Citation
Continuous High Pressure Carbon Dioxide Processing of Watermelon Juice

Material Information

Title:
Continuous High Pressure Carbon Dioxide Processing of Watermelon Juice
Creator:
LECKY, MATTHEW
Copyright Date:
2008

Subjects

Subjects / Keywords:
Carbon dioxide ( jstor )
Flavors ( jstor )
Food ( jstor )
High pressure ( jstor )
Juices ( jstor )
Microorganisms ( jstor )
pH ( jstor )
Water temperature ( jstor )
Watermelons ( jstor )
Yeasts ( jstor )
Miami metropolitan area ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Matthew Lecky. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2006
Resource Identifier:
495636957 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

CONTINUOUS HIGH PRESSURE CA RBON DIOXIDE PROCESSING OF WATERMELON JUICE By MATTHEW LECKY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Matthew Lecky

PAGE 3

This document is dedicated to my pa rents, sister, wife and children.

PAGE 4

ACKNOWLEDGMENTS I would like to thank my major advisor, Dr. Murat Balaban, for his continuing support and friendship. Without his support and encouragement this study could not have been completed. I would also like to thank my committee members, Dr. Maurice Marshall, Dr. Amy Simonne, and Dr. Allen Wysocki, for all their guidance throughout my research. I would also like to thank my parents for their unwavering support and love. I also want to acknowledge their constant financial support during both my undergraduate degree and parts of my graduate degree. Lastly I want to thank my wife and children. Without their smiles, kisses and hugs everyday when I came home I would have never had the courage and motivation to complete my graduate research. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ...........................................................................................................viii LIST OF FIGURES .............................................................................................................x ABSTRACT ......................................................................................................................xii CHAPTER 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................3 Origin of Watermelon (Citrullus lanatus)....................................................................3 Watermelon Production................................................................................................3 USA Production.....................................................................................................3 Watermelon Composition.............................................................................................5 Watermelon Grades...............................................................................................5 Maturity and Ripeness...........................................................................................5 Composition..........................................................................................................6 Color and Lycopene......................................................................................................7 Carotenoids............................................................................................................7 Watermelon Color.................................................................................................8 Lycopene...............................................................................................................8 Stability of Watermelon................................................................................................9 Flavor Evaluation................................................................................................10 Taste....................................................................................................................10 Aroma..................................................................................................................11 Sensory................................................................................................................13 Processing of Beverage Products................................................................................13 HTST and LTLT..................................................................................................13 Ultrapasteurization..............................................................................................13 Ultra High Temperature (UHT)...........................................................................14 Non-Thermal Methods........................................................................................14 High Pressure Carbon Dioxide Processing.................................................................15 Effect on Microorganisms...................................................................................17 Micro-Bubble technique...............................................................................20 v

PAGE 6

Effect on spores............................................................................................21 Mode of Cellular Destruction..............................................................................21 Explosive decompression.............................................................................21 pH Lowering effect......................................................................................23 Inactivation of enzymes...............................................................................25 Extraction of cellular material......................................................................26 Effect on Flavor and Aroma................................................................................28 High Pressure Carbon Dioxide Processing of Watermelon Juice..............................28 Objectives...................................................................................................................29 3 MATERIALS AND METHODS...............................................................................30 Experimental Equipment............................................................................................30 Continuous high-pressure CO 2 Pasteurizer.........................................................30 Heat Pasteurization..............................................................................................32 Carbonation.........................................................................................................32 Watermelon Juice.......................................................................................................33 Watermelon Juicing.............................................................................................33 Aged Sample Procedure......................................................................................34 Modified Watermelon Juice................................................................................34 Experimental Design..................................................................................................35 First Cleanability Study.......................................................................................35 Second Cleanability Study..................................................................................36 High Pressure CO 2 Experimental Procedure.......................................................37 Pressure and Temperature Design.......................................................................37 Carbon Dioxide Concentration and Retention Time Design...............................38 Pressure and Temperature Design Using Modified Watermelon Juice...............38 Storage Study.......................................................................................................39 Analysis of Treated Samples......................................................................................40 Microbial Evaluation...........................................................................................40 Brix....................................................................................................................40 pH........................................................................................................................40 Titratable Acidity.................................................................................................40 Color....................................................................................................................41 Flavor...................................................................................................................41 Sensory Panels.....................................................................................................43 Lycopene.............................................................................................................44 4 RESULTS AND DISCUSSION.................................................................................46 Microbial Reduction Experiments..............................................................................46 Effects of Pressure and Temperature...................................................................46 Effects of Changing Levels of Carbon Dioxide..................................................48 Effects of Changing Residence Time..................................................................49 Effects of Pressure and Temperature on Modified Watermelon Juice................50 Microbial Growth During Storage Study............................................................52 vi

PAGE 7

Effect of High Pressure Carbon Dioxide on Physical Properties...............................55 Effect on Brix.....................................................................................................55 Effect on pH........................................................................................................56 Effect on Titratable Acidity.................................................................................58 Effect of High Pressure Carbon Dioxide on Watermelon Juice Aroma and Flavor...59 Sensory Panel Evaluation....................................................................................59 Flavor Evaluation by Gas Chromatography-Olfactometry and Mass Spectrometry....................................................................................................63 Effect of High Pressure Carbon Dioxide on Color.....................................................72 Effect of High Pressure Carbon Dioxide on Lycopene..............................................76 5 CONCLUSION AND RECOMMENDATIONS.......................................................79 APPENDIX A RAW EXPERIMENTAL DATA...............................................................................82 B SENSORY BALLOTS...............................................................................................86 LIST OF REFERENCES...................................................................................................89 BIOGRAPHICAL SKETCH.............................................................................................96 vii

PAGE 8

LIST OF TABLES Table page 2-1 Fresh watermelon nutrition........................................................................................6 4-1 Effects of pressure and temperature on aerobic organisms survivability of aged watermelon juice......................................................................................................46 4-2 Effects of pressure and temperature on yeast and mold survivability of aged watermelon juice......................................................................................................48 4-3 Effect of % CO 2 on aerobic organisms survivability of aged watermelon juice......49 4-3 Effect of % CO 2 on yeast and mold survivability of aged watermelon juice...........49 4-4 Effect of retention time on aerobic organisms survivability of aged watermelon juice..........................................................................................................................50 4-5 Effect of retention time on yeast and mold survivability of aged watermelon juice50 4-6 Effects of pressure and temperature on aerobic organisms survivability of modified aged watermelon juice..............................................................................................51 4-7 Effects of pressure and temperature on yeast and mold survivability of modified aged watermelon juice..............................................................................................52 4-8 Total aerobic growth during the 8 week storage study............................................53 4-9 Yeast and mold growth during the 8 week storage study........................................54 4-10 Change in Brix during the 8 week storage study....................................................56 4-11 Change in pH during the 8 week storage study........................................................57 4-12 Titratable acidity (% malic acid) during the 8 week storage study..........................58 4-13 Retention times, aroma descriptors of watermelon juice.........................................64 4-14 Aroma compounds in watermelon juice...................................................................65 A-1 Flavor difference from control evaluation of the hidden control, HPCD treated and heat pasteurized treated samples throughout the storage study................................82 viii

PAGE 9

A-2 Aroma difference from control evaluation of the hidden control, HPCD treated and heat pasteurized treated samples throughout the storage study................................82 A-3 Overall likeability mean value of all treatments during storage study.....................82 A-4 FID peak areas for aroma active compounds for each treatment through storage study.........................................................................................................................83 A-5 Aroma peak areas for aroma active compounds for each treatment through storage study.........................................................................................................................84 A-6 Sediment weight, lycopene concentration and color values for HPCD treated sample throughout storage study..............................................................................84 A-7 Sediment weight, lycopene concentration and color values for heat pasteurized sample throughout storage study..............................................................................85 A-8 Sediment weight, lycopene concentration and color values for untreated sample throughout storage study..........................................................................................85 ix

PAGE 10

LIST OF FIGURES Figure page 2-1 Molecular Structure of Lycopene...............................................................................8 2-2 Phase diagram of CO 2 ..............................................................................................15 2-3 CO 2 solubility in water at different temperatures and pressures..............................16 2-4 Dissociation of Carbonic Acid.................................................................................23 3-1 Schematic diagram of the Continuous High Pressure CO 2 Pasteurizer...................31 4-1 Effect of heat and pressure on the log reduction of total aerobes............................47 4-2 Total aerobic during storage study...........................................................................54 4-3 Yeast count during storage study.............................................................................55 4-4 Brix during storage study.........................................................................................56 4-5 pH during storage study...........................................................................................57 4-6 Titratable acidity (% malic acid) during storage study............................................59 4-7 Aroma evaluation of the hidden control, HPCD treated and heat pasteurize treated samples throughout the storage study......................................................................60 4-8 Flavor evaluation of the hidden control, HPCD treated and heat pasteurize treated samples throughout the storage study......................................................................61 4-9 Hedonic rating for all treatments at each time period through storage....................62 4-10 Overall likeability of each treatment during storage................................................63 4-11 FID peak areas for the untreated watermelon juice at 0 and 4 weeks......................66 4-12 Aroma peak areas for the untreated watermelon juice at 0 and 4 weeks.................67 4-13 FID peak area for the HPCD treated watermelon juice at weeks 0, 4 and 8............68 4-14 Aroma peak area for the HPCD treated watermelon juice at weeks 0, 4 and 8.......69 x

PAGE 11

4-15 FID peak area for the heat pasteurized juice at weeks 0, 4 and 8............................70 4-16 Aroma peak area for the heat pasteurized juice at weeks 0, 4, and 8.......................70 4-17 FID peak areas at week 0 for all three treatments....................................................71 4-18 Aroma peak area at week 0 for all three treatments.................................................72 4-19 Visual representation of the CIE color scale............................................................73 4-20 Watermelon juice L* after treatment and during storage.........................................74 4-21 Watermelon juice a* after treatment and during storage.........................................74 4-22 Watermelon juice b* after treatment and during storage.........................................75 4-23 Heat pasteurized watermelon juice a* when compared to the amount of sediment.75 4-24 Heat pasteurized watermelon juice L* when compared to the inverse amount of sediment...................................................................................................................76 4-25 Example lycopene chromatogram for an untreated sample at week 0.....................77 4-26 Example of the lycopene spectrum while it is eluting.............................................77 4-27 Amount of lycopene (ug/ml juice) in treated samples through the storage study....78 B-1 Sample taste test ballot.............................................................................................86 xi

PAGE 12

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science CONTINUOUS HIGH PRESSURE CARBON DIOXIDE PROCESSING OF WATERMELON JUICE By Matthew Lecky August 2005 Chair: Murat Balaban Major Department: Food Science and Human Nutrition In 1994, the total world watermelon production was 38 million metric tons. However, 10 years later the industry had almost tripled in size to 93 million metric tons. Watermelons are almost exclusively used for fresh market, and this leads to a very short shelf life, 2 to 3 weeks after harvest. In addition, many watermelons are left to rot in the field because of the economics when prices drop below the break-even line. Therefore, watermelons need an alternative market and/or product, such as juice. The processing of watermelon into juice could add strength and stability to the market. Processing of juices also increases their shelf life while offering consumers more availability and variety. High pressure carbon dioxide (HPCD) processing is a non-thermal method for “cold” pasteurizing beverages. This technology has shown that pathogens and spoilage organisms can be killed to make a safe and extended shelf life product. HPCD processing is the mixing of liquid carbon dioxide (CO 2 ) and juice at controlled pressures xii

PAGE 13

and temperatures. This technology has shown that many important sensory aspects can be maintained in their “like-fresh” quality. In this study, HPCD technology was evaluated to process watermelon juice. It was determined on fresh watermelon juice that using 34.4 MPa, 40C, 10% CO 2 in juice, and 5 min residence time could reduce native aerobic microorganisms by 6 log cycles. In an acidified, sweetened and carbonated product (pH 4.3, Brix 10.5), using the same processing parameters above could achieve a 4.5 log reduction. This level of treatment allowed the juice to be stored for 8 weeks at refrigerated conditions without major spoilage or unacceptable taste panel results. During the 8-week storage study, Brix, pH , titratable acidity, color, lycopene, flavor and aroma of the HPCD treated sample were evaluated and compared to a flash pasteurized (74C for 15 sec) sample. B rix, pH and titratable acidity did not change after treatment or during storage. The color data showed that HPCD treated juice L*, a*, b* values were about 31, 27, 24 respectively, and stayed stable over the total shelf life. The heat pasteurized samples had average L*, a*, b* values of about 36, 19, and 18 respectively. On Week 0, the taste panel determined that the taste and aroma of the heat-treated and HPCD treated samples had the same difference from control; they could also be differentiated from the hidden control. However on Week 8, the taste panel could not differentiate the taste or aroma of the treated samples from the hidden control. These results suggest that HPCD processing offers a viable alternative to heat pasteurization. This new technology has the potential to bring watermelon juice to the market with or without carbonation, and with or without acidification and sweetening. xiii

PAGE 14

CHAPTER 1 INTRODUCTION Watermelon (Citrullus lanatus) are part of the dessert melon category and they are thought to have originated from the Middle-East and tropical African regions. In 2004, the total world production of watermelons was 93 million metric tons (Mt) (FAOSTAT, 2004). In the United States, Florida is a leading state in watermelon production. In 2002, Florida was the number one watermelon producer in the country, producing about 344,000 Mt (USDA, 2002). Watermelon farming is a risky business. Later in the season the price can be so low that breaking even is difficult because of dropping prices and the reduced yield of additional harvests. This combined with the expensive manual labor to pick watermelons cause many to be left in the fields to rot when the price drops too low. Many “undesirable” (because of esthetic qualities, odd color, shape, size, etc.) watermelons are picked and transported to the packinghouse. This adds an additional expense because these “undesirable” melons cannot be sold in fresh market. Fresh watermelon has a very short shelf life, two to three weeks after harvest, and that of sliced or cubed flesh is even shorter (Vasquez and Nesheim, 2000). Value-added processing of watermelon juice can increase the value many folds thus adding strength and stability to the market. Processing of juices also increases their shelf life. Typical juice processing requires heat pasteurization. This type of processing may have negative effects on flavor, aroma and appearance (Zeuthen, 1984). In flash pasteurization, juice is usually heated to 74 – 85 C for 5 – 30 seconds depending on the 1

PAGE 15

2 process design. A 5-log reduction of pathogens is required in fruit juice processing to prevent disease outbreaks (FDA, 2001). Heating watermelon juice may adversely affect quality; off-flavors, off-aromas may form, and color may be affected. Non-thermal processing is an alternative to maintain the fresh flavor, aroma, texture and appearance of watermelon juice while increasing shelf life and maintaining safety. Past research has determined that high pressure carbon dioxide (HPCD) processing can reduce microbial populations, inactivate enzymes, and maintain fresh-like qualities in orange juice (Kincal, 2000). In HPCD processing liquid carbon dioxide (CO 2 ) and juice are mixed at controlled pressures and temperatures. As pressure increases so does the solubility of CO 2 into the juice. In this study high pressure carbon dioxide (HPCD) technology was evaluated using watermelon juice. The potential microbial reduction of this technology was examined first. Once a processing condition was determined an 8-week storage study was conducted. During the storage study, HPCD juice was compared to a flash pasteurized juice. Consumer taste panels were used to evaluate flavor and aroma. Juices were examined using gas chromatography-olfactometry and gas chromatragraphy-mass spectroscopy. Color, Brix, pH, titratable acidity, lycopene and microbial levels were monitored throughout the study. The objectives of this study were to determine effects of a continuous high pressure CO 2 processing on watermelon juice. The specific objectives were: 1. To determine effects of high pressure CO 2 on microorganisms 2. To pasteurize juice using non-thermal high pressure CO 2 technology to maintain aroma, flavor, appearance, and other attributes. 3. To increase the value of watermelon juice by extending shelf life

PAGE 16

CHAPTER 2 LITERATURE REVIEW Origin of Watermelon (Citrullus lanatus) The dessert melon category includes both muskmelons (Cucumis) and watermelons (Citrullus). Melons are part of the Cucurbitaceae family which includes cucumbers and squashes. Melons are thought to have originated from the Middle-East and tropical African regions. They spread to the Mediterranean, India and Asia. In 1994, the total world watermelon production was 38 million metric ton. However 10 years later, the industry has almost tripled in size to 93 million metric ton (Mt) in 2004 (FAOSTAT, 2004). The top five watermelon producing countries are: China with 68 million Mt, Turkey with 4 million Mt, Iran with 1.9 million Mt, USA with 1.7 million Mt, and Egypt with 1.6 million Mt (FAOSTAT, 2004). China with its 68 million Mt accounts for more than 70% of the worldwide watermelon production. Watermelons typically have a mottled green skin color and red flesh color. Watermelon size varies widely. There are both the diploid seeded and triploid seedless strains. Watermelon Production USA Production In 1996, the yearly watermelon consumption rate reached 16.8 pounds per person. The majority (85%) was purchased from retail stores for home use. Demographics show that the demand for watermelon is strong in middle-income families, with increased demand by Hispanics and Asians, which are two of the fastest growing ethnic groups in the US (Lucier and Lin, 2001). 3

