzymatic browning, and they also concluded that in fruit juices such as lemon, orange, and pineapple, nonenzymatic brownin g is related to ascorbic acid and sugar degradations. Sulfur containing amino acids have antioxidative and antitoxic effects due to their ability to act as reducing agents, oxygen scavengers, and good nucleophiles (Friedman and Molnar Perl 1990) These amino acids can b e used to inhibit enzymatic and nonenzymatic browning (Friedman and Molnar Perl 1990; Molnar Perl and Friedman

PAGE 33

33 1990; Haleva Toledo and others 1999) Formation of (hydroxymethyl) furfural (HMF) and methylfurfural (MF), both related to browning in stored citrus juices, was studied in 1 to 10 mM L cysteine (Cys) and N acetyl L cysteine (AcCys) buffered solutions (pH 3, 5, or 7) and incubated at 70 C for 48 h. HMF and MF formation was inhibited significantly ( P < 0.01) at pH 3 and 5 in the presence of Cys and AcCys; but at the same pH, browning (measured as absorbance at 420 nm) was promoted (Haleva Toledo and others 1999). In another study (Naim and oth ers 1994) Cys and AcCys (0.5 to 2.5 mM) reduced browning (optical density at 420 nm) significantly ( P < 0.05) in SSOJ stored for 14 d at 45 C. The increase in browning at higher concentrations of Cys and AcCys was attributed to products formed from gluc ose or rhamnose Cys interaction at low pH and was not related to the presence of oxygen, since the same results were obtained under anaerobic conditions (Haleva Toledo and others 1999). Enzymatic browning Enzymatic browning consists of the oxidation of phe nolic compounds to o quinones that combine with amines and sulfur groups from proteins and reducing sugars to produce brown colored polymers (melanines). Enzymatic browning occurs when the cell structure is damaged and the phenolic compounds located in the vacuole are brought in contact with the enzymes responsible for the oxidation of phenols, which are located in plastids (Arts and others 1998; Burton 2003; Acree and Arn 2004; Franck and others 2007) In fruits a nd vegetables, the enzymes responsible for enzymatic browning are polyphenol oxidases (PPO), although a synergistic effect between PPO and peroxidases (POD) has been suggested. There are 2 types of PPO: catechol oxidase (EC 1.10.3.1) and laccase (EC 1.10.3 .2). Both contain copper in their structure and use

PAGE 34

34 molecular oxygen as a secondary substrate (Martinez and Whitaker 1995; Arts and others 1998; Acree and Arn 2004; Gmez and others 2006; Dhuique Mayer and others 2 007; Franck and others 2007) Catechol oxidase catalyses 2 reactions referred to as having cresolase and catecholase activity. They differ in that cresolase c onverts monophenols into ortho phenols and catecholase oxidizes o phenols to o quinones. (Figure 1 2), while laccase activity oxidizes p phenols and o phenols (Figure 1 3) (Arts and others 1998; Dhuique Mayer and others 2007). Enzymatic browning is not a concern in most fruit juices because the pasteurization process, when applied at the proper temper ature time conditions, inactivates the enzymes present in the raw fruit. Since this review is focused on pasteurized fruit juices, no further discussion of this is included here. Aroma The mixture of volatile compounds with aroma activity is responsible fo r the characteristic aroma of fruit juices. Dissolved oxygen may af fect the flavor of fruit juices by formation of precursors such as dehydroascorbic acid, a precursor of aldehydes formed during the Strecker degradation. It is difficult to quantify the eff ect of DO in aroma because there are biochemical pathways occurring during the ripening of fruit that do not involve oxygen and also produce aroma active volatile compounds. Perez and Sanz (2008) describ ed these biochemical pathways in detail and no further discussion is included in this review. Reactions in which oxygen is the reactant or that are consecutive to those affected by the presence of DO and produce aroma active compounds are described subsequ ently. The oxidation of L AA to DHA as previously mentioned produces color and flavor changes. DHA contains an dicarbonyl structure that participates in the Strecker

PAGE 35

35 degradation. Strecker degradation is the reaction between the dicarbonyl group from DHA and a primary amino acid to form an aldehyde with a carbon chain that is 1 carbon shorter than the precursor amino acid chain and that is aroma active. Well known examples of Strecker aldehydes are methional and acetaldehyde. Methional is derived is an important contributor to the cooked sensory attribute of canne d orange juices (Bradley and Min 1992; Reineccius 2006; Perez Cacho and others 2007) Acetaldehyde is formed from the amino acid alanine and has a fruity aroma (Cremer and Eichner 2000; Rega and others 2003) Depending on fruit cultivar and extraction method, the amount of peel oil present in or ange juice can be as high as 0.06%. But because the excess of peel oil produces a sensation, the desired concentration of peel oil in orange juice is between 0.02% and 0.03% (Braddock 1999) About 90% of citrus oil is d limonene. Because d limonene has a high aroma threshold, it doe s not contribute to the overall aroma perception, but serves as a carrier for other aroma active compounds (Perez Cacho and Rouseff 2008b) Even d limonene oxidation in the presence of molecular oxygen produces carveol and carvone (Kutty and others 1994) Of these 2 compounds, only the presence of carvone i n orange juice has been reported and a contribution to the aroma has not been attributed to it, even if the amount of carvone (110 ppb) was higher than its sensory threshold in water (6.7 ppb) (Moshonas and Shaw 1994) In a citral buffer containing d limonene and in the presence of oxygen, a decrease of 20% in d limonene concentration accounted for an increase of carvone and carveol of 200 and 100 ppm, respectively, duri ng the first 2 wk of storage (Kutty and others 1994).

PAGE 36

36 Trammell and others (1986) studied the effect of oxygen on SSOJ taste. The juice was pasteurized, aseptically filled in sterile glass bottles in a nitrog en atmosphere and stored at 22 C for 22 wk prote cted from light. Sensory evaluation was performed in juices with DO concentrations between 0.6 and 10.1mgL during storage. Juices with the lower DO concentration (0.6 mgL ) were statistically preferred over the high DO concentration (10.1 mgL ) until week 4 but not later. Researchers concluded that based on taste, deaeration did not extend the shelf life of the SSOJ. Methods of Measurement of DO The U.S. geological survey (USGS) standard field methods to measure dissolved oxygen are amperometric, luminescence, Winkler titration, and colorimetric methods (Table 1 4). Several limitations were associated wit h some of these methods. The industry standard technique for DO measurement uses a conventional Clark type electrode (amperometric method). Limitations associated with electrochemical probes include drifts in calibration due to oxygen consumption by the el ectrode, need for turbulent flow to prevent the formation of concentration gradients, a nd electrical interference. The most recently developed methods are solid state luminescence methods based on fluorescence peak quenching or multi frequency phase shift. Typically, fluorophore coatings are deposited on the tip of fiber optic probes. The most common fluorophores are ruthenium or platinum complexes (Sang Kyung and Icharo 1997) because they are excited and emit at wavelengths in the visible light range and they have high sensitivity to oxygen and large Stokes shifts. When oxygen molecules collide with a fluorophore in its excited sta te, the energy is transferred to the oxygen molecule in a nonradiative way, decreasing the fluorescence intensity. Ruthenium complex based sensors are the most common and are used for aqueous

PAGE 37

37 liquids and vapors. Sensors having platinum complex as a fluorop hore are the most sensitive and, therefore, are used for low levels of DO. Tang and others (2003) reported that Tris (4,7 diphenyl 1,10 phenanathroline) ruthenium(II) (Ru[dpp]32+) immobilized in Octyl triEOS/TEOS sol gel has a nitrogen to oxygen intensity ratio ( I N2/ I O2) of 16.48 Whereas, Yeh and others (2006) reported that oxygen sens itive dyes Platinum tetrakis pentafluorophenylporphine (PtTFPP) and platinum octaethylporphine (PtOEP) immobilized in same matrix Octyl triEOS/TEOS sol gel has I N2/ I O2 of 22 and 47, respectively. There are some commercially available ruthenium based sensor s that can be used for the determination of DO in organic liquids and vapors. Chu and Lo (2007) reported that PtTFPP and PtOEP immobilized in a support matrix n propyl TriMOS/TFP TriMOS has higher sensitivities compared to Octyl triEOS/TEOS sol gel. When calibrating a fiber optic probe, standards should have the same refractive index as the sample. Limitations of fiber optic sensors are sensitivity to surrounding light, photo bleaching of the fluorophore, and signal intensity dep endence of fiber bending Attempts to isolate the fluorophore coating from surrounding light using a dark overcoat resulted in slow response time, poor reproducibility and fouling. Another way of measuring oxygen using luminescent based sensor system is us ing phase modulation techniques. The phase shift between the exciting and the emitted light source is correlated with DO concentration (Barnes and others 1990; Jiang and others 2002) Phase measurements are more stable and essentially independent of the luminescence dye concentration and photobleaching which affect signal intensity (Ogurtsov and Papkovsky 1998)

PAGE 38

38 Methods of Removal of DO Methods of oxygen removal such as vacuum deaeration, nitrogen sparging, membrane deaerators, enzyme based deaerators, and oxygen scavengers are discussed in the subsequent sections. Vacuum D eaeration Vacuum deaeration is based on the reduction of the pressure of the gas above the juice. A s pressure in the juice headspace decreases, dissolved air is released from the liquid phase following deaeration method and it performs the dual role of removing oxygen and removing excess peel oil from orange juice before pasteurization (Braddock 1999). This operation juice undergoes as it enters the deaeration tank, producing a practically instantaneous separ ation. Figure 1 3 shows a diagram of a vacuum deaeration system that consists of a vacuum system, a vacuum chamber, juice feeding with a spray inlet, discharge of juice in the bottom of the tank and air and volatile compounds outlet at the top of the tank. In the citrus industry, to ensure a high elimination of air, juice is preheated to 50 to 60 C before deaeration. The pressure in the vacuum chamber and the inlet temperature are adjusted so that the inlet temperature is 2 to 5 C above the boiling point of the OJ at that pressure (Ringblom 2004). The 1st type of deaerator in the juice industry was the batch type that was later substituted by continuous centrifugal, spray, or film deaerators. In centrifugal deaerators, the juice is fed to a vacuum chamber containing a revolving pan, which lets the juice flow through small orifices in the rim. Spray deaerators also feed a chamber, but in this case juice is passed at high velocity through a s mall nozzle to disperse it into

PAGE 39

39 small droplets, as Figure 1 4 shows. In the film type deaerator, juice is dispersed by spreading a jet into a number of thin layers (Joslyn 1961). U.S. patent 5006354 established a deaeration method for fruit juices in which water and a concentrated source of fruit juice are deaerated separa tely and subsequently mixed. Water is deaerated at a temperature range of 50 to 60 C, with vacuum residence times between 1 and 3 min to reach a final DO concentration between 0 and 0.4 mg L Fruit juice concentrate is deaerated under milder conditions to avoid damaging volatile components and vitamins, at temperatures in the range of 0 to 5 C, a vacuum residence time of 3 to 6 min, and a final DO concentration lower than 3 mg L The vac uum pressure suggested is 0.1 bar (Rahrooh and Engel 04 09 1991) Gas Sparging Gas sparging consists of displacing the DO with another gas such as nitrogen or carbon di oxide. The gas is bubbled into the liquid or meets countercurrently with the liquid in a chamber. Figure 1 5 depicts gas being bubbled into the juice. The size of the bubbles of gas, the height of the chamber, and the flow rates of gas and liquid determine the rate and extent of oxygen removal. For example in OJ, around 99% of O 2 was removed using N 2 sparging compared with 77% of O 2 r emoved with vacuum deaeration (Joslyn 1961). Sparging with helium at a flow rate of 40 mL s for 30 s in one liter of aqueous buffer was more effective at removing DO than vacuum deaeration at 90 kPa (Degenhardt and others 2004) The disadvantage of this deaeration process is that flavor volatile components are removed and lost to the atmosphere (Jordan and others 2003)

PAGE 40

40 Membrane Deaerators Membrane gas separation is based on the use of a membrane as a barrier between the liquid and the gas phase; for example, between contaminated water and a sweeping air phase (Mahmud and others 1998) A membrane separator is a device that performs mass transfer between a gas and a liquid or 2 liquids without dispersion of t he phases on each other by passing the fluids on opposite sides of a microporous membrane. The membrane configuration can be a flat sheet or a hollow fiber (Gabelman and Hwang 1999) Hollow fiber configuration has a high surface to volume ratio making the membrane more efficient by providing a large contact surface area compared with the flat sheet configuration (Mahmud and others 1998). In wa ter deaeration, hollow fiber configuration membrane has a surface to volume ratio higher than conventional packed tower air stripping (Mahmud and others 1998). Since the 1970s, the capacity of selective hollow fibers to remove oxygen has been studied. It w as demonstrated that a polypropylene hollow fiber separator of 7.7 f t 3 volume removed 96% of DO in a flux of 37 854 L d 1 of water (Cole and Genetelli 1970) Applications in the food industry of hollow fiber membrane contactors are carbonation of beverages, nitrogena tion of beer to give a dense foam head, deoxygenation of beer to preserve flavor, as a low cost alternative to supercritical extraction, extraction of ethanol and other metabolites during fermentation, separation of products during enzymatic transformation protein extraction, osmotic distillation to concentrate fruit and vegetable juices, and in the production of low alcohol wine (Gabelman and Hwang 1999) This technology is commercially available; for example, the Liqui Cel TM membrane contactors schematized in Figure 1 6, consist of a membrane separator with several hollow fiber membranes knitted into a fabric and wrapped around a center tube called the

PAGE 41

41 distribution/collection tube. Liquid flows into the distribution tube and is forced to pass radially through the membrane, where a vacuum or a swept gas is applied to remove the oxygen from the liquid. The membrane is made of porous polypropyle ne and polyolefin. These hydrophobic membranes allow only gases to pass through membrane pores. As in any type of membrane, fouling can be a problem in hollow fiber separators. Fouling in water treatment applications has been reported due to suspended soli ds accumulation, growth of bacteria and slime development, calcium carbonate deposition, and iron oxide fouling (Mahmud and others 1998). Of these mentioned, suspended solid accumulation and growth of bacteria can be a problem when processing fruit juices, making this technology unsuitable for nonclarified fruit juices. Enzyme Based Deaeration Glucose oxidase (GOx) catalase complex has the ability to remove oxygen from solutions and headspace based on the reactions shown in Eq. 1 14 and 1 15: ( 1 14) ( 1 15) GOx catalase catalyzes the oxidation of D glucose to gluconic acid in the presence of diatomic oxygen. The by product hydrogen peroxide is converted by catalase to water and oxygen. Thus, the overall effect of the enzymatic reaction is removal of 0.5 mol of oxygen per mole of oxidized D glucose (Sagi and Mannheim 1990). At pH 4 to 7, GOx from Aspergillus niger has at least 90% of activity compared to its optimal activity at pH 5.5 to 6 (Kalisz and others 1990). Catalase, from Micrococcus sp., has a working pH of 3 to 9 and the pH optimum is 5.5 (Biocatalysts 2004; Schomburg 2007) and catalase from E. coli has a working pH of 6 to 8 (Varnado and

PAGE 42

42 others 2004) GOx catal ase complex showed a maximum activity at pH 6 and 30 C in a model salad dressing (Mistry and Min 1992) GOx catalase has been studied in orange, grape, apple, and pear (Sagi and Mannheim 1990; Pickering and others 1998; Par pinello and others 2002) In orange juice, 17 units L 1 of a GOx catalase enzymatic preparation were enough to reduce the DO concentration to less than 1 ppm in 30 min at 20 C, but also an important loss of vitamin C was reported. This loss was attribute d to the reaction of L AA with hydrogen peroxide (Sagi and Mannheim 1990). In Riesling grape (pH 3.08, free SO 2 24.5 mg L 1 and bound SO 2 85.5 mg L 1 ) treated with GOx catalase, the low pH inhibited the enzyme reaction. At the same GOx catalase concentrat ion (2 g L 1 ), the gluconic acid produced at pH 6 (80 g L 1 ) was almost 3 times the gluconic acid produced at pH 3.1 (30 g L 1 ). In the previously described study, the enzyme was used to reduce the amount of glucose available for fermentation to produce re duced alcohol wine (Pickering and others 1998). A mixture of GOx catalase called GOX100 (Gist Brocades, Lille, France) was compared with ascorbic acid and peroxidase in terms of oxygen removal and browning control in apple and pear purees. The unit of acti vity of GOx was defined as 1 mol of gluconic acid formed per min, at 35 C, pH 5.1, and for catalase as 1 mol of hydrogen peroxide consumed per min, at 25 C, pH 7. In the activity assay at different pH values, GOx and catalase showed its maximum activity at 50 and 12 units mg 1 at pH 5.5 and 6.5, respectively (Parpinello and others 2002). In this study, the reduction to pH 4 decreased GOx and catalase activity to 15 and 5 units mg 1 respectively. Ascorbic acid was the most effective in preventing nonenzymatic browning. The main di sadvantage of GOx catalase is that its activity is reduced at low pH values limiting its application to fruit

PAGE 43

43 juices. Therefore, studies have been done on incorporating GOx into the package to interact with the juice but preventing migration into the food. The catalytic efficiency, defined as k cat K m 1 is a measure of the substrate specificity and catalytic ability of the enzyme and was compared between immobilized GOx and GOx in solution. Immobilized GOx embedded on a UV polymerizable resin had similar (no significant difference) k cat K m 1 values to GOx in solution at 30, 25, and 10 C, but was 100 times lower (significant difference) at 5 C (Kothapalli and others 2007) More studies need to be done to improve the catalytic activity of immobilized GOx. Oxygen Scavengers Kinetic studies of DO in liquid foods describing oxygen consumption have been done in packaged and u npackaged products. Ahrne and others (1997) developed a mathematical model for pa ckaged liquid products that consider both oxygen consumption and oxygen transfer through the package. The model considered oxygen consumption as a 1st order kinetic reaction and oxygen transfer as a 3 step process: (1) oxygen from the atmosphere being abso rbed in the packaging material; (2) diffusion of oxygen through the package; (3) desorption of oxygen from package and solubilization into the food. Equation 1 16 describes both processes: ( 1 16) where and are the DO concentration in the food inside the package at time t the DO concentration at equilibrium, and the initial DO concentration, respectively. K is the rate constant estimated for this mathematical model. ( 1 17)

PAGE 44

44 ( 1 18) where K R K D and K are reaction rate constants for oxygen consumption, DO mass transfer coefficient, and rate constant of the model, respectively, all of them in (d ). is a ratio of partition coefficient and is considered equal to 1. For orange juice, K values were between 2.69 and 0.167 d a nd E a w as 46 kJ mol The model adjusted well to the reported data for orange and apple juices (Ahrne and others 1997). Oxygen sca vengers are commonly used as the basis for active packaging to remove residual oxygen from food and extend shelf life. Commonly used oxygen scavengers are based on iron powder oxidation (Kerry and others 2006) ascorbic acid oxidation (Brody 2005) photosensitive dye oxid ation (Vermeiren and others 1999) enzyme mediated oxidation (glucose oxidase catalase and alcohol oxidase) (Andersson and others 2002; Lpez Rubio and others 200 4; Ozdemir and Floros 2004) unsaturated fatty acids oxidation, and polymers oxidation (Solis and Rodgers 2001) Oxygen scavenging materials are added to the package as sachets, adhesive labels, cap liners, incorporated into the matrix of plastic containers during the container fabrication, or as a multilayer active film (Vermeiren and others 1999). A list of oxygen scavenging systems patented from 1994 to 2002 can be found in Ozdemir and Floros (2004). Oxygen scavenger performances can be measured in terms of their absorption capacity and absorption rate. Absorption capacity refers to t he amount of oxygen that can be removed. Absorption rate of iron oxidation and enzyme mediated oxidation oxygen scavengers have been described as a 1st order reaction. Reported values for iron oxidation based oxygen scavengers at temperatures between 5 and 35 C are 0.011 and 0.082 h 1 g 1 of oxygen scavenger ( Charles and others 2006) and between

PAGE 45

45 0.26 and 2.46 h 1 for temperatures of 1.5 to 25 C (Tewari and others 2002) E a for different brands of iron oxidation based oxygen scavengers had values between 44.08 and 49.91 kJ mol 1 a nd for an enzyme catalyzed oxidation based oxygen scavenger, a value of 44.72 kJ mol 1 ( Tewari and others 2002; Charles and others 2006). Migration of the oxygen scavenger material to the packaged food is a concern, especially when using nonedible subs tances (iron salts) in liquid foods. Lopez Cervantes and others (2003) compared migration in a plastic cup and a sachet iron based oxygen scavenger. The overall migration values were 5.35 (plastic cup) and 44.9 (sachet) mg L 1 in water and 35.35 and 63.15 mg L 1 in a 3% acetic acid solution. The study concluded that the use of sachets in acidic liquid foods is not desirable because the overall migration of iron (63.15 mg L 1 ) is higher than 60 mg L 1 the maximum allowed valu e established by the European Union (EU). Incorporation of the oxygen scavenger during container fabrication into the matrix of plastic containers or as a multilayer active film is desirable for liquid foods because there are some materials that do not ha ve migration problems. Oxygen removal via polymer oxidation is based on the use of an oxidizable plastic and cobalt as catalyst. The typical oxidizing plastics are polyethylene terephthalate (PET) and polyamides such as MXD 6 nylon, and the reaction is tri ggered by humidity. The Chevron Phillips OSP TM or oxygen scavenger polymer is a system that consists of 90% ethylene methylacrylate cyclohexenylmethyl acrylate (EMCM) and 10% cobalt salt with a proprietary photoinitiator. The advantages of this system are that it is inactive until exposed to UV light, it has FDA approval, can be used in rigid and flexible packaging, and it is recommended for meats, cheese, snacks, bread, coffee, tea, nuts, candies, fruit juices,

PAGE 46

46 beers, and wines. The OSP system has a scaven ging capacity of 45 to 70 mL O 2 g 1 of scavenger. Comparing orange juice packaged in regular cartons and cartons with OSP, the latter retained higher concentrations of vitamin C and showed a lower browning index (Solis and Rodgers 2001) Table 1 5 shows a list of oxygen scavengers commercially available and recommended to use on liquid f oods. Most of them are based in polymer oxidation with moisture or UV light as the activating agent, and in lower proportion they are based on ascorbic acid and GOx. There are also O 2 indicators that are useful to detect O 2 inside the package due to low ac tivity of the O 2 scavenger or to O 2 entering the package via leaks. Commercial brands are, for example, Ageless Eye TM (Mitsubishi Gas Chemical Co. Inc.,Tokyo, Japan) used for the master package of O 2 scavenging sachets (Ageless Eye) with color changes due to the presence (blue) or absence (pink) of O 2 O 2 indicators and absorbers are used together for chilled and shelf stable ready made packaged foods (Rodrigues and Han 2003) Current research on oxygen scavengers focuses on the development of packages with oxidizing polymers and biological entities such as enzymes and microorganisms; for example, the incorporation of GOx in a package, using UV polymerization (Kothapalli and others 2007) ; and the entrapment of aerobic microorganisms ( Kocuria varians and Pichia subpelliculosa) into a polymeric matrix (Altieri and others 2004) Oxygen is incorporated into fruit juices during their preparation and processing and remains dissolved during storage, affecting parameters correlated with the quality of the juice such as L AA concentration, color, and aroma. The direct aerobic oxidation of L AA indirectly affects color and aroma profile. L AA oxidation occurs at the beginning of