PAGE 17

4 In the United States, watermelon is the number one melon crop in terms of per capita consumption, acreage and production. However, cantaloupe with its higher per unit value is the leader in terms of crop value. Between the years 1998 and 2000, the average farm value of watermelon in the United States was $262 million per year (Lucier and Lin, 2001). Florida is a leading state in watermelon production, producing about 344,000 metric ton (Mt) in 2002. Other major states were Texas with 302,000 Mt, California with 273,000 Mt, and Georgia with 231,000 Mt in watermelon production during 2002. In the 2002 growing season, Florida’s watermelon crop value was $62 million from 23,000 acres cultivated (USDA, 2002). Florida Harvest Watermelons in South Florida are usually harvested in April and May. Production in central Florida is harvested in May. In north-central Florida, they are harvested in June and July. Small amounts of watermelons are harvested the entire year, and between December through April, Florida is the only US producer. Growing watermelon in South Florida is roughly twice as expensive as growing them in North Florida but South Florida has the benefit of reaching the market 2 months before North Florida. This early market can see prices that are 2-3 times higher (Hochmuth et al., 1997). Watermelons are harvested by hand but not all watermelons in a field will ripen at the same time. This requires multiple harvests. Each additional harvest will have smaller yields than the previous (USDA, 1994). The watermelon market fluctuates widely depending on supply, weather and time of year. Early crops can be very profitable but once the crops in other states begin to be harvested, prices usually drop (USDA, 1994). Often later in the season, the price can be so low that breaking even is difficult because of

PAGE 18

5 the dropping prices and the reduced yield of additional harvests. Therefore, watermelons are left in the fields to rot. Florida seeded varieties include: Carnival, Fiesta, Mardi Gras, Regency, Royal Flush, Royal Star, Royal Sweet, Sangria, StarBright, Stars-n-Stripes, and Summer Flavor 500. Florida seedless varieties include: Constitution, Crimson Trio, Freedom, Genesis, King of Hearts, Millionaire, Revere, Scarlet Trio, Summer Sweet 5244, Summer Sweet 5544, and Tri-X-313 (Hochmuth et al., 1999). Watermelon Composition Watermelon Grades Watermelon quality has three United States Grade Standards: U.S. Fancy, U.S. No.1, and U.S. No.2. U.S. Fancy is the highest of the grades. These grades are mostly based on appearance. The conditions of importance are maturity, shape, ripeness, disease and physical damage. However, there are also optional internal quality standards for sweetness. These include “Very Good” having 10% soluble solids or more and “Good” having at least 8% soluble solids (Sargent, 2000). Maturity and Ripeness Maturity and ripeness are both difficult to determine visually. The conditions that experienced handlers look for are color of the bloom, change in tendril color from green to brown, ground spot having a yellow color, and a hollow sound when the melons are thumped (Sargent, 2000). Watermelons are usually harvested fully ripe, unless they are being marketed far away, where the less ripe firmer fruit fairs better during shipment. This is because the flesh color and sugar development will usually stop once the fruit is removed from the vine. Watermelons are sensitive to ethylene, which in low concentration causes

PAGE 19

6 deleterious effects (tissue deterioration, etc.). However, watermelons do not have an autocatalytic ethylene production ability. Once removed from exogenous ethylene, the melon respiration returns to normal. Therefore, watermelons are considered non-climacteric (Elkashif et al., 1989). Composition High sugar content is very important to the quality of watermelon fruit. The vast majority of watermelon’s sweetness is from sucrose (Pratt, 1971). Fructose can also be present in some watermelon cultivars (Elmstrom and Davis, 1981). Work conducted by Chisholm and Picha (1986) determined that the predominant organic acid is malic, with lower concentrations of citric acid also being found. Watermelon is high in dietary fiber, Vitamins: C, B6 (pyridoxine), A, B1 (thiamin), B5 (pantothenic acid), H (biotin), and minerals magnesium, potassium (Table 2-1). Table 2-1. Fresh watermelon nutrition 100 grams of watermelon flesh Nutrient Unit Value Water g 91.45 Energy kcal 30 Protein g 0.61 Total Lipid g 0.15 Ash g 0.25 Carbohydrates g 7.55 Fiber g 0.4 Calcium mg 7 Iron mg 0.24 Magnesium mg 10 Phosphorus mg 11 Potassium mg 112 Sodium mg 1 Zinc mg 0.10 Copper mg 0.042 Manganese mg 0.038 Selenium mcg 0.4 Vitamin C mg 8.1 Thiamin mg 0.033 Riboflavin mg 0.021

PAGE 20

7 Table 2-1 Continued Nutrient Unit Value Niacin mg 0.178 Pantothenic acid mg 0.221 Vitamin B-6 mg 0.045 Folate mcg 3 Vitamin B-12 mcg 0 Vitamin A IU 569 Vitamin E mg 0.05 USDA, 2004 Color and Lycopene Carotenoids Carotenoids are pigments synthesized by plants, algae and fungi. They are the source of red, orange and yellow colors for many fruits and vegetables including watermelon. These chemicals occur in both the photosynthetic and non-photosynthetic tissues. In the photosynthetic tissues, the degradation of the green pigment chlorophyll is often required for the carotenoids color to be visible. The consumption of carotenoids can act as colorants in some animals. For example, salmon flesh or the color of flamingo feathers require astaxanthin. There are more than 600 types of carotenoids, the most widely known are -carotene and -carotene (Stahl and Sies, 1996). For humans, carotenoids have many functions, some serve to form vitamin A and most have antioxidant potential. Also carotenoids are used as non-toxic natural food colorants. Carotenoid’s basic structure is a tetraterpene made up of two C 20 units linked tail to tail. This long structure of conjugated bonds is responsible for their color, and their susceptibility to oxidation and free radicals thus their antioxidant activity (Young and Britton, 1993).

PAGE 21

8 Watermelon Color Watermelon flesh comes in a variety of flesh colors (red, orange and yellow). The vast majority of watermelons harvested are red-fleshed. The red color of watermelon is attributed to the carotenoid lycopene. The other flesh colors orange and yellow are created by mixture of -carotene and xanthophylls (Watanabe et al., 1987). Lycopene Lycopene is a fat soluble carotenoid, which has an acyclic structure that contains 11 conjugated double bonds usually found in the linear all-trans formation consisting of only carbon and hydrogen (Figure 2-1) (Stahl and Sies, 1996). Lycopene does not have provitamin A activity. It happens to have the highest singlet oxygen-quenching rate of all carotenoids, having twice the antioxidant potential of -carotene (Di Mascio et al., 1989) This phytochemical also has the highest TEAC (Trolox-equivalent radical scavenging capacity) value of all carotenoids (Rice-Evans et al., 1997). Figure 2-1. Molecular Structure of Lycopene Numerous studies have been released that inversely link lycopene with some chronic diseases, namely certain cancers and coronary heart disease (Bramley, 2000). In an epidemiological study by Giovannucci et al. (1995), they discovered an inverse relation with the consumption of tomato products, which are high in lycopene, and the occurrence of prostate cancer. High levels of serum lycopene were associated with a reduced risk of bladder cancer (Helzlsouer et al., 1989). Higher levels of lycopene in the adipose tissue were associated with reduced risk of cardiovascular disease (Kohlmeier et

PAGE 22

9 al., 1997). Cancer cell growth in tissue cultures was shown to be inhibited by lycopene (Dorgan et al., 1998). In the United States, 80-85% of our lycopene come directly from tomatoes and tomato products (Chug-Ahuja et al., 1993; Bramley, 2000). The remainder comes from watermelon, guava, pink grapefruit, apricots, persimmons and papaya. However, watermelon consumption in Spanish diets can account for as much as 50% of their dietary lycopene (Granado et al., 1996). Fresh watermelon (48.7 g/g) has roughly 60% more lycopene than raw tomato (30.3 g/g) (Holden et al., 1999). Stability of Watermelon In retail stores, watermelon can be found as whole fruit or it maybe found in slices (half, quarters or sections). The melon is also found as cubes alone or mixed with other fresh fruits and vegetables. Fresh whole watermelon has a short shelf life; two to three weeks after harvest. Cut watermelons shelf life is even shorter (Vasquez and Nesheim, 2000). Watermelons can also be damaged by external ethylene and are prone to chill injury when stored below 10 C (Sargent, 2000). By processing watermelons into juice, a value-added product can be produced. This may increase its value many folds and thus add strength and stability to the watermelon market. Processing watermelon into juice will also increase its shelf life, and possibly create a product that would be available year around.

PAGE 23

10 Flavor Evaluation Flavor quality can arguably be the most important attribute when consumers evaluate food for purchase. But how do we evaluate and understand flavor? Flavor evaluations are very subjective and thus difficult to measure. Flavor is the combination of taste and aroma. There are four major tastes; sweet, sour, bitter, and salty. These taste compounds are non-volatile and can be detected in the parts per hundred range. However, aroma compounds can be detected in the parts per billion range and consist of volatile compounds which are detected by the olfactory nerve in the nose (DeRovira, 1997). Human perceptions of flavor, however are not that simple. Since both taste and aroma are integrated in the brain to determine flavor it is often difficult to determine one from the other, and people sometimes interpret a change in aroma as a change in taste (O’Mahony, 1995). The perception of flavor can further be misunderstood because of synergistic and/or masking effects that some chemical combinations have. Finally, there are also anesthetization effects that can occur when our olfactory nerve becomes overloaded (DeRovira, 1997). The use of human taste panels can help us understand the preference and acceptance of food based on flavor. Taste Taste is the combination of four perceptions; sweet (sugars), sour (hydronium ions), salty, and bitter. The tongue perceives for instance sodium ions as salty, and alkaloids or glucosides as bitter. But more important in fruit is the perception of sweet and sour. Sweetness is created with sugars and sugar substitutes. Important sugars in fruit are sucrose, fructose and glucose. Overall sugar content is typically evaluated using a

PAGE 24

11 soluble solids content (SSC, Brix) measurement. Higher SSC equates to higher sweetness. Refractometers are used to determine total SSC. However, depending of the product, the relationship is not always linear and sugar substitutes minimally influence SSC. Sourness is caused by hydronium ions which, come from organic acids in fruits. Some of the well known fruit acids are citric acid found in citrus and tomatoes, malic acid found in apples, and tartaric acid, a key acid in grapes. Each acid has a different sour taste profile based on its chemical structure. Chemical structures with higher molecular weight or hydrophobicity tend to increase the perception of sourness. However, lower sourness is perceived with an increase in carboxyl groups (Hartwig and McDaniel, 1995). A common way to measure acids is using titratable acidity (TA) or by measuring pH. In order to better understand the sourness and sweetness profile, a ratio of SSC/TA is used to compare fruits and fruit juices. The higher the ratio the higher sweetness to sourness taste profile. Aroma Aroma is created by the volatile compounds in food and is perceived as flavor. Cell disruption of fruits often releases compartmentalized substrates and enzymes, which can now interact and create aroma compounds (Buttery, 1993). Aroma compounds can also be linked to sugars, which can be released by heat or enzymes. The breakdown of pigments, lignin, lipids, and amino acids also creates aroma active compounds (Buttery and Ling, 1993) Aroma compounds generally fall into two categories: light aroma and heavy aroma. The lighter aromas can be characterized as top-notes, being low molecular weight, polar and hydrophilic. These compounds tend to be the characteristic aromas of the food being

PAGE 25

12 evaluated, having a large impact on perception. Whereas, heavier aroma compounds are generally high molecular weight, non-polar, hydrophobic compounds are considered background notes (DeRovira, 1997). The measurement of volatile compounds has evolved over the last few years. Aroma compounds are present in low concentrations thus they become difficult to capture and analyze. Once the compounds are concentrated, they can be analyzed using gas chromatography mass-spectrometry (GC-MS). Steam distillation and solvent extraction are classic isolation/concentration procedures, however internal standards are needed and this method becomes laborious for large sample numbers. The true flavor profile is considered to be closely reflected in the static headspace of a food and so headspace methods have been developed (Baldwin, 2002). Purge and trap headspace sampling methods concentrate volatiles on a solid support, which are later released by heat for analysis by GC-MS (Baldwin, 2002). This method works well for quantifying and identifying aroma compounds (Schamp and Dirinck, 1982). Cold trap isolation methods have been developed to reduce the damaging effects of heat on flavor compounds (Teranishsi and Kint, 1993). The most recent method is solid phase micro extraction (SPME), where a coated fiber adsorbs volatile compounds from the headspace and desorbs them in the GC injection port. This method has been used on apples, tomatoes and strawberries (Song et al., 1997). SPME is proving to be a very rapid method for volatile analysis. Linking volatile chemical concentrations to actual flavor intensity still proves very difficult. Gas chromatography olfactometry (GC-O, GC with sniff port) allows for the separated compounds to be evaluated for relative intensity by the human nose (Baldwin, 2002).

PAGE 26

13 Sensory Aroma, flavor and taste are often evaluated by human perception via taste panels. There are two typical types of taste panels, untrained panels and trained panels. Generally untrained panels consist of more than 50 panelists, who usually rank perceptions on a traditional 9-point hedonic scale. Acceptance and difference tests are often conducted by untrained panelists. Trained panelists can be used to evaluated individual flavor attributes and their intensities (Baldwin, 2002). Processing of Beverage Products Typical juice processing requires heat pasteurization. There are a few forms of pasteurization for juice and dairy beverages, high temperature short time (HTST), low temperature long time (LTLT), ultrapasteurization and ultra high temperature (UHT). These types of processing causes negative effects on flavor, aroma and appearance (Zeuthen, 1984). Because of the damaging effects of heat to food products, many so-called non-thermal methods are now being explored and developed. HTST and LTLT High temperature short time (HTST) processing usually heats products to 74-85 C for 5-30 seconds depending on the process design. Low temperature long time processing usually heats products to 63 C for 30 min. These treatments are used to inactivate pathogenic vegetative microorganisms. However, they might not eliminate all spoilage-causing microorganisms, nor any spore formers. These processes result in a refrigerated product with a 2-3 week shelf life. Ultrapasteurization This is pasteurization using very high temperatures. During ultrapasteurization temperatures will reach 138 C or higher for a minimum time of 2 seconds or longer.

PAGE 27

14 This method creates a refrigerated product with an extended shelf life of 6 to 8 weeks. The increased heat treatment results in the reduction of a larger number of spoilage microorganisms. Ultra High Temperature (UHT) Ultra high temperature processing uses very high temperatures of 129 to 146 C for 2 to 45 seconds. The product is then aseptically filled. The result is a commercially sterile shelf stable product that can last from 1 to 2 years at ambient temperature. The shelf life is determined more by the package (barrier) characteristics, oxygen and moisture transmission, which cause physicochemical changes, than by the contamination of microorganisms (David et al., 1985). Non-Thermal Methods Research is currently being conducted to develop non-thermal processing methods. Many of these methods are product dependent and have not proven themselves like the traditional heat pasteurization processes. Some of the non-thermal processing methods are hydrostatic pressure, pulsed electric fields, oscillating magnetic fields and irradiation (Barbosa-Canovas, 1998). The major advantage of non-thermal processing is the potential for minimal effects from heat, while achieving microbial reduction. Heat has many effects on food products, some good and some deleterious. Heat destroys microorganisms and undesirable enzymes, but it also causes increased browning formation (sometimes desirable; as in baked products), texture changes, nutritional degradation, color changes, and flavor changes. Past research has determined that high pressure carbon dioxide processing can reduce microbial populations, inactivate some enzymes, and maintain fresh-like qualities in orange juice (Arreola et al., 1991b; Kincal, 2000). Work by Folkes (2000) found that a

PAGE 28

15 consumer taste panel found negligible differences between high pressure carbon dioxide treated beer and fresh beer. High Pressure Carbon Dioxide Processing High pressure CO 2 processing is the mixing of liquid CO 2 and juice at controlled pressures and temperatures. Carbon dioxide is a great gas for food use because it is physiologically safe, inexpensive, and easily available in high purity and in large quantities. Also, CO 2 has a low critical pressure (72.8 atm) and low critical temperature (31.1 C), see phase diagram in Figure 2-2. These low critical parameters allow researchers to achieve and examine suband supercritical CO 2 effects on liquid products. Supercritical CO 2 acts as a solvent of lipophilic materials, which could possibly lead to more desirable microbial effects. Supercritical fluids have liquid like density and gas-like diffusion and viscosity properties. Figure 2-2. Phase diagram of CO 2

PAGE 29

16 The solubility of CO 2 in water is of critical importance during high pressure carbon dioxide processing. By increasing pressure, we are able to increase the amount of dissolved CO 2 into the process liquid. However, increasing temperature generally decreases the solubility of CO 2 in water, see Figure 2-3. Figure 2-3. CO 2 solubility in water at different temperatures and pressures, From Dodds et al., 1956.