PAGE 47

47 storage. Most of the kinetic studies of the effect of DO in fruit juices are on L AA degradation and browning development and only a few of them describe the kinetics of DO consumption. It is necessary to develop models that accurately describe the ascorbic acid and oxygen interaction during oxidation due to the importance of these reactions. Determination of kinetic parameters i n real systems instead of models or pure solutions provides apparent kinetic parameters that, although highly variable due to intrinsic raw material variability, are more likely to be of practical value to juice processors. Mechanisms and compounds directl y related to DO and aroma compounds also need to be studied. Oxygen removal methods suitable for pulpy or cloudy juices are vacuum deaeration and gas sparging, but these methods have the main disadvantage of removing important volatile aroma compounds. The methods that can remove dissolved oxygen but preserve the aroma compounds are membrane deaerators and GOx catalase but only the 1st one is currently used in the fruit juice industry. Oxygen can also be removed during storage by adding enzymes or an oxygen scavenger material inside the package or incorporating them into the packaging material. Specific Objectives The central hypothesis of this research is that current processing conditions deteriorate orange juice quality and can be improved The overall ob jective of this research is to understand the effect of initial dissolved oxygen (DO) and temperature in vitamin C, color and aroma of not from concentrated orange juice to increase fresh like quality retention for industrial processing and storage. The fi rst specific objective is t o understand the effect of selected initial DO levels in vitamin C, aroma, volatile compounds and color of pasteurized commercial orange juice during storage at 5 C

PAGE 48

48 The second specific objective is t o understand the effect of s elected initial DO on the rate of DO consumption, vitamin C, volatile compounds, and color of fresh and pasteurized orange juice during storage at selected temperatures to increase fresh like quality retention and identify opportunities for processing imp rovement. The working hypothesis for both specific objectives is that DO decreases vitamin C, deteriorates aroma and color of orange juice during pasteurization and storage. Table 1 1 1 ) Orange juice, frozen concentrate, unsweetened, undiluted 137.9 White guava, frozen pulp 136.0 Red guava, frozen pulp 56.0 Grapefruit juice, white, frozen concentrate, unsweetened, undiluted 1 19.8 Papaya raw 61.8 Orange juice, raw 50.0 Pineapple and grapefruit juice drink, canned 46.0 Lemon juice, raw 46.0 Cranberry juice cocktail, bottled 42.3 Grapefruit juice, pink and white, raw 38.0 Orange juice, canned, unsweetened 34.4 Orange juic e, chilled, includes from concentrate 32.9 Grape drink, canned 31.4 Grapefruit juice, white, canned, unsweetened 29.2 Mango raw 27.7 Lemon juice, canned or bottled 24.8 Pineapple and orange juice drink, canned 22.5 Tangerine juice, canned, sweetened 22.0 Apricot raw 10.0 Pineapple juice, canned, unsweetened, without added ascorbic acid 10.0 Kiwi raw 9.3 Lime juice, canned or bottled, unsweetened 6.4 Passion fruit frozen pulp 4.3 Prune juice, canned 4.1 Apple juice, canned or bottled, unsweetene d, without added ascorbic acid 0.9 Modified from (Aymoto Hassimotto and others 2005; Holden 2006; Genovese and others 2008)

PAGE 49

49 Table 1 2 Reaction rate constants for AA degradation obtained under reducing and oxi dizing conditions at 90 C K1 (h 1 ) K2 (h 1 ) K3 (h 1 ) Control: L AA solution 1 ) 0.218 + 4.66 x 10 3 0.249 x 10 8 + 6.34 x 10 7 0.218 + 5.35x 10 3 L AA solution under reducing conditions: cysteine added 1 cysteine 0.0682 + 1.05 x 10 3 0.161 + 0.164 x 10 4 0.0286 + 3.68 x 10 3 L AA solution under oxidizing conditions: Fe 3+ added 1 Fe 3+ 0.449 + 6.30 x 10 4 0.212 x 10 8 + 0.285 x 10 8 0.449 + 1.19 x 10 3 Data from (Serpen and Gkmen 2007)

PAGE 50

50 Table 1 3. Summary of selected studies of ascorbic acid degradation in juices in which the effec t of oxygen was assessed. Product DO concentration 1 ) K inetic order Rate constant k Ea v alue 1) T (C) p p H Stora ge time (d) Other variables Reference Initial Final Orange juice ND < 0.4 ND ND ND 4 and 25 ND 160 to 300 Packaging mat erial, light absence (Ros Chumillas and others 2007) Concentrated orange juice 7.47 0.01 First 5.1 x10 4 to 4.6 x10 2 s 1 ND 25 to 88 N D 0 Processed with TASTE C evaporator (Braddock and Sadler 1989) M odel solution Anaerobic First 1.58 x10 5 to 3.98 x10 8 s 1 54 to 63 24 to 90 3. 5 ND Nonelectrolytes and electrolytes (Rojas and Gerschenson 2001) Orange juice 2.7 < 0.04 ND ND ND 4 and 25 N D 373 Packaging material (Zerdin and others 2003) FCOJ a < 1 ND ND ND ND 20 3.9 270 Packaging (Berlinet and others 2006) Citrus like beverage ND First 2.64 x10 8 1.38 x10 8 s 1 ND 5 and 35 3.18 120 Packaging material perm eability (Baiano and others 2004) FCOJ 1 ~0 ND ND ND 8 N D 52 Presence of light, packaging material (Solomon and others 1995) Clarified orange juice Anaerobic First 7.17 x 10 7 to 1.33 x 10 4 s 1 124.35 (serum) 70.3 to 97.6 3.6 0.125 Brix content (Johnson and others 1995) Orange juice Anaerobic First 7.5 x 10 7 to 8.15 x 10 5 s 1 115.56 3.5 Pasteurized lemon juice 0.41 1.44 3.74 0.12 0.14 0.14 Fi rst 1.93 x 10 7 1.91 x10 7 2.08 x10 7 s 1 ND 36 2.6 42 light absence (Robertson and Samaniego 1986)

PAGE 51

51 Table 1 3. Continued. Product DO concentra tion 1 ) K inetic order Rate constant k Ea v alue 1) T (C) p p H Stora ge time (d) Other variables Reference Initial Final Canned orange juice 3.5% < 0.1% Zero 0.28 0.73 mg L 1 h 1 ND 30 3.5 to 3.9 0 to 4 Different methods to remove DO can material (Kefford and others 1959) SSOJ D 4.7 to 5.67 Weibull model 1.79 x 10 5 s 1 38.6 20 to 45 3 .73 1. 5 Light absence (Manso and others 2001) Ascorbic acid solutions 0.5 to 1 atm 0.4 atm Pseudo f irst 0.16 x 10 4 to 2.31 x 10 4 s 1 ND 25 and 0.4 3 to 3.85 ND pH absence and presence of Cu 2+ and Fe 3+ (Khan and Martell 1967) Orange and tangerine juice mixture 2.5 1.6 First 1.04 x 10 5 to 6.2 x 10 5 s 1 35.9 50 to 100 3.7 ND ND (Dhuique Mayer and othe rs 2007) Canned orange juice 3.5% < 0.1% Zero 0.28 0.73 mg L 1 h 1 ND 30 3.5 to 3.9 0 to 4 Different methods to remove DO, can material (Kefford and others 1959) Polynomial 2.1 x10 2 to 4.7 x10 2 mg L 1 h 1 (Linear) 5 x10 4 to 2.1 x10 3 1 h 1 (Quadratic) 8 to 375 A naerobic First 1.0 x10 8 to 1.3 x10 8 s 1 0 to 375 SSOJ 7.0 ~0 First 4.89 x 10 8 s 1 ND 32 ND 35 Presence of Cu 2+ (Rassis and Saguy 1995) Concentrated orange juice 6.0 1.0 First 1.09 x 10 7 s 1

PAGE 52

52 Table 1 3. Continued. Product DO concentration 1 ) K inetic order Rate constant k Ea v alue 1) T (C) p p H Stora ge time (d) Other variables Reference Initial Final Ascorbic acid aqueous solutions 7.5 (at 760 mm Hg) Constant bubbling of air Model of consecutiv e zero and first Zero: 81.3 mg L 1 h 1 First: 4.69 x 10 5 s 1 ND 30 ND 0.46 ND (Sakai and others 1987) FCOJ 4.5 < 0.5 to 2 Zero 2.29 x 10 5 to 4.34 x 10 5 mg L 1 s 1 ND 4 to 105 ND 64 Packaged in TetraB rik cartons (Kennedy and others 1992) First 9.26 x 10 8 to 3.36 x 1 0 7 s 1 Acetate/aceti c buffer solution 9.2 to 8.4 2.9 to <1.0 Pseudo First 9.0 x 10 6 to 1.06 x 10 4 s 1 53.54 to 105.70 26.5 to 33 3 to 5 0.46 Sucrose concentration, pH (Hsieh and Harris 1993) Second 2.38 x 10 4 to 3.27 x 10 3 M 1 s 1 N D SSOJ 0.6 1.8 6.5 10.1 9.8 6.0 1.6 0.7 ND ND ND 22 N D 150 Pasteurization method (Trammell and others 1986) Infant formula 1 to 8.71 <1.0 First 5.5 6 x 10 6 to 6.67 x 10 5 s 1 12.42 to 47.76 7.2 to 23.9 ND 1 No quantification of ascorbic acid (Mack and others 1976) Infant formu la 1 to 8.71 0.4 Second 1.27 x 10 4 to 18.16 10 4 L mg 1 h 1 ND 7.2 ND 1 Light intensities (Singh and others 19 76) a FCOJ= from concentrated orange juice; b SSOJ =single strength orange juice; c TASTE=thermally accelerated short time evaporator; ND= not determined.

PAGE 53

53 Table 1 4. Comparison of different methods used to measure DO Method Advantages Disadvantages Electr ochemical probe Variety of shapes, sizes and consumables Some probes have build on thermistor for temperature compensation (Lewis 2006a) O2 consumption.Electrolyte solution should be free of bubbles to avoid in terferences Membrane fouling, stagnation or rupture Optical probe Ease of miniaturization Do not consume oxygen Sensitive to low levels of oxygen Fast response time: 0.2 1 s (Choi and Xiao 2000) Continuous measurements. Sensitive to ambient light Photo bleaching of the fluorophore Position dependence of optic fiber Winkler titration Official reference method (ASTM D 888 92) Highly dependent on analyst experience. Contamination with atmospheric O2 during titration. Chemical and color interferences Colorimetric methods Official methods (ASTM D 5543 94). Kit available. Able to measure ppb of O2. Not for contin uous measurements

PAGE 54

54 Table 1 5 Selected commercially available O 2 scavengers used in liquids Trade name Manufacturer Country Scavenger mechanism Incorporation method Activating agent Used in Shelf plus Ciba Specialty Chemicals United States PET copo lyester oxidation Into the plastic film Moisture Soups, beverages PureSeal W.R. Grace Co. Ltd United States Ascorbate/ metallic salts On bottle crowns NS a Liquid food ZerO2 CSIRO Food Science Australia/ Visy Industries Australia Polymer oxidation & p hotosensitive dye compound Into the plastic film UV light Beverages O2 displacer closure system Alcoa CSI United States NS Shell molded liner NS Beer Amosorb 3000 BP Amoco P.L.C United States Organic oxygen absorber for polyester Into the PET bottles Metal salt Beverages Starshield TM Crown Cork & Seal Co., Inc. United States Polymer oxidation Into the PET bottles Moisture Wines Cryovac OS1000 & 2000 Cryovac Sealed Air Corporation United States Polymer oxidation Into the plastic film UV light Not specified Oxyguard Toyo Seikan Kaisha, Inc Japan Iron oxidation Cap liner, into the plastic cup NS Soft drinks, beer, wine, soup

PAGE 55

55 Table 1 5. Continued Trade name Manufacturer Country Scavenger mechanism Incorporation method Activating agent Used in Tri SO2RB EVA Tri Seal International, Inc United States Polymer oxidation Closure liner NS Beverages OSPTM Chevron Phillips Chemical Co. United States Polymer oxidation Blended or as a layer in package UV light Fruit juices, beers, wines Oxbar, Starsh ield bottles Constar United States, France, UK Cobalt catalyzed oxidation of MXD 6 Into the plastic bottle NS For wines, beer, sauces, fruit juices and flavored alcoholic beverages Oxycap Standa Industrie France NS Cap liner NS Wine, beer, fruit juice, sauces, soft drinks Darex technology Grace Darex United States Ascorbic acid oxidation Into the plastic film NS Juices, tea, beer, sport drinks a NS = not specified

PAGE 56

56 Figure 1 1. Ascorbic acid degradation mechanism modified from Davey and others (2000) Figure 1 2. Reactions catalyzed by catechol oxidases.

PAGE 57

57 Figure 1 3. Reactions catalyzed by laccases

PAGE 58

58 Figure 1 4. Vacuum deaerator scheme.

PAGE 59

59 Figure 1 5. Gas sparging method scheme.

PAGE 60

60 Figure 1 6. Hollow fiber membrane separator (modified from (Polypore 2007)

PAGE 61

61 CHAPTER 2 ASSESSMENT OF DISSOL VED OXYGEN EFFECT ON VITAMIN C AND AROMA COMPOUNDS IN PASTEUR IZED NO T FROM CONCENTRATE ORA NGE JUICE Introduction Orange juice was the most co nsumed fruit juice in the 2007/08 season, representing 50% (14.4 L, single strength e quivalent per capita ) of the total annual consumption of fruit juices (USDA 2008a) Although pasteurized orange juice lacks of the freshness of non pasteurized orange juice and has cooked like aromas, i t is commercially preferred because pasteurization inactivates enzymes and destroy s microorganisms increasing the shelf life of the juice (Perez Cacho and Rouseff 2008a) Another factor that is thought to affect quality of orange juice is the dissolved oxygen that is incorporated duri ng extraction which may promote oxidation of ascorbic acid (Ringblom 2004) Ascorbic acid can be oxidized to dehydroascorbic acid, which may participate in Strecker degradation producing aroma active aldehydes such as methional (cooked potato aroma) and acetaldehyde (fruity aroma) (Perez Ca cho and Rouseff 2008a). Orange juice is considered an important source of ascorbic acid. Therefore, preventing ascorbic acid degradation during processing is necessary in order to protect nutritional value and aroma of the juice. Ascorbic acid degradation is a two step process: 1) ascorbic acid is oxidized to dehydroascorbic acid in a reversible reaction and 2) the dehydroascorbic acid is irreversibly hydrolyzed to 2,3 diketogulonic acid. In the citrus industry, deaeration before pasteurization has been implemented to remove the air incorporated into the juice. However, important volatile compounds such as alcohols, aldehydes, ketones, esters and terpene hydrocarbons are also removed during these procedures reducing the aroma intensity of the orange juice (Braddock

PAGE 62

62 1999; Jordan and others 2003; Ringblom 2004; Perez Cacho and Rouseff 2008a; Garca Torres and others 2009) Studies on the effect of oxygen during storage of citrus juices have focused on changes in vit amin C content (Kennedy and others 1992; Dhuique Mayer and others 2007) browning (Trammell and others 1986; Rassis and Saguy 1995; Solomon and others 1995) volatile sulfur compounds (Jab alpurwala and others 2010) and packaging (Zerdin and others 2003) M ost of the se parameters have been stud ied separate ly. Comprehensive studi es of changes in orange juice characteristics need to be performed to have a full assessment of t he effect of oxygen in fruit juices. The objective of this study was to correlate the changes in vitamin C content and aroma profile of orange juice with diffe rent dissolved oxygen (DO) levels during storage at refrigeration conditions (5 C). Materials and Methods Orange J uice P reparation and C haracterization Not from c oncentrate (NFC) pasteurized early mid V alencia orange juice was obtained from Citrosuco S.A. ( Lake Wales FL, U.S.A.) Soluble solids pH, tritatable acidity and suspended pulp content were measured in the oran ge juice before treatments. Soluble solids were measured using a digital refractometer (Abbe Mark II, Reichert Depew, NY ) pH using a pH meter (Orion 5 S tar,Thermo Fishe r Scientific, Beverly, MA ), titratable acidity (TA) by titration with 0.1 N NaOH using an automatic titrator ( Schott, Elmsford, NY ). Suspended pulp was measured by centrifugation of 50 mL of orange juice during 10 min at 365 x g and calculations were done as percentage of volume of pulp precipitated pe r total volume of centrifuged sample. S creened pulp was measured by passing 500 mL of orange juice through a 20 mesh screen using a

PAGE 63

63 FMC FoodTech quick fiber device (JBT Food Tech, Lakeland, FL) that shakes the screen during 2 min, until the pulp is free of excess of juice. Results are expressed as weight of pulp recovered in the screen by volume of juice used for the test (FMC 2002) Sample P reparation and Storage Study Sixteen liters of orange juice were divided in four portions of 4 L each : A nitrogen atmosphere was created by filling a vacuum atmospheric chamber (Spilfyter Green Bay, WI) with nitrogen. The f irst sample portion was bubbled with industrial grade nitrogen (Praxair, Inc. Danbury, CT) in the nitrogen atmosphere; the second portion was bubbled with ultra high purity oxygen (Praxair, Inc. Danbury, CT) ; the third and fourth portion s were not bubbled. Gas bubbli ng consisted of passing the gas at a flow of 1.5 L min 1 for 10 min thr ough 700 mL of orange juice in 1 L gas w ashing bottle (Chemglass, Vineland, NJ) as appendix C shows. All orange juice portions were poured in to 30 mL glass amber vials No headspace was left in vials filled with samples from first, second, and third portions. Air headspace equivalent to one thi rd of container total volume (10 mL) was left in vials filled with samples from the fourth portion. For samples bubbled with nitrogen, vials were filled in the nitrogen atmosphere. Storage of all samples was at 5 C during 60 days. Single analysis of thre e separate vials from the same batch of juice w as done Dissolved oxygen (DO) and vitamin C were measured during the storage. Volatile analysis was performed at day 1 of storage for all the juice portions and at day 60 only nitrogen bubbled and headspace s amples were available for analysis Microbial A nalysis Orange juice was received in aseptic packages that were kept at 4 C during 3 days until sample preparation, all the containers and utensils used to manipulate and

PAGE 64

64 store the juice were sterilized using an autoclave (40 min at 121 C ) orange juice was poured into final containers for storage within a l aminar flow hood to minimize product contamination because in preliminary tests we observed some contamination Microbial population was qua ntified in all juices during storage to insure that observed changes in DO and volatile compounds were not the result of microbiological activity. Microbial analysis was performed at days 1, 6, 15, 30 and 60. Aerobic plate counts were condu cted using plate count agar ( P CA), acidophilic bacteria associated with spoilage of citrus products and molds, yeasts were quantified using orange serum agar (OSA) and acid ified potato dextrose agar ( PDA) respectively. Samples were serially diluted with sterile 0.1% peptone as needed and poured in the plates by duplicate. PCA, OSA, PDA plates were incubated 24 h at 35 C 24 h at 30 C and 2 days at 25 C, respectively. DO M easurement DO concentration was measured under a nitrogen atmosphere using an electrochemical probe model Orion 083005D and pH meter model Orion 5 Star (Thermo Fisher Scientific, Beverly, MA) that was inserted in the vials containing orange juice then the cap was tightly attached to avoid entry of air and no headspace was left Constant stirring was maintained during the measurements and DO concentration values were collected digitally using a computer program written in LabVIEW 7.1 (National Instruments, Austin, TX). Calibration of t he probe and measurement of DO was done at the same as the storage temperature (5 C) for each sample A s tandard curve was generated by correlation of the DO probe current response expressed in n anoamperes vs DO concentration expressed in M from an oxyg en saturated and an oxygen free standards at the same temperature. Oxygen saturated standard was air saturated water

PAGE 65

65 for which DO concentration has been reported by Lewis (2006b) at different pressure and temperature conditions. Oxygen free standard was a 13% glucose solution with 0.05% of the enzyme glucose oxidase that was all owed to consume ox ygen for 5 min before measuring oxygen and DO concentration was assumed to be 0 M. T he use of small amount of enzyme was enough to decrease DO content. U se of glucose o xidase instead of sodium sulfite to prepare the 0% DO standard had th e advantage of not fouling the probe membrane Vitamin C M easurement Vitamin C was quantified as ascorbic (AA) and dehydroascorbic (DHA) acid. The capillary electrophoresis (CE) system used was a model P/ACE TM MDQ with DAD with the software Karat 32 ver sion 5.0 from Beckman Coulter (Fullerton, CA, U.S.A.). CE system was run at 21 kV, 23 C, detection at 263 nm and 35 mM sodium borate with 5% (v/v) acetonitrile as running buffer according to the method described by Cancalon (2001) The capillary was bare fused silica from Polymicro Technologies (Phoenix, AZ, U.S.A) 50 m m id, 56 cm total length. Calibration curve was done with standards containing ascorbic acid concentrations of 284, 852, 1419, 1987, 2555 and 3123 M. Samples were diluted (1:4) with a 1000 mg L 1 Ethylenediamine tetra acetic acid dipotassium salt ( EDTA ) s olution and filtered through a 0.45 m filter. To determine dehydroascorbic acid (DHA), L dithiothreitol (DTT) was added at a 13 mM (0.2 %) final concentration in samples in order to reduce DHA to AA. The DHA content of the sample was calculated by subtrac ting the AA content before adding DTT from the total AA content after adding DTT. Tryptophan was used as internal standard at a 200 mg L 1 final concentration and was detected at 214 nm.