PAGE 30

17 Effect on Microorganisms The phenomenon of bacterial cell disruption by pressure and carbon dioxide has been of interest since the 1950’s, when Fraser (Fraser, 1951) and Foster (Foster et al., 1962) first reported that high pressure CO 2 seemed to have a bactericidal effect. In the 1980s, researchers reported the bacteriostatic action of CO 2 on the growth and metabolism of some microorganisms (Doyle, 1983; Enfors and Molin, 1980; Jones and Greenfield, 1982). It was not until the late 80s, when researchers began examining processing conditions and determining lethality, that this field took off. Work investigating the effects of carbon dioxide on baker’s yeast, Escherichia coli, Staphylococcus aureus, and Aspergillus niger was conducted by Kamihira et al., (1986). They determined that inactivation was possible with supercritical CO 2 at 200 atm and 35 C on cells with water content between 70 and 90%. However, dry cells with water content between 2 and 10% could not be sterilized. Further confirmation that moisture has a large influence on the success of high pressure carbon dioxide came from Haas et al., (1989). His research with high pressure carbon dioxide (6.2 MPa, 23 C, and 2 hour treatment) determined that with high moisture flour (28%) a 99.6% reduction on Standard Plate Counts could be achieved, but with low moisture flour (12%) no bactericidal effect could be found. He further found that up to a four log reduction of Escherichia coli, and Staphylococcus aureus in nutrient broth or isotonic saline solution resulted with 6.2 MPa, room temperature, and a treatment time of 2 hours (Haas et al., 1989). Later research provided confirmation that Escherichia coli suspension could be reduced by 4 logs within 14 min with CO 2 at 5 MPa and 45 C in a study by Ballestra et al., (1996). Haas and colleagues had also effectively

PAGE 31

18 inactivated the microorganisms in fresh herbs using 5.5 MPa CO 2 at 45 C for 2 hours (Haas et al., 1989). The aerobic total plate count of fresh-squeezed orange juice was reduced by 2 log using CO 2 at 25 C and 33 MPa for 45 min in a static batch system (Arreola et al., 1991a). Wei et al., (1991) found that a CO 2 treatment of 6.18 MPa and 35 C for 2 hours could kill Listeria monocytogenes in water. They also found that CO 2 at 13.7 MPa and 35 C for 2 hours could reduce Salmonella spiked in chicken meat and Listeria spiked in shrimp meat (Wei et al., 1991). Saccharomyces cerevisiae, a yeast, was investigated using high pressure carbon dioxide by Lin et al., (1992a ). They found that this yeast could be inactivated by both supercritical and subcritical CO 2 , however supercritical was found to be more effective (Lin et al., 1992a). Staphylococcus aureus in broth could be inactivated by CO 2 at 8 MPa and 25 C in 2 hours, however S. aureus in whole and skim milk was more difficult to inactivate in a study by Erkmen (1997). He determined that CO 2 at 14.6 MPa for 5 hours was necessary to inactivate S. aureus in whole milk. Erkmen concluded that the fat content might have a protective effect possibly slowing down the penetration of CO 2 through the cell membrane (Erkmen, 1997). A buffer solution of Lactobacillus plantarum was reduced by 5 logs using CO 2 at 45 C, 13.8 MPa and a process time of 5 min in research completed by Hong et al., (1999). Dillow and associates (1999) found that many microorganisms could be inactivated with CO 2 at 20.5 MPa, between 25 – 40 C and between 0.6 – 4 hours treatment time. However, they used a technique of depressurizing and repressurizing 5 times per hour during treatment. They found that Bacillus cereus, Legionella dunnifii,

PAGE 32

19 Staphylococcus aureus, Listeria innocua, Salmonella salfor, Proteus vulgaris, Pseudomonas aeruginosa, and Escherichia coli all could be inactivated by CO 2 depending on the conditions used (Dillow et al., 1999). In a series of experiments with CO 2 conducted by Erkmen, he found that Salmonella typhimurium in saline could be inactivated (6.08 MPa, 35 C for 15min), Listeria monocytogenes in saline could be inactivated (6.08 MPa, 45 C for 60 min), Enterococcus faecalis in saline could be inactivated (6.05 MPa, 35 C for 18 min), and Escherichia coli in nutrient broth could be inactivated (10 MPa, 30 C for 50 min) (Erkmen, 2000a; Erkmen, 2000b; Erkmen, 2000c; Erkmen, 2001). Research conducted by Kincal (2000) found that in a continuous HPCD system native aerobes in orange juice could be reduced by 3 logs. She also found that the pathogenic organisms Escherichia coli O157:H7, Salmonella tyhimurium and Listeria monocytogenes could be reduced by 5 logs. Most recently in a semi-continuous system, Spilimbergo et al., (2003) found that the yeast cell Saccharomyces cerevisiae could be reduced by 7 logs within 7.5 min with CO 2 at 8 MPa and 38 C. They also found that the gram positive cell Bacillus subtilis took 2.5 min at 7.4 MPa and 38 C to be reduced by 7 logs but the gram-negative microorganism Serratia marcescens only took a 0 min hold time (pressurized up and then immediately depressurized) at the same conditions to be reduced by 7 logs. Their work indicates that using a semi-continuous system can dramatically reduce the processing time needed (Spilimbergo et al., 2003). Work conducted by Folkes (2004) developed a model that predicted a maximum reduction of beer yeasts cell at 7.38 log using a continuous system with the following conditions, at 26.5 MPa, 21C, 9.6% CO 2 , and 4.77 minutes. Up until the work conducted by Kincal,

PAGE 33

20 the majority of high pressure carbon dioxide research was conducted on batch type equipment. Research has shown that increasing temperature generally increases the effects of high pressure CO 2 on microorganisms (Haas et al., 1989; Arreola et al., 1991a; Lin et al., 1992a; Lin et al., 1994; Ballestra et al., 1996; Hong et al., 1999; Erkmen, 2000a; Erkmen, 2000b, Erkmen, 2000c). It is believed that the higher temperatures stimulate diffusion of CO 2 into the cells and also increase the fluidity of the cell membrane, thus making penetration of CO 2 easier. It has also been seen that increasing pressures leads to more effective death rates of microbial cells (Lin et al., 1992a; Lin et al., 1994; Ballestra et al., 1996; Erkmen, 1997; Hong et al., 1997; Hong et al., 1999; Erkmen, 2000a; Erkmen, 2000b, Erkmen, 2000c; Hong et al., 2001; Erkmen, 2001). With increased pressure comes increased solubility of CO 2 into the media and cells, therefore possibly producing more contact time between CO 2 and microorganisms. Micro-Bubble technique Work conducted by Ishikawa et al., 1995b, determined that adding a porous stainless steel (10m pore size) filter forced the CO 2 to create “micro-bubbles,” which could increase the concentration of dissolved CO 2 from 0.4 to 0.92 mol/l at time 0 just after the process reached parameters (25 MPa and 35 C) in the liquid. In a later study, they found that this filter would enhance the reduction of both Lactobacillus brevis and Saccharomyces cerevisiae at subcritical and supercritical CO 2 . This enhancement allowed for shorter treatment times (15-30 min) than seen from the previous research. The filter had its largest effect with supercritical CO 2 . They also showed that high density supercritical CO 2 above 0.9 g/cm 3 could cause sterility in only 30 min (Ishikawa et al., 1995b).

PAGE 34

21 Effect on spores Minimal reduction of Bacillus subtilis and Bacillus stearothermophilus spores in distilled water treated with CO 2 at 20.3 MPa for 2 hr at 35 C was seen by Kamihira et al., (1987). A study conducted by Haas et al. determined that Penicillium roqueforti spores in a yeast and mold broth at pH 3.5 could be reduced by 5 logs with a CO 2 treatment of 5.5 MPa, 45 C for 2 hr (Haas et al., 1989). Further work by Haas et al. observed that a spore suspension of Clostridium sporogenes could be inactivated by CO 2 at 70 C, 5.5 MPa, treatment time of 2 hr when the pH of suspension was 2.5 to 3.0 (Haas et al., 1989). Research conducted by Enomoto et al., (1997a) saw that Bacillus megaterium spore cells could be reduced by 10 7 with high pressure CO 2 using the optimal conditions 5.9 MPa, 60 C, and 30 hr. Interestingly, pressures higher than 5.9 MPa did not increase the inactivation of this spore. However, increasing time and, more importantly, temperature did increase the inactivation of Bacillus megaterium. Enomoto et al. speculated that moderate heat and a low pH has a large impact on the reduction of spores (Enomoto et al., 1997a). Various Bacillus spores could be inactivated at the following conditions: 45 – 60 C, 30 MPa, 30-60min, by using supercritical CO 2 “micro-bubble” technique in work conducted by Ishikawa et al., (1997). Mode of Cellular Destruction Explosive decompression This process involves the penetration of liquid or supercritical fluid through a cells membrane at high pressures. Upon the release of pressure, the liquid fluid turns to gas and rapidly expands, thus rupturing the cells and releasing intercellular material. The

PAGE 35

22 benefits of this method is that it does not damage enzymes and functional properties of proteins (Lin et al., 1991). Early work conducted by Fraser to gather intracellular material determined that CO 2 was effective at rendering Escherichia coli 95-99 % non-viable, by releasing a suspension at 3.4 MPa to atmospheric pressure “as rapidly as possible” (Fraser, 1951). Fraser also noted that only 48-57% of the cells where ruptured. He further speculated that the intact cells, which were non-viable, may have been killed by the increase in acidity from CO 2 in water at 3.4 MPa (Fraser, 1951). Foster and colleagues later determined that explosive decompression using nitrogen at 12 MPa could only reduce Serratia marcescens by 31-59% and Brucella abortus and Staphylococcus aureus by 10-25% (Foster et al., 1962). The use of nitrogen for rapid decompression had little effect on the destruction of yeast cells in work by Nakamura et al., (1994) and Enomoto et al., (1997b). More recently, yeast cells (Saccharomyces cerevisiae) were evaluated using explosive decompression with CO 2 in a study by Lin et al., (1992b). They concluded that yeast cells were more difficult to rupture than microbial cells because of their thicker cell wall. However, increased temperature (35 C) and pressure (20.7 MPa) enhanced the penetration of CO 2 and/or relaxed the cell walls to allow for penetration. Their research determined that at 35 C, 20.7 MPa, for 3.5 hr, 80% of the yeasts cells could be ruptured upon decompression, whereas over 80% of non-yeast cells could be ruptured in under an hour (Lin et al., 1992b). Baker’s yeast at 40 C, 4.05 MPa, held for more than 3 hours in CO 2 were destroyed when pressure was suddenly released in work conducted by Nakamura et al., (1994). However, they speculated that the killing action is more than

PAGE 36

23 just rupturing of the cells, and that possible physiological damage occurs (Nakamura et al., 1994). In a study by Folkes (2004), yeast in beer was exposed to 27.5 MPa, 10% CO 2 , 21 C, for 5 min in a continuous system. The yeast were examined using scanning electronic microscopy and showed that the treated cells were shrunken, had divots and some were even exploded. In a study by Arreola et al., (1991a), rapid decompression was not determined to have a significant effect in reducing aerobic counts in orange juice. To further understand the effect of rapid decompression on yeast cells, Enomoto et al., (1997), conducted a comparative study involving rapid and slow decompression. Rapid decompression was characterized as a rate of 48 atm/min while slow decompression rate was 0.33 atm/min. Their work determined that the survival rate of yeast treated with CO 2 at 4.05 MPa, 40 C for 4 hr was the same for rapid and slow decompression (1.0 x 10 -7 ). They determined that the yeast cells appear to have been killed during the pressurization and that ruptured cells might not have been caused mainly by rapid decompression but some other effect (Enomoto et al., 1997b). pH Lowering effect During high pressure CO 2 processing, carbonic acid is formed as the CO 2 dissolves into water and temporarily lowers pH. Carbonic acid is a weak acid, which can dissociate into bicarbonate and carbonate ions in water, see Figure 2-4 below. Figure 2-4, Dissociation of Carbonic Acid CO 2 ( aq ) + H 2 O( l ) H 2 CO 3 ( aq ) (carbonic acid) H 2 CO 3 H + + HCO 3 (bicarbonate) pK1 = 6.57 HCO 3 H + + CO 3 2 (carbonate ion) pK1=10.62

PAGE 37

24 Many authors have suggested that the formation of carbonic acid that occurs during high pressure CO 2 has a destructive effect on microorganisms (Daniels et al., 1985; Kamihira et al., 1987; Lin et al., 1993; Lin et al., 1994; Ballestra et al., 1996; Hong et al., 1999; Dillow et al., 1999; Erkmen, 2000a; Hong et al., 2001). Generally, by increasing the solubility of CO 2, we increase the formation of carbonic acid. Increasing pressure will increase the solubility of gaseous CO 2 into water, however liquid and supercritical CO 2 is not affected by pressure as much. Solubility of CO 2 is also increased by lowering temperature. Work conducted by Toews et al., (1995) using pH indicator and UV-vis spectrophotometer showed that the pH of water in equilibrium with CO 2 at 25 C and 7.09 MPa was 2.83. They also found that at 70 C and 20.3 MPa, the pH of water under CO 2 was 2.84. They concluded that within those parameters, pH of water in equilibrium with CO 2 changes little due to temperature and pressure (Toews et al., 1995). Meyssami et al., (1992) found that in pure water systems the lowing of pH due to CO 2 was a function of pressure more than temperature. They found that in ascorbic acid, citric acid, model systems, the pH was insensitive to pressure and was a function of acid concentration. In order to determine if a reduction of pH can destroy microorganisms, a number of researchers have artificially reduced the pH by added acids. Haas and colleagues (1989) found that E. coli in nutrient broth reduced to pH 3.2 at atmospheric pressure had little effect on reducing microorganisms when compared to the sterilizing effect of CO 2 at 6.2 MPa. Wei et al., (1991) found similar results when they reduced a Listeria solution to pH 3.02 with 0.1 N HCL. They found that the increased acidity samples had no effect on the microorganism population (Wei et al., 1991). However, the internal pH of the

PAGE 38

25 microorganism is the cause of cellular destruction, not the external pH (Kashket, 1987). Carbon dioxide is able to reduce the cell’s internal pH and thus affect metabolic activities (Daniels et al., 1985). It is believed that a higher acidic environment increases the cell’s permeability to dissolved CO 2 (Lin et al., 1993; Hong et al., 1999). Therefore, as external pH decreases due to the formation of carbonic acid, more dissolved CO 2 is able to penetrate into the microorganism and thus form carbonic acid inside the cell. Lin and coworkers (1994) speculate that CO 2 under high concentration would form bicarbonate inside the cell. Upon the release of process pressure, the bicarbonate will change to carbonate ion and precipitate intercellular calcium and other ions causing irreversible damage to the cells. With the penetration of CO 2 into the microorganisms and the lowering of intercellular pH, it is possible that the cell’s cytoplasmic pool is unable to maintain proper cellular pH (Hong et al., 1999; Erkmen, 2000a; Hong et al., 2001). The reduced intercellular pH may cause essential metabolic systems to fail (Ballestra et al., 1996). Inactivation of enzymes The inactivation of intracellular enzymes by high pressure CO 2 has been proposed by many researchers (Daniels et al., 1985; Balaban et al., 1991; Ballestra et al., 1996; Erkmen, 1997; Hong et al., 2001). CO 2 may directly bind to enzymes and cause inactivation (Chen et al., 1992). Kamihira et al., (1987), Hutkins and Nannen (1993) and Erkmen (2000a) have all speculated that the lowering of cell’s internal pH might inactivate essential enzymes that are used for metabolic and regulation processes. Ishikawa et al. found that by using the micro-bubble technique, the higher density CO 2 (above 0.9 g/cm 3 ) could inactivate enzymes (Ishikawa et al., 1995a). They believed that