PAGE 66

66 Color Measurement A Minolta colorimeter model CR 400 (Konica Minolta, Japan) was used to measure Hunter parameters and L, a, b were calculated with respect to control orange juice. A roma A nalysis Headspace s ampling Under a nitrogen atmosphere, 10 mL of sample were placed into a 40 mL glass vial containing a stirring bar and tightly closed with a screw cap with Teflon coated septum. The sample was equilibrated at 40 C for 20 min in a water bath at slow stirring, and a 1 cm 50/30 m (DVB/Carboxen/PDMS) SPME stable flex fiber (Supelco, Bellefonte, PA, U.S.A.) was inserted into the vial and exposed for 30 min. The fiber was then inserted into the injector port of the GC at 220 C for 5 min to allow the volatiles to be desorbed. Gas Chromatography FID/ O lfactometry Aroma intensity of volatile compounds was analyzed using an HP 6890A GC (Hewlett Packard Inc., Palo Alto, CA U.SA. ) equipped with a sniffing port. Samples were run on a polar column (DB wax, 30 m x 0.32 mm i.d. x 0.5 m; J&W Scientific, Folsom, CA U.S.A. ) The oven temperature was programmed from 40 to 240 C at 7 C min 1 with a holding time of 5 min at the maximum temperature. Helium was used as c arrier gas at flow rate of 2 mL min 1 Injector and detector temperatures were 200 and 290 C respectively. The column effluent was split with one third flow directed to FID and the other two thirds to the humidified olfactory port for sniffing. Two trained assessors evaluated each sample in duplicate. A time intensity method using a linear potentiometer as previously described (Bazemore and others 1999) was used to record

PAGE 67

67 their sensory impressions (retention times, aroma descriptors and intensities). Olfactory assessor and FID responses were separately recorded and integrated usin g two channels with ChromPerfect 5.5.5 ( Justice Labs, Melbourne, FL, U.S.A.) software. Intensity of each odor active compound was expressed as normal ized peak intensity. Peak norma lization, consisted in doing an average of all intensities of all odor activ e compounds for each sample and then normalize to a value of 10 given for the highest peak among all samples (peak of octadien 3 one, 1,5(z) in deaerated juice after 60 days of storage) A peak was considered to be aroma active if detected 2 times or more out of total four responses for each sample. Aroma active compounds were identifi ed by matching odor descriptors with linear retention index (LRI) values on GC O and GC MS Confirmation of identity was done by comparison with those reported on the literatu re Gas Chromatography Mass Spectrometry (GC MS). Aroma active compounds were identified using GC MS. A Perkin Elemer Clarus 500 quadrupole mass spectrometer with Turbo Mass software version 5.4.0 (Perkin Elmer, Shelton, CT, U.S.A.) a polar column (Stab ilwax; 60 m x 0.25 mm i.d. x 0.5 m film thickness, Restek, Bellefonte, PA, U.S.A.) was used. Helium was used as carrier gas in constant flow mode of 2 mL min. Source and injection port temperatures were maintained at 200C and the transfer line temperatur e was 260C. The oven was programmed from 40 to 240C at 7C min. Electron impact ionization (70 eV) was operated in the total ion current mode. Helium was used as carrier gas at flow rate of 1.5 mL min 1 Mass spectra matches were made by comparison of NI ST 2005 version 2.0 standard spectra (NIST, Gaithersburg, MD, U.S.A.). Only compounds with spectral fit values > 900 were considered positive identification. Samples were run in triplicate

PAGE 68

68 each replicate was collected from a different SPME headspace sampl e Identity of compounds was verified by the linear retention index (LRI) values, database matching and by comparison with those reported on the literature (Acree and Arn 2004; Rouseff 2008). Rate of R eaction Zero, 0.5, Pseudo first and second order rate model were applied to DO data. Statistical A nalysis One way ANOVA analysis was employed to determine significant differences in dissolved oxygen and vitamin C (AA and DHA) A two way ANOVA was used to find significant differences in volatile compounds bet ween the dif ferent treatments PCA analysis was performed for volatile compounds using MATLAB. Results and Discussion Control not from concentrate orange juice had a pH of 3.7 + 0.04, 10.3 + 0.2 Brix, 14.3 + 1.0 pulp and 0.7 + 0.02 acidity perc entage First, second, third and fourth portions will be referred as deaerated, oxygen saturated, control and air headspace orange juice respectively. Microbial A nalysis No microbial growth was detected in the NFC orange juice during the 60 days of storage for all the treatments except deaerated orange juice that had a final microbial count of 1.5, 1.8 and 1 .8 log CFU mL 1 for aerobic pla te counts (PCA) acidophilic bacterias (OSA) and molds and yeast (PDA) respectively as T able 2 1 shows The microbial pop ulation in deaerated orange juice was below the spoilage level of 5 log CFU mL 1 and it was not considered to affect aroma, volatiles, vitamin C or color of the orange juice.

PAGE 69

69 Dissolved Oxygen (DO) Since sample preparation (gas sparging and pouring of vial s) required one day, DO measurement started on day one of storage. At day one DO concentration in deaerated orange juice was reduced by 76 % with respect to control. In oxygen saturated orange juice DO concentration increased 278 % compared with control. As shown in F igure 2 1, deaerated control and oxygen saturated juices reached the lowest constant DO concentration of around 15 23 M from day 17 until the end of the storage period A similar fast re duction of DO concentration of seven days was observed i n lemon juice stored at 36 C in glass containers tightly sealed and without headspace (Robertson and Samaniego 1986) In this study, for samples with air headspace DO concentration in the orange juice remained between 220 125 M until day 38 and thereafter until reaching 35 M except for day 51 when DO content was higher. There is no reasonable explanation other than manipulation error and thus can be considered an outlier (F igure 2 1) These results suggest that the presence of air in the headspace has more effect in the DO concentration than bubbling oxygen possibly due to the aw (Ringblom 2004) Thus, as oxygen is consumed in the orange juice it is replenished with oxygen from air in the headspace to reach equilibrium. This equilibrium can be kept as long as the oxygen sup ply of the headspace lasts. Following the ideal gas law, the amount of oxygen present in the headspace at 5 C was estimated as 9.21 x10 5 moles that are approximately 10 times higher than 7.98 x10 6 moles in oxygen saturated orange juice From the kinetic m odels applied to DO data, pseudo first and second order had the best fit (Figure 2 2 ) Data was fitted to first and second order models Second order model

PAGE 70

70 for dissolved oxygen and ascorbic acid described better AA oxidation (Singh and others 1976) However in our experiments AA was in excess. The pseudo first order model applies when one of the reactants is present in excess over the other and the rate of reaction is proportional to the concentration of the reactant not in excess (D issolved oxygen) In a second order reaction th e rate of reaction is proportional to the square of the concentration of one of the reactants (Dissolved oxygen) First order r ate constant values of 1.3 x 10 1 2.3 x 10 2 2.8 x 10 1 3.8 x 10 2 and 2.8 x 10 1 4.1 x 10 2 d 1 were obtained for deaerated, control an d oxygen saturated juices It was expected to have the same reaction rate for all the juices since initial DO should not affect the rate reaction and the lower rate reaction for deaerated juice is attributed to having insufficient data points available for the rate reaction calculations. These values were about double the value of rate constants of DO consumption calculated in C hapter 3 for NFC orange juice from Hamlin oranges stored at 5 C ( Table 3 2). This difference in rate constant values was attribut ed to different processing conditions and pulp content, because the screened pulp content of the juice used on this study was 0% compared to 5 % scree ned pulp in the juice used for C hapter 3. Vitamin C On day 1, AA concentration in oxygen saturated orange juice was significantly ( =0.05) lower than that of the other samples (Figure 2 3 A ) that corresponded to the maximum initial DO concentration. This result suggests AA oxidation. On day 17, air headspace orange juice reached AA concentration as low as oxygen saturated juice ( Figure 2 3 A ) Finally, on day 60, samples with air headspace showed the low est AA

PAGE 71

71 concentration (Figure 2 3 A ) DHA concentration was higher in the samples with more oxidized AA (oxygen saturated and air hea dspace) during the first 30 d of storage (Figur e 2 3 B). Although for oxygen saturated orange juice DHA concentration had a c onstant decrease over the time. After 60 d of storage, orange juice with air headspace lost the most AA (42 %). Our results agree d with Beltrn Gonzlez and others (2008) that observed that higher oxygen content in headspace correlated with higher reduct ion in AA concentration in mandarin juice. However, in the absence of air in the headspace, DO can oxidize only a small amount of AA since the DO to AA proportion was 0.25 (495 M DO/1940 M AA). In this experiment t he initial DO concentration ranged from 495 to 42 M ( F igure 2 1 ) wher e as ascorbic acid ranged from 1940 to 1282 M (Figure 2 3 A ) for all the juices at day one of storage and approximately 13 % of AA was degraded by day 60 for samples without headspace independent of the initial DO content In th e case of air headspace samples, the DO/AA ratio (0.1) was about the same during most of the storage. However, at day 60 it dropped to less than half (0.04) because DO reached its lowest level (35 M) Pseudo first rate constant of AA oxidation was 1.9x10 2 3.0x10 3 d 1 Finally besides deaeration, avoiding air in headspace of orange juice containers is critical to prevent AA degradation. Color No change in colo r was visually detected among treatments and over time L, a and b values appear in Table 2 4. L and a parameters did not change over the time, b parameter changed slightly from yellow to blue.

PAGE 72

72 Aroma A ctive C ompounds The proportion of aroma active compounds in NFC orange juice at day one of storage is shown in T able 2 5 Aldehydes had the highes t i mpact in aroma profile (29 51%) for all the treatments and in deaerated orange juice, terpenes were similarly important (24%). About 12 to 23% of aroma active compounds could not be identifie d because only a polar column was used (DBWax). The use of a non polar column (DB5) in addition to a polar column will allow to identify polar and non polar compounds in the juices. Our r esults agree with other researchers that observed that aldehydes have an important contribution to orange juice aroma (Perez Cac ho and Rouseff 2008). L imonene was present in orange juice at the highest amount but the peak could not be integrated because it was out of the chromatogram scale, this compound is not aroma active and its function is only as a car rier of aroma active com pounds (Braddock 1999). Twenty four aroma active compounds were detected using GC O analysis (Tables 2 6 and 2 7 ) Table 2 6 shows the aroma active compounds identified in the NFC orange juice at day one of storage at 5 C Nonanal described as piney and herbal had the main contribution to aroma intensity of control, oxygen saturated and air headspace juices presenting a normalized peak height of 8.2, 9.5 and 9.1 for each juice respectively. Figure 2 4 shows the aroma active compounds organized by aroma d escriptor By comparison with control orange juice, deaerated orange juice cooked descriptor was the 34% of aroma profile compared to 17% in control orange juice (Figure 2 4 A and B) this corresponded to a peak intensity of 8.8 and 3.5 in deaerated and con trol orange juice for an unknown compound with a cooked rice descriptor. Oxygen saturated juice aroma profile was 44% a s herbal piney compared with 33% for control (Figure 2 4 C and B) due to the higher intensity of nonanal and terpinolene in

PAGE 73

73 the oxygen saturated juice of 9.5 and 9.2 compared with 7.2 and 7.3 for control respectively. Cooked note in orange juice with air in the headspace was 26% and in control 17%. This difference is due to methional being perceived as 7.5 of peak intensity in juice store d with air headspace but not perceived in control juice. (Table 2 6). Methional was also detected in deaerated and oxygen saturated orange juice with a peak intensity of 5.7 and 7.2. The presence of methional in oxygen saturated and air headspace orange ju ice is explained by the accumulation of DHA via AA oxidation that can promote the formatio n of methional from methionine (Perez Cacho and Rouseff 2008). But the presence of methional in deaerated orange juice was not expected since deaeration removed the o xygen to prevent oxidation of AA. Volatile Compounds A lcohols, aldehydes and terpenes were the volatile compounds detected at the highest percentage of 11 17, 9 13 and 6 10% respectively, for all the total volatile compounds in all the samples. Esters and ketones constituted less than 5% o f the total volatile compounds ( T able 2 5 ). The proportion of volatile compounds in this study is under 17, 39, 14, 19 and 11% of alcohols, aldehydes, terpenes, esters and ketones respectively, reported from a consensus of 36 aroma active compounds in fresh squeezed orange juice (Perez Cacho and Rouseff 2008b) Deaeration either by gas sparging or vacuum heating is known to remove the most volatile alcohols, aldehydes and terpene hydrocarbons (Ringblom 2004). Jordan and others (2003) observed significant losses of 50% or more in volatile compounds of fresh squeezed orange juice after deaeration. On this study n one of the detected compounds decrease d significantly in content after 10 min of either nitrogen or oxygen

PAGE 74

74 bubbling (Table 2 7) probably because an important portion of volatiles was lost during the previous storage of the juice at the juice processing pla nt. Ca rvone was significantly lower in control orange juice but similar in deaerated and oxygen saturated juices since carvone is a product of limonene oxidation it was expected that this compound content be higher in oxygen saturated juice and lower in d eaerated juice with respect to control. From the compounds with significant differences at day one of storage at 5 C, terpinene and cis terpineol were detected in all juices except in control juice. elemene, selinene and 3 carene had the highest amount in deaerated orange juice compared with the other juices. Acetaldehyde, sabinene, cymene, terpinolene and limonene oxide were detected in deaerated and control juices but not in oxygen saturated and air headspace juices. At day 60 of storage th e content of most of the compounds decreased with respect to day 0 (Table 2 8) except for E 2 hexenal and limonene oxide whose content increased in air headspace orange juice and 1 hexanol that increased at day 60 of storage in deaerated orange juice. Prin cipal Component Analysis (PCA) applied to the volatile compounds, explained the 63 % of variation (F igure s 2 5 and 2 6 ) among the juices. It was possible to differentiate between deaerated and headspace orange juice at day one and 60 of storage as indicated by separated well defined clusters. C ontrol and deaerated juices at day one of storage grouped together. According to the PCA loading plot (Figure 2 6) a ir headspace orange juice a t day one had significant higher amounts of nonanal, octanal and decanal fr om air headspace juice at day 60 of storage Carvone was present in significant higher amounts in oxygen saturated and air headspace orange juices compared wi th deaerated and control juices at day one of storage. Acetaldehyde was

PAGE 75

75 present only in deaerated and control orange juice and was the only significantly different as indicated with an asterisk in Figure 2 6 These trends could be used as reference for future studies, a lthough it should be kept in mind th at other reactions such as acid catalyzed hydrat ion may have a greater impact than oxygen in changes of orange juice aroma (Shaw and others 1993). Examp les of compounds formed by acid catalyzed hy dration reactions are terpineol, cis and trans 1,8 p methanediol (Marshall and others 1986) Conclusions At low storage temperature ( 5 C ) and 60 days of storage DO did not produce changes in color that could be visually perceived. The amount of DO present in orange juice, even at saturated levels only accounted for a small consumption of AA at the beginning of the storage. And the presence of air in headspace of containers caused detectable AA degradation. From the aroma active compounds, m ethional was perceived at higher intensity in air headspa ce orange juice with respect to control only and nonanal had the main contribution to aroma intensity of contro l oxygen saturated and air heaspace orange juices At day one of storage elemene, selinene and 3 carene had the highest amounts in deaerat ed orange juice compared with the other juices At day 60 of storage, the content of most of the compounds decreased with respect to day 0 except for E 2 hexenal and limonene oxide whose content increased in air headspace orange juice.

PAGE 76

76 Table 2 1 Microb ial counts in NFC orange juice during 60 days of storage at 5 C Deaerated Control Oxygen saturated Air headspace Day PCA a OSA b PDA c PCA OSA PDA PCA OSA PDA PCA OSA PD A 1 ) 1 6 15 30 1. 5 0.7 1.1 0.7 1.4 0. 5 60 1.5 0.8 1.8 0.5 1. 8 0.4 a aerobic plate counts, b acidophilic bacterias, c molds and yeast not detected

PAGE 77

77 Table 2 2 Pseudo first and second order rate of DO consumption in not from concentrated orange juice during storage at 5 C data points used Pseudo first order k (d 1 ) Second order k ( M d) 1 Deaerated 3 1.3x10 1 5.2x10 2 r 2 =0.87 4 1.3x10 1 2.3x10 2 6.0x10 3 9.4x10 4 r 2 =0.94 r 2 =0.95 Control 4 2.8x10 1 3.8x10 2 7.5x10 3 7.9x10 4 r 2 =0.96 r 2 =0.98 Oxygen Saturated 4 2.8x10 1 4.1x10 2 2.6x10 3 4.5x10 4 r 2 =0.96 r 2 =0.92 Table 2 3. Pseudo fi rst rate constant of AA oxidation and DH A formation in not from concentrated orange juice during storage at 5 C Air headspace data points used Pseudo first order k (d 1 ) AA 5 1.7x10 2 4.7x10 3 r 2 =0.81 6 1.9x10 2 3.0x10 3 r 2 =0.91 DHA 3 2.3x10 2 7.7x10 3 r 2 =0.90 4 1.9x10 2 3.4x10 3 r 2 =0.94 Table 2 4 Color change in NFC orange juice during storage at 40 C and expressed as L, a and b with respect to control orange juice. Deaerated Oxygen saturated Air headspace Time (d) L a b L a b L a b 1 1.9 0.3 0.6 1.9 0.2 0.7 3.6 0.7 1.9 9 1.7 0.0 0.5 2.6 0.2 1.5 4.3 0.7 2.1 16 9.6 2.0 8.2 9.2 1.8 7.5 11.4 2.3 8.6 30 2.5 0.0 0.5 1.9 0.4 1.1 4.7 0.1 1.3 51 1.5 0.1 2.1 3.1 0.5 3.8 1.0 0.4 3.4

PAGE 78

78 Table 2 5 Proportion of a roma active and volatile compounds in NFC orange juice at day 1 of storage at 5 C Aroma active compounds % peak height Deaerated Control Oxygen saturated Air headspace alcohols 9 16 16 0 aldehydes 29 29 34 51 ketones 3 0 0 0 terpenes* 24 26 19 9 esters 13 18 17 23 unknowns 23 12 14 16 Volatile compounds % peak area Deaerated Control Oxygen saturated Air headspace alcohols 11 13 18 17 aldehydes 13 10 9 9 ketones 1 1 1 1 terpenes a 10 10 8 6 esters 2 3 4 2 a Wit h out limonene

PAGE 79

79 Table 2 6 Aroma ac tive compounds identified in commercial NFC orange juice with different levels of DO after 1 day of storage at 5 C Values are average of normalized peak height standard erro r. Different letters indicates si gnificant difference (Duncan t est =0.05) LRI a Compound Descriptor Normalized peak height Deaerated Control Oxygen saturated Air headspace 1681 a seinene grains, musty 7.5 0.2 a 0.0 b 3.7 0.3 ba 0.0 b 1706 unknown grains,musty, cooked potatoe 3.8 1.8 ba 0.0 b 0.0 b 6.3 0.3 a 16 59 unknown roasted pecan, medicine, egg yolk, cooked rice 8.8 4.0 a 3.5 1.4 ba 4.3 0.4 ba 0.0 b 1192 sabinene minty,musty 6.8 0.7 a 5.2 0.1 a 4.0 2.3 a 0.0 b 1284 unknown cooked egg 8.4 1.6 a 0.0 b 0.0 b 7.2 1.4 a 1443 methional cooked potatoe 5.7 2.1 a 0.0 b 7.2 2.5 a 7.5 0.7 a 1805 decadienal (E,E) 2,4 cooked potatoe, cooked 0.0 b 4.7 0.2a 0.0 b 6.6 1.7 a 1527 decanal cooked,mushroom, cooked, melon 4.6 1.2 a 5.4 0.5 a 0.0 b 0.0 b 1228 ethyl hexanoate fruity 5.9 0.2 a 5.8 0.3 a 7.0 0.8 a 6.0 1.1 a 1 540 linalool fruity, bubble gum 8.8 2.2 a 7.3 2.5 a 9.4 1.7 a 0.0 b 1033 ethylbutanoate fruity, bubble gum, tutti fruity 8.6 1.7 a 8.2 0.7 a 7.6 2.1 a 7.8 0.6 a

PAGE 80

80 Table 2 6. Continued LRI a Compound Descriptor Normalized peak height Deaerated Contr ol Oxygen saturated Air headspace 1332 ethyl heptanoate fruity, sweet, 4.9 0.2 a 0.0 b 5.6 1.2 a 5.7 2.3 a 1051 unknown fruity, floral, citrus 7.7 2.4 a 5.3 1.0 ba 5.5 1.7 ba 0.0 b 998 acetaldehyde orange peel, fruity, cooked orange 6.5 2.0 a 3.8 1.4 a 4.3 1.6 a 6.0 0.9 a 1817 nerol citrus lemonade, citrus, green 0.0 b 0.0 b 5.8 0.2 a 0.0 b 1366 nonanal herbal, piney, 8.1 1.8 a 8.2 1.5 a 9.5 1.2 a 9.1 0.8 a 1489 nonenal (z) 2 herbal, fatty, medicine, piney, 8.7 0.1 a 0.0 b 8.5 1.8 a 0.0 b 1717 a terpineol herbal, green 0.0 b 0.0 b 4.5 0.3 a 0.0 b 1076 hexanal herbal, green, grape 4.8 0.2 a 0.0 b 5.9 0.7 a 6.7 0.2 a 1620 1 octanol herbal, musty, floral,soapy 5.0 2.2 a 4.7 0.1 a 0.0 b 0.0 b 1263 a terpinolene herbal, piney 7.2 1.8 a 7.2 1.2 a 9.2 1.1 a 0.0b 1848 Perilla aldehyde herbal,lemon grass, herbal 6.8 0.3 a 0.0 b 5.7 0.7 a 6.7 0.7 a 1383 unknown plastic, herbal 6.1 0.7 a 0.0 b 6.7 0.7 a 0.0 b a DBWax

PAGE 81

81 Table 2 7 Volatile comp ounds identified by GC MS in NFC orange juice with different le vels of DO after 1 day of storage at 5 C. Values are average of normalized peak area standard error. Different letters indicates significant difference s =0.05) LRI a C ompound Deaerated b Control b Oxygen saturated b Air Headspace b Aldehydes (8) 719 Acetaldehyde 6.50E 01 2.90E 01 a 1.86E 01 6.09E 02 a c 1095 Hexanal 8.66E 02 1.94E 02 a 9.05E 02 6.94E 03 a 7.21E 02 1.06E 02 a 1.06E 01 1.05E 02 a 1240 E 2 hexenal 2.76E 02 8.49E 04 1303 Octanal 3.01E 01 8. 16E 02 a 2.84E 01 1.42E 02 a 1.81E 01 1.76E 03 a 3.43E 01 2.74E 02 a 1412 Nonanal 5.87E 02 1.23E 02 a 5.18E 02 1.09E 03 a 3.90E 02 6.68E 03 a 6.06E 02 4.20E 03 a 1457 E 2 octenal 2.74E 03 3.87E 04 a 4.90E 03 4.90E 03 a 1519 Deca nal 1.47E 01 2.56E 02 a 1.28E 01 7.18E 03 a 8.55E 02 1.50E 02 a 1.53E 01 1.41E 02 a 1835 Perillaldehyde 2.29E 02 2.95E 03 a 1.53E 02 8.80E 04 a 1.51E 02 9.26E 04 a 2.49E 02 2.06E 03 a Esters (5) 1048 Ethylbutanoate 1.12E 01 4.08E 02 a 1.10E 01 1.06E 02 a 7.09E 02 8.23E 03 a 8.85E 02 6.13E 03 a 1198 methyl hexanoate 6.04E 03 7.50E 04 1239 ethyl hexanoate 3.05E 02 5.13E 03 a 4.04E 02 7.85E 03 a 2.54E 02 4.79E 03 a 3.48E 02 5.16E 03 a 1444 Ethyloctanoate 2.96E 03 2.64E 04 1701 ethyl 3 hydroxyhexanoate 5.74E 02 7.70E 03 a 4.07E 02 5.86E 03 a 4.12E 02 1.57E 03 a 4.82E 02 2.42E 03 a Terpenes (17) 1030 pinene 2.51E 02 7.37E 03 a 1.43E 02 3.73E 03 a 6.31E 03 5.23E 04 a 1.65E 02 6.97E 03 a 1162 myrcene 2.85E 01 1.09E 01 a 2.51E 01 2.37E 02 a 1.14E 01 3.31E 03 a 2.26E 01 4.86E 02 a

PAGE 82

82 Table 2 7 Continued LRI a compound Deaerated b Control b Oxygen saturated b Air Headspace b 1172 3 carene 8.86E 03 5.48E 03 1182 terpinene 1 .04E 02 4.58E 03 a 2.47E 03 4.38E 04 a 6.66E 03 1.99E 03 a 1186 humulene 7.90E 03 1.82E 03 1210 Limonene 6.112.13 a 5.132.76 a 6.912 .59 a 4.971.21 a 1216 Sabinene 3.54E 02 1.20E 02 a 7.31E 02 4.49E 02 a 1282 cymene 9.67E 03 5.52E 03 a 7.68E 03 1.19E 03 a 1290 terpinolene 6.09E 03 2.46E 03 a 3.19E 03 3.19E 03 a 1611 elemene 3.27E 03 9.18E 04 a 1651 limonene oxide 3.46E 02 7.19E 03 a 2.25E 02 2.97E 03 a 1752 Vale ncene 5.09E 01 1.32E 01 a 3.43E 01 1.55E 02 a 2.27E 01 3.84E 02 a 2.32E 01 2.70E 02 a 1760 selinene 3.74E 02 7.41E 03 a 1.84E 02 1.50E 04 ba 1.36E 02 3.56E 03 b 1.14E 02 2.72E 04 b 1779 Carvone 8.04E 02 1.28E 02 a 3.94E 02 3.93E 03 b 6.46E 02 5.09E 03 ba 7.33E 02 4.71E 03 a 1801 panasinsen 3.27E 02 8.04E 03 a 2.20E 02 4.64E 04 a 1.41E 02 1.60E 03 a 1.67E 02 2.37E 03 a Alcohol terpenes(6) 1555 Linalool 4.79E 01 1.04E 01 a 5.13E 01 3.69E 02 a 3.88E 01 5.14E 03 a 7 .35E 01 4.34E 02 a 1586 cis terpineol 3.64E 03 7.12E 04 a 1.76E 03 1.76E 03 a 3.52E 03 2.65E 04 a 1629 4 terpineol 2.54E 01 3.98E 02 a 1.84E 01 1.64E 02 a 1.89E 01 2.86E 03 a 2.29E 01 1.30E 02 a

PAGE 83

83 Table 2 7 Continued LRI a compound D eaerated b Control b Oxygen saturated b Air Headspace b 1720 terpineol 8.19E 02 1.30E 02 a 9.65E 02 4.65E 03 a 7.23E 02 7.92E 04 a 1.05E 01 5.57E 03 a 1857 cis/trans 1 carveol 2.06E 02 4.13E 03 a 1.74E 02 1.54E 03 a 1.74E 02 1.96E 03 a 2.24E 02 4.57E 04 a 1774 citronellol 3.38E 03 3.38E 03 b 7.04E 03 4.76E 04 b 7.80E 03 1.13E 03 ba 8.63E 03 3.89E 04 a Alcohols(6) 949 Ethanol 2.09E 01 6.61E 02 a 1.81E 01 2.64E 02 a 1.11E 01 7.91E 03 a 2.01E 01 3.45E 02 a 1465 1 heptano l 3.11E 03 1.46E 03 a 6.56E 03 1.30E 03 a 2.03E 03 2.03E 03 a 6.82E 03 1.07E 03 a 1567 Octanol 3.95E 02 7.75E 03 a 4.41E 02 3.77E 03 a 3.96E 02 4.86E 03 a 5.91E 02 4.42E 03 a 1668 1 nonanol 6.17E 03 8.15E 04 a 3.76E 03 5.02E 04 a 5.83 E 03 1.62E 03 a 6.53E 03 4.90E 04 a 1814 Nerol 9.00E 03 1.95E 03 a 7.28E 03 2.06E 04 a 6.72E 03 4.55E 05 a 1.08E 02 7.70E 04 a a DBWax, b average std error, c Not detected, tentatively identified

PAGE 84

84 Table 2 8 Volatile compounds identified b y GC MS in NFC orange juice with different levels of DO after 60 days of storage at 5 C Values are average of normalized peak area standard error. D ifferent letters indicates significant differen ce (Duncan test =0.05) LRI a Compound Deaerated b Air head space b Aldehydes (7) and Ketones (1) 719 acetaldehyde c 1095 hexanal 4.45E 02 1.23E 02 a 1240 E 2 hexenal 3.59E 02 1.06E 02 a 1303 octanal 2.43E 02 6.35E 03 a 5.92E 02 4.80E 03 a 1412 nonanal 1.11E 02 2.79E 03 a 1519 decanal 4.48E 02 1.28E 02 a 1.99E 02 2.23E 03 a 1835 Perilla aldehyde 1.08E 02 4.18E 04 a 1365 6 methyl 5 hepten 2 one 1.12E 02 3.14E 03 a 6.12E 03 2.06E 03 a Esters (3 ) 1048 ethylbutanoate 3.86E 02 7.10E 03 a 4.56E 02 2.74E 03 a 1239 ethyl hexanoate 5.26E 03 5.32E 04 1701 ethyl 3 hydroxyhexanoate 3.63E 02 2.98E 03 a 3.09E 02 1.00E 03 a Terpenes (11 ) 1162 myrcene 8.15E 02 3.94E 02 a 1.94E 02 8.86E 03 a 1172 3 carene 1.10E 02 1.10E 02 1210 limonene 2.25 1.07 8.64E 01 1.39E 01 1216 sabinene 1.09E 02 6.59E 03 3.59E 03 5.69E 04 1752 valencene 1.50E 01 5.04E 02 a 4.09E 02 1.54E 03 b 1760 selinene 1.02E 02 4.54E 03 1779 carvone 4.44E 02 2.14E 03

PAGE 85

85 Table 2 8 Continued LRI a Compound Deaerated b Air headspace b 1801 panasinsen 8.34E 03 2.93E 03 1651 limonene oxide 1.80E 02 9.05E 04 Alcohol terpenes(6) 1555 linalool 3.31E 01 2.10E 02 3.94E 01 1.56E 02 1629 4 terpineol 1.78E 01 2.22E 02 1.40E 01 9.94E 03 1720 terpineol 5.89E 02 2.91E 03 5.47E 02 2.18E 03 1857 cis/trans 1 carveol 1.41E 02 1.49E 03 Alcohols(2 ) 949 ethanol 3.45E 01 2.45E 02 1.40E 01 1.19E 03 1373 1 hexanol 2.36E 02 3.25E 03 1814 nerol 7.20E 03 3.96E 04 1567 octanol 2.42E 02 2.98E 04 3.12E 02 1.79E 04 a DBWax, b average + std error, c Not detected

PAGE 86

86 Figure 2 1. Dissolved oxygen (DO) concentration in NFC orange juice during 60 days of storage at 5 C. ( ) deaerated, ( ) control, ( ) oxygen saturated, ( ) air h eadspace

PAGE 87

87 Figure 2 2 Kinetic models fitted to dissolved oxygen (DO) concentration in NFC orange juice during 60 days of storage at 5 C

PAGE 88

88 Figure 2 3 Changes in (A) ascorbic acid and (B) dehydroascorbic acid in NFC orange juice during 60 days of stor age at 5 C. ( ) deaerated, ( ) control, ( ) oxygen saturated, ( ) air headspace

PAGE 89

89 Figure 2 4. Aroma active compounds organized by aroma descriptor in NFC orange juice at day 1 of storage at 5 C. (A) deaerated, (B) control, (C) oxygen saturated and (D) air headspace orange juices.

PAGE 90

90 Figure 2 5 PCA score plot of volatile compounds in orange juice at different DO concentrations at day 1 and 60 of storage at 5 C ( deaerated, control, oxygen saturated and air headspace)

PAGE 91

91 Figure 2 6 Loading plot of volatile compounds in orange juice at different DO concentrations at day 1 and 60 of storage at 5 C *significantly different.

PAGE 92

92 CHAPTER 3 IMPACT OF DISSOLVED OXYGEN LEVEL AND TEM PERATURE ON VOLATILE COMPOUNDS, VITAMIN C AND COLOR IN PASTEURIZED NOT FROM CONCENTRATE ORANGE J UICE Introduction Deaeration of orange juice in citrus industry is applied in the production of not from concentrate orange juice before pasteurization with the purpose of removing oxygen and essential oil (Braddock 1999; Ringblom 2004) The importance of the deaeration step to the citrus industry has relied on the assumption that if oxygen (21% in air) is present during pasteurization and storage, it causes ascorbic acid degradation, as well as chan ges in aroma and color (Ringblom 2004) The main disadvantage of deaeration is that volatile compounds that give the fresh like aroma to the juice are also removed (Jordan and others 2003) during this step. Ascorbic acid (AA) degradation o f citrus juices in the presence of oxygen an d during storage has been described under different initia l dissolved oxygen (DO) concentrations (Kefford and others 1959; Robertson and Samaniego 1986) storage temperatures (Braddock and Sadler 1989) and packaging material s (Kennedy and others 1992; Zerdin and others 2003; Ros Chumillas and others 2007) AA degradation in orange juice has been described as first order for concentrated orange juice (Burdurlu and others 2006) ; for membrane clarified concentrated orange juice; for fresh squeezed orange juice (Dhuique Mayer and others 2007) ; as well as zero and first order, for deaerated and non deaerated si ngle strength from concentrated orange juice respectively, (Soares and Hotchkiss 1999) Volatile compounds in orange juice thought to be a product of oxidation include carvone and carveol from limonene oxi dation (Kutty and others 1994) and met hional

PAGE 93

93 from dehydroascorbic acid participation in Strecker degradation of methionine (Perez Cacho and Rouseff 2008a) Time temperature effects on aroma of orange juice during storage can be assessed following terpineol production due to linalool and limonene degradation by acid catalyzed hydration dehydration pathway (ACHD) and not oxidation (Marcotte and others 1998) The i nfluence of DO in non enzymatic browning has been suggested as a consequence of AA oxidation but no significant effects o n col or of orange juice has been observed (Rassis and Saguy 1995) Oxidation reactions can be accelerated in the juice depending on storage and packaging conditions. Previous studies have focused on individual compounds (such as vitamin C) that are affected by time, packaging material, or storage conditions instead of doing comprehensive studies. There is a lack of information on the single effect of dissolved oxygen in vitamin C degradation, excluding any other variables such as packaging material. Therefore, to imp rove processing conditions, it is important to better understand the mechanisms of how oxygen deteriorates orange juice quality. The objective of this study was to elucidate the changes in vitamin C, aroma and color of pasteurized not from concentrate (NFC ) orange juice with selected initial levels of DO during storage at selected temperatures without headspace. We also compared the changes in AA, color and aroma in the presence of air headspace under the same storage conditions. Materials and Methods Orang e Juice Preparation and Characterization Oranges (Hamlin var.) were obtained from CREC groves (Lake Alfred, FL) and Soluble solids content (SSC) pH, tit r atable acidity, suspen ded and sc reened pulp

PAGE 94

94 content were measured in orange juice. SSC were measured using a digital refractometer (Abbe Mark II, Reichert Depew, NY ) pH using a pH meter (Orion 5 S tar,Thermo Fishe r Scientific, Beverly, MA ) titratable acidity (TA) by titratio n with NaOH 0.1 N until reaching a pH of 8.2 and using an automatic titrator ( Schott, Elmsford, NY ) S uspended pulp was measured by centrifugation of 50 mL of orange juice during 10 min at 365 x g and calculations were done as percentage of volume of pulp precipitated per total volume of centrifuged sample. S creened pulp was measured by passing 500 mL of orange juice through a 20 mesh screen using a FMC FoodTech quick fiber device (JBT Food Tech, Lakeland, FL) that shakes the screen during 2 min, until the pulp is free of excess of juice. Results are expressed as weight of pulp recovered in the screen by volume of juice used for the test (FMC 2002) Sample Preparation and Storage Study O range juice was divided in to four portions: A nitrogen atmosphere was created by filling a vacuum emptied glove bag Spilfyter Green Bay, WI) with nitrogen. The f irst sample portion was bubbled with industrial grade nitrogen (Praxair, Inc. Danbury, CT) in the nitrogen atmosphere; the second portion was bub bled with ultra high purity oxygen (Praxair, Inc. Danbury, CT) ; the third and fourth portion s were not bubbled. Gas bubbli ng consisted of passing the gas at a flow of 1.5 L min 1 for 10 min thr ough 700 mL of orange juice in 1 L gas washing bottle (Chemglas s, Vineland, NJ) Gas flow was regulated with a valve and time monitored with a timer All orange juice portions were pasteurized in separate batches at 90 C for 24 s using a HTST pa steurizer (Microthermics, Raleigh, NC). O range juice portions were poured in to 30 mL amber vials one per sampling time No headspace was left in vials filled with juice from the first, second, and third portions. An a ir headspace equivalent to one third (10 mL) of

PAGE 95

95 container total volume ( 3 0 mL) was left in vials filled with ju ice from the fourth portion Figure 3 1 shows the juices preparation process For juices bubbled with nitrogen, vials were filled in a nitrogen gas atmosphere. Control, oxygen saturated and air headspace juices were filled in a laminar hood. A ll juice port ions were poured to a sterile glass beaker from the glass bottle used to collect the juice from pasteurizer and then amber vials were completely submerged into the beaker and capped submerged to avoid air bubbles. Glass amber vials were stored under regul ar atmospheric conditions (air) at 5 13, 21.5, 30.5 and 40 C during 60 12, 12, 6 and 6 days respectively. Only two replicates of each treatment were prepared from the same fruit batch b ecause Hamlin var. oranges were not available to prepare more juice later in the season DO and AA were measured at the times indicated in Table 3 1. Because c olor changes were detected only in air headspace sample at 40 C and volatile compounds at 40 C did not show a significant difference. Color and volatile compounds at storage temperatures lower than 40 C were not measured. Microbial A nalysis In order to assure the orange juice shelf life, especially at intermediate temperatures (13 to 30.5 C) fruit was properly washed, the extractor was sanitized according to manu facturer instructions, all the containers and utensils used to manipulate and store the juice were sterilized, and the pasteurizer was cleaned using a CIP protocol before and after processing the juice. Oranges were extracted in a clean room and juice was poured in final containers for storage under a laminar flow hood to minimize product contamination because in preliminary tests we observed some contamination The clean in place (CIP) protocol used was performed at a set flow of 1.4 L min 1 and the follow ing solutions were passed through the pasteurizer

PAGE 96

96 1. 20 min washing with 2% NaO H solution at 80 C 2. 10 min water rinsing at 80 C 3. 20 min 1% phosphoric acid at 30 C 4. 15 min water rin sing at 30 C Microbial population was quan tified in all juices during storage to insure that observed changes in DO and volatile compounds were not the result of microbiological activity. Microbial analysis was performed at days 1, 6, 15, 30 and 6 0 at 5 C ; at day 1, 6, 12 at 13 and 21.5 C at day 1 and 6 at 30.5 C; at day 1 and 3 at 40 C Aerobic plate counts were conducted using plate count agar ( A PCA), acidophilic bacteria associated with spoilage of citrus products and molds, yeasts were quantified using orange serum agar (OSA) and acidified potato dextrose agar (APDA) respec tively Samples were serially diluted with sterile 0.1% peptone as needed and poured in the plates by duplicate. PCA, OSA, PDA p lates were incubated 24 h at 35 C 24 h at 30 C and 2 days at 25 C respectively. DO Measurements DO concentration was measur ed under a nitrogen atmosphere using an electrochemical probe model Orion 083005D and pH meter model Orion 5 Star (Thermo Fisher Scientific, Beverly, MA) that was inserted in the vials containing orange juice then the cap was tightly attached to avoid entry of air. Constant stirring was maintained during the measurements and DO concentration values were collected digitally using a computer program written in LabVIEW 7.1 (Na tional Instruments, Austin, TX). Calibration of the probe and measurement of DO was done at the same as the storage temperature (5 40 C) for each sample A s tandard curve was generated by correlation of the DO probe response expressed in nA vs DO conce ntration expressed in M of an

PAGE 97

97 oxygen saturated and an oxygen free standards at the same temperature The o xygen saturated stand ard was air saturated water whose DO concentration have been reported by Lewis (2006b) for different pressure and temperature conditions. Oxygen free standard was a 13% glucose solution with 0.05% of the enzyme glucose oxidase that was allowed to consume oxygen during 5 min before measuring oxygen and DO concentration was assumed to be 0 M T he use of small amount of enzyme was enough to decrease DO content. U se of glucose oxidase instead of sodium sorba te to prepare the 0% DO standard had the advantage of not fouling the probe membrane Vitamin C Measurement Vitamin C was quantified as ascorbic (AA) and dehydroascorbic (DHA) acid. A capillary electrophoresis (CE) MDQ with a DAD and Karat 32 version 5.0 software from Beckman Coulter (Fullerton, CA). The CE system was run at 21 kV, 23 C, detection at 263 nm and 35 mM sodium borate with 5% (v/v) acetonitrile as background electrolyte running buffer according to the method de scribed by Cancalon (2001) The capillary was bare fused silica fr om Polymicro Technologies (Phoenix, AZ) 50 m id, 56 cm total length. A calibration curve was done with standards containing AA concentrations of 284, 852, 1419, 1987, 2555 and 3123 M. Samples were diluted (1:4) with a 1000 mg L 1 EDTA solution and filter ed through a 0.45 m syringe filter. To determine DHA, L dithiothreitol (DTT) was added at a 13 mM (0.2%) final concentration in samples in order to reduce DHA to AA. The DHA content of the sample was calculated by subtracting the initial AA content from t he total AA content after the conversion. Tryptophan was used as an internal standard at a 200 mg L 1 final concentration and was detected at 214 nm

PAGE 98

98 Color Measurement A Minolta colorimeter model CR 400 (Konica Minolta, Japan) was used to measure Hunter pa rameters and L a b were calculated with respect to control pasteurized orange juice. Oil in Juice Assay Scott Oil method was used to determine oil content in SSOJ and consists of a distillation and a titration. To 25 g of SSOJ 25 mL of water and the s ame amount of isopropanol are added and mixed in a 300 mL round distillation flask. An antifoam agent and boiling chips are added to the mixture The mixture is distillated until all isopropanol carrying the oil, is recovered. Then 10 mL of 4 M HC l and a d rop of 1% methyl orange indicator are added to the distillated sample. Titration to a colorless endpoint usi ng 0.025 N potassium bromide bromate solution is performed. Percentage of oil in the juice is calculated as (3 1) Volatile Compounds Analysis Headspace sampling Under a nitrogen atmosphere, 10 mL of sample were placed into a 40 mL glass vial containing a stirring bar and tightly closed with a screw cap with Teflon coated septum. Each sample was equilibrated at 40 C for 20 min in a water bath with slow stirring, and a 1 cm 50/30 m (DVB/Carboxen/PDMS) StableFlex SPME fiber (Supelco, Bellefonte, PA) was inserted into the vial and exposed for 30 min. The fiber was then inserted into the injector port of the GC at 220 C for 5 min to allow the volatiles to be desorbed.

PAGE 99

99 Gas Chromatography Mass Spectrometry (GC MS). Volatile compounds were identified using GC MS. A Perkin Elemer Clarus 500 qua drupole mass spectrometer with Turbo Mass software version 5.4.0 (Perkin Elmer, Shelton, CT) with a polar column (Stabilwax; 60 m x 0.25 mm i.d. x 0.5 m film thickness, Restek, Bellefonte, PA). Helium was used as carrier gas in constant flow mode of 2 mL min 1 Source and injection port temperatures were maintained at 200C and the transfer line temperature was 260 C. The oven was programmed from 40 to 240 C at 7 C min. Electron impact ionization (70 eV) was operated in the total ion current mode. Heli um was used as carrier gas at flow rate of 1.5 mL min 1 Mass spectra matches were made by comparison of NIST 2005 version 2.0 standard spectra (NIST, Gaithersburg, MD). Only compounds with spectral fit values > 900 were considered positive identification. Samples were run in triplicate each replicate was collected from a d ifferent SPME headspace sample Identification of compounds was verified by the linear retention index (LRI) values, GC MS identification and by comparison with those reported on the lit erature (Acree and Arn 2004; Rouseff 2008). Peak areas were normalized using 4 heptadecanone as internal standard. Peak area values of each volatile compound were divided between the peak area of the internal standard for the same replicate Then average and standard error of the three replicates was calculated. Rate of R eaction Zero, 0.5, first and second order rate model were applied to DO data Statistical A nalysis Analysis of variance ( ANOVA ) of two ways and Duncan test s at =0.05 using SAS software (Cary, NC) were performed to determine significant differences in volatile compounds between the different treatments

PAGE 100

100 Results and Discussion The orange juice ( cultivar Hamlin ) used for the storage study had a 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 and 0.7 0.02 % titratable acidity expressed as citric acid Microbial A nalysis Microbial population in not pasteurized fresh squeezed orange juice was 2.2 0.07, 2.3 0. 02 and 1.9 0.03 log CFU mL 1 in aerobic plate counts (APCA), molds and yeasts (APDA) and acidophilic bacteria (OSA) respectively. These values are around half of 4.5 log CFU mL 1 reported by Sadler and others (1992) in non pasteurized fresh squeezed orange juice. During 60 days of storage a t 5 C microbial counts in APCA were lower than 1 log CFU mL 1 for all the juices ; in APDA and OSA microbial counts were between 1.4 1.8 and 1.1 1.7 log CFU mL 1 respectively for day 60 of storage No microbial growth was ob served in orange juice stored at 40 C but juice stored at 13, 21.5 and 30. 5 C presented microbial populations of 1.7 to 3 log CFU mL 1 from the first day of storage reaching counts of 5 log CFU mL 1 at days 12 (13 and 21.5 C) and 6 (30.5 C). Due to ju ices stored at 13, 21.5 and 30.5 C reaching 5 log CFU mL 1 microbial populations, considered as microbial spoilage, no further analyse s were performed onto these juices. DO Consumption DO data fitted pseudo first and second order kinetic model (Figure 3 2) but only pseudo first order rate constant was calculated for all the temperatures because DO consumption has been modeled as first order with respect to oxygen (Ahrne and others 1997) and a second order kinetic model with respect to oxygen and ascorbic acid (Singh and others 1976) Pseudo first order adjusted well for control and oxygen

PAGE 101

101 satu rated juices. However f or deaerated juices, no t enough data was available to have a good adjustment for the model (Fig ure 3 3). For the juice with air headspace DO concentration remained mostly unchanged during storage and a rate constant could not be calculated Rate c onstant values are reported in T able 3 2. Control and oxygen saturated have similar rate constants at t he same temperature but deaerated juice is not in agreement probably because more data points are needed to do a better model adjustment At 5 C k values were 0.05, 0.14 and 0.13 d 1 for deaerated, control and oxygen saturated juices that compared with 0 .128, 0.283 and 0.285 d 1 k values calculated for deaerated, control, oxygen saturated juices in the C hapter 2 are about half of the values. Orange juices in both cases were stored at 5 C during 60 days and these differences are attributed to the differen t processing conditions (deaeration, pasteurization) and pulp content. Commercial NFC orange juice used for chapter 2 had 0% screened pulp and Ham lin NFC orange juice had 5 % screened pulp Our results for control and oxygen saturated juices are in agreeme nt with those reported by Ponagandla (2010) who used an optic fiber for the determinations instead of an polarographic probe as it was done on this experiment. Ponagandla (2010) reported pseudo first order rate constants for pineappl e juice, which DO content was not modified, of 0.77, 1.12, 2.69 and 4.1 d 1 at 13, 21.5, 30.5 and 40 C respectively. Our results for control and oxygen saturated juices are in agreement with the previous results except at 13 C where the rate constant for pineapple juice is more than double (0.77 d 1 ) that the value reported on this study for control orange juice (0.29 d 1 ) maybe because at low temperatures the rate of DO consumption is very slow. DO consumption in AA solutions at 30 C, with or without sugar had a first order rate of 1.4 and 0.9 d 1