PAGE 39

26 high density CO 2 increases absorption of CO 2 into intracellular and membrane bound enzymes, which results in inactivation (Ishikawa et al., 1995b). Ballestra’s group (1996) found that enzymatic inactivation was selective. They saw that alkaline phosphatase and -galactosidase disappeared after treatment with high pressure CO 2 , yet acid phosphatase and naphtol-AS-BI phosphohydrolase were minimally affected. Ballestra speculated that the acidification of the cell’s internal pH might precipitate enzymes that have acid isoelectric points. Alkaline phosphatase and -galactosidase both have acid isoelectric points (Ballestra et al., 1996). However, Hong and Pyun (2001) found differing results. They found that the following enzymes were minimally affected by high pressure CO 2 : lipase, leucine arylamidase, -galactosidase, acid and alkaline phosphatase, and naphthol AS-BI-phosphohyrolase. Yet cystine arylamidase, -galactosidase, and -glucosidase, and N-acetyl--galactosidase lost significant activity after treatment. Even though reduced activity of cellular enzymes could be found in high pressure CO 2 treated cells, it is not clear if changes in enzyme activity is the major cause of cell inactivation (Hong and Pyun, 2001). Extraction of cellular material The extraction of membrane or intercellular material by high pressure CO 2 could have a deleterious effect to microorganisms. The solvent-like behavior of CO 2 helped by possibly allowing CO 2 to diffuse into the cell and remove vital components (Isenschmid et al., 1995). It has been speculated by many researchers that high pressure CO 2 can extract intercellular material or cell membrane phospholipids and thus inactivate the microorganism (Kamihira et al., 1986; Lin et al., 1992a; Ballestra et al., 1996; Erkmen,

PAGE 40

27 1997; Erkmen, 2000b). It has also been speculated that damage to cellular material by high pressure CO 2 might cause cell death without rupturing or cell leakage (Lin et al., 1992a; Ballestra et al., 1996; Erkmen, 2000b). However, membrane damage and/or extraction of cell wall constituents would cause leakage of cellular material into the surrounding environment (Hong and Pyun, 2001). In response to injury, microorganisms may leak into the environment amino acids, peptides, lipids and ions (Hurst, 1977). Work conducted by Dillow et al., (1999) found very little difference in the cell wall of before and after high pressure CO 2 treated Staphylococcus aureus and Pseudomonas aeruginosa when looking at scanning electron microscope images (SEM). They also found that gram-negative cells appeared to have slightly more defects after treatment than gram-positive cells, possibly due to the thicker gram-positive cell resisting treatment better. However, since the majority of the cells had intact cell walls, they determined that extraction of cell wall lipids was not the primary cell deactivation mechanism (Dillow et al., 1999). Hong and Pyun (2001) took a different approach, by measuring the presence of increased cellular material in the extracellular environment. They identified an increase of intracellular ions (Mg and K ions) and UV-absorbing materials in the extracellular environment after Lactobacillus plantarum was exposed to high pressure CO 2 . The proton permeability of the cytoplasmic membrane barrier function was also impaired after treatment. They also stained the treated cells with Phloxine B. This staining showed significant membrane integrity losses. From the results above and SEM images showing “wrinkles” and “holes”, Hong and Pyun determined that there was substantial

PAGE 41

28 damage to the microorganisms cell membrane after treatment with high pressure CO 2 (Hong and Pyun, 2001). Effect on Flavor and Aroma Heat processing usually causes a quality reduction in flavor, aroma, color or texture (Gould, 1989). Heat may also alter a product’s color, particular darkening. High pressure CO 2 processing can be performed at ambient conditions, therefore having no heat damaging effects on the juice. This technology allows for cold pasteurization of beverages. In a study conducted by Arreola and colleagues, they found that high pressure CO 2 treated orange juice’s flavor, aroma and overall acceptability were not altered. They also found the pH and Brix were not changed. Surprisingly, they found that the orange juice color was improved after treatment (Arreola et al., 1991b). The vast majority of research on high pressure CO 2 has been conducted in an attempt to understand its effect on microorganisms. There is minimal amount of research completed on understanding the effects of high pressure CO 2 on flavor and aroma. It is possible that flavor striping might occur as CO 2 may extract some aromatics. Also during the de-gassing step, where excess CO 2 is removed by vacuum, highly volatile top-notes could be lost. It is important to understand if there is a loss of flavors and aroma so that possible new advances and remedies can be explored. High Pressure Carbon Dioxide Processing of Watermelon Juice Watermelon juice is a novel product of great interest. As a juice product, it needs to be produced safely and must have high sensory qualities. The FDA currently requires that a fruit juice process be capable of achieving a 5-log reduction of a target pathogen to prevent disease outbreaks (FDA, 2001). Past research with high pressure carbon dioxide

PAGE 42

29 shows that it is possible to achieve at least a 5-log reduction. Past research has also shown that high pressure CO 2 can eliminate many types of microorganisms, including yeasts. What is of great interest is how high pressure CO 2 will increase shelf life of a potentially marketed watermelon juice product and the quality aspects of a product processed in this manner. Traditionally heat treatment methods are used to reduce pathogens and increase the shelf life of juices. These heat treatments have many adverse effects on flavor, color and aroma. Since high pressure CO 2 does not use heat, it is possible that the processed juice will maintain many fresh like characteristics such as flavor, aroma, texture, and appearance while still maintaining safety. However, it is necessary to understand the relationship between processing conditions, pressure, temperature, residence time and concentration of CO 2 on the effectiveness of high pressure CO 2 to accomplish the above. Objectives The objectives of this study were to determine effects of a continuous high pressure CO 2 processing on watermelon juice. The specific objectives were: 1. To determine effects of high pressure CO 2 on microorganisms 2. To pasteurize juice using non-thermal high pressure CO 2 technology to maintain aroma, flavor, appearance, and other attributes. 3. To increase value of watermelon juice by extending shelf life

PAGE 43

CHAPTER 3 MATERIALS AND METHODS Experimental Equipment Continuous high-pressure CO 2 Pasteurizer A continuous high-pressure carbon dioxide pasteurizer has been developed and provided by Praxair (Burr Ridge, IL). This unit is capable of max pressure of 69 MPa, and flow-rate of about 0.5 liters per minute. The unit consists of a feed tank, CO 2 supply tanks, CO 2 chiller and pump, product supply pump, high-pressure pump, mixing and holding coils (which can be heated), depressurization valve, vacuum tank (product tank) and vacuum pump. Temperature can be measured throughout the process by K-type thermocouples. Pressure and flow rate can also be measured by gauges instrumented throughout. The equipment first pressurizes the feed juice to 6.9 MPa with a variable speed-reciprocating pump. A backpressure regulating value modulates this set pressure as needed. Two siphon tube CO 2 tanks supply liquid CO 2 . This liquid CO 2 is chilled to 4 C before it is pressurized to 6.9 MPa by a small reciprocating pump and mixed with the juice feed. The temperature and pressure of the system must be such that the CO 2 remains in liquid state while being pumped. The second reciprocating high-pressure pump pressurizes the juice and CO 2 mixture to the operating pressure (usually 10.3 .5 MPa). Placed directly after the second pump is a reduced diameter tube, this creates turbulent flow, which insures proper mixing of the liquid CO 2 and juice. After the reduced diameter tube is the holding tube (79.2 m, 0.635 cm ID). The holding tube has 30

PAGE 44

31 electrical heating tape wrapped around it. This heating tape allows the holding tube to be heated above room temperature (usually 25-45 C). A backpressure regulator value placed at the end of the holding tube modulates the pressure as required by the experiment. After the backpressure regulating value is the exit for the system where the juice steam can be collected into a desired vessel. P Main Pump Juice stream Vacuum Heating systemHold tube TreatedCO2juiceCO tank2 Expansion valve Pump Pump Chiller 4 1 2 3 5 6 7 8 9 Figure 3-1. Schematic diagram of the Continuous High Pressure CO 2 Pasteurizer All data from the temperature probes, flow meters, and pressure gauges are collected into a Fluke Hydra Series II (Fluke Corporation, Everett, WA). This data is collected in real time (for monitoring) and saved as an Excel file to be examined at a later time.

PAGE 45

32 Heat Pasteurization A Peristaltic pump (Cole Parmer, Chicago, IL) pumped watermelon juice through two 5.47m of copper tubing (4.76mm ID), which was placed in a water bath (Precision Scientific Group, Chicago, IL). The first copper tubing was necessary to heat up the sample from room temperature to water bath temperature and the second copper tubing had the proper volume and flow rate to create a 15 sec hold time at the desired water bath temperature. The water bath temperature was set at 75 C. After the heating step, watermelon flowed through 11m of copper tubing (4.76mm) placed in ice slush and the exiting juice temperature was 10 C. Flow rate was no faster than 400 ml/min. At these conditions, the juice was heated for at least 15 sec at 74 C or slightly higher. After juice was collected, it was immediately placed in a cold room (4 C). Carbonation Carbonating the watermelon juice was necessary in order to bring all samples to a chosen CO 2 level. The HPCD treaded, heat treated and untreated (control) juice needed to have the same level of carbonation to minimize bias in any qualitative and sensory results. In order to do this, a Zahm & Nagel Pilot Plant Carbonator (Zahm & Nagel Co., Inc., Buffalo, NY) was used. This is a batch type carbonator with a volume of about 7.5 liters. Before carbonation, all juice was refrigerated to 4 C. The carbonation equipment was cleaned with soap and alcohol. All carbonation was conducted inside of a 4 C cold room. The following parameters were used: juice temperature 4 C, pressure 55 kPa, CO 2 bubbling time 10 min. Those conditions would have about 2.2 liters of CO 2 per liter of water (Zahm & Nagel, Volumes of CO 2 gas dissolved in water). The equipment filled

PAGE 46

33 750 ml champagne bottles. All filled and capped (metal crown) bottles were stored at 4 C. Watermelon Juice Watermelon Juicing Through contact with Mr. Gene McAvoy from the Hendry county Extension office in Florida, and Eugene Tolar of Red Star Farms Inc. (Hendry county, Florida) 2000 kg of cull watermelons were obtained. These melons were unmarketable due to size, shape or color. The watermelons were of the Sentinel variety and were harvested at about 90 days. About 100 watermelons were juiced the first day and the remaining 200 juiced the next day. First the melons were washed with tap water and scrubbed with a soft brush to remove dirt and debris. Scrubbed watermelons were then dipped into a 300 ppm chlorine solution for 30 sec, then placed in a fresh water dip. Then the melons were weighed and recorded. Next the melons were sliced open into fourths. Then the flesh was removed by cutting and scooping out with a large spoon. The rinds were thrown away and the flesh was placed aside for squeezing. A bladder press (Willmes Presser, type WP100, Bensheim, Hessen, Germany) was lined with a filter cloth and the watermelon flesh was placed inside. The flesh was squeezed with 700 kPa air pressure. After each squeeze, the juice was taken to a large stainless steel tank in a cold room, 1.67 C. Once this tank was half full, the filling of containers started. Two-liter glass containers were filled with about 1.7 L using a peristaltic pump with Tygon tubing. Thirteen and one-third liter plastic pails were filled to about 11.4 L in the same fashion. Both the glass containers and the plastic pails were placed into frozen storage at C.

PAGE 47

34 Aged Sample Procedure When aging was necessary for the experimental procedure the following steps where taken. First the 13.3-liter pail was removed from frozen storage and placed in the cold room (4 C) for one week to thaw. After one week, the pail was shaken and placed on the bench top at room temperature (23 C) for 44 hours with periodic shaking. The purpose was to increase the microbial load to ascertain the effectiveness of the process. Modified Watermelon Juice It was necessary to acidify the watermelon juice for two reasons. First watermelon juice has a pH of about 5.5 and as such is considered a low acid beverage. Secondly, acidification was needed because the watermelon juice product would remain carbonated after processing. Carbonation when bottled creates a reduced oxygen environment. Low acid and low oxygen products are problematic because of the potential to harbor Clostridium botulinum, a dangerous anaerobic spore forming pathogen. The potential neurotoxin that Clostridium botulinum creates is not heat stable, however a beverage is not cooked before consumption. By creating a high acid (pH less than 4.6) product we overcome this potential problem. Food grade powder malic acid (Presque Isle Wine Cellars, North East, PA) was used to reduce the watermelon juice’s pH. The target pH value was 4.3, and a concentration of 1.7 grams of malic acid per liter juice achieved this. Because of the increased sourness of the product, it was then necessary to sweeten the juice. High fructose corn syrup (HFCS) is a commonly used sweetener in the industry. Through taste tests with various concentrations of added HFCS, it was determined that 35 grams of Isoclear 55 HFCS (Cargill, Dayton, OH) per liter of watermelon juice created a

PAGE 48

35 balanced (acid to sweetness) beverage. The added sweetener increased the watermelon juice’s Brix to about 10.5. Any experiment with modified watermelon juice used the above-mentioned concentration of malic acid and HFCS. Experimental Design First Cleanability Study The ability to clean the Praxair High Pressure CO 2 equipment was very important to ensure no contamination during experimentation. The cleaning chemicals used were Oxonia and Principal (Ecolab, St. Paul, MN). The non-pathogenic organism used during this study was Lactobacillus fermentum (ATCC, Manassas, VA). This organism was shipped as a freeze dried sample and was rehydrated in 100 ml of Lactobacillus MRS broth (Difco Laboratories, Sparks, MD). The culture was grown at 37 C and then diluted down with 6 L of Butterfield’s phosphate buffer to create approximately 10 6 colony forming units (CFU) per ml. The high pressure equipment was cleaned first with a solution of Principal (50 ml) in 19 L of warm tap water and then rinsed with 19 L of tap water. After the rinse, a solution of Oxonia (946 ml) in 19 L of cold water was pumped through the equipment. Oxonia solution remained (7.6 L) in the feed tank and throughout the equipment over night. The following day, the remaining Oxonia was pumped through the equipment and 6 L of sterile water was used to wash the equipment of all residual chemicals. The 6 L of 10 6 CFU/ml Lactobacillus fermentum in Butterfield’s phosphate buffer was then pumped through the equipment, capturing all the exiting buffer. The exiting buffer was plated to measure the recovered microbial load, which was about 10 5 CFU/ml recovered.

PAGE 49

36 After the equipment was contaminated with Lactobacillus fermentum, it was necessary to clean and sanitize the equipment. As per normal use, water was used to rinse the equipment and then the sanitizing procedure above was used. The Oxonia remained in the equipment for 2 hours before continuation. Six liters of sterile water were pumped to rinse the chemical and then 6 L of sterile Butterfield’s Phosphate Buffer were pumped through and collected into sterile bottles. Sterile 47mm water membrane (0.45 micron) mixed cellulose filters (Fisher Scientific, Pittsburg, PA) were used to filter all 6 L of collected buffer. The filters were placed on Lactobacillus MRS agar (Difco Laboratories, Sparks, MD) and incubated for 24 hr at 37 C. After incubation all filters showed zero growth, confirming the efficacy of the cleaning procedure. Second Cleanability Study Even though the first cleanability design was able to sanitize the equipment, it was determined that such a high concentration of Oxonia could damage the pumps’ plungers. Ecolab was contacted and a representative visited to help develop an appropriate cleaning schedule. With the help of Ecolab, a concentration of 0.5% Principal solution and a 0.28% Oxonia solution was developed to clean and protect the equipment. With an Ecolab chemical titration test kit, is was determined that 100 ml Oxonia in 24.6 L water created a solution with 2900 ppm concentration and 100 ml of Principal in 28.4 L water created a solution of 4700 ppm, and this was deemed acceptable. On a periodic basis, sterile water would be pumped through after cleaning and plated on 3M Aerobic Petrifilm to confirm that the equipment was indeed clean.

PAGE 50

37 High Pressure CO 2 Experimental Procedure For every experiment involving the Praxair high-pressure CO 2 equipment, the following procedure was used. First the equipment, was cleaned as described above the day before. Then on the day of the experiment the equipment was rinsed with 12 L of sterile deionized water. During the last 6 L of the wash, the parameters (pressure, % CO 2 , time, temperature) of the run were set up and a steady state achieved. Once steady state was achieved and the feed tank was empty, watermelon juice was added. The first 5 L (about 2 hold up volumes) of watermelon juice exiting the equipment were drained due to possible dilution effect from the rinse water. After the diluted juice was drained, a sample could be taken. Once the first sample was taken and the equipment parameters changed all juice exiting would go to drain. After steady state at the new parameters was achieved, 5 more liters of juice would be drained before a sample could be taken. This procedure of waiting for steady state and then draining the first 5 L of product was performed before taking any sample in all experiments. The samples would then be placed into sterile glass bottles and quickly placed into refrigerated storage. Pressure and Temperature Design The first series of experiments were designed to understand the relationship between changes in pressure and temperature and the reduction of total aerobic plate count and yeast and mold plate count. A full design was used to evaluate all 9 combinations of the following, pressures (10.3, 20.6, 34.4 MPa) and temperature (room temperature 24 C, 30 C, and 40 C). The retention time and % CO 2 remained constant during these experiments (5 min, and 10% CO 2 ), these were middle values and not believed to effect the microbial reduction. These experiments were conducted on aged juice (juice aged for 44 hours at room temperature to increase microbial load). The

PAGE 51

38 microbial load of the aged juice was about 10 9 aerobic CFU/ml. This series of nine experiments were duplicated on a separate day. Microbial analysis (total aerobic, and yeast and mold) was conducted on these samples. Carbon Dioxide Concentration and Retention Time Design The second series of experiments were designed to evaluate the effect of % CO 2 and retention time on the reduction of total aerobic plate count and yeast & mold plate count. During these experiments, pressure and temperature were held constant at the most lethal parameters found from the above experiment (see Pressure and Temperature Design), pressure set at 34.4 MPa and temperature set a 40 C. These experiments were conducted on aged juice. The microbial load of the aged juice was about 10 9 aerobic CFU/ml. Three CO 2 concentrations 5%, 10%, and 15% were examined. During the CO 2 concentration experiments, pressure, temperature, and retention time were held constant (34.4 MPa, 40 C, and 5 min). Microbial analysis (total aerobic, and yeast and mold) was conducted on these samples. This experiment was duplicated. To test the effect of retention time on microorganisms, three hold times 4, 5, and 6 min were evaluated. The pressure, temperature and CO 2 concentration were held constant at 34.4 MPa, 40 C, and 10% CO 2 . Microbial analysis (total aerobic, and yeast and mold) was conducted on these samples. This experiment was duplicated. Pressure and Temperature Design Using Modified Watermelon Juice It was important to understand if the modified watermelon juice (added acid and sugar) would behave in the same manner as fresh (unchanged) watermelon juice. Pressure and temperature had the largest effect on the reduction of total aerobic counts

PAGE 52

39 with the unchanged watermelon juice. Therefore, this experiment was designed to test pressure and temperature on the modified watermelon juice. Acid and sugar were added as depicted in the Modified Watermelon Juice section. These experiments were conducted on aged juice. The addition of acid and sugar occurred before aging. A full design was used to evaluate all 9 combinations of the following pressures (10.3, 20.6, 34.4 MPa) and temperatures (ambient, 30 C, and 40 C). The retention time and % CO 2 remained constant during these experiments (5 min, and 10% CO 2 ). The microbial load of the aged juice was about 10 7 aerobic CFU/ml. This series of nine experiments were duplicated on a separate day. Microbial analysis (total aerobic, and yeast and mold) was conducted on these samples. Storage Study The storage study was designed to evaluate microbial and qualitative aspects of HPCD treated, pasteurized and untreated watermelon juice. The samples were held for a total of eight weeks. Microbial counts, Brix, pH, TA, flavor, color and lycopene were evaluated once a week. On weeks 0, 2, 4, 6 and 8, samples were evaluated for sensory data with 60 untrained panelists. Approximately 60 liters of modified watermelon juice were HPCD treated with the following conditions: 34.4 MPa, 40 C, 10% CO 2 , and 5 min retention time. About 50 L of modified watermelon juice were heat pasteurized (see Heat Pasteurization section). All juice was carbonated (see Carbonation section) and filled into 750 ml glass champagne bottles with metal crown tops and stored at 4 C inside cardboard boxes.

PAGE 53

40 Analysis of Treated Samples Microbial Evaluation Watermelon juice samples were plated on 3M Petrifilm Plates for aerobic count and yeast and mold (3M Microbiology, St. Paul, MN). NaOH 1N was added to raise the watermelon juice pH to 7.0 before plating on aerobic Petrifilm. The juice was serially diluted using 9 ml of Butterfield’s phosphate buffer. All dilutions were done in triplicate. Aerobic Petrifilm was incubated at 35 C for 48 hours. Yeast and mold was incubated at 25 C for 5 days. Brix All Brix measurement were determined with a Reichert Abbe Mark II refractometer (Buffalo, NY) and/or a Fisherbrand Handheld Refractometer with a 0 – 18 Brix scale (Fisher Scientific, Pittsburgh, PA). Equipment was calibrated using distilled water. Disposable plastic pipettes were used to deliver 2-3 drops of sample onto the prism. pH Metrohm 632 pH-Meter and Metrohm pH Probe (Brinkmann Instruments Co., Westbury, NY) was used to measure initial pH. The pH meter was calibrated with pH 4 and pH 7 standard solutions first on each test day. A 20 ml sample was placed in a 50 ml beaker with a magnetic stir bar and measurement was recorded after pH was stable. Titratable Acidity A Brinkmann Instruments Co. setup including Metrohm 655 Dosimat, Metrohm 614 Impulsomat, Metrohm 632 pH-Meter and Metrohm pH Probe (Brinkmann Instruments Co., Westbury, NY) was utilized for titratable acidity. Titratable acidity was determined using 0.1 N NaOH and titrating 20 ml of watermelon juice to an end point of

PAGE 54

41 pH 8.2. The percent acid values were based on malic acid (0.067 mg/meq). The following formula was used: Titratable acidity (malic acid g/L) = (mL of base needed)(N base)(0.067)(1000) / (mL sample). Color Color of juice was determined with the Colorgard 14 system (Colorgard system, BYK-Gardner Inc., Columbia, MD). Equipment was allowed to warm up for 10 min prior to use. For calibration, black and white tiles were used (BYK Gardner, white 94.31 L*, -0.92 a*, -0.50 b*, Black zero reference standard). The watermelon juice was well mixed before a sample was poured. A glass sample cup was filled with a 50 ml sample and placed on the sample port. The calibration white tile was placed white side down on the sample cup. Each sample was evaluated on an L (lightness), a* (redness), b* (yellowness) scale at time 0 and after sitting for 10 min. Between samples, the glass sample cup was cleaned with distilled water and Kimwipes. A 2854 K quartz halogen lamp was used as the light source for the Colorgrad 14 system. Flavor Gas chromatography-olfactometry (GC-O) was utilized to evaluate flavor profile of watermelon juice. At each test day during the storage study, 7 ml of sample was placed into 22 ml amber vials with Teflon septas and frozen at C until analyzed. The GC-O equipment included a HP-5890 GC (Palo Alto, CA) with a sniffing port (DATU, Geneva, NY) and a flame ionization detector (FID). Both a ZB-5 (30m x 0.32 mm ID x 0.5 m film thickness) from J&W Scientific (Folson, CA) and a Stabilwax column (30m x 0.32 ID x 0.5 m film thickness) were used to analyze the watermelon juice.

PAGE 55

42 The solid phase micro-extraction (SPME) technique was used for sample preparation. The SPME fiber used was a StableFlex Divinylbenzene/Carboxen/Polydimethylsiloxane (DVB/CAR/PDMS) (Supleco, Bellefonte, PA), which is designed for flavor analysis. The samples were heated to 40 C with stirring and the fiber was optimized to a 30 min exposure time in the vial’s headspace. The fiber was conditioned for 1 hour at 270 C before use. The operation parameters for the GC for both columns were as follows: injector port temperature of 240 C, detector port temperature of 250 C, initial oven temperature of 40 C, ramp rate of 7 C per min, with a hold time of 5 mins at the final temperature of 250 C. Alkane (C5 C20) standard solution (0.5 L) was also injected. The GC-O split the effluent between the FID detector (getting one third) and the olfactometer sniffer (getting two thirds). Two trained evaluators were used to sniff the eluting compounds. The evaluators were trained based on aroma active compounds in watermelon juice. Standard aroma mixtures were also used to exposed assessors to a variety of aromas and descriptors. Each evaluator analyzed each sample twice. Upon identifying an aroma each assessor had to determine its intensity. The intensity was inputted into the computer by a device, with a slider bar. The left side represented no aroma, the right side represented extreme aroma. Once an aroma was perceived the bar would be moved to the proper intensity level and the evaluator would manually write down the aroma descriptor and the time. During the evaluation of the results the aromagrams and chromatograms would be studied. A matching FID peak and aroma peak were tentatively identified using the aroma descriptors and the alkane standard with linear retention indexe (LRI) values.

PAGE 56

43 Watermelon samples were also analyzed using gas chromatography mass spectroscopy (GC-MS). A Perkin Elmer Clarus 500 GC and Clarus 500 MS with a DB-wax column (60m x 0.25mm ID x 0.5 um) and Turbomass software were utilized to evaluate and identify compounds. Sample preparation was conducted in the same way as the GC-O method. Sensory Panels Sensory panels were conducted at the University of Florida’s taste panel facility (Bldg 120, University of Florida, Gainesville, FL). The taste panel facilities contain 10 privacy booths with computers in each booth for data entry. Every other week (weeks 0, 2, 4, 6, and 8), 60 untrained panelists would evaluate the storage study watermelon juice. The panelists were presented with unsalted crackers and water to cleanse the palate before each sample. Four samples were given to each panelist; a fresh thawed control (as a reference), hidden control, HPCD treated, and a heat pasteurized sample. The samples were randomly assigned codes and they were placed in different orders (all possible combinations) on a white tray. The panelists were asked a few demographic questions, and then asked to determine the difference from control on aroma, taste and carbonation. The difference from control tests used a continuous line scale with values from 0-15, 0 meaning no difference and 10 meaning extreme difference. An overall hedonic likeability test was conducted as well, on a scale of 1 to 9, 1 being extremely dislike and 9 being extremely like. All juice bottles were held in the refrigerator until opened for pouring into sample cups. About 30ml of juice per sample cup was used. After pouring the samples were immediately served. Once the bottles were opened, they were placed on ice to retain the

PAGE 57

44 proper temperature and carbonation level. The lights were controlled inside the booths, red bulbs were used to eliminate any possible difference because of sample color. Compusense five system (Compusense, Guelph, Ontario, Canada) was used for data entry and to analyze the results. A Tukey’s mean separation was conducted on all data. Lycopene The determination of lycopene was conducted using a high performance liquid chromatograph (HPLC) setup consisting of a Waters 2695 separations module (Franklin, MA) and a Waters 997 photodiade array detector (Franklin, MA) with a C 30 column (Prontosil 200-5-C30, 5.0m, 4.6 x 250mm, Chadds Ford, PA). To prepare the watermelon samples, the following procedure was used. Watermelon juice was mixed thoroughly with sonication for 30 min and a 30 ml sample was placed into 35 ml centrifuge tubes. Sample was spun for 15 min at 7000g using a Beckman Coulter J2-HC centrifuge (Fullerton, CA). The supernatant was removed and sample (pellet) washed with methanol. Methanol wash was used to remove residual sugars. After methanol wash, the sample was centrifuged again for 15 min at 7000g, and the methanol was poured off. Methylene chloride was used to transfer the pellet to a glass tube. Methylene chloride was blended with the pellet and homogenized (Powergen 700D, Fisher Scientific, Pittsburgh, PA) at 3600 rpm for 30 sec. The mixture was allowed to stand to separate the layers. The bottom layer was placed in a 25 ml volumetric flask. The extraction was repeated until the volumetric flask final volume was 25 ml. The 25 ml sample was then poured into a 100 ml beaker and about 3 g of sodium sulfate was added to remove water. Then 10 ml of sample was added to a 50 ml beaker and the methylene chloride was evaporated with nitrogen using a N-EVAP 116

PAGE 58

45 (Organomation Associates, Inc., Berlin, MA). A few drops of Tetrahydrofuran were added to the beaker, and the sample was transferred to a 5 ml volumetric flask. The beaker was washed repeatedly with mobile phase (70: 30, Methanol: tert-butyl methyl ether) without ethyl acetate and added to 5 ml volumetric flask until 5 ml was achieved. The 5 ml sample was loaded into a syringe and filtered into a HPLC vial. The vials were then loaded into the HPLC autosampler carousel. The HPLC auto sampler injected 20 l and a mobile phase of 50% methanol, 40% tert-butyl methyl ether and 10% ethyl acetate was pumped through the column at 1 ml/min isocratically (Ishida, 2001). The equipment was setup with a sample chamber temperature of 5 C and a 28 C column temperature. The detection max peak was set at 450nm.

PAGE 59

CHAPTER 4 RESULTS AND DISCUSSION Microbial Reduction Experiments Effects of Pressure and Temperature Aging the juice increased the watermelon juice total aerobic plate count to about 9 logs per ml. Three temperatures (ambient, 30 C, 40 C) and three pressures (10.3, 20.6, 34.4 MPa) were evaluated while other parameters were held constant at 10% CO 2 and 5 min residence time. Ambient temperature was about 22 C. The effect of the treatments can be seen in Table 4-1. Total log reductions ranged from 0.08 to a maximum of 6.82. The effects of temperature appear to be larger than those of pressure. Increasing temperature above room temperature to 30 C increased the kill by about 2 logs. Another 2 logs were achieved by increasing from 30 C to 40 C. Table 4-1. Effects of pressure and temperature on aerobic organisms survivability of aged watermelon juice. Percent CO 2 was constant at 10% while residence time was constant at 5 min Rep 1 Rep 2 Aerobic Log Counts* Log Reduction Aerobic Log Counts* Log Reduction Control Juice 9.29 0.02 9.03 0.01 10.3 MPa, room temperature 9.18 0.08 0.1 6.56 0.06 2.5 20.6 MPa, room temperature 8.06 0.07 1.2 6.16 0.04 2.9 34.4 MPa, room temperature 7.16 0.03 2.1 5.94 0.02 3.1 10.3 MPa, 30 C 5.71 0.04 3.6 4.67 0.04 4.4 20.6 MPa, 30 C 5.02 0.03 4.3 4.66 0.05 4.4 34.4 MPa, 30 C 4.26 0.01 5.0 4.35 0.03 4.7 10.3 MPa, 40 C 2.77 0.02 6.6 2.94 0.01 6.1 20.6 MPa, 40 C 2.59 0.04 6.7 2.81 0.03 6.2 34.4 MPa, 40 C 2.48 0.01 6.8 2.74 0.02 6.3 * Mean of three values Std Error 46

PAGE 60

47 Figure 4-1 shows that the slope of changing temperature is much steeper than the slope of changing pressure. This further illustrates the increased effect of temperature. The maximum log reduction on total aerobic counts was 6.82 during the first repetition at 34.4 MPa and 40 C. 213040 10.3MPa20.6MPa34.4MPa 01234567Log ReductionTemperature (C)Pressure (MPa) 6-7 5-6 4-5 3-4 2-3 1-2 0-1 Figure 4-1. Effect of heat and pressure on the log reduction of total aerobes The effects of pressure and temperature were also evaluated on native yeast and mold counts. The same juice above, which was aged and evaluated, also was enumerated for yeast and mold. After the aging step, the total yeast and mold could was about 2000 cfu/ml. Table 4-2 shows the counts and log reductions on yeast and mold for all treatment conditions. No trend was found between treatments and a one-log reduction could be obtained. Because the initial juice load had such high total aerobic counts and

PAGE 61

48 so few yeast and mold counts the fact that only a one-log reduction was achieved on yeast and mold is probably of little concern. Table 4-2. Effects of pressure and temperature on yeast and mold survivability of aged watermelon juice. Percent CO 2 was constant at 10% while residence time was constant at 5 min Rep 1 Rep 2 Yeast & Mold Log Reduction Yeast & Mold Log Reduction Control Juice 2200 328 1730 266 10.3 MPa, room temperature 780 29 0.5 88 18 1.3 20.6 MPa, room temperature 38 15 1.8 153 19 1.1 34.4 MPa, room temperature 135 18 1.2 100 21 1.2 10.3 MPa, 30 C 85 12 1.4 115 6 1.2 20.6 MPa, 30 C 90 9 1.4 125 13 1.1 34.4 MPa, 30 C 4 1 2.7 155 35 1.1 10.3 MPa, 40 C 145 21 1.2 33 5 1.7 20.6 MPa, 40 C 58 16 1.6 40 11 1.6 34.4 MPa, 40 C 168 13 1.1 78 22 1.4 * Mean of three values Std Error Effects of Changing Levels of Carbon Dioxide It was anticipated that the effects of different percent added CO 2 would be minimal at best. Therefore, a limited experiment was conducted instead of a full factorial design. Table 4-3 below shows that changes in percent CO 2 (5%, 10%, 15%) had minimal effect when the other conditions were held at 34.4 MPa, 40 C and 5 min residence time. Again the watermelon juice was aged to achieve higher microbial numbers. Temperature and pressure parameters were chosen based on the most effective conditions found in previous work, (Table 4-1).