PAGE 102

102 (Hsieh and Harris 1993) that was lower than the rate (2.86 d 1 ) obtained in this study for oxygen saturated orange juice at a similar temperature (30.5 C ). The effect o f temperature assessed by ac tivation energy ranged from 68.66 0.47 to 81.2 0.13 kJ mol 1 and was in agreement with 58.6 to 92.5 kJmol 1 reported for AA solutions with or without sugar at a temperature range of 26.5 to 33 C (Hsieh and Harris 1993) However, it was higher compared to pineapple juice ( E a =48.6 kJmol 1 ) at the same temperature range, suggesting than orange juice is more sensitive to temperature effec ts than pineapple, which is probably due to the adsorption of O 2 onto the pulp. Glass vials were used in this study, to avoid gas excha nge with the surroundings. Soares and Hotchkiss (1999) f ound that deaeration did not make a difference in DO concentration of orange jui ce during 60 days of storage at 7 C when stored in containers with oxygen permeability between 0.35 1.6 mL O 2 day 1 per container of 229 cm 3 of volume. AA Degradation The stoichiometry of AA reaction with oxygen is 1:1 (Hsieh and Harris 1993) and from the data obtained for control orange juice at 5 C 152 M of DO and 2060 M of AA the r atio between DO/AA is 0.07. Therefore, with respect to its initial concentration, only a small amount of AA is needed to consume the entire amount of DO present in the orange juice. This is confirmed as no changes in AA concentration were observed in contr ol orange juice during the period of time ( 1 day at 40 C ) when DO was depleted ( Figure 3 5 B ). During pasteurization the loss of AA increased as init ial DO increased and the values were 12, 14 and 19 % for deaerated, control and oxygen satu rated orange jui ce respectively (Table 3 3). In some cases, at the end of the storage AA concentration was higher compared to the beginning of the storage, this was cause for

PAGE 103

103 an experimental error during AA measurement (capillary and equipme nt variability) and in that cas e the loss of AA could not be calculated. AA loss during 3, 6, 12, 12 and 60 days of storage at 40, 30.5, 21.5 13 and 5 C of orange juices without headspace was between 3 and 23 % AA loss in control and oxygen saturated orange juices was similar for the s ame temperature although it was not the case for 5 and 40 C possibly due to experimental errors during measurement of AA For air headspace stored 60 days at 5 C 6 days at 30.5 C and 3 days at 40 C the loss was 100, 54 and 73%, respectively (Table 3 3 ). As discussed in chapter 2, oxygen from the headspace is constantly being supplied to the orange juice promoting AA oxidation Loss of AA in NFC orange juices studi ed in C hapter s 2 and 3 under same storage conditions (60 days of storage at 5 C ) was 42 a nd 100 % respectively The AA decrease in NFC orange juice from Chapter 2 and 3 was 688 and 2000 M respectively. The difference in AA loss could be due to the different thermal treatment and previous storage time temperature conditions used for each juic e. Figures 3 4 and 3 5 sh ow that AA decreased the most in air headspace juices (Figure 3 4 D and 3 5 D) stored at 5 (Figure 3 4D) and 40 C (Figure 3 5D) perhaps due to the presence of a considerable amount of DO during storage (350 to 150 M DO) Therefo re, to protect AA during storage, headspace should b e filled with an inert gas such as nitrogen which is common practice in industry. Figures 3 4 and 3 5 contrast concentration changes in DO and AA during storage at 5 or 40 C respectively. The overall tre nd for AA in deaerated, control and oxygen saturated orange juices is to remain about the initial concentration while DO content reaches the lowest concentration at the beginning of the storage time. In juice with air headspace AA and

PAGE 104

104 DO have a similar tre nd for the decrease in concentration during storage (Figure 3 4D and 3 5D). The p seudo first order r ate constants of AA degradation were calculated only for air headspace samples (Table 3 4) and the values ranged from 0.018 0.20 d 1 at temperatures b etween 5 40 C These values are higher than 0.0080 0.029 d 1 reported for single strength from concentrate orange juice (SSOJ) stored at 4 37 C (Kennedy and others 1992) ; than 0.0039 0.036 d 1 reported for concentrated orange juice stored at 28 45 C (Burdurlu and others 2006) ; and 6.84x10 5 0.0045 d 1 membrane clarified concentrated orange juice stored at 4 24 C (Lee and Chen 1998) .All the mentioned studies used a first model. Activation energy for AA degradation calculated in study (12.12 kcal mol 1 ) was lower than concentrated orange juices: 12.77 25.39 kcal mol 1 (Burdurlu and others 2006) 34.3 kcal mol 1 (Lee and Chen 19 98) and SSOJ 36 kJ mol 1 (Dhuique Mayer and others 2007) Color Hunter parameters a, b and L were measured in orange juice stored at 40 C (Table 3.5) and o nly air headspace orange juice stored at 40 C showed a darker color after 3 days Th e rest of the juices stored at less than 40 C were not measured because no color change was visually observed. Although in future experiments visual observations should be confirmed with instrumental measurements. For the air headspace orange juice at 40 C, a increased toward more positive direction corresponding to more redness and b became more negative corresponding to less yellowness. These values are exactly opposite of color changes in membrane clarified concentrated orange juice (Lee and Chen 1998) Furfural, produced from AA

PAGE 105

105 degradation has been correlated to browning in orange juice model solutions (Shinoda and others 2005) In this study furfural was identified in all orange juice samples at day 6 of storage at 40 C as explained in the volatile compounds section. Volatile Compounds Thirty five volatile compounds were identified in NFC orange juice and since each replicate was obtained from a different batch of juic e importan t variability between replicates of the same juice was indicated as raw pea k area values in Table 3 7 O il content is shown in Table 3 6 similar values of 0.04% were observed for all three batches of control orange juice. Although oil content in deaerate d and oxygen saturated juices were lower than the control for the same batch of juice ranging between 0.02 to 0.03% probably because limonene was removed during the nitrogen/oxygen bubbling Measurement of limonene in control orange juice from the differen t batches by SPME and GC MS was explored in order to explain samples variability, the SPME conditions of 5 min equilibration at 40 C and 10 s of fiber exposition and same GC MS conditions used for samples analysis were applied to obtain a limonene peak t hat could be completely integrated, limonene raw peak areas obtained for control o range juice were 2.88 x 10 9 2.35 x 10 9 2.91 x 10 9 for batch 1, 2 and 3 respectively but limonene peak was still not well defined in the chromatogram The SPME and GC MS approa ch still needs to be improved in SPME conditions for an accurate measurement of limonene. In order to reduce variability between replicates, t hree normalizations were applied to volatiles data : with internal standard, with tot al peak area and with day 0. Calculations used for the different normalizations are presented in Table 3 7 Normalization with internal standard compensates for experimental error, normalization with total peak area compensates for differences between the replicates and normalization with day 0

PAGE 106

106 compensates for initial differences in replicates. Relative standard deviation s (RSD) of selected compounds in control orange juice at day 0 of storage at 40 C were compared between normalizations (Table 3 8), normalizatio n with internal stand ard had RSD values between 31 and 36, normalization with total peak had RSD values of 15 to 25 and RSD could not be calculated for normalization at day 0 because the assumption of all peaks at day 0 have a value of 100. Normalization with total peak area w as selected for PCA analysis and further discussion because the lower RSD values. Normalized peak values and PCA analysis of the other two normaliz ations methods are included in A ppendix D. Tables 3 9 to 3 11 shows average, ups ( =0.05) for the normalization with total peak area of the three replicates of juices at day 0, 0.5 and 6 respectively. Figure 3 6 shows the proportion of terpenes, alcohols, aldehydes and esters present in all the juices at day 0 and 6 T erpenes were the volatile s present in highest proportion at both storage times even though limonene was excluded from the calculations. At day 0 (Figure 3 6A) not pasteurized fresh squeezed orange juice had the highest proportion of 0.8 for terpenes and the lowest of 0.07 for aldehydes compared with the pasteurized juices Among the pasteurized juices, deaerated juice showed the same trend than fresh squeezed with respect to the rest of the juices. A low proportion of aldehydes in fresh squeezed orange juice was unex pected since aldehydes have been reported as important contributors to orange juice flavor (Perez Cacho and Rouseff 2008b) At day 6 of storage (Figure 3 6B) deaerated orange juice had about the same 0.25 proportion of terpenes, alcohols, aldehydes and alcohol terpenes. Oxygen saturated had about the

PAGE 107

107 similar proportion of volatile compounds than deaerated juice exce pt that terpenes represented a proportion of 0.4 of the total composition. Principal component analysis ( PCA ) is a tool that helps to exam ine large data set reducing them by associating variables (individual volatiles) to lower dimensions or principal components. PCA help to identify p atterns in the data and highlight similarities and differences among samples (different levels of DO) and help to identify those volatiles that are more dif ferent than the entire data set and that explain the differences between samples. 54% of variation b etween j uice samples was explained by PCA and distribution of samples in the loading plot (Figure 3 7) was by regions depending of the storage time going in diagonal from the right up corner to the left bottom corner from day 0 to day 6. Only replicates of deaerated and air headspace juices were clustered in well defined groups (Figure 3 7 ). If Figures 3 7 and 3 8 are overlapped, t he compounds in the loading plot (Figure 3 8) are in higher amount in samples located in the same section of the plot in Figure 3 7 and that compounds explain the differences between samples. In this case, pinene, limonene, myrcene, octanal, n onanal, decanal, elemene, geranial, ethyl octanoat e and oct yl acetate are higher in juices at day 0 compared to day 6. T he opposite tre nd was observed for furfural, terpineol terpineol that were only detected at day 6, and 1 hexanol which content increased over the time. The mentioned compounds were significant different except limonene, nonanal, decanal, geranial, ethyl octanoate an d octyl acetate. From Du groups reported in T ables 3 9 to 3 11, it was observed that among juices pinene content was higher in air headspace juice, myrcene and elemene were lower in

PAGE 108

108 control and oxygen saturated juices and octanal was lower in deaerated and oxygen saturated juices. Conclusions Orange juice samples prepared with a ir in the headspace produced the highest AA degradation during storage as observed in the previous chapter DO consumption in control and oxygen saturated orange jui ce had similar kinetic rate values under the same storage conditions and fit a pseudo first order model. For deaerated juice, insufficient sampling in DO determination did not allow accurate determination of the rate constant. Experimental error in AA mea surement has to be address by running an AA standard the same day that orange juice samples, to compensate by differences among days of testing. AA and DHA measurements of orange juice without headspace should be repeated to have enough data to calculate p ercentage of AA loss and rate constant of AA oxidation. Using the normalization with total peak, PCA and ANOVA, pinene myrcene, octanal elemene, are higher in juices at day 0 compared to day 6. F urfural, terpineol, terpineol were only detected at day 6 1 hexanol increased over the time. Being all this differences significant over the time but not between the juic es at different levels of DO. Additional studies are needed to confirm if DO has a direct effect on volatile compounds. 9 to 3 11, it was observed that among juices pinene content was higher in air headspace juice, myrcene and elemene were lower in control and oxygen saturated juices and octanal was lower in deaerated and oxygen saturated juices.

PAGE 109

109 Table 3 1 Dissolved oxygen and ascorbic acid sampling schedule during storage study T (C) DO AA 5 Every 3 d fo r 2 wk and 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 and at day 6 and 12 30.5 At 6 h, 12 h, 18 h, 1 d, 3 d, 6 d 40 2 h, 6 h, 12 h, 24 h, 3 d Table 3 2. Pseudo first order rate constant (d 1 ) of dissolved oxygen cons umption in NFC orange juice T ( C) Deaerated Control Oxygen saturated Control w/ air headspace 5 0.05 0.01 (r 2 =0.85) 0.14 0.02 (r 2 =0.99) 0.13 0.03 (r 2 =0.89) b 13 0.37 0.03 (r 2 =0.96) 0.29 0.08 (r 2 =0.97) 0.24 0.02 (r 2 =0.95) b 21.5 a 0.96 0.03 (r 2 =0.94) 1.00 0.06 (r 2 =0.99) b 30.5 a a 2.86 0.1 (r 2 =0.91) b 40 a 6.61 0.8 (r 2 =0.98) 5.01 0.7 (r 2 =0.99) b E a (kJmol 1 ) 81.24 0.13 68.66 0.47 b a Insufficient data to determine rate constant at low incubation time, a [ DO] did not vary significantly during this experiment

PAGE 110

110 Table 3 3. AA loss due to pasteu rization and storage at selected temperatures AA l oss due to pasteurization with respect to not pasteurized fresh squeezed (%) T (C) storage time (d) Deaerated Co ntrol Oxygen saturated Control w/ air headspace 5 0 12 4 14 4 19 3 14 4 AA loss due to storage after pasteurization (%) T ( C) storage time (d) Deaerated Control Oxygen saturated Control w/ air headspace 5 60 12 5 5 3 23 2 100 0.0 13 12 ND 12.0 5 7 1 17 9 21.5 12 ND 9 3 10 3 27 3 30.5 6 3 1 19 3 20 4 54 2 40 3 ND 4 2 18 4 73 11 a not calculated due to experimental errors Table 3 4. Pseudo first order rate constant (d 1 ) of AA degr adation in NFC orange juice with air headspace T ( C ) k (d 1 ) r 2 5 0.018 0.96 13 21.5 0.027 0.88 30.5 0.093 0.9 8 40 0.203 0.96 E a 12.12 kcal mol 1 *Insufficient data to determine rate constant at low incubation time

PAGE 111

111 Table 3 5. Color change in NFC orange juice durin g storage at 40 C and expressed as L, a and b with respect to control orange juice Deaerated Oxygen saturated Air headspace Time (d) L a b L a b L a b 0.08 0.3 0.3 1.0 0.3 1.2 4.4 2.4 0.6 1.4 0.25 2.3 0.4 2.0 0.3 1.5 4. 7 2.5 0.1 0.3 0.5 1.1 0.03 0.8 2.0 1.0 3.9 0.9 0.8 1.4 1 1.1 0.2 0.3 0.6 0.5 2.9 3.5 0.3 0.3 3 2.3 0.9 4.2 2.7 1.4 5.2 3.0 2.4 3.4 Table 3 6. Oil content in NFC orange juice at day 0 of storage at 40 C calculated using Scot t oil method O il content (%) Orange juice batch control deaerated oxygen saturated 1 4.5E 02 2.6E 03 3.3E 02 2.6E 03 3.1E 02 1.4E 03 2 4.1E 02 2.2E 03 3.2E 02 a 2.4E 02 1.6E 03 3 4.1E 02 0 1.8E 02 1.2E 03 1.6E 02 2.0E 04 a sample was av ailable for only one replicate

PAGE 112

112 Table 3 7. Normalization methods applied to volatile co mpounds of NFC orange juice Normalization Calculation of normalized peak area Internal standard Total peak area Day 0 A value of 100 is assigned to peak area at day 0 of all the compounds

PAGE 113

113 Table 3 8 Comparison of r aw peak area and two different normalizations applied to selected volatile compounds in control not from concentra ted orange juice at day 0 of storage at 40 C. C ontrol orange juice at day 0 R aw peak area Compound Replicate 1 2 3 RSD linalool 2.01E+08 7.60E+07 9.55E+07 54.2 octanal 1.92E+08 5.67E+07 8.73E+07 63.3 nonanal 5.36E+07 2.16E+07 2.68E+07 50 .5 4 heptadecanone (internal standard) 1.26E+09 4.01E+08 3.34E+08 77.6 Normalization with internal standard Compound Replicate 1 2 3 RSD linalool 0.15 0.14 0.26 35.8 octanal 0.04 0.05 0.08 32.8 nonanal 0.16 0.19 0.29 31.1 Norm alization with total peak area Compound Replicate 1 2 3 RSD linalool 1.25E+08 1.06E+08 1.43E+08 15.0 octanal 1.19E+08 7.88E+07 1.31E+08 24.9 nonanal 3.32E+07 3.01E+07 4.01E+07 14.9 Normalization with day 0, All values were 100 and RSD could not be calculated

PAGE 114

114 Table 3 9 Volatile compounds identified in NFC orange juice at different DO content at day 0 of storage at 40 C. Values are average and standard error of areas normalized with total peak area Values followed by differen t letters are =0.05). LRI ( ref )* LRI (DBWax) compound F resh squeezed D eaerated Control Oxygen saturated A ir headspace ALDEHYDES (9 ) 698 ( f ) 722 acetaldehyde 9.74E+07 4.43E+07 a 8.05E+07 3.63E+07 a 2 .96E+08 2.10E+08 a 1.82E+08 1.20E+08a 1297 ( c ) 1217 octanal 5.28E+06 2.76E+06 1.04E+08 1.14E+07 ba 1.10E+08 1.58E+07 a 4.40E+07 2.21E+07 b 1.44E+08 4.96E+06a 1411 ( b,c ) 1317 nonanal 2.42E+06 7.20E+05 3.99E+07 5.53E+05 a 3.45E+07 2.97E+06 a 3.5 7E+07 1.15E+07 a 4.35E+07 1.44E+06a 1433 ( d ) 1396 furfural 1501 ( b,c ) 1415 decanal 5.67E+06 2.77E+05 2.84E+08 3.16E+06 a 2.02E+08 1.12E+07 a 2.12E+08 3.47E+07 a 2.79E+08 1.68E+07a 1630 ( g ) 1585 neral 1.27E+07 5.36E+05 1.78E+07 1. 00E+06 ba 3.06E+06 9.76E+05 b 1.63E+07 4.45E+06 a 2.10E+07 5.78E+06a 1728 ( f ) 1619 dodecanal 1.10E+07 1.06E+06 3.87E+07 1.23E+06 a 2.13E+07 8.70E+05 a 4.45E+07 2.14E+07 a 3.10E+07 1.58E+06a 1766 ( a ) 1651 geranial 1.32E+07 1.69E+06 2.42E+07 7.00 E+05 a 1.74E+07 1.18E+06 b 1.61E+07 2.16E+06 b 2.62E+07 5.16E+06b 1835 ( f ) 1713 perillaldehyde 2.27E+06 2.72E+05 1.96E+07 2.51E+06 a 7.89E+06 3.25E+06 a 2.03E+07 6.73E+06 a 1.73E+07 2.42E+06a ESTERS (4 ) 1049 ( b,c,e ) 978 ethyl butanoate 1.78E+07 4.32E+06 1.72E+07 1.53E+06 a 2.82E+07 8.51E+06 a

PAGE 115

115 Table 3 9. Continued LRI ( ref )* LRI (DBWax) compound F resh squeezed D eaerated Control Oxygen saturated A ir headspace 1448 ( b,c,e ) 1344 ethyl octanoate 1.17E+07 1.18E+06 1.14E+07 9.16E+05 a 5.50E+06 2.72E+05 a 7.48E+06 6.77E+05 a 6.73E+06 2.54E+05a 1485 ( c ) 1383 octyl acetate 7.45E+06 3.25E+06 3.16E+07 7.12E+05 a 1.38E+07 1.26E+06 b 2.38E+07 3.49E+06 ba 2.08E+07 8.78E+05ba 1706 ( f ) 1626 neryl acetate 1.08E+07 1.21E+06 a 2.29E+0 6 8.51E+05 b 8.91E+06 2.35E+06 a TERPENES (10 ) 1025 ( a,c,e ) 964 pinene 2.13E+07 2.32E+06 1.59E+08 1.37E+07 b 1.22E+08 1.09E+07 b 8.87E+07 2.91E+07 b 1.66E+08 1.49E+07a 1188 ( a,c,d,e ) 1139 L imonene 4.90E+09 3.91E+08 1.85E+10 3.18E+08 a 1.24E +10 1.78E+08 b 1.09E+10 1.21E+09 b 1.56E+10 3.23E+08a 1148 ( a,b,d,e ) 1082 myrcene 4.70E+07 4.55E+07 1.24E+09 7.73E+07 a 7.02E+08 1.00E+08 ba 6.72E+08 1.58E+08 b 1244 ( e ) 1195 cymene 1.47E+06 7.85E+05 3.21E+06 3.74E+05 a 1290 ( c,d, e ) 1204 terpinolene 3.93E+06 4.57E+05 3.22E+07 1.22E+06 a 1.05E+07 2.47E+06 a 5.61E+07 3.68E+07 a 8.56E+06 6.98E+05a 1624 ( a ) 1501 elemene 8.13E+06 3.09E+06 1.26E+07 9.69E+05 a 4.93E+06 4.46E+05 b 7.31E+06 4.54E+05 ba 2.91E+08 1.46E+07b 17 31 ( d,e ) 1636 V alencene 3.77E+08 6.69E+06 4.63E+08 2.20E+07 a 4.18E+06 1.22E+06 b 2.71E+08 2.19E+07 a 1137 ( a,e ) 1142 S abinene 2.50E+07 4.94E+05 2.00E+08 1.24E+08 a 8.38E+08 1.96E+08a

PAGE 116

116 Table 3 9. Continued LRI ( ref )* LRI (DBWax) compoun d F resh squeezed D eaerated Control Oxygen saturated A ir headspace 1248 ( a,e ) 1169 ocimene 3.07E+07 2.96E+06 6.17E+07 2.56E+05 a 1689 ( e ) 1642 selinene 3.19E+07 4.96E+06 1.99E+07 1.21E+06 a 1.23E+07 2.90E+06 a 8.49E+07 7.16E+07 a 1.42E+07 3.63E+06a ALCOHOL TERPENES (5) 1560 ( a,c,d ) 1449 L inalool 1.27E+07 1.58E+06 1.74E+08 1.90E+07 a 1.24E+08 1.08E+07 a 3.16E+08 2.03E+08 a 1.68E+08 1.21E+07a 1572 ( e ) 1521 4 terpineol 6.95E+07 3.61E+06 a 5.01E+07 4.98E+06 a 7.04E+07 2.24E+07 a 7.06E+07 3.57E+06a 1616 ( d ) 1543 terpineol 1661 ( d ) 1608 terp ineol 8.04E+06 1.90E+06 a 1777 (a,c) 1655 citronellol 4.75E+06 4.27E+05 1.02E+07 1.04E+06 a 9.88E+06 1.84E+06 a 1.40E+07 6.62E+06 a 1.10E+07 1.92E+06a ALCOHOLS (7 ) 949 ( a,e ) 887 E thanol 4.41E+07 8.50E+06 a 1.56E+08 3.51E+07 a 1.44E+08 1.03E+07 a 2.08E+08 1.14E+08 a 1.42E+08 9.12E+06a 1364 (a) 1265 1 hexanol 2.77E+06 4.01E+05 1351 ( d ) 1298 3 hexen 1 ol 2.65E+06 3.55E+05 2.91E+06 5.82E+05 a 1534 ( e ) 1458 1 octanol 7.08E+06 8.14E+05 8.40E+07 1.32E+07 a 5.44E+ 07 3.11E+06 a 1.19E+08 6.51E+07 a 7.44E+07 5.54E+06a

PAGE 117

117 Table 3 9. Continued LRI ( ref )* LRI (DBWax) compound F resh squeezed D eaerated Control Oxygen saturated A ir headspace 1640 ( e ) 1558 1 nonanol 1.36E+07 1.61E+06 a 7.60E+06 2.80E+05 a 1.75E+0 7 8.68E+06 a 1.06E+07 7.52E+05a 1745 ( e ) 1654 1 decanol 3.82E+07 1.82E+06 a 1.05E+07 4.66E+06 a 3.35E+07 8.78E+06 a 2.86E+07 2.87E+06a 1818 ( a,d ) 1736 nerol 4.35E+06 4.30E+05 1.14E+07 1.13E+06 a 6.31E+06 4.01E+05 a 1.32E+07 5.61E+06 a 8.15E+06 6. 89E+05a a (Jabalpurwala 2009) ; b (Arena and others 2006) ; c (Berlinet and others 2007) ; d ( Berlinet and others 2006) ; e (Brat and others 2003) ; f (Rouseff 2008) ; g (Mottram 2010) not detected.