PAGE 62

49 Table 4-3. Effect of % CO 2 on aerobic organisms survivability of aged watermelon juice. Temperature was constant at 40 C, pressure was constant at 34.4 MPa, and residence time was at 5 min Rep 1 Rep 2 Aerobic Counts Log Reduction Aerobic Counts Log Reduction Control Juice 1.04E+09 4.17E+07 1.04E+09 4.17E+07 5% CO 2 5400 400 5.3 870 70 6.0 10% CO 2 4900 270 5.3 840 58 6.1 15% CO 2 5400 200 5.3 2500 300 5.6 * Mean of three values, Std Error The effect of percent CO 2 on the reduction of yeast and mold can be seen in Table 4-3. The aged control juice had a very low count of total yeast and mold per ml. The changing levels of CO 2 did not have much effect. Because the total counts were so low it is difficult to evaluate whether changing percent CO 2 has an impact on yeast and mold. Table 4-3. Effect of % CO 2 on yeast and mold survivability of aged watermelon juice. Temperature was constant at 40 C, pressure was constant at 34.4 MPa, and residence time was at 5 min Rep 1 Rep 2 Yeast & Mold Log Reduction Yeast & Mold Log Reduction Control Juice 250 29 250 9 5% CO 2 285 22 -0.1 13 5 1.3 10% CO 2 215 16 0.1 158 24 0.2 15% CO 2 922 43 -0.5 323 22 -0.1 * Mean of three values Std Error Effects of Changing Residence Time The effect of residence time on aged juice was also of importance. To reduce the variables and model size, a small study of three different residence times (4, 5, 6 min) was conducted. The other parameters were set at 34.4 MPa, 40 C and 10% CO 2 . As shown in Table 4-4, little effect was seen by increasing residence time from 4 min to 6 min. This is more than likely due to the fact that in a well mixed (i.e. turbulent flow)

PAGE 63

50 hold tube, time is not a critical factor once a minimum time is reached for saturation. In our case 5 min was chosen for safety and economical reasons. Table 4-4. Effect of retention time on aerobic organisms survivability of aged watermelon juice. Temperature was constant at 40 C, pressure was constant at 34.4 MPa, and % CO 2 was constant at 10% Rep 1 Rep 2 Aerobic Counts Log Reduction Aerobic Counts Log Reduction Control Juice 1.04E+09 4.17E+07 1.04E+09 4.17E+07 4 min residence time 7380 660 5.2 570 59 6.3 5 min residence time 4800 270 5.3 840 59 6.1 6 min residence time 870 51 6.1 310 37 6.5 * Mean of three values Std Error The effect of increasing residence time on the reduction of yeast and mold counts can be found in Table 4-5. Again, we observe low initial yeast and mold counts, therefore it was difficult to evaluate the effect of residence time. With the watermelon juice, yeast and mold do not seem to be a point of concern. Table 4-5. Effect of retention time on yeast and mold survivability of aged watermelon juice. Temperature was constant at 40 C, pressure was constant at 34.4 MPa, and % CO 2 was constant at 10%, Rep 1 Rep 2 Yeast & Mold Log Reduction Yeast & Mold Log Reduction Control Juice 250 29 250 29 4 min residence time 1160 40 -0.7 90 20 0.4 5 min residence time 215 16 0.1 157 24 0.2 6 min residence time 60 11 0.6 15 3 1.2 * Mean of three values Std Error Effects of Pressure and Temperature on Modified Watermelon Juice The watermelon juice was modified to maintain safety for the storage and taste panel studies. It was necessary to reduce the pH from 5.6 to 4.3 with malic acid. This reduction of pH increased the level of sourness to an unacceptable level, therefore high fructose corn syrup which increased Brix from 8.4 to 10.5 was used. This modified

PAGE 64

51 juice was aged in the same manner as the unmodified juice. However, the modified juice did not reach the same total aerobic cfu per ml level than the unmodified juice. This is attributed to the potential lag as the native microorganisms readjusted to the new environment. After aging, the total aerobic count was about 6 logs per ml. The effect of temperature (ambient, 30 C, 40 C) and pressure (10.3, 20.6, 34.4 MPa) on log reduction can be seen in Table 4-6. All the combinations had about the same influence on total aerobic reduction, about 4.5 logs. It is unclear why in the non-modified juice the temperature and pressure had a large effect while in the modified juice the lowest pressure and temperature had the same effect as the highest temperature and pressure. Microorganisms in their growth phase are more susceptible to environmental stress, and it is possible that these microorganisms were in their growth phase. Also, the initial yeast and mold count was much higher in the modified juice than in the unmodified juice. Therefore, we know that the microbial population is different here than in the non-modified juice. There could be vastly different populations and sub-populations that grew and out competed others at this lower pH juice. Table 4-6. Effects of pressure and temperature on aerobic organisms survivability of modified aged watermelon juice. Percent CO 2 was constant at 10% and residence time was constant at 5 min Rep 1 Rep 2 Aerobic Counts* Log Reduction Aerobic Counts* Log Reduction Control Juice 2580000 320000 7800000 560000 10.3 MPa, room temperature 268 28 4.5 208 21 4.6 20.6 MPa, room temperature 225 24 4.6 213 18 4.6 34.4 MPa, room temperature 243 41 4.5 163 26 4.7 10.3 MPa, 30 C 250 29 4.5 175 29 4.7 20.6 MPa, 30 C 850 120 4.0 285 38 4.4 34.4 MPa, 30 C 245 13 4.5 228 22 4.5 10.3 MPa, 40 C 508 36 4.2 190 12 4.6 20.6 MPa, 40 C 208 19 4.6 178 29 4.6 34.4 MPa, 40 C 740 39 4.0 155 21 4.7 * Mean of three values Std Error

PAGE 65

52 In Table 4-7, we see the effect of pressure and temperature on the reduction of yeast and mold counts in modified watermelon juice. The modified aged juice was able to promote the growth of higher concentrations of yeasts than previously found in non-modified juice. The lower pH might be giving the yeasts a competitive edge over other native organisms. From the results, the yeast and mold counts can be reduced by as much as 4 logs, however no trend was determined. Table 4-7. Effects of pressure and temperature on yeast and mold survivability of modified aged watermelon juice. Percent CO 2 was constant at 10% and residence time was constant at 5 min Rep 1 Rep 2 Yeast & Mold* Log Reduction Yeast & Mold* Log Reduction Control Juice 1550000 114000 55000 7200 10.3 MPa, room temperature 113 4 4.1 233 45 2.4 20.6 MPa, room temperature 80 8 4.3 280 45 2.3 34.4 MPa, room temperature 223 40 3.8 218 17 2.4 10.3 MPa, 30 C 255 10 3.8 353 81 2.2 20.6 MPa, 30 C 2325 165 2.8 295 35 2.3 34.4 MPa, 30 C 80 14 4.3 235 25 2.4 10.3 MPa, 40 C 385 12 3.6 263 35 2.2 20.6 MPa, 40 C 105 12 4.2 270 31 2.3 34.4 MPa, 40 C 3775 180 2.6 248 38 2.4 * Mean of three values Std Error Microbial Growth During Storage Study It was determined to use the best combinations of parameters to treat watermelon juice for the storage study. The best conditions for non-modified juice occurred at 34.4 MPa and 40 C. All pressure-temperature combinations on the modified juice had the same values. Therefore, the parameters where set at 34.4 MPa, 40 C, 10% CO 2 , and 5 min residence time for the HPCD treated juice. The flash pasteurized juice was treated at 74 C for 15 sec.

PAGE 66

53 In Table 4-8, the total aerobic counts during each week are shown. Week 0 shows the difference between each treatment. Between 2 and 3 log reductions of total aerobic counts were achieved by both the HPCD and the heat pasteurized juice. It is believed that because of difficult transportation, facilities and equipment, contamination occurred after processing in both treatments. After processing, the juice needed to be transported from processing equipment to carbonator and then bottled. These steps add additional sources of contamination. Every effort in maintaining sanitation was made to minimize this effect. However, contamination would most likely affect each sample in the same way. Table 4-8. Total aerobic growth during the 8 week storage study: untreated week 0 represents the control Week Untreated* HPCD Treated* Pasteurized* 0 223333 73333 747 17 330 32 1 66667 7753 710 191 610 163 2 45667 5812 197 18 147 28 3 19000 3215 233 22 150 6 4 9500 1343 193 26 157 23 5 N/A 170 21 230 10 6 N/A 160 10 140 6 7 N/A 180 6 177 17 8 N/A 340 12 447 227 * Mean of three values Std Error The total aerobic population remained steady in the HPCD and heat pasteurized juices for the duration of the study. The untreated juice total aerobic counts actually decreased during storage (Figure 4-2). It is believed that because all the juice was carbonated and stored at 4 C until evaluation, that the growth of surviving microorganisms would be hindered. The addition of acid and then the removal of oxygen by carbonation may have lead the native organisms into a state where their growth rates were very slow.

PAGE 67

54 11010010001000010000010000000246810weekscfu/ml HPCD Pasteurized untreated Figure 4-2. Total aerobic during storage study Yeast counts in HPCD and heat pasteurized juice dropped by about 1 log just after processing. The yeast counts remained fairly stable throughout the study with small fluctuations, believed to be caused by sampling errors (Table 4-9). Table 4-9. Yeast and mold growth during the 8 week storage study: untreated week 0 represents the control Week Untreated* HPCD Treated* Pasteurized* 0 4300 557 250 30 703 23 1 6500 436 923 321 2600 203 2 6533 549 2467 88 1133 784 3 4767 285 700 176 2033 58 4 4800 436 3133 84 973 240 5 N/A 1200 41 1183 208 6 N/A 1433 173 1100 203 7 N/A 1500 33 767 58 8 N/A 1800 88 1333 58 * Mean of three values Std Error From Figure 4-3, the fluctuations can be seen; however the trend is a flat line. Again, the suppression of growth is most likely caused by the refrigerated storage and lack of oxygen.

PAGE 68

55 1101001000100000246810time (weeks)counts per ml, Y&M HPCD Pasteurized Untreated Figure 4-3. Yeast count during storage study Effect of High Pressure Carbon Dioxide on Physical Properties During the storage study, samples were evaluated for Brix, pH, titratable acidity, flavor, aroma, color and lycopene concentration. HPCD treated samples processed at 34.4 MPa, 40 C, 10% CO 2 and 5 min residence time were compared to typical flash pasteurized and an untreated watermelon juices. Effect on Brix Table 4-10 shows the Brix for HPCD, heat pasteurized and untreated samples throughout the storage study. HPCD and heat pasteurized treatments did not have an effect on watermelons Brix values.

PAGE 69

56 Table 4-10. Change in Brix during the 8 week storage study: untreated week 0 represents the control Week Untreated* HPCD Treated* Pasteurized* 0 10.4 0.03 10.4 0.03 10.5 0.03 1 10.5 0.03 10.4 0.03 10.4 0.03 2 10.4 0.03 10.3 0.03 10.5 0.03 3 10.4 0.03 10.4 0.03 10.4 0.03 4 10.5 0.03 10.5 0.03 10.4 0.00 5 N/A 10.4 0.03 10.4 0.00 6 N/A 10.5 0.03 10.4 0.00 7 N/A 10.5 0.00 10.4 0.03 8 N/A 10.5 0.03 10.4 0.00 * Mean of three values Std Error Figure 4-4 shows a graphical version of the data above. This is as expected; previous research has found that supercritical CO 2 and high pressure CO 2 do not affect the Brix values (Arreola et al., 1991b; Kincal, 2000). 99.51010.51111.5120246810time (weeks)Brix HPCD Pasteurized Untreated Figure 4-4. Brix during storage study Effect on pH Table 4-11 shows the pH values for each sample during the storage study. At week 0, the difference between treatments is negligible, therefore HPCD and heat

PAGE 70

57 pasteurization did not have an affect on watermelon juice pH. Throughout the 8-week storage study, the pH did not change. Table 4-11. Change in pH during the 8 week storage study: untreated week 0 represents the control Week Untreated* HPCD Treated* Pasteurized* 0 4.35 0.003 4.32 0.015 4.35 0.003 1 4.26 0.000 4.21 0.003 4.29 0.003 2 4.24 0.000 4.22 0.003 4.24 0.003 3 4.34 0.003 4.33 0.003 4.34 0.006 4 4.32 0.006 4.33 0.003 4.32 0.003 5 N/A 4.33 0.005 4.31 0.003 6 N/A 4.31 0.007 4.32 0.003 7 N/A 4.32 0.012 4.30 0.000 8 N/A 4.29 0.006 4.30 0.003 * Mean of three values Std Error Figure 4-5 below shows a graphical version of the data above. This is as expected; previous research has found that supercritical CO 2 and high pressure CO 2 do not affect the pH value of juices (Arreola et al., 1991b; Kincal, 2000). 3.53.73.94.14.34.54.74.95.15.35.50246810time (week)pH HPCD Pasteurized Untreated Figure 4-5. pH during storage study

PAGE 71

58 Effect on Titratable Acidity Table 4-12 shows the titratable acidity (TA) values of HPCD, heat pasteurized and the untreated sample at week 0 and throughout an 8-week storage study. The titratable acidity for the HPCD and heat pasteurized sample is the same as the untreated sample at week 0, about 0.28% malic acid equivalent. Table 4-12. Titratable acidity (% malic acid) during the 8 week storage study: untreated week 0 represents the control Week Untreated* HPCD Treated* Pasteurized* 0 0.29 0.006 0.28 0.002 0.28 0.006 1 0.29 0.002 0.29 0.004 0.28 0.001 2 0.28 0.002 0.28 0.002 0.29 0.003 3 0.30 0.004 0.31 0.002 0.29 0.001 4 0.31 0.005 0.32 0.003 0.30 0.004 5 N/A 0.30 0.001 0.28 0.004 6 N/A 0.29 0.003 0.29 0.003 7 N/A 0.29 0.001 0.29 0.002 8 N/A 0.32 0.002 0.32 0.004 * Mean of three values Std Error Figure 4-6 shows that the titratable acidity fluctuates somewhat over the length of the storage study. However, no trend is perceived and on any given test week the treatments were very close in value. Therefore, the fluctuation is most likely due to sample variation, and not treatment or storage effects. Kincal (2000) found that after treatment titratable acidity values increased in orange juice. However, research conducted by Arreola et al. (1991b) found that treated juice had a significant decrease in titratable acidity.

PAGE 72

59 0.1500.2000.2500.3000.3500.4000.4500246810Time (week)% Malic Acid HPCD Pasteurized Untreated Figure 4-6. Titratable acidity (% malic acid) during storage study Effect of High Pressure Carbon Dioxide on Watermelon Juice Aroma and Flavor Sensory Panel Evaluation Difference from control tests were used to understand the effect of HPCD on watermelon juice. Consumer taste panels where conducted on the following intervals during the storage study: week 0, 2, 4, 6, and 8. Each panel consisted of 60 untrained panelists, which were different from week to week. Panelists where asked to rate the amount of difference from a reference (i.e. the control). The scale used was a 0 to 15 intensity line scale where 0 = no difference and 15 = extreme difference. Raw data can be found in the appendix A. Sample ballot can be found in appendix B. Figure 4-7 shows the difference from control aroma evaluation throughout the storage study. After processing, the HPCD juice and the heat pasteurized sample had the same difference from control. This represents that the magnitude of difference was equal, however it does not specify that the cause of the difference was the same. Throughout the storage study the means fluctuated slightly, always staying below a

PAGE 73

60 difference intensity of 4 on a scale that goes to 15. This means that there were small differences from the reference control. The panelists were also unaware that there was a hidden control among the three samples. If they were informed of this the treatment means would have been even lower. At the end of the storage study, week 8, the panelist could not differentiate the hidden control from either treated sample. 012345678910111213141502468Time (Week)Difference Intensity Hidden HPCD Heat Figure 4-7. Aroma evaluation of the hidden control, HPCD treated and heat pasteurize treated samples throughout the storage study (Tukey’s separations labeled) Similar results were found during the difference from control flavor evaluation (Figure 4-8). Initially the HPCD and heat pasteurized treatments rated differently than the hidden control. For the most part the treatments were the same and different from the control until week 8 where all three were the same. Again the magnitude of difference is very small considering the hidden control (with an average score of 3) has exactly the same value.

PAGE 74

61 012345678910111213141502468Time (week)Difference Intensity Hidden HPCD Heat Figure 4-8. Flavor evaluation of the hidden control, HPCD treated and heat pasteurize treated samples throughout the storage study (Tukey’s separations labeled) Overall likeability was evaluated during the storage study. Panelist were asked to rate the juice on a scale of 1 to 9, 1 = extreme dislike and 9 = extreme like. Figure 4-9 shows the effect of storage time on the mean values of all treatments in a given week. As no trend was apparent during the eight weeks, it is believed that storage time had little influence on likeability. Since week 8 has the lowest rating, it is possible that it is the start of a downward trend but without further information this cannot be determined.

PAGE 75

62 012345678902468Time (week)Hedonic Rating Figure 4-9. Hedonic rating for all treatments at each time period through storage (Duncan’s Multiple Range Test) Each week the panelists chose the hidden control as the most liked product, as expected. However the HPCD and heat pasteurized samples went back and forth for second place (Figure 4-10). The treatment likeability ratings are very close to the untreated hidden control, which shows that both the heat pasteurization and HPCD treatment have very little effect on overall likeability. It is more than likely that given a longer storage study the two treatments would have spread farther from the hidden control.

PAGE 76

63 Overall Likeability of each Treatment During Storage012345678902468Time (week)Hedonic Rating Hidden HPCD Heat Figure 4-10. Overall likeability of each treatment during storage (Tukey’s mean separation labeled) Flavor Evaluation by Gas Chromatography-Olfactometry and Mass Spectrometry Watermelon samples where evaluated using SPME on both a ZB-5 (30m x 0.32 mm ID x 0.5 m film thickness) and Stabilwax (30m x 0.32 mm ID x 0.5 m film thickness) column with an FID detector and a sniff port. Upon evaluating both columns, it was clear that the Stabilwax separated the aroma peaks more than the ZB-5. Therefore, the bulk of the research was conducted on the wax column. Odor descriptors, retention times of aroma and FID active compounds of watermelon juice for the wax column are listed on Table 4-13. Two trained assessors developed these descriptors during training and analysis. Raw data can be found in appendix A.

PAGE 77

64 Table 4-13. Retention times, aroma descriptors of watermelon juice Retention time (min) A ssessors Descriptor on Stabilwax 8.2 Gra ssy-green 11.4 Mush room, Rindy 12.3 Fruity-citric 13.3 Fruity Watermelon 14.3 Rindy-Floral 14.5 Garli c-Onion-Rindy 15.43 Fruity Watermelon 16.5 Earthy, Rindy 17.1 Rindy 17.95 gra ssy-Green 19.36 Tra shy 19.9 Rindy-green 20.47 Fruity, Honey Eleven of the thirteen aroma active compounds were identified using linear retention indices (LRI) and mass spectro scopy. Each aroma active compound had a calculated LRI value and the assessor’s descriptors. These two pieces of information can be looked up on flavor databases for possibl e tentative identification. After tentative identification GC-MS can be used to confirm the identification. In Table 4-14, the compounds in red have been confirmed using mass spectroscopy.

PAGE 78

65 Retention time My Descriptor LRI Value Flavor Database** Database Descriptor* LRI Wax* 8.2 Grassy-green 1098 Hexanal (913) fatty, green, grassy, powerful 1092 11.4 Mushroom, Rindy 1271 E-2-Hexenal (899) green, banana-like 1236 12.3 Fruity -citric 1317 Octanal (850) fatty, tallowy, citrus-like 1302 13.3 Fruity Watermelon 1368 6-Methyl-5-hepten-2-one (937) 14.3 Rindy-Floral 1419 Nonanal (940) p iney, floral, citrusy 1409 14.5 Garlic-Onion-Rindy 1429 ethyl octanoate fruity,floral 1444 15.43 Fruity Watermelon 1477 6-Nonenal (931) floral 1479 16.5 Earthy, Rindy 1533 mercapto-4 methylpentan-2-ol,4 grapefruit, flowery 1534 17.1 Rindy 1565 E-2-Nonenal (934) fatty, tallowy, soapy 1560 17.95 grassy-Green 1612 E,Z-2,6-Nonadienal (923) cucumber, green 1611 19.36 Trashy Rindy 1694 Z-3-Nonenol (937) 19.9 Rindy-green 1727 Z-6-Nonenol (912) 20.47 Fruity, Honey 1762 E,Z-3,6-Nonadienol (897) Table 4-14. Aroma compounds in watermelon juice, Compounds in red confirmed by GC-MS * Citrus Flavor and Color, Flavor Database ** Goodness of identification based on 1000. Anything over 700 is good, 800 is great and over 900 is a sure thing

PAGE 79

66 Figure 4-11 displays the peak areas of the aroma active compounds in untreated watermelon juice at Week 0 and Week 4. No t much significant change occurred during those 4 weeks. An increase in 6-methyl-5-hepten-2-one was seen from the untreated watermelon juice in Week 4. The trained asse ssors described this as a fruity watermelon smell. Since the watermelon juice remained at 4C and bottled under carbon dioxide, it is possible that chemical and biological reactions would be slowed. However for safety concerns, the original experiment was not designed to have the untreated juice taste panel and therefore samples were only monitored for 4 weeks. 0100000200000300000400000500000600000700000800000900000HexanalE-2-HexenalOctanal6-Methyl-5-hepten-2-oneNonanal6-NonenalE-2-NonenalE,Z-2,6-NonadienalZ-3-NonenolZ-6-NonenolE,Z-3,6-NonadienolPeak Area FID Week 0 FID Week 4 Figure 4-11. FID peak areas for the untreated watermelon juice at 0 and 4 weeks The peak areas from the aromagrams were also evaluated to investigate the possible trends. Figure 4-12 shows the peak areas of untreated watermelon juice at Weeks 0 and 4. In general, the aroma active compounds in week zero had larger peak areas, except for

PAGE 80

67 octanal and 6-methyl-5-hepten-2-one. Both compounds also had a higher FID peak area to match. Octanal was described as having a citrus like aroma. 0500000100000015000002000000250000030000003500000HexanalE-2-HexenalOctanal6-Methyl-5-hepten-2-oneNonanal6-NonenalE-2-NonenalE,Z-2,6-NonadienalZ-3-NonenolZ-6-NonenolE,Z-3,6-NonadienolPeak Area Untreated week 0 Untreated week 4 Figure 4-12. Aroma peak areas for the untreated watermelon juice at 0 and 4 weeks The HPCD treated juice’s aroma active chemical distribution over storage time was also very stable (Figure 4-13). The FID detector determined that many compounds had roughly the same concentrations throughout the storage study. Two compounds could not be detected by FID on weeks 4 and 8: ethyl octanoate and mercapto-4 methlypentan-2-ol,4. However both those compounds’ identification could not be confirmed on GC-MS, therefore their impact is unknown. On week 8, z-6-nonenol could not be detected. The assessors described z-6-nonenol as a rindy green aroma.

PAGE 81

68 Figure 4-14 shows the aroma peak areas for the HPCD watermelon juice at week 0, 4 and 8. In general, there seemed to be a reduction of intensity as the juice aged. The perceived intensity of 6-nonenal was much higher on week 0 than on 4 or 8. 6-Nonenal has a very strong watermelon floral aroma. 050000100000150000200000250000300000350000400000450000HexanalE-2-HexenalOctanal6-Methyl-5-hepten-2-oneNonanalethyl octanoate6-Nonenalmercapto-4 methylpentan-2-ol,4E-2-NonenalE,Z-2,6-NonadienalZ-6-NonenolE,Z-3,6-NonadienolPeak Area Week 0, FID Week 4, FID Week 8, FID Figure 4-13. FID peak area for the HPCD treated watermelon juice at weeks 0, 4 and 8

PAGE 82

69 050000010000001500000200000025000003000000350000040000004500000HexanalE-2-HexenalOctanal6-Methyl-5-hepten-2-oneNonanalethyl octanoate6-Nonenalmercapto-4 methylpentan-2-ol,4E-2-NonenalE,Z-2,6-NonadienalZ-6-NonenolE,Z-3,6-NonadienolPeak Area Week 0 Aroma Week 4 Week 8 Figure 4-14. Aroma peak area for the HPCD treated watermelon juice at weeks 0, 4 and 8 The FID and aroma intensity for the heat pasteurized juice had some fluctuation (Figures 4-15 and 4-16). No trends could be determined; however there were some changes. 6-Methyl-5-hepten-2-one was larger in week 4 than week 0, but on week 8, it could not be detected. 6-nonenal was significantly lower in week 0 than week 4 or 8. Both of those observation were backed up by the aroma intensity data from Figure 4-16. Week 0 heat pasteurized watermelon juice lacked e,z-2,6-nonadienal (grassy-green), z-6-nonenol (rindy-green), and e,z-3,6-nonadienol (fruity); however on week 4 and 8, these compounds could be found. The aroma intensity from Figure 4-16 follow the same trends as the FID peak areas in Figure 4-15.

PAGE 83

70 0100000200000300000400000500000600000700000800000HexanalE-2-HexenalOctanal6-Methyl-5-hepten-2-oneNonanal6-Nonenalmercapto-4 methylpentan-2-ol,4E-2-NonenalE,Z-2,6-NonadienalZ-3-NonenolZ-6-NonenolE,Z-3,6-NonadienolPeak Area Heat Week 0, FID Heat Week 4, FID Heat Week 8, FID Figure 4-15. FID peak area for the heat pasteurized juice at weeks 0, 4 and 8 0500000100000015000002000000250000030000003500000HexanalE-2-HexenalOctanal6-Methyl-5-hepten-2-oneNonanal6-Nonenalmercapto-4 methylpentan-2-ol,4E-2-NonenalE,Z-2,6-NonadienalZ-3-NonenolZ-6-NonenolE,Z-3,6-NonadienolPeak Area Heat week 0, Aroma Week 4 week 8 Figure 4-16. Aroma peak area for the heat pasteurized juice at weeks 0, 4, and 8

PAGE 84

71 The FID peak area for each treatment at week 0 was graphed in Figure 4-17. This figure was designed to show any difference caused by the HPCD or the heat pasteurization process. The data shows that HPCD FID peak areas match closely to the untreated results. HPCD has two extra small peaks that the untreated does not have, the unconfirmed compounds, ethyl octanoate and mercapto-4 methlypentan-2-ol,4. Also, HPCD is lacking the z-3 nonenol compound, which was an unpleasant rindy-trashy aroma. The heat pasteurized treatment had larger FID peak areas than the untreated or HPCD treated juice for 6-nonenal, e-2-nonenal, e,z-2,6 –nonadienal and z-6-nonenol. The heat pasteurized juice’s FID peak areas did not match the untreated as closely as the HPCD samples. The aroma intensity peak areas for each treatment at week 0 can be seen in Figure 4-18. 0100000200000300000400000500000600000700000800000HexanalE-2-HexenalOctanal6-Methyl-5-hepten-2-oneNonanalethyl octanoate6-Nonenalmercapto-4 methylpentan-2-ol,4E-2-NonenalE,Z-2,6-NonadienalZ-3-NonenolZ-6-NonenolE,Z-3,6-NonadienolPeak Area Untreated FID CO2, FID Heat, FID Figure 4-17. FID peak areas at week 0 for all three treatments

PAGE 85

72 050000010000001500000200000025000003000000350000040000004500000HexanalE-2-HexenalOctanal6-Methyl-5-hepten-2-oneNonanalethyl octanoate6-Nonenalmercapto-4 methylpentan-2-ol,4E-2-NonenalE,Z-2,6-NonadienalZ-3-NonenolZ-6-NonenolE,Z-3,6-NonadienolPeak Area Untreated Aroma CO2 Treated Aroma Heat treated Aroma Figure 4-18. Aroma peak area at week 0 for all three treatments Effect of High Pressure Carbon Dioxide on Color The CIE color scale was used for the watermelon juice color measurements. The L* value represents the “lightness” scale from 0 black to 100 white, a* value represents the “redness greeness”, and the b* values represents the “yellowness blueness” (Figure 4-19).

PAGE 86

73 Figure 4-19. Visual representation of the CIE color scale During the measurement of the L*, a*, b* values, it was noticed that the watermelon’s red sediment was not uniform among samples. The red sediment settled out in a few hours if left undisturbed. Therefore, it was necessary to blend two sample bottles (from each test day) together. However, the heat pasteurized juice samples still had high levels of variation. The untreated sample was not measured after week 4. Raw data can be found in appendix A. Figure 4-20 shows the L* values for each sample during the storage study. Note that after week 0, the HPCD treated juice levels off and stays steady at about 31. The heat pasteurized juice fluctuates more but on average it has a value of 36. A higher value on the lightness scale implies a paler product.

PAGE 87

74 051015202530354045500246810Time (week)L* Value CO2 HEAT untreated Figure 4-20. Watermelon juice L* after treatment and during storage The a* value for HPCD treated sample was steady at about 27 and maintained that level throughout the storage study (Figure 4-21). The heat pasteurized juice varied widely but on average its values were around 19. The untreated sample started higher than the HPCD but then leveled off at the same value. 051015202530350246810time (week)a* Value CO2 Heat Untreated Figure 4-21. Watermelon juice a* after treatment and during storage Just after treatment and during storage, the b* value of HPCD and the untreated samples were about equal (24), Figure 4-22. The heat pasteurized juice had a

PAGE 88

75 dramatically lower value of 18. This shows that the HPCD treated juice had a darker, deeper red and yellow values than that of the pasteurized juice. 0510152025300246810time (week)b* Value CO2 Heat Untreated Figure 4-22. Watermelon juice b* after treatment and during storage To further understand the relationship between sediment and color, the amount of sediment (mg/ml) in the heat pasteurized juice was graphed against it’s a* value, Figure 4-23. The points follow each other up and down, therefore there must be a direct relationship between sediment and a*. 051015202530350246810time (week)a* Value00.511.522.533.544.5Sediment (mg/ml) Heat a* Heat Sediment wt Figure 4-23. Heat pasteurized watermelon juice a* when compared to the amount of sediment

PAGE 89

76 The amount of sediment in the heat pasteurized juice was also graphed against the L* value. Since the higher the sediment, the lower the L* value, the sediment is on an inverse scale. Again, the figure shows that the amount of sediment and L* are related. 051015202530354045500246810Time (week)L* Value00.511.522.533.544.5Sediment (mg/ml) Heat a* Heat Sediment Figure 4-24. Heat pasteurized watermelon juice L* when compared to the inverse amount of sediment Effect of High Pressure Carbon Dioxide on Lycopene The amount of lycopene in each treatment each week during the storage study was measured. An example chromatogram (Figure 4-25) shows a week 0 untreated juice sample. This figure shows all the absorbance at 450nm wavelength during the run. Lutein and beta carotene can be found as the first two peaks. Figure 4-26 shows the scanning wavelengths (200-700nm) at the specific time that lycopene elutes, this shows a very distinguishable profile. The untreated sample was not measured after week 4. Raw data can be found in the appendix A.

PAGE 90

77 AU 0.00 0.02 0.04 0.06Minutes 5.00 10.00 15.00 20.00 25.00 30.00 35.0 0 L y co p ene Beta-carotene Lutein Figure 4-25. Example lycopene chromatogram for an untreated sample at week 0 wavelength was 450nm AU -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18nm 300.00 400.00 500.00 600.00 471.6 25 min (lycopene) Figure 4-26. Example of the lycopene spectrum while it is eluting. Figure 4-27 shows the actual amount of lycopene in the watermelon juice for each treatment throughout the storage study. There is much variation with the lycopene values. This is believed to be caused by differing levels of sediment in the samples. However except for week 4, HPCD lycopene levels remain > 4 ug/ml throughout the

PAGE 91

78 storage study. From weeks 4 through 8, the amount of lycopene in the HPCD treated juice is higher than that in the heat pasteurized juice. 0246810121416012345678Time (week)Lycopene (ug/ml) HPCD Pasteurized Untreated Figure 4-27. Amount of lycopene (ug/ml juice) in treated samples through the storage study. Untreated not measured after week 4.