PAGE 118

118 Table 3 10. Volatile compounds identified in NFC orange juice at different DO content at day 0.5 of storage at 40 C. Values a re average and standard error of areas normalized with total peak area V alues followed by different =0.05) LRI ( ref )* LRI (DBWax) C ompound D eaerated Control Oxygen saturated A ir headspace ALDEHYDES (9 ) 698 ( f ) 722 acetaldehyde 1.06E+08 3.50E+07 7.56E+06 3.41E+0 6 1.50E+08 4.00E+07 3.83E+08 3.23E+08 1297 ( c ) 1217 octanal 4.68E+07 4.51E+05 5.91E+07 2.64E+07 5.68E+07 4.49E+06 8.03E+07 4.03E+07 1411 ( b,c ) 1317 nonanal 1.67E+07 3.30E+05 2.05E+07 4.40E+06 1.87E+07 2.04E+06 3.63E+07 1.92E+07 1433 ( d ) 1396 furfural 1501 ( b,c ) 1415 decanal 1.21E+08 4.46E+06 1.28E+08 3.10E+07 1.19E+08 1.31E+07 2.77E+08 1.65E+08 1630 ( g ) 1585 neral 4.67E+06 4.94E+05 4.71E+06 1.70E+06 5.38E+06 5.53E+05 1.44E+07 1.05E+07 1728 ( f ) 1619 dodecanal 1.65E+07 9.45E+05 1.36E+07 4.22E+06 1.36E+07 2.73E+06 8.98E+06 4.71E+06 1766 ( a ) 1651 geranial 1.62E+07 4.43E+06 6.00E+06 1.57E+06 1835 ( f ) 1713 perillaldehyde 8.23E+06 1.66E+06 1.31E+07 3.56E+06 9.86E+06 1.17E+06 3.69E+07 2.84E+07

PAGE 119

119 Table 3 10. Continued LRI (ref)* LRI (DBWax) Compound Deaerated Control Oxygen saturated Air headspace ESTERS (4) 1049 (b,c,e) 978 ethyl butanoate 1.01E+07 3.27E+06 1.55E+07 2.91E+06 1448 (b,c,e) 1344 ethyl octanoate 6.24E+06 2.9 2E+05 4.26E+06 8.89E+05 6.66E+06 7.48E+05 9.56E+06 5.55E+06 1485 (c) 1383 octyl acetate 1.77E+07 1.73E+06 9.81E+06 3.99E+06 1.96E+07 2.68E+06 2.16E+07 1.12E+07 1706 (f) 1626 neryl acetate 5.22E+06 1.20E+06 3.41E+06 1.36E+06 6.77E+06 7.44E +05 TERPENES (10) 1025 (a,c,e) 964 pinene 8.65E+07 1.50E+07 2.28E+07 9.46E+05 9.69E+07 1.58E+07 1.04E+08 1.05E+07 1188 (a,c,d,e) 1139 L imonene 1.17E+10 1.07E+08 4.95E+09 3.99E+08 1.23E+10 4.72E+08 7.50E+09 3.48E+09 1148 (a,b,d,e) 1082 myrcene 2.42E+06 6.42E+05 1244 (e) 1195 cymene 2.72E+07 9.38E+06 3.56E+06 1.44E+06 1.79E+07 2.64E+06 1290 (c,d,e) 1204 terpinolene 4.94E+06 9.20E+05 1.37E+06 5.71E+05 6.62E+06 6.46E+05 8.11E+06 4.07E+06 1624 (a) 1501 elemene 2.67E+08 1.67E+07 1.51E+08 5.41E+07 3.27E+08 1.45E+07 3.25E+08 1.57E+08 1731 (d,e) 1636 V alencene 3.62E+07 5.66E+06 7.45E+07 3.07E+06

PAGE 120

120 Table 3 10. Continued LRI (ref)* LRI (DBWax) Compound Deaerated Control Oxygen saturated Air headspace 1137 (a,e) 1142 S abinene 6.44E+08 1.16E+08 7.53E+07 6.38E+07 2.83E+08 2.78E+08 5.00E+08 3.79E+07 1248 (a,e) 1169 ocimene 2.85E+07 7.82E+06 1689 (e) 1642 selinene 2.02E+07 2.16E+06 3.58E+06 1.79E+06 2.13E+07 1.44E+06 1.20E+07 3.87E+06 ALCOHOL TERPENES (5) 1560 (a,c,d) 1449 L inalool 9.52E+07 7.28E+06 1.98E+08 6.66E+07 1.12E+08 7.66E+06 2.76 E+08 1.91E+08 1572 (e) 1521 4 terpineol 3.52E+07 2.96E+06 5.05E+07 1.53E+07 4.16E+07 4.53E+06 1.09E+08 7.40E+07 1616 (d) 1543 terpineol 1.92E+06 1.12E+05 1661 (d) 1608 terpineol 1777 (a,c) 1655 citronellol 9.74E+06 1.42E+06 6.41E+06 2.08E+06 6.78E+06 5.16E+05 2.10E+07 1.46E+07 ALCOHOLS (7) 949 (a,e) 887 E thanol 1.07E+08 1.40E+07 1.14E+08 3.95E+07 1.12E+08 1.72E+07 1.68E+08 8.52E+07 1364 (a) 1265 1 hexanol 2.03E+06 5.27E+05 3.03E+06 1.25E+06 2.44E+ 06 3.91E+05 1351 (d) 1298 3 hexen 1 ol 4.80E+06 3.11E+06 2.06E+06 7.10E+05 2.16E+06 2.49E+05

PAGE 121

121 Table 3 10. Continued LRI (ref)* LRI (DBWax) Compound Deaerated Control Oxygen saturated Air headspace 1534 (e) 1458 1 octanol 4.52E+07 5. 85E+06 9.50E+07 2.08E+07 4.85E+07 2.51E+06 1.24E+08 8.47E+07 1640 (e) 1558 1 nonanol 1.32E+07 2.27E+06 1.05E+07 3.82E+06 7.81E+06 4.00E+05 2.24E+07 1.65E+07 1745 (e) 1654 1 decanol 1.23E+07 7.10E+06 1.56E+07 4.63E+06 4.54E+07 3.04E+07 1818 (a,d) 1736 N erol 5.97E+06 7.87E+05 6.88E+06 2.74E+06 7.74E+06 8.65E+05 2.27E+07 1.73E+07 *a (Jabalpurwala 2009) ; b (Arena and others 2006) ; c (Berlinet and others 2007) ; d (Berlinet and others 2006) ; e (Brat and others 2003) ; f (Rouseff 2008) ; g (Mottram 2010) not detected.

PAGE 122

122 Table 3 11. Volatile compounds identified in NFC orange juice at different DO content at day 6 of storage at 40 C. Values are average and standard error of areas normalized with total peak area. Values followed by different =0.05) LRI (ref)* LRI (DBWax) Compound Deaerated Control Oxygen saturated A ir headspace ALDEHYDES (9 ) 698 ( f ) 722 A cetaldehyde 1.56E+08 3.87E+07 1.12E+08 4.63E+06 1.80E+08 1.35E+08 1.39E+08 1.78E+07 1297 ( c ) 1217 octanal 1.65E+07 7.59E+06 4.44E+07 4.11E+06 1.37E+07 4.71E+06 2.59E+07 3.04E+06 1411 ( b,c ) 1317 no nanal 8.51E+06 2.90E+06 1.16E+07 6.22E+05 8.18E+06 2.05E+06 8.37E+06 3.21E+05 1433 ( d ) 1396 furfural 2.10E+06 8.25E+05 1.91E+06 2.03E+05 2.63E+06 1.22E+06 2.43E+06 9.84E+05 1501 ( b,c ) 1415 decanal 5.54E+07 1.13E+07 7.73E+07 3.49E+06 4.52 E+07 8.80E+06 6.68E+07 1.04E+06 1630 ( g ) 1585 neral 1.00E+06 2.81E+05 7.29E+06 2.75E+06 1728 ( f ) 1619 dodecanal 1.06E+07 2.20E+06 3.81E+07 2.12E+07 6.63E+06 4.87E+06 1766 ( a ) 1651 geranial 1.25E+06 3.20E+05 1835 ( f ) 1713 perillaldehyde 8.52E+06 3.52E+06 8.43E+06 6.61E+05 7.77E+06 2.28E+06 4.66E+06 5.39E+05

PAGE 123

123 Table 3 11. Continued LRI (ref)* LRI (DBWax) Compound Deaerated Control Oxygen saturated Air headspace ESTERS (4 ) 1049 ( b,c,e ) 978 ethyl butanoate 1 .06E+07 1.12E+06 1.38E+07 2.36E+06 1.25E+07 4.61E+06 1448 ( b,c,e ) 1344 ethyl octanoate 4.84E+06 5.98E+05 4.85E+06 3.85E+05 4.35E+06 5.04E+05 4.73E+06 6.07E+05 1485 ( c ) 1383 octyl acetate 1.55E+07 3.39E+06 1.27E+07 1.82E+06 1.34E+07 3.89 E+06 1.16E+07 1.50E+06 1706 ( f ) 1626 neryl acetate 5.98E+06 1.73E+06 3.50E+06 1.25E+06 4.01E+06 4.54E+05 2.18E+06 2.20E+05 TERPENES (10 ) 1025 ( a,c,e ) 964 pinene 2.24E+07 2.85E+06 9.73E+07 6.41E+06 2.63E+07 7.37E+06 1.05E+08 1.82E+07 11 88 ( a,c,d,e ) 1139 L imonene 5.35E+09 4.70E+08 1.23E+10 9.44E+07 5.17E+09 4.20E+08 1.26E+10 6.89E+07 1148 ( a,b,d,e ) 1082 myrcene 2.32E+06 8.57E+05 1244 ( e ) 1195 cymene 1.05E+07 2.05E+06 5.01E+07 2.74E+07 8.46E+06 1.50E+06 1.97E+0 7 5.83E+06 1290 ( c,d,e ) 1204 terpinolene 2.31E+06 5.52E+05 2.57E+06 5.63E+05 3.18E+06 1.08E+06 1624 ( a ) 1501 elemene 1.84E+08 2.58E+07 2.20E+08 2.00E+07 2.10E+08 5.05E+07 2.65E+08 1.45E+07 1731 ( d,e ) 1636 V alencene 2.68E+07 6.52 E+06 3.63E+06 6.02E+05

PAGE 124

124 Table 3 11. Continued LRI (ref)* LRI (DBWax) Compound Deaerated Control Oxygen saturated Air headspace 1137 ( a,e ) 1142 S abinene 4.95E+06 9.38E+05 7.71E+08 9.25E+07 2.46E+08 2.85E+07 6.21E+08 4.79E+07 1248 ( a,e ) 1169 ocimene 1.81E+06 1.71E+05 1689 ( e ) 1642 selinene 1.31E+07 1.77E+06 1.79E+07 2.90E+06 1.77E+07 4.61E+06 7.94E+06 5.32E+06 ALCOHOL TERPENES (5) 1560 ( a,c,d ) 1449 L inalool 1.68E+08 6.10E+07 1.00E+08 8.83E+06 1.28E+08 4.14E+07 6.68 E+07 7.53E+06 1572 ( e ) 1521 4 terpineol 4.24E+07 1.36E+07 3.92E+07 3.06E+06 3.66E+07 1.15E+07 3.44E+07 1.57E+06 1616 ( d ) 1543 terpineol 1.98E+07 8.89E+06 8.77E+06 1.65E+06 2.35E+07 1.10E+07 9.93E+06 2.77E+06 1661 ( d ) 1608 terpineol 8.04E+06 1.90E+06 1777 (a,c) 1655 citronellol 1.14E+07 4.22E+06 9.84E+06 3.26E+06 1.20E+07 3.88E+06 ALCOHOLS (7 ) 949 ( a,e ) 887 E thanol 1.68E+08 4.87E+07 1.30E+08 1.67E+07 1.18E+08 2.96E+07 1.09E+08 1.08E+07 1364 (a) 1265 1 hexanol 7.28E+06 4.06E+06 1.64E+06 3.91E+05 1.52E+06 1.69E+05 1351 ( d ) 1298 3 hexen 1 ol 5.22E+06 2.08E+06 1.83E+06 4.29E+05 1.69E+06 1.54E+05

PAGE 125

125 Table 3 11. Continued LRI (ref)* LRI (DBWax) Compound Deaerated Control Oxygen saturated A ir headspace 1534 ( e ) 1458 1 octanol 1.07E+08 4.19E+07 5.01E+07 1.69E+06 6.97E+07 1.98E+07 3.89E+07 2.55E+06 1640 ( e ) 1558 1 nonanol 1.67E+07 7.04E+06 7.04E+06 4.29E+05 9.70E+06 2.35E+06 5.80E+06 3.00E+05 1745 ( e ) 1654 1 decanol 3.80E+0 7 1.57E+07 1.73E+07 2.44E+06 2.23E+07 4.76E+06 3.54E+06 1.08E+06 1818 ( a,d ) 1736 N erol 1.51E+07 4.88E+06 7.82E+06 1.73E+06 1.43E+07 5.24E+06 9.55E+06 1.69E+06 *a (Jabalpurwala 2009) ; b (Arena and others 2006) ; c (Berlinet and others 2007) ; d (Berlinet and others 2006) ; e (Brat and others 2003) ; f (Rouseff 2008) ; g (Mottram 2010) not detected.

PAGE 126

126 Figure 3 1. Flow diagram of sam ple preparation of not from concentrated orange juice (Hamlin var.)

PAGE 127

127 Figure 3 2 Different kinetic models applied to DO data of not from concentrate orange juice ( cultivar Hamlin ) stored at 5 C.

PAGE 128

128 Figure 3 3. First order kinetic data of not from con centrate orange juice (Hamlin var.) stored at 5 C. ( )deaerated, ( ) control, ( )oxygen saturated, ( )air headspace

PAGE 129

129 Figure 3 4. ( )ascorbic acid and ( )dissolved oxygen content in not from concentrated orange juice (Hamlin var) during storage at 5 C A) Deaerated, B) control, C) Oxygen saturated, D) air headspace.

PAGE 130

130 Figure 3 5. ( )ascorbic acid and ( ) dissolved oxygen content in not from concentrated orange juice (Hamlin var) during storage at 40 C A) deaerated, B) control, C) oxygen saturated, D) air headspace.

PAGE 131

131 Figure 3 6 P roportion of v olatile compounds identified in NFC orange juice stored at 40 C. (A) 0 days of storage and fresh squeezed orange juice (B) 6 days of storage. P roportion of terpenes were calculated excluding limonene.

PAGE 132

132 Figure 3 7. Score plot of PCA appl ied to volatile compounds from NFC orange juice stored at 40 C normalized with total peak area of volatile compounds

PAGE 133

133 Figure 3 8. Loading plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C normalized with total peak area of volatile compounds significant differences

PAGE 134

134 CHAPTER 4 COMPREHENSIVE RESULT S AND FUTURE WORK This research described the changes in vitamin C content, color and aroma in not from concentrate orange juice during processing and storage at diff erent temperatures. The first chapter presents the knowledge already available about oxygen interactions with vitamin C, color and aroma in different fruit juices and the methods used for DO removal. In most of these previously published studies the oxidat ion of AA due to presence of oxygen is assumed without monitoring changes in DO concentration during the experiment. This research hypothesized that DO decreases vitamin C and deteriorates aroma and color of orange juice during storage. The second and thir d chapter confirmed that the amount of oxygen present in orange juice even at saturation levels is about 10 times smaller than vitamin C content and just a small portion of vitamin C reacted with DO oxygen when no additiona l oxygen was supplied to the samp le then dissolved oxygen did not account for important loss of ascorbic acid. The effect of having air in the headspace in the loss of ascorbic acid was also studied and the loss of asco rbic acid ranged from 42 to 100 % after two months of storage at 5 C Under this experiment conditions, oxygen dissolved into the juice or present in the headspace of containers did n ot seem to have a direct effect on volatile compounds since significant dif ferences could not be correlated to DO content. Although volatile c ompounds such as furfural, produced due to ascorbic acid degradation were identified Color change was visually detected only in orange juice with air headspace after 3 days of storage at 40 C For the rest of the juices no color change was observed mainl y because storage times were short, less than 15 days at temperatures of 20.5 40 C and 60 days at 5 C

PAGE 135

135 The s econd chapter showed that DO content reached its lowest concentration after 7 days of storage at 5 C Considering the shelf life of not from co ncentrate orange juice is two months at 5 C oxidation can happen only at the beginning of the storage. Presence of air in the headspace of containers produced a larger AA degradation (42%) than oxygen saturated juices without air in headspace during 60 da ys of storage at 5 C From the aroma active compounds, methional was perceived at higher intensity in air headspace orange juice with respect to control only and nonanal had the main contribution to aroma intensity of control, oxygen saturated and air hea space orange juices. At day 1 of storage elemene, selinene and 3 carene had the highest amounts in deaerated orange juice compared with the other juices. At day 60 of storage, the content of most of the compounds decreased with respect to day 0 excep t for E 2 hexenal and limonene oxide whose content increased in air headspace orange juice. In the third chapter, a pseudo first order was used to calculate rate constant and activation energy of DO consumption in NFC orange juice. Loss of AA due to pasteu rization and due to storage was quantified separately Loss of AA due to pasteurization was less than 19% for all the juices and during storage losses of AA between 17 to 100% happened in juice containing air headspace Change in color was visually detecte d only in orange juice with headspace at the highest temperature ( 40 C). Different normalizations were applied to volatiles data in order to compensate for differences between the replicates. Differences between replicates were not detected by Scott oil m easurement Normalization with total peak are a gave relative standard deviations smaller than normalization with internal standard and with day 0 and it was

PAGE 136

136 selected for further ANOVA and PCA analysis. observed that among j uices pinene content was higher in air headspace juice, myrcene and elemene were lower in control and oxygen saturated juices and octanal was lower in deaerated and oxygen saturated juices. F urfural, terpineol, terpineol were only detected at day 6, and 1 hexanol content increased over the time. To address the remaining questions from this research, the effect of oxygen on each quality parameter (vitamin C, aroma and color) should be studied separate ly before performing a storage study monitoring all the parameters. To calculate an accurate pseudo first order rate constant for deaerated orange juice a storage study with frequent sampling times at the beginning of storage and longer sampling times towar the end of storage is required. Moreover a me chanism for filling the vials after pasteurization that avoids reentry of air needs to be implemented. Storage times of at least 6 months are required to detect changes in AA, DHA and color in juices without headspace, but since microbial growth is a conce rn at temperatures between 13 and 30.5 C a filling method that increases shelf life of orange juice at these temperatures such as hot filling instead of cold filling should be utilized Confirmation of identity of aroma active com pounds should be done usi ng a second column (DB5) in order to verify the results obtained in the second chapter Identification of volatile compounds should be repeated using replicates from the same batch of orange juice, pouring the juice immediately after gas sparging (nitrogen or oxygen) directly into the vials to use for the analysis and flushing the headspace with nitrogen to have only the effect of dissolved oxygen

PAGE 137

137 The effect of headspace composit ion in orange juice can be studied by having glass vials with a headspace volu me of 3% of the total container volume filled with different gases such as nitrogen, air and oxygen. Future studies on the effect of DO o n aroma could be conducted relating to packaging of orange juice, since the oxygen permeability of packaging material, volume and composition of headspace of packages affect the amount of dissolved oxygen in the juice

PAGE 138

138 APPENDIX A DISSOLVED OXYGEN CON TENT IN NFC OJ (HAML IN CULTIVAR)

PAGE 139

139

PAGE 140

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

PAGE 141

141 APPENDIX C GAS SPARGING EXPERIM ENTAL SETUP Sparging gas (O 2 or N 2 ) Orange juice

PAGE 142

142 APPENDIX D NORMALIZATION OF VOLATILE COMPOUND S IN NFC OJ ( HAMLIN CULTIVAR) AND PCA ANALYSIS NORMALIZED WITH INTERNAL STANDARD Table D 1. Volatile compounds identified in NFC orange juice at different DO content at day 0 of storage at 40 C. Values are average and standard error of normalized peak area using 4 heptadecanone as internal standard. Values =0.05). LRI (ref) LRI (DBWax) C ompo und Fresh squeezed Deaerated Control Oxygen saturated Air headspace ALDEHYDES (10) 698 ( f ) 722 acetaldehyde 1 8.77E 02 7.06E 02 9.07E 02 7.45E 02 1.29E 01 6.61E 02 2.99E 01 2.16E 01 1100 ( a,d ) 1024 hexanal 5.08E 03 5.15E 04 129 7 ( c ) 1217 octanal 5.53E 03 4.20E 03 1.71E 012.95E 02 1.47E 01 5.69E 03 9.67E 02 8.29E 03 1.74E 01 1.02E 02 1411 ( b,c ) 1317 nonanal 6.33E 03 6.34E 04 6.17E 021.50E 02 4.83E 02 5.69E 03 3.56E 02 4.48E 03 5.15E 02 8.50E 05 1433 ( d ) 1396 fu rfural 1501 ( b,c ) 1415 decanal 1.19E 02 1.00E 03 4.36E 01 1.00E 01 2.93E 01 4.74E 02 2.61E 01 2.80E 02 3.22E 01 1.19E 02 1630 ( g ) 1585 neral 2.46E 02 1.62E 03 2.87E 02 7.11E 03a 4.67E 03 3.19E 03b 1.73E 02 4.12E 04ba 2.81E 02 1 .15E 02a 1728 ( f ) 1619 dodecanal 2.20E 02 2.87E 03 5.95E 02 1.21E 02 3.19E 02 6.46E 03 3.38E 02 7.67E 04 3.63E 02 1.67E 03

PAGE 143

143 Ta ble D 1 Continued LRI (ref) LRI (DBWax) C ompound Fresh squeezed Deaerated Control Oxygen saturated Air headspace 1766 ( a ) 1651 geranial 2.33E 02 8.30E 04 3.80E 02 1.05E 02 2.80E 02 3.04E 03 2.67E 02 8.32E 04 3.28E 02 1.09E 02 1835 ( f ) 1713 perillaldehyd e 4.35E 03 1.00E 03 3.41E 02 1.13E 02 1.01E 02 8.26E 03 1.99E 02 1.78E 04 1.85E 02 2.51E 03 EST ERS (5 ) 1049 ( b,c,e ) 978 ethyl butanoate 4.09E 02 1.44E 02 2.48E 02 5.65E 04 2.92E 02 5.78E 03 1246 ( b,c,d,e ) 1155 ethyl hexanoate 4.76E 02 1.04E 02 1448 ( b,c,e ) 1344 ethyl octanoate 2.59E 02 2.19E 03 1.62E 02 1.98E 03a 8.19E 03 1.80E 03b 1.01E 02 1.03E 03ba 7.97E 03 2.45E 04b 1485 ( c ) 1383 octyl acetate 2.16E 02 1.17E 03 4.94E 02 1.31E 02 2.32E 02 3.92E 03 3.06E 02 4.11E 03 2.44E 02 9.49E 05 1706 ( f ) 1626 neryl acetate 1.46E 02 3.55E 03a 2.83E 03 1.98E 03b 9.66E 03 1.30E 03ba TERPENES (11 ) 1025 ( a,c,e ) 964 pinene 4.31E 02 9.74E 03 2.23E 01 2.63E 02 1.92E 01 8.29E 03 1.73E 01 1.22E 02 1.89E 01 2.02E 02 1188 ( a,c,d,e ) 1139 L imonene 1.04E+01 1.45E+00 2.87E+01 7.40E+00 1.92E+01 3.65 E+00 1.77E+01 7.23E 01 1.88E+01 2.99E 01 1244 ( e ) 1195 p cymene 3.54E 03 2.34E 03 5.50E 03 1.92E 03

PAGE 144

144 Table D 1 Continued Fresh squeezed Deaerated Control Oxygen saturated Air headspace LRI ( ref ) LRI (DBWax) compound 1290 ( c, d,e ) 1204 terpinolene 8.38E 03 1.03E 03 5.08E 02 1.49E 02 1.34E 02 5.37E 03 2.82E 02 2.44E 04 1624 ( a ) 1501 elemene 2.24E 02 1.16E 03 2.06E 02 6.88E 03 8.14E 03 9.24E 04 1.11E 02 4.60E 04 9.73E 03 7.46E 04 1565 ( e ) 1517 caryophyllene 3.56E 02 4.58E 03 1731 ( d,e ) 1636 valencene 7.59E 01 5.70E 03 7.43E 01 2.11E 01a 5.68E 03 3.43E 03b 4. 27E 01 3.14E 03ba 3.41E 01a 1.69E 02 1137 ( a,e ) 1142 sabinene 5.03E 02 3.63E 03 1.11E 01 9.69E 04 1148 ( a,b,d,e ) 1082 myrcene 1.36E 01 1.32E 01 1.79E+00 4.78E 01 1.02E+00 8.83E 02 1.20E+00 7.50E 02 1.26E+00 1.01E 01 1248 ( a,e ) 1 169 ocimene 5.66E 02 6.33E 03 9.38E 02 2.27E 02 1689 ( e ) 1642 selinene 5.44E 02 1.51E 03 3.24E 02 8.84E 03 1.72E 02 8.74E 03 1.94E 02 7.29E 04 1.92E 02 7.07E 03 ALCOHOL TERPENES (5) 1560 ( a,c,d ) 1449 linalool 2.87E 02 3.48E 0 4 2.95E 01 7.79E 02 1.75E 01 1.47E 02 1.65E 01 9.03E 03 1.91E 01 2.06E 03 1572 ( e ) 1521 4 terpineol 1.10E 01 2.28E 02 7.20E 02 4.28E 05 7.02E 02 7.64E 04 8.48E 02 7.24E 03

PAGE 145

145 Table D 1. Continued Fresh squeezed Deaerated Control Oxygen sa turated Air headspace LRI ( ref ) LRI (DBWax) compound 1616 ( d ) 1543 b terpineol 1661 ( d ) 1608 terpineol 1.10E 02 5.45E 03 1777 ( a ) 1655 b citronellol 9.35E 03 1.76E 03 1.73E 02 4.95E 03 1.37E 02 5.62E 03 1.08E 02 1.09E 03 1.46E 02 3.52E 03 ALCOHOLS (7 ) 949 ( a,e ) 887 ethanol 9.05E 02 3.31E 02 2.33E 01 2.62E 02 2.05E 01 2.28E 02 1.38E 01 1.28E 02 1.63E 01 3.70E 03 1364 ( a ) 1265 1 hexanol 6.28E 03 1.01E 03 1351 ( d ) 1298 3 hexen 1 ol 5. 46E 03 1.44E 03 4.89E 03 2.80E 04 1534 ( e ) 1458 1 octanol 1.48E 02 3.30E 03 1.48E 01 3.29E 02 7.94E 02 7.12E 03 7.88E 02 4.72E 03 8.43E 02 7.56E 04 1640 ( e ) 1558 1 nonanol 2.29E 02 4.58E 03 1.14E 02 1.21E 03 1.30E 02 6.13E 04 1.21 E 02 3.02E 04 1745 ( e ) 1654 1 decanol 5.98E 02 1.18E 02a 1.79E 02 7.53E 03b 3.62E 02 8.67E 04ba 3.57E 02 6.08E 03ba 1818 ( a,d ) 1736 nerol 7.93E 03 5.28E 04 1.90E 02 3.97E 03 9.35E 03 5.63E 04 1.10E 02 4.78E 04 9.25E 03 7.74E 04 a (Jabalpurwala 2009) ; b (Arena and others 2006) ; c (Berlinet and others 2007) ; d (Berlinet and oth ers 2006) ; e (Brat and others 2003) ; f (Rouseff 2008) ; g (Mottram 2010) 1 Peak not detected

PAGE 146

146 Table D 2 Volatile compounds identified in NFC orange juice at different DO content at day 0.5 of storage at 40 C. Values are average and standard error of normalized peak area using 4 heptadecanone as internal standard. =0.05). LRI ( ref ) LRI (DBWax) C ompound Deaerated Control Oxygen saturated A ir headspace ALDEHYDES (9 ) 698 ( f ) 722 A cetaldehyde 1.33E 01 4.25E 02 1.91E 02 6.40E 03 2.13E 01 8.92E 02 1.35E 01 2.91E 02 1297 ( c ) 1217 O ctanal 6.13E 02 6.93E 03 1.43E 01 7.25E 02 8.08E 02 2.25E 02 7.60E 02 5.22E 02 1411 ( b,c ) 1317 N onanal 2.19E 02 2.45E 03 5.57E 02 1.23E 02 2.53E 02 4.78E 03 2.80E 02 1.00E 02 1433 ( d ) 1396 F urfural 1501 ( b,c ) 1415 D ecanal 1.59E 01 1.84E 02 3.46E 01 8.44E 02 1.61E 01 2.98E 02 1.88E 01 6.03E 02 1630 ( g ) 1585 N eral 6.10E 03 9.32E 04 1.21E 02 4.61E 03 7.21E 03 1.02E 03 6.99E 03 1.05E 03 1728 ( f ) 1619 D odecanal 2.14E 02 2.14E 03 3.70E 02 1.24E 02 1.81E 02 4.24E 03 1.84E 02 9.22E 03 1766 ( a ) 1651 G eranial 2.03E 02 4.53E 03 1.58E 02 3.68E 03 1835 ( f ) 1713 P erillaldehyde 1.10E 02 3.00E 03 3.40E 02 8.07E 03 1.39E 02 3.35E 03 1.60E 02 1.50E 03 ESTERS (6) 1049 ( b,c,e ) 978 ethyl butanoate 2.60E 02 8.11E 03 2.35E 02 9.19E 03

PAGE 147

147 Table D 2. Continued LRI ( ref ) LRI (DBWax) C ompound Deaerated Control Oxygen saturated A ir headspace 1448 ( b,c,e ) 1344 ethyl octanoate 8.12E 03 7.24E 04 1.21 E 02 3.17E 03 8.89E 03 1.22E 03 6.76E 03 2.44E 03 1485 ( c ) 1383 octyl acetate 2.34E 02 4.27E 03 2.25E 02 7.19E 03 2.71E 02 6.23E 03 1.64E 02 5.22E 03 1706 ( f ) 1626 neryl acetate 6.50E 03 6.93E 04 7.84E 03 1.21E 03 9.04E 03 1.21E 03 TE RPENES (12) 1025 ( a,c,e ) 964 pinene 1.17E 01 2.96E 02 7.06E 02 2.38E 02 1.39E 01 3.86E 02 1.52E 01 6.93E 02 1244 ( e ) 1195 cymene 2.99E 03 4.25E 04 1290 ( c,d,e ) 1204 terpinolene 3.31E 02 7.34E 03 8.75E 03 3.80E 03 2.42E 02 4. 75E 03 1624 ( a ) 1501 elemene 6.59E 03 1.71E 03 3.73E 03 1.70E 03 9.22E 03 2.03E 03 6.19E 03 2.36E 03 1731 ( d,e ) 1636 Valencene 3.46E 01 2.37E 02 4.47E 01 1.75E 01 4.68E 01 1.20E 01 2.58E 01 9.29E 02 1137 ( a,e ) 1142 Sabinene 1.05E 01 2.73E 02 1.04E 01 2.17E 02 1148 ( a,b,d,e ) 1082 myrcene 8.66E 01 2.24E 01 1.36E 01 8.75E 02 2.35E 01 2.27E 01 7.83E 01 3.87E 01 1248 ( a,e ) 1169 ocimene 3.95E 02 1.32E 02 1689 ( e ) 1642 selinene 2.60E 02 2.39E 03 1.10E 02 7.36E 03 3.08E 02 8.78E 03 1.38E 02 7.76E 03

PAGE 148

148 Table D 2. Continued LRI ( ref ) LRI (DBWax) C ompound Deaerated Control Oxygen saturated A ir headspace ALCOHOL TERPENES (5) 1560 ( a,c,d ) 1449 Linalool 1.25E 01 1.89E 02 5.04E 01 1.66E 01 1.60E 01 4. 54E 02 1.52E 01 3.65E 02 1572 ( e ) 1521 4 terpineol 4.64E 02 7.50E 03 1.33E 01 4.10E 02 5.88E 02 1.44E 02 6.02E 02 1.25E 02 1616 ( d ) 1543 terpineol 6.01E 03 2.05E 03 1661 ( d ) 1608 terpineol 1777 (a,c) 1655 citronell ol 1.24E 02 9.14E 04 1.63E 02 4.96E 03 9.50E 03 2.11E 03 1.17E 02 3.83E 03 ALCOHOLS (8) 949 ( a,e ) 887 Ethanol 1.44E 01 3.12E 02 2.85E 01 9.59E 02 1.66E 01 5.98E 02 1.36E 01 5.42E 02 1364 (a) 1265 1 hexanol 2.69E 03 8.68E 04 6.98E 03 2.5 1E 03 3.59E 03 1.08E 03 0.00E+00 0.00E+00 1351 ( d ) 1298 3 hexen 1 ol 5.44E 03 2.99E 03 5.25E 03 1.84E 03 3.06E 03 7.58E 04 1534 ( e ) 1458 1 octanol 6.01E 02 1.26E 02 2.56E 01 5.37E 02 6.98E 02 1.85E 02 6.87E 02 1.61E 02 1640 ( e ) 1558 1 nonanol 1.67E 02 1.76E 03 2.61E 02 9.24E 03 1.08E 02 2.20E 03 1.08E 02 1.83E 03 a (Jabalpurwala 2009) ; b (Arena and ot hers 2006) ; c (Berlinet and others 2007) ; d (Berlinet and others 2006) ; e (Brat and others 2003) ; f (Rouseff 2008) ; g (Mottram 2010) 1 Peak not detected

PAGE 149

149 Table D 3 Volatile compounds identified in NFC orange juice at different DO content at day 6 of storage at 40 C. Values are average and standard error of normalized peak area using 4 heptadecanone as internal standard. Values =0.05) Deaerated Control Oxygen saturated Air headspace LRI ref LRI (DBWax) co mpound ALDEHYDES (10) 698 f 722 acetaldehyde 9.49E 02 2.74E 02 1.84E 01 1.02E 02 3.00E 01 2.71E 01 2.04E 01 3.77E 02 1297 c 1217 octanal 1.32E 02 1.14E 02b 6.78E 02 8.37E 04a 2.21E 02 1.16E 03b 3.79E 02 2.28E 03b 1411 b, c 1317 non anal 7.46E 03 4.60E 03 1.97E 02 1.37E 03 1.22E 02 1.03E 04 1.33E 02 6.40E 05 1433 d 1396 furfural 2.12E 03 1.01E 03 3.28E 03 5.08E 05 4.56E 03 1.28E 03 2.37E 03 1.73E 04 1501 b, c 1415 decanal 4.78E 02 1.86E 02b 1.32E 01 2.40E 02a 6.49E 02 3.85E 03ba 1.08E 01 5.22E 03a 1630 g 1585 neral 8.96E 04 5.59E 05b 1.12E 02 2.71E 03a 1728 f 1619 dodecanal 9.34E 03 5.21E 04 8.55E 02 6.29E 02 1.03E 02 1.03E 02 1766 a 1651 geranial 9.33E 04 5.28E 04 1835 f 1713 perillaldehy de 8.79E 03 4.08E 03 1.30E 02 2.56E 03 1.19E 02 5.91E 04 6.75E 03 9.78E 05 ESTERS (6) 1049 b,c,e 978 ethyl butanoate 8.70E 03 1.31E 03 2.10E 02 3.84E 03 1.91E 02 8.29E 03 1448 b,c,e 1344 ethyl octanoate 4.06E 03 6.83E 04b 8.26E 03 1.84 E 04a 5.73E 03 1.47E 04b 7.96E 03 2.16E 03a

PAGE 150

150 Table D 3. Continued Deaerated Control Oxygen saturated Air headspace LRI ( ref ) LRI (DBWax) compound 1485 c 1383 octyl acetate 1.40E 02 2.01E 03 2.39E 02 3.22E 03 1.87E 02 5.24E 03 2.14E 02 3.05E 03 1706 f 1626 neryl acetate 5.31E 03 5.00E 04 6.75E 03 4.21E 03 5.30E 03 6.58E 05 3.63E 03 3.96E 04 TERPENES (12) 1025 a,c,e 964 a pinene 1.45E 02 1.22E 03b 1.69E 01 3.30E 02a 2.27E 02 1.76E 03b 1.92E 01 2.38E 02a 1188 a,c,d,e 1 139 limonene 3.67E+00 7.07E 01b 2.02E+01 2.29E+00a 5.72E+00 7.66E 02b 2.08E+01 1.31E+00a 1148 a,b,d,e 1082 myrcene 5.39E 03 2.44E 03b 3.75E+00 1.52E+00a 8.22E 01 2.98E 01ba 1.89E+00 6.55E 01a 1244 e 1195 p cymene 2.10E 03 4.02E 04 1290 c,d,e 1204 terpinolene 7.46E 03 3.37E 03b 3.79E 02 9.48E 03a 1.19E 02 2.92E 05ba 2.57E 02 1.17E 02a 1624 a 1501 elemene 1.92E 03 2.31E 04 3.78E 03 1.68E 03 6.64E 03 2.43E 03 1731 d,e 1636 valencene 1.53E 01 1.08E 02b 3.92E 01 8.44E 02a 3.06E 01 2.29E 02ba 4.11E 01 1.20E 02a 1137 a,e 1142 sabinene 2.09E 02 1.12E 02 4.71E 03 7.4 7E 04

PAGE 151

151 Table D 3. Continued Deaerated Control Oxygen saturated Air headspace LRI ( ref ) LRI (DBWax) compound 1248 a,e 1169 b ocimene 1.31E 03 3.74E 04 1689 e 1642 a selinene 1.10E 02 2.62E 03 3.24E 02 1.13E 02 2.56E 02 3.66 E 03 1.63E 02 1.24E 02 ALCOHOL TERPENES (5) 1560 a,c,d 1449 linalool 1.68E 01 1.22E 02 1.51E 01 4.33E 03 2.03E 01 6.91E 03 1.22E 01 9.71E 03 1572 e 1521 4 terpineol 4.20E 02 9.25E 03 6.21E 02 1.43E 02 5.71E 02 2.36E 03 5.47E 02 4.93E 04 1616 d 1543 b terpineol 2.08E 02 1.05E 02 1.55E 02 6.47E 03 4.10E 02 1.06E 02 1.22E 02 3.55E 03 1661 d 1608 a terpineol 1.09E 01 4.94E 02 1.03E 01 4.51E 02 1.36E 01 1.35E 01 9.08E 02 2.54E 02 1777 a 1655 b citronellol 1.13E 02 7.10E 05 1.57E 02 2.51E 03 1.52E 02 7.97E 03 ALCOHOLS (8) 949a,e 887 ethanol 1.59E 01 1.22E 02 2.07E 01 2.12E 02 1.74E 01 4.18E 02 1.62E 01 1.65E 02 1364a 1265 1 hexanol 7.23E 03 2.27E 03a 3.31E 03 1.00E 03a 2.76E 03 3.08E 04a 1351d 1298 3 hexen 1 ol 5.34E 03 3.04E 04a 3.54E 03 1.65E 04b 3.01E 03 3.47E 05b 1534e 1458 1 octanol 1.07E 01 7.17E 04 8.51E 02 9.80E 03 1.07E 01 4.85E 04 6.38E 02 3.55E 03 1640e 1558 1 nonanol 1.64E 02 1.83E 03 1.22E 02 1.18E 03 1.43E 02 4.01E 0 4 9.80E 03 6.36E 05

PAGE 152

152 Table D 3. Continued Deaerated Control Oxygen saturated Air headspace LRI (ref) LRI (DBWax) compound 1745 e 1654 1 decanol 3.61E 02 6.37E 03 3.24E 02 4.24E 03 3.14E 02 3.15E 03 5.46E 03 3.24E 03 1818 a,d 173 6 nerol 1.50E 02 2.90E 03 1.54E 02 5.12E 03 2.35E 02 2.72E 03 1.39E 02 3.21E 03 a (Jabalpurwala 2009) ; b (Arena and o thers 2006) ; c (Berlinet and others 2007) ; d (Berlinet and others 2006) ; e (Brat and others 2003) ; f (Rouseff 2008) ; g (Mottram 2010) 1 Peak not detected

PAGE 153

1 53 Figure D 1. Score plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C normalized with inter nal standard.

PAGE 154

154 Figure D 2. Loading plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C normalized with internal standard.

PAGE 155

155 NORMALIZED WITH DAY O OF STORAGE Table D 4 Volatile compounds identified in NFC orange juice at d ifferent DO content at day 0 of storage at 40 C. Values are average and standard error of normalized peak area with day 0. Values followed by letters are significantly =0.05) LRI Compound Deaerated Control Oxyge n saturated Air headspace 716 acetaldehyde 100 0 100 0 100 0 100 0 891 ethanol 1000 1000 1000 1000 968 a pinene 1000 1000 1000 1000 984 ethylbutanoate 1000 1000 1000 1000 1100 b myrcene 1000 1000 1000 1000 1148 limonene 1000 1000 6 7 33 1000 1142 sabinene 1000 1000 1000 1000 1179 b ocimene 1000 1000 1000 1000 1204 o cymene 1000 1000 1000 1000 1213 te rpinolene 1000 1000 1000 1000 1219 octanal 1000 1000 1000 1000 1269 1 hexanol 1000 1000 1000 1000 1301 3 hexen 1 ol, (z) 1000 1000 1000 1000 1318 nonanal 1000 1000 1000 1000 1348 ethyl octanoate 1000 1000 1000 1000 1387 octyl acetate 1000 1000 1000 1000 Not detected

PAGE 156

156 Table D 5 Volatile compounds identified in NFC orange juice at different DO conten t at day 0.5 of storage at 40 C. Values are average and standard error of normalized peak area with day 0. Values followed by letters are =0.05) LRI compound Deaerated Control Oxyge saturated Air headspace 716 acetaldehyde 20263 5040 13648 9142 891 ethanol 6618 3512 211107 6437 968 a pinene 8614 8221 984 ethylbutanoate 6014 138 129129 6630 1100 b myrcene 9627 5218 7526 1148 limonene 6329 6923 6635 4333 1142 sabinene 7931 246109 1000 1179 b ocimene 14854 17477 1000 1204 o cymene 5218 936 756 4521 1213 terpinolene 6821 12414 11920 7133 1219 octanal 5316 987 757 4923 1269 1 hexanol 7229 22836 8733 6524 1301 3 hexen 1 ol, (z) 7535 27322 8327 6624 1 318 nonanal 6321 15325 8817 6019 1348 ethyl octanoate 329 307129 462 243 1387 octyl acetate 9141 5410 Not detected

PAGE 157

157 Table D 6 Volatile compounds identified in NFC orange juice at different DO content at day 6 of storage at 40 C. Valu es are average and standard error of normalized peak area with day 0. Values followed by letters are significantly =0.05) LRI C ompound Deaerated Control Oxyge n saturated Air headspace 716 acetaldehyde 24192 103 2 9736 15573 891 ethanol 148 9313 7038 649 968 a pinene 1000 899 7227 984 ethylbutanoate 12818 7434 8417 1100 b myrcene 3420 370 178 368 1000 1148 limonene 1811 471 117 193 1142 sabinene 1000 1000 1000 1179 b ocimene 1082 1000 100 0 1204 o cymene 188 395 366 204 1213 terpinolene 3510 10216 9021 8028 1219 octanal 166 459 339 266 1269 1 hexanol 627 917 8222 4313 1301 3 hexen 1 ol, (z) 7910 10613 10619 5511 1318 nonanal 413 9112 7715 518 1348 ethyl octano ate 41 6413 1387 octyl acetate 53 Not detected

PAGE 158

158 Figure D 3 Score plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C normalized with day 0

PAGE 159

159 Figure D 4. Loading plot of PCA applied to volatile compounds from NFC orange juice stored at 40 C normalized with day 0.

PAGE 160

160 APPENDIX E POINTS TO CONTROL FO R REDUCING VARIABILI TY DURING SAMPLE PREPARATION Points list 1. To harvest fruit at the same maturity stage, size and to extract the juice at the same time for all the r eplicates. 2. Set the gas flow using water and before pouring the orange juice into the bottles used for bubbling. 3. Set flow and a constant temperature in microthermics before starting pasteurization. 4. To avoid reentry of air to orange juice after pasteurizati on by filling the vials under a nitrogen atmosphere or using a mechanism that avoid the contact of

PAGE 161

161 air with the juice. To fill vials with the minimum manipulation of the juice, using a laminar hood and sterile vials, gloves and any equipment in contact wit h the orange juice. Hot filling may be used to have a more stable orange juice at temperatures between 13 and 30.5 C For volatile analysis juice should be poured directly in the vials that will be used for the analysis (glass vials with septum) to avoid disturbing the sample. 5. To set a constant temperature the day before preparing the juice and to monitor temperature constantly during storage, especially if a power outage occurs. 6. To keep juice at the same storage temperature after sampling and until freezi ng for further analysis. 7. Juice has to be transferred and manipulated under a nitrogen atmosphere to avoid entry of air. To measure DO immediately after sampling and to freeze juices to keep for further analysis of AA, DHA and color. Posterior thawing of s amples should be as quick as possible avoiding temperature abuse of samples.