PAGE 92

CHAPTER 5 CONCLUSION AND RECOMMENDATIONS This study evaluated the potential of a continuous high pressure carbon dioxide non-thermal pasteurizer to process watermelon juice. The watermelon juice was a product of cull fruit, which would have been left to rot in the fields because currently there is no market to support the fresh market rejects. The vast majority of watermelons are sold as whole or semi-whole fruit and offer a very short shelf life. As the consumer constantly demands variety, safety and availability, it was important to evaluate this processing technique with watermelons. It was determined that with the HPCD technology, temperature and pressure had the largest effect on reducing native microbial populations in unmodified aged watermelon juice, with log reduction up to 6.8. However, once the juice was acidified and sweetened that effect was lost and all combinations of pressure and temperature created the same log reductions (about 4.5 log reductions). Using this technology, watermelon juice was carbonated, bottled and stored for 8 weeks without any major flavor changes or microbial growth during storage. Brix, titratable acidity, and pH remained the same before and after treatment and then throughout the storage study. Color, one of the more appealing attributes, remained bright and vivid throughout the storage study. Another unique aspect of watermelon juice is the lycopene content. This carotenoid causes the red color and is a powerful antioxidant. The lycopene remained steady at about 4-5 ug/ml throughout the storage study. 79

PAGE 93

80 The extent of aroma and taste difference from before and after HPCD treatment and during storage was difficult to evaluate. Taste panels could detect a difference just after processing, however at the end of storage (week 8), they could no longer determine a difference when comparing it to a fresh frozen modified watermelon juice. GC-O was able to monitor the fluctuation in identified aroma active compounds, however the real world importance of this is still unclear. The magnitude of the differences during the taste panel shows us that only slight differences could be determined at some test weeks and other weeks there appeared to be no difference. It is important to understand that not all combinations of process conditions could be tested through a storage study. The treatment with the largest microbial reduction was used. It is entirely possible when working with modified watermelon juice that the following conditions could be used for safety, 10.3 MPa, room temperature, 10% CO 2 , 5 min residence time. These low settings may have different flavor and aroma aspects, which maybe more appealing to panelists and will have an economic advantage. Further product development with acids, sweeteners, with and without carbonation should be conducted to determine the marketable success of a watermelon juice product processed in this way. There are very few if any carbonated juice products on the market today and once the carbonation level is optimized the juice could be HPCD treated and then immediately bottled or packaged at that carbonation level. If carbonation is not desired then further engineering and research needs to be conducted to figure out how to remove residual CO 2 without removing aroma. Also, no attention was given to the initial fruit quality, as they were donated. Creating quality controls on the incoming cull watermelons could offer premium raw

PAGE 94

81 material and those could offer the very best to discriminating consumers. It is also quite possible that the shelf like of HPCD watermelon juice could go well beyond 8 weeks. It would be interesting to see at what week past 8 that the taste panel and microbial counts would obviously show an unacceptable product when compared to heat pasteurized juice. The relationship between lycopene, color and watermelon juice sediment needs to be evaluated further. During this study, it was observed that there is a relationship where higher sediment lends itself to higher lycopene, lower L* value and higher a*. If this is a true relationship then a mathematical model could be developed to predict it. The initial fruit-squeezing step also would need to be evaluated for optimal levels of sugar, water, flavor, color, sediment, etc.

PAGE 95

APPENDIX A RAW EXPERIMENTAL DATA Table A-1. Flavor difference from control evaluation of the hidden control, HPCD treated and heat pasteurized treated samples throughout the storage study Week Hidden* Std error (SE) CO 2 * SE Heat* SE 0 2.38 0.29 4.24 0.32 5.02 0.35 2 2.89 0.30 4.05 0.33 4.10 0.35 4 3.20 0.33 4.36 0.35 3.95 0.33 6 2.62 0.32 4.44 0.31 3.74 0.36 8 3.57 0.33 4.62 0.34 4.55 0.36 * Difference intensity 0-15 scale, 0 = no difference, 15 = extreme difference Table A-2. Aroma difference from control evaluation of the hidden control, HPCD treated and heat pasteurized treated samples throughout the storage study Week Hidden Control* SE CO 2 * SE Heat* SE 0 1.94 0.26 3.26 0.32 3.25 0.34 2 2.84 0.35 3.59 0.35 2.98 0.33 4 2.64 0.33 3.74 0.35 3.02 0.36 6 2.06 0.28 3.94 0.39 2.22 0.31 8 2.96 0.35 3.45 0.31 2.70 0.34 * Difference intensity 0-15 scale, 0 = no difference, 15 = extreme difference Table A-3. Overall likeability mean value of all treatments during storage study Week Mean* SE 0 4.51 0.16 2 4.32 0.15 4 4.05 0.15 6 4.36 0.14 8 3.51 0.13 * Hedonic scale, 0 = extreme dislike, 9 = extremely like 82

PAGE 96

83 Table A-4. FID peak areas for aroma active compounds for each treatment through storage study Untreated Untreated HPCD HPCD HPCD HEAT HEAT HEAT week 0 SE week 4 SE week 0 SE week 4 SE week 8 SE week 0 SE week 4 SE week 8 SE Hexanal 64,350 15,917 68,824 25,336 61,450 5,217 66,894 4,118 79,780 8,266 79,937 9,244 82,473 4,259 91,116 2,399 E-2-Hexenal 15,181 5,332 20,612 2,191 19,135 1,659 14,788 5,015 17,400 1,668 23,822 1,987 21,806 1,358 21,468 837 Octanal 0 15,101 5,383 0 7,153 4,131 0 20,982 3,052 14,751 5,351 15,026 5,488 6-Methyl-5-hepten-2-one 176,555 28,565 233,661 7,312 169,761 6,123 194,480 3,079 218,094 11,099 166,537 9,997 228,498 5,423 0 Nonanal 327,101 112,292 365,424 89,375 247,620 36,602 295,417 37,742 283,798 45,719 389,039 84,494 337,474 81,263 327,996 73,282 ethyl octanoate 0 0 24,822 15,890 0 0 6-Nonenal 335,604 89,593 378,050 49,781 277,616 22,414 286,022 15,238 261,011 27,128 59,726 6,230 405,792 54,342 387,715 42,707 mercapto-4 methylpentan-2-ol,4 0 0 39,407 6,277 0 449,736 40,930 0 #DIV/0! 0 E-2-Nonenal 263,992 56,731 245,554 15,879 217,153 30,376 218,495 23,826 192,184 25,210 423,349 56,003 510,936 36,340 435,839 17,229 E,Z-2,6-Nonadienal 180,125 32,316 171,284 19,010 156,653 29,412 151,604 20,326 134,995 22,435 0 390,601 19,043 310,089 8,096 Z-3-Nonenol 600,410 78,108 580,031 206,956 0 0 0 133,470 11,736 0 #DIV/0! 664,199 12,759 Z-6-Nonenol 97,132 12,007 128,058 14,665 85,181 8,787 92,943 7,392 0 0 99,719 7,391 0 E,Z-3,6-Nonadienol 405,372 41,520 462,479 49,810 382,485 41,508 389,933 33,065 361,597 48,376 0 404,494 23,454 410,688 11,637

PAGE 97

Table A-5. Aroma peak areas for aroma active compounds for each treatment through storage study 84 Untreated Untreated HPCD HPCD HPCD HEAT HEAT HEAT week 0 SE week 4 SE week 0 SE week 4 SE week 8 SE week 0 SE week 4 SE week 8 SE Hexanal 479,938 60,779 272,646 194,994 595,923 191,505 331,641 118,675 119,593 69,050 273,448 121,739 367,122 96,479 332,269 132,334 E-2-Hexenal 395,123 239,237 411,068 252,391 430,254 273,132 447,175 284,321 207,077 120,579 432,964 378,855 276,391 196,045 439,999 283,523 Octanal 247,438 152,487 461,583 189,738 267,939 160,355 392,055 148,784 171,606 103,053 481,312 294,210 6-Methyl-5-hepten-2-one 434,686 262,720 963,598 569,980 584,707 246,167 312,032 220,236 434,040 306,297 268,388 227,142 716,856 544,559 0 Nonanal 928,429 350,826 885,730 219,350 1,355,342 441,137 1,136,497 78,080 711,444 190,985 1,334,102 136,166 1,022,524 416,883 1,147,703 291,347 ethyl octanoate 0 0 951,750 370,744 0 0 0 0 0 6-Nonenal 2,295,134 504,973 1,768,407 324,931 3,340,011 780,611 1,858,812 271,351 1,457,188 310,909 315,520 178,071 1,647,439 510,643 1,699,926 198,072 mercapto-4 methylpentan-2-ol,4 224,458 154,239 1,692,509 679,118 0 0 E-2-Nonenal 2,177,974 845,135 1,982,104 778,620 1,797,293 566,726 1,851,131 497,659 1,104,415 361,739 1,883,941 595,031 2,107,539 957,997 2,286,279 752,796 E,Z-2,6-Nonadienal 2,221,036 506,879 1,682,877 691,849 2,234,980 985,714 1,899,343 532,580 1,737,922 633,063 0 1,808,032 667,726 2,020,678 594,109 Z-3-Nonenol 833,566 348,130 0 0 0 0 426,823 247,051 0 611,843 122,927 Z-6-Nonenol 1,130,463 295,357 643,379 266,523 983,926 384,618 767,927 255,297 0 0 756,881 503,166 0 E,Z-3,6-Nonadienol 737,084 230,961 626,600 230,255 637,690 235,662 136,296 91,963 387,298 144,848 0 197,215 114,291 360,576 177,814 Table A-6. Sediment weight, lycopene concentration and color values for HPCD treated sample throughout storage study Week sediment (mg/ml)* SE Lycopene (ug/ml)* SE L* SE a* SE b* SE 0 2.23 0.007 3.91 0.480 33.71 0.2 22.72 0.36 21.99 0.26 1 2.74 0.003 5.29 0.550 30.77 0.1 26.26 0.06 21.24 0.02 2 3.60 0.080 5.24 1.740 30.27 0.16 28.17 0.12 22.28 0.29 3 2.96 0.053 4.34 1.300 30.59 0.04 27.5 0.01 22.46 0.05 4 3.13 0.063 1.34 0.220 30.38 0.11 28.03 0.06 23.47 0.06 5 3.25 0.020 4.51 0.512 31.22 0.11 26.81 0.13 25.9 0.07 6 3.19 0.067 3.92 0.407 30.59 0.14 27.98 0.05 25.38 0.03 7 3.20 0.108 5.06 0.628 32.35 0.08 25.41 0.06 26.48 0.05 8 3.58 0.143 4.37 1.463 29.83 0.02 29.22 0.04 23.66 0.12 * mean of three values

PAGE 98

85 Table A-7. Sediment weight, lycopene concentration and color values for heat pasteurized sample throughout storage study week sediment (mg/ml)* SE Lycopene (ug/ml)* SE L* SE a* SE b* SE 0 2.41 0.030 2.30 0.114 41.06 0.34 11.32 0.15 12.48 0.03 1 3.73 0.119 7.59 1.778 29.96 0.14 29.35 0.03 16.65 0.07 2 1.78 0.050 1.09 0.034 43.42 0.09 8.7 0.08 10.8 0.04 3 4.08 0.152 6.39 0.958 30.28 0.03 26.54 0.02 17.66 0.02 4 1.55 0.082 2.19 1.088 37.88 0.17 16.71 0.12 20.38 0.06 5 2.47 0.147 2.65 0.205 36.38 0.08 19.09 0.08 18.58 0.02 6 1.65 0.086 2.84 0.155 32.9 0.05 23.3 0.07 21.34 0.05 7 1.72 0.133 2.10 0.121 39.87 0.02 14.26 0.02 19.35 0.06 8 2.94 0.057 2.53 0.672 31.35 0.1 25.58 0.17 21.66 0.12 * mean of three values Table A-8. Sediment weight, lycopene concentration and color values for untreated sample throughout storage study week sediment (mg/ml)* SE Lycopene (ug/ml)* SE L* SE a* SE b* SE 0 4.08 0.409 5.91 1.658 27.4 0.37 32.65 0.17 21.23 0.14 1 3.60 0.006 10.12 3.774 28.76 0.16 31.67 0.1 22.9 0.04 2 3.36 0.012 4.42 1.776 30.07 0.08 26.19 0.06 22.81 0.04 3 1.13 0.063 3.10 1.261 31.23 0.21 24.28 0.31 23.75 0.07 4 2.01 0.028 2.87 0.204 30.64 0.06 26.31 0.06 23.5 0.04 * mean of three values

PAGE 99

APPENDIX B SENSORY BALLOTS Figure B-1. Sample taste test ballot Carbonated Watermelon Juice To start the test, click on the Continue button below: Panelist Code: ________________________ Panelist Name: ________________________________________________ Question # 1. Please indicate your gender. Male Female Question # 2. Male: Please indicate your age range. Under 18 18-29 30-44 45-65 Over 65 Question # 3. Female: Please indicate your age range. Under 18 18-29 30-44 45-65 Over 65 86

PAGE 100

87 Question # 4 Sample ______ Review Instructions Do not taste any of the samples at this time. The first test will be smelling the samples. Please read the directions on the next screen. You are being presented with a reference sample marked 000. Please SMELL the reference sample. Then SMELL sample %01 and compare it to the reference sample. Please mark how different the sample SMELLS from the reference sample on the line scale. Sample Aroma Not Different Very At All Different Question # 5 Sample ______ Review Instructions Please use the water and cracker provided to rinse your mouth between samples. Please read the directions on the following screen. You are being presented with a reference sample marked 000. Please TASTE this sample. Then TASTE sample %01 and compare it to the reference sample. Then mark how different the sample TASTES from the reference sample on the line scale. Taste Difference Not Different Very At All Different Question # 6 Sample ______ How different is the amount of bubbliness (carbonation) in sample %01 from the reference sample 000? Carbonation Difference Not Different Very At All Different

PAGE 101

88 Question # 7 Sample ______ How much do you like the sample %01 OVERALL? Sample %01 dislike neither like extremely like nor extremely dislike 1 2 3 4 5 6 7 8 9 Question # 8. How often do you buy watermelon? Never Once or Twice a Year Once a Month Once a Week Twice a Week This is the end of the test. Do NOT click on 'Continue' on the next screen. Please lift the window until it latches to let the server know you have finished. Thank you for participating. You just helped someone get closer to getting their MS degree.

PAGE 102

LIST OF REFERENCES Arreola, A.G., Balaban M.O., Wei C.I., Peplow A., Marshall M. 1991a. Effect of supercritical carbon dioxide on microbial populations in single strength orange juice. Journal of Food Quality. 14: 275-284. Arreola, A.G., Balaban M.O., Marshall M.R., Peplow A.J., Wei C.I., Cornell J.A. 1991b. Supercritical carbon dioxide effects on some quality attributes of single strength orange juice. Journal of Food Science. 56, 1030-1033. Balaban, M.O., Arreola A.G., Marshall M., Peplow A., Wei C.I., Cornell J. 1991. Inactivation of pectinesterase in orange juice by supercritical carbon dioxide. Journal of Food Science. 56, 743-747 Baldwin, E.A. 2002. Fruit and Vegetable Flavor. USDA ARS, Citrus and Subtropical Products Laboratory, Winter Haven, FL. Ballestra, P., Abreu Da Silva A., Cuq J.L. 1996. Inactivation of Escherichia coli by carbon dioxide under pressure. Journal of Food Science. 61: 829-836. Barbosa-Canovas G.V., U.R. Pothakamury, E. Palou, and B.G. Swanson. 1998. Nonthermal Preservation of Foods. Mercel Dekker, Inc., New York. Bramley, P.M., 2000. Is lycopene beneficial to human health? Phytochemistry. 54, 233-236. Buttery, R.G. 1993. Quantitative and sensory aspects of flavor of tomato and other vegetables and fruits. In Flavor Science: Sensible Principles and Techniques, ACS Books, Washington DC. Buttery, R.G., Ling L.C. 1993. Volatiles of tomato fruit and plant parts; Relationship and biogenesis. ACS Books, Washington DC. Chen, J.S., M.O. Balaban, C.I. Wei, M.R. Marshall and W.Y. Hsu. 1992. Inactivation of polyphenol oxidase by high-pressure carbon dioxide. Journal of Agricultural and Food Chemistry 40, 2345-2349 Chisholm, D.N., Picha D.H., 1986. Distribution of sugars and organic acids within ripe watermelon fruit. Hortscience. 21: 501-503. 89

PAGE 103

90 Chug-Ahuja, J.K., Holden J.M., Forman M.R., Mangels A.R., Beecher G.R., Lanza E. 1993. The development and application of a carotenoid database for fruits , vegetables, and selected multicomponent foods. J. Am. Diet. Assoc. 93: 318-323. Daniels, J.A., Krishnamurthi R., Rizvi S.S. 1985. A review of effects of carbon dioxide on microbial growth and metabolism of micro-organisms. Journal of Food Protection. 48: 532-537. David, J.R.D., Graves R.H., Carlson V.R. 1985. Aspetic Processing and Packaging of Food. CRC Press, New York. DeRovira, D. 1997. Manual – Flavor nomenclature workshop: An odor description and sensory evaluation workshop. Flavor Dynamics, Inc., Somerset NJ. Di Mascio, P., Kaiser S., Sies H. 1989. Lycopene as the most effective biological carotenoid singlet oxygen quencher. Arch. Biochem. Biophys. 274: 532-538. Dillow, A.K., Dehighani F., Hrkach J.S., Foster N.R., Langer R. 1999. Bacteria inactivation by using nearand supercritical carbon dioxide. Proc. Natl. Acad