PAGE 162

162 LIST OF REFERENCES Acree T, Arn H. 2004. Flavornet. Available from http://www.flavornet.org/index.html A ccesed April 30 2009. Adams JB, Brown HM. 2007. Discoloration in raw and processed fruits and vegetables. Critical Reviews in Food Science and Nutrition 47:319 333. Ahrne LM, Oliveira FAR, Manso MC, Drumond MC, Oste R, Gekas V. 1997. Modelling of dissolv ed oxygen concentration during storage of packaged liquid foods. Journal of Food Engineering 34(2):213 224. Alaka OO, Aina JO, Falade KO. 2003. Effect of Storage conditions on the chemical attributes of Ogbomoso mango juice. European Food Research Technol ogy 218:79 82. Altieri C, Sinigaglia M, Corbo MR, Buonocore GG, Falcone P, Nobile MAD. 2004. Use of entrapped microorganisms as biological oxygen scavengers in food packaging applications. Lebensm Wiss Technol 37 (2004) 9 15 37:9 15. Andersson M, Anderss on T, Adlercreutz P, Nielsen T, Hornsten EG. 2002. Toward an enzyme based oxygen scavenging laminate. Influence of industrial lamination conditions of the performance of glucose oxidase. Biotechnology and Bioengineering 79(1):37 42. Arena E, Guarrera N, Campisi S, Asmundo CN. 2006. Comparison of odour active compounds detected by gas chromatography olfactometry between hand squeezed juices from different orange varieties. Food Chemistry 98(1):59 63. Arts F, Castaer M, Gil MI. 1998. Review: Enzymatic browning in minimally processed fruit and vegetables. Food Science and Technology International 4(6):377 389. Aymoto Hassimotto NM, Genovese MI, Lajolo FM. 2005. Antioxidant activity of dietary fruits,vegetables, and commercial frozen fruit pulps. J Agri c Food Chem 53(8):2928 2935. Baiano A, Marchitelli V, Tamagnone P, Del Nobile MA. 2004. Use of active packaging for increasing ascorbic acid retention in food beverages. J Food Sci 69(9):E502 E508. Barnes RH, Davis RC, Nicholson WB, inventors; MEDEX IN C, assignee. 1990 04 11 Oxygen sensor using phase shift. USA patent EP0363219. Bazemore R, Goodner K, Rouseff R. 1999. Volatiles from unpasteurized and excessively heated orange juice analyzed with solid phase microextraction and GC olfactometry. Journal of Food Science 64(5):800 803.

PAGE 163

163 Beltrn Gonzlez F, Prez Lpez AJ, Lpez Nicols JM, Carbonell Barrachina AA. 2008. Effect of packaging materials on color, vitamin C and sensory quality of refrigerated mandarin juice. Journal of Food Quality 31(5):596 611. Berlinet C, Brat P, Brillouet JM, Ducrue V. 2006. Ascorbic acid, aroma compounds and browning of orange juices related to PET packaging materials and pH J Sci Food Agric 86(13):2206 2212. Berlinet C, Guichard E, Fournier N, Ducruet V. 2007. Effect of p ulp reduction and pasteurization on the release of aroma compounds in industrial orange juice. Journal of Food Science 72(8):S535 S543. Biocatalysts. 2004. Data sheet catalase C641L Available from www.biocatalyst s.com Accesed 01 25 07. Braddock RJ. 1999. Handbook of citrus by products and processing technology. New York: John Wiley & Sons. Braddock RJ, Sadler G. 1989. Chemical changes in citrus juices during concentration processes. In: Jen JJ, editor. Quality factors of fruits and vegetables: Chemistry and technology American Chemical Society. p. 293 304. Bradley D, Min D. 1992. Singlet oxygen oxidation of foods. Crit Rev Food Sci Nut 31(3):211 236. Brat P, Rega B, Alter P, Reynes M, Brillouet J M. 2003. Di stribution of Volatile Compounds in the Pulp, Cloud, and Serum of Freshly Squeezed Orange Juice. J Agric Food Chem 51:3442 3447. Brody AL. 2005. Micro oxygen packaging retains quality in beer. Food Technology 59(4):69 71. Burdurlu HS, Koca N, Karadeniz F. 2006. Degradation of vitamin C in citrus juice concentrates during storage. Journal of Food Engineering 74:211 216. Burton SG. 2003. Laccases and phenol oxidases in organic synthesis a review. Current Organic Chemistry 7:1317 1331. Cmara M, Matal lana MC, Snchez Mata MC, Ayu RL, Labra E. 2003. Lycopene and hydroxymethylfurfural (HMF) evaluation in tomato products. In: Bieche B & Branthome X, editors. 8th International Symposium on the Processing Tomato Istanbul, Turkey: ISHS. p. 365 371. Cancal on PF. 2001. Routine analysis of ascorbic acid in citrus juice using capillary electrophoresis. Journal of AOAC International 84(3):987 991.

PAGE 164

164 Charles F, Sanchez J, Gontard N. 2006. Absorption kinetics of oxygen and carbon dioxide scavengers as part of activ e modified atmosphere packaging. Journal of Food Engineering 72(1):1 7. Choi MMF, Xiao D. 2000. Single standard calibration for an optical oxygen sensor based on luminescence quenching of a ruthenium complex. Analytica Chimica Acta 403:57 65. Chu CS, Lo YL. 2007. High performance fiber optic oxygen sensors based on fluorinated xerogels doped with Pt(II) complexes. Sensors and Actuators B Chemical 124(2):376 382. Cole CA, Genetelli EJ. 1970. Decarbonation and deaeration of water using selective hollow fib ers. Environmental Science & Technology 4(6):514 517. Cremer DR, Eichner K. 2000. Formation of volatile compounds during heating of spice paprika (Capsicum annuum) powder. J Agric Food Chem 48:2454 2460. Davey MW, Montagu MV, Dirk Inz MS, Kanellis A, Sm irnoff N, Benzie IJJ, Strain JJ, Favell D, Fletcher J. 2000. Plant L ascorbic acid: chemistry, function, metabolism, bioavailability and effects of processing. Journal of the Science of Food and Agriculture 80(7):825 860. Degenhardt OS, Waters B, Rebelo C ameirao A, Meyer A, Brunner H, Toltl NP. 2004. Comparison of the effectiveness of various deaeration techniques. Dissolution Technologies 11(1):6 11. Dhuique Mayer C, Tbatou M, Carail M, Caris Veyrat C, Dornier M, Amiot MJ. 2007. Thermal degradation of an tioxidant micronutrients in citrus juice: Kinetics and newly formed compounds. J Agric Food Chem 55:4209 4216. Eddy CW. 1936. Absorption rate of oxygen by orange juice. Industrial and engineering chemistry 28(4):480 483. FMC F. 2002. Procedures for analy sis of citrus products 4 ed. Lakeland, FL: FMC Technologies, Inc. Franck C, Lammertyn J, QT H, P V, Verlinden B, Nicolai B. 2007. Browning disorders in pear fruit. Postharvest Biology and Technology 43(1):1 13. Friedman M, Molnar Perl I. 1990. Inhibition of browning by sulfur amino acids. 1. Heated amino acid glucose systems. Journal of Agricultural and Food Chemistry 38(8):1642 1647. Gabelman A, Hwang S T. 1999. Hollow fiber membrane contactors. Journal of Membrane Science 159:61 106.

PAGE 165

165 Garca Torres R, P onagandla NR, Rouseff RL, Goodrich Schneider RM, Reyes De Corcuera JI. 2009. Effects of dissolved oxygen in fruit juices and methods of removal. Comprehensive Reviews in Food Science and Food Safety 8:409 423. Genovese MI, Da Silva Pinto M, De Souza Schmi dt Gonalves AE, Lajolo FM. 2008. Bioactive compounds and antioxidant capacity of exotic fruits and commercial frozen pulps from Brazil. Food Sci Tech Int 14(3):207 214. Gill CB, Menneer ID. 1997. Advances in gas control technology in the brewery. The B rewer Feb:76 84. Gmez PA, Geysen S, Verlinden BE, Arts F, Nicola BM. 2006. Modelling the effect of superatmospheric oxygen concentrations on in vitro mushroom PPO activity. J Sci Food Agric 86:2387 2394. Haleva Toledo E, Naim M, Zehavi U, Rouseff RL 1999. Effects of l cysteine and N acetyl l cysteine on 4 hydroxy 2,5 dimethyl 3(2H) furanone (furaneol), 5 (hydroxymethyl)furfural, and 5 methylfurfural formation and browning in buffer solutions containing either rhamnose or glucose and arginine. J Agri c Food Chem 47(10):4140 4145. Holden J. 2006. USDA National Nutrient Database for Standard Reference. USDA. Hsieh Y HP, Harris ND. 1993. Effect of sucrose on oxygen uptake of ascorbic acid in a closed aqueous system. J Agric Food Chem 41:259 262. Hughes DE. 1985. Irreversible reaction kinetics of the aerobic oxidation of ascorbic acid. Anal Chem 57(2):555 558. Jabalpurwala F. 2009. Grapefruit juice and citrus blossom volatiles. Food Science and Human Nutrition Gainesville: University of Florida. p. 117. Jabalpurwala F, Gurbuz O, Russell LR. 2010. Analysis of grapefruit sulphur volatiles using SPME and pulsed flame photometric detection. Food Chemistry 120:296 303. Jiang D, Chen X, Liu E, Huang J. 2002. Preparation and properties of sensing membrane for fiber optic oxygen sensor. Journal of Wuhan University of Technology Materials Science Edition 17(2):51 53. Johnson JR, Braddock RJ, Chen CS. 1995. Kinetics of ascorbic acid loss and nonenzymatic browning in orange juice serum: experimental rate constant s. J Food Sci 60(3):502 505. Jordan MJ, Goodner KL, Laencina J. 2003. Deaeration and pasteurization effects on the orange juice aromatic fraction. Lebensm Wiss Technol 36:391 396.

PAGE 166

166 Joslyn MA. 1961. Physiological and enzymological aspects of juice productio n. In: Tressler DK, editor. Fruit and vegetable juice processing technology Westport, Conn: Avi Pub Co. Kefford JF, Mckenzie HA, Thompson PCO. 1959. Effects of oxygen on quality and ascorbic acid retention in canned and frozen orange juices. J Sci Food A gric 10:51 63. Kennedy JF, Rivera ZS, Lloyd LL, Warner FP, Jumel K. 1992. L ascorbic acid stability in aseptically processed orange juice in TetraBrik cartons and the effect of oxygen. Food Chemistry 45:327 331. Kerry JP, O'Grady MN, Hogan SA. 2006. Past current and potential utilisation of active and intelligent packaging systems for meat and muscle based products: A review. Meat Science 74(1):113 130. Khan MMT, Martell AE. 1967. Metal ion and metal chelate catalyzed oxidation of ascorbic acid by molec ular oxygen. I. Cupric and ferric ion catalyzed oxidation. J A Chem Soc 89(16):4176 4185. Kothapalli A, Hayes K, Sadler G, Morgan M. 2007. Comparison of kinetic profile of free and immobilized glucose oxidase, immobilized on low density polyethylene usi ng UV polymerization. J Food Sci 72(9):C478 C482. Kutty V, Braddock RJ, Sadler G. 1994. Oxidation of d limonene in presence of low density polyethylene. J Food Sci 59(2):402 405. Lee HS, Chen CS. 1998. Rates of Vitamin C Loss and Discoloration in Clear O range Juice Concentrate during Storage at Temperatures of 4 24 C. J Agric Food Chem 46:4723 4727. Lee SK, Kader AA. 2000. Preharvest and postharvest factors influencing vitamin C content of horticultural crops. Postharvest Biology and Technology 20(3):20 7 220. Lewis ME. 2006a. National Field Manual. Chapter 6.2 Dissolved oxygen. Available from Accesed Lewis ME. 2006b. National field manual. Chapter 6.2 Dissolved oxygen. Available from http://water.usgs.gov/owq/FieldManual/Chapter6/6.2_contents.html Accesed 01 15 2008. Lewis VM, McKenzie HA. 1947. Amperometric Determination of Dissolved Oxygen in Orange Juice. Anal Chem 19(9):643 646.

PAGE 167

167 Lin SH, Agalloco J. 1979. Degradation k inetics of ascorbic acid. Process Biochemistry 14(9):22 26. Lopez Cervantes J, Sanchez Machado DI, Pastorelli S, Rijk R, Paseiro Losada P. 2003. Evaluating the migration of ingredients from active packaging and development of dedicated methods: a study of two iron based oxygen absorbers. Food Additives and Contaminants 20(3):291 299. Lpez Rubio A, Almenar E, Pilar Hernandez Muoz, Lagarn JM, Catal R, Gavara R. 2004. Overview of active polymer based packaging technologies for food applications. Food R eviews International 20(4):357 387. Mack TE, Heldman DR, Singh RP. 1976. Kinetics of oxygen uptake in liquid foods. J Food Sci 41:309 312. Mahmud H, Kumar A, Narbaitz RM, Matsuura T. 1998. Membrane air stripping: A process for removal of organics from aqueous solutions. Separation Science and Technology 33(14):2241 2255. Manso MC, Oliveira FAR, Oliveira JC, Fras JM. 2001. Modelling ascorbic acid thermal degradation and browning in orange juice under aerobic conditions. Int J Food Sci Technol 36(3):3 03 312. Marcotte M, Stewart B, Fustier P. 1998. Abused Thermal Treatment Impact on Degradation Products of Chilled Pasteurized Orange Juice. J Agric Food Chem 46:1991 1996. Marshall M, Nagy S, Russell LR. 1986. Factors impacting on the quality of stored citrus fruit beverages. In: Charalambous G, editor. The shelf life of foods and beverages Amsterdam, Netherlands: Elsevier. p. 237 254. Martinez MV, Whitaker JR. 1995. The biochemistry and control of enzymatic browning. Trends in Food Science & Technolog y 6:195 200. Min DB, Boff JM. 2002. Chemistry and reaction of singlet oxygen in foods. Comprehensive Reviews in Food Science and Food Safety 1(2):58 72. Mistry B, Min DB. 1992. Reduction of dissolved oxygen in model salad dressing by glucose oxidase catalase dependent on pH and temperature. J Food Sci 57(1):196 199. Molnar Perl I, Friedman M. 1990. Inhibiton of browning by sulfur amino acids. 2. Fruit juices and protein containing foods. Journal of Agricultural and Food Chemistry 38(8):1648 1651.

PAGE 168

168 M oshonas MG, Shaw PE. 1994. Quantitative determination of 46 volatile constituents in fresh, unpasteurized orange juices using dynamic headspace gas chromatography. J Agric Food Chem 42(7):1525 1528. Mottram R. 2010. LRI and odour database. Available from http://www.odour.org.uk/index.html Accesed June 27 2010. Naim M, Zehavi U, Zuker I, Rouseff R, Nagy S. 1994. Effect of L cysteine and N acetyl L cysteine on off flavor formation in stored citrus products. In: Mussinan CJ & Keelan ME, editors. Sulfur compounds in foods Washington, DC: American Chemical Society. p. 80 89. Nelson KA. 2005. Photochemically induced flavor changes in orange juice exposed to light in glass and polyethylene terephthalate at 4C. Food Science and Human Nutrition Gainesville: University of Florida. p. 82. Ogurtsov VI, Papkovsky DB. 1998. Selection of modulation frequency of excitation for luminescence lifetime based oxygen sensors. Sensors and Actuators B Chemical 51(1 3):377 381. Ozdemir M, Floros JD. 2004. Active food packaging technologies. Critical Reviews in Food Science and Nutrition 44(3):185 193. Parpinello GP, Chinnici F, Versari A, Riponi C. 2002. Preliminary study on glucose oxidase catalase enzyme system to control the browning of apple and pear purees. Lebensm Wiss Technol 35(3):239 243. Patkai G, Kormendy I, Kormendy Domian A. 2002. Vitamin C decomposition kinetics in solutions, modelling citrus juices. Acta Alimentaria 21(2):125 147. Perez Cacho PR, Mahattanat awee K, Smoot JM, Rouseff R. 2007. Identification of sulfur volatiles in canned orange juices lacking orange flavor. Journal of Agricultural and Food Chemistry 55(14):5761 5767. Perez Cacho PR, Rouseff R. 2008a. Processing and storage effects on orange ju ice aroma: A review. J Agric Food Chem 56(21):9785 9796. Perez Cacho PR, Rouseff RL. 2008b. Fresh squeezed orange juice odor: A review. Crit Rev Food Sci Nut 48(7):681 695. Perez AG, Sanz C. 2008. Formation of fruit flavor. In: Brckner B & Wyllie SG, ed itors. Fruit and Vegetable Flavour, recent advances and future prospects Cambridge, England: Woodhead Publishing Ltd. p. 41 70.

PAGE 169

169 Pickering GJ, Heatherbell DA, Barnes MF. 1998. Optimising glucose conversion in the production of reduced alcohol wine using g lucose oxidase. Food Research International 31(10):685 692. Polypore. 2007. Liqui Cel membrane contactor: product tour. Available from Accesed Ponagandla NR. 2010. Kinetics of dissolved oxygen consumption and deoxygenation of pineapple juice and model s olutions using a thin film enzyme reactor Food Science and Human Nutrition Gainesville: University of Florida. p. 68. Rahrooh G, Engel RA, inventors. 04 09 1991. Deaeration process for the manufacture of fruit juices. US patent 5006354. Rassis D, Saguy IS. 1995. Oxygen effect on nonenzymatic browning and vitamin C in commercial citrus juices and concentrate. LWT Food Science and Technology 28(3):285 290. Rega B, Fournier N, Guichard E. 2003. Solid phase microextraction (SPME) of orange juice flavor: o dor representativeness by direct gas chromatography olfactometry (D GC O). J Agric Food Chem 51:7092 7099. Reineccius G. 2006. Flavor chemistry and technology 2th ed. Boca Raton, Fl: Taylor & Francis Group. Ringblom U. 2004. The orange book 2 ed. Lund, Sweden: Tetra Pak. Robertson GL, Samaniego CML. 1986. Effect of initial dissolved oxygen levels on the degradation of ascorbic acid and the browning of lemon juice during storage. J Food Sci 51(1):184 192. Rodrigues ET, Han JH. 2003. Intelligent packa ging. In: Dennis H, editor. Encyclopedia of Agricultural, Food and Biological Engineering New York: Marcel Dekker. p. 528 535. Rojas AM, Gerschenson LN. 2001. Ascorbic acid destruction in aqueous model systems: an additional discussion. Journal of the Sc ience of Food and Agriculture 81:1433 1439. Ros Chumillas M, Belissario Y, Iguaz A, Lopez A. 2007. Quality and shelf life of orange juice aseptically packaged in PET bottles. Journal of Food Engineering 79(1):234 242. Rouseff R. 2008. Citrus flavor datab ase. Available from http://www.crec.ifas.ufl.edu/crec_websites/Rouseff/index.htm Accesed April 30 2009.

PAGE 170

170 Rouseff RL, Nagy S, Naim M, Zahavi U. 1992. Off flavor development in citr us juice products. In: Charalambous G, editor. Off flavors in foods and beverages Amsterdam: Elsevier. p. 211 227. Sadler GD, Parish ME, Wicker L. 1992. Microbial, enzymatic and chemical changes during storage of fresh and processed orange juice. J Food Sci 57(5):1187 1191. Sadler GD, Roberts J, Cornell J. 1988. Determination of oxygen solubility in liquid foods using a dissolved oxygen electrode. J Food Sci 53(5):1493 1496. Sagi I, Mannheim H. 1990. The effect of enzymatic oxygen removal on quality of u npasteurized and pasteurized orange juice. Journal of Food Processing and Preservation 14:253 266. Sakai Y, Watanabe H, Takai R, Hasegawa T. 1987. A kinetic model for oxidation of ascorbic acid and beta carotene. Journal of Food Processing and Preservatio n 11:197 207. Sang Kyung L, Icharo O. 1997. Optical sensor for oxygen using a porphyrin doped Sol gel glass. The Analyst 122:81 84. Schomburg D. 2007. Brenda. The Comprehensive Enzyme Information System Institut fuer Biochemie, Universitaet zu Kln, Zlp icher Str. 47, 50674 Kln, Germany. Serpen A, Gkmen V. 2007. Reversible degradation kinetics of ascorbic acid under reducing and oxidizing conditions. Food Chemistry 104:721 725. Shaw PE, Nagy S, Russell LR. 1993. The shelf life of citrus products. In: C haralambous G, editor. Shelf life studies of foods and beverages: Chemical, biological, physical and nutritional aspects Amsterdam: Elsevier. p. 755 778. Shinoda Y, Komura H, Homma S, Murata M. 2005. Browning of model orange juice solutions: factors affe cting the formation of decomposition products. Biosci Biotechnol Biochem 69(11):2129 2137. Singh RP, Heldman DR, Kirk JR. 1976. Kinetics of quality degradation: ascorbic acid oxidation in infant formula during storage. J Food Sci 41:304 308. Soares NFF, Hotchkiss JH. 1999. Comparative Effects of De aeration and Package Permeability on Ascorbic Acid Loss in Refrigerated Orange Juice. Packag Technol Sci 12:111 118. Solis JA, Rodgers BD. 2001. Factors affecting the performance of new oxgyen scavenging polym er for packaging applications. Journal of Plastic Film & Sheeting 17:339 349. Solomon O, Svanberg U, Sahlstrm A. 1995. Effect of oxygen and fluorescent light on the quality of orange juice during storage at 8 C. Food Chem 53:363 368.

PAGE 171

171 Tang Y, Tehan EC, T ao Z, Bright FV. 2003. Sol gel derived sensor materials that yield linear calibration plots, high sensitivity, and long term stability. Anal Chem 75:2407 2413. Tewari G, Jayas DS, Jeremiah LE, Holley RA. 2002. Absorption kinetics of oxygen scavengers. Int ernational Journal of Food Science and Technology 37:209 217. Trammell DJ, Dalsis DE, Malone CT. 1986. Effect of oxygen on taste, ascorbic acid loss and browning for HTST pasteurized, single strength orange juice. J Food Sci 51(4):1021 1023. USDA. 2007. Fruit and tree nuts situation and outlook yearbook Washington, DC, USA: Economic Research Service, USDA. p. 200. USDA. 2008a. Fruit and tree nuts situation and outlook yearbook Washington, DC, USA: Economic Research Service, USDA. p. 200. USDA. 20 08b. Vegetables and melons situation and outlook yearbook. VGS 2008 Washington, DC, USA: Economic Research Service, USDA. Varnado CL, Hertwig KM, Thomas R, Roberts JK, Goodwin DC. 2004. Properties of a novel periplasmic catalase peroxidase from Escheri chia coli O157:H7. Arch Biochem Biophys 421:166 174. Vermeiren L, Devlieghere F, Beest Mv, Kruijf Nd, Debevere J. 1999. Developments in the active packaging of foods. Trends in Food Science & Technology 10:77 86. Weiss RF. 1970. The solubility of nitrog en, oxygen and argon in water and seawater. Deep Sea Research 17:721 735. Yeh TS, Chu CS, Lo YL. 2006. Highly sensitive optical fiber oxygen sensor using Pt(II) complex embedded in sol gel matrices. Sensors and Actuators B Chemical 119(2):701 707. Zerdin K, Rooney ML, Vermue J. 2003. The vitamin C content of orange juice packed in an oxygen scavenger material. Food Chemistry 82:387 395.

PAGE 172

172 BIOGRAPHICAL SKETCH Rosalia was born and raised in Puebla, M xico. She completed her B.S in food engineering and he r M.S. in food science at University of Amricas Puebla, Mxico Rosalia was awarded a scholarship from CONACyT, Mexico to pursue a Ph.D. in food science in the University of Florida. Rosalia served as the Citrus Products Division IFT student representativ e in 2008 and as team member of the Danisco product development team in 2009. including new processing technologies, quality, safety and packaging. She is interested in the chemic al changes that occur during postharvest storage and processing that deteriorate acceptability of fresh and processed food products. She specialized in the effect of oxygen in orange juice nutritional content (vitamin C), color and aroma. Upon completion o f the Ph.D. program Rosalia intends to seek a postdoctoral position in order to gain research and teaching experience to eventually become a food science professor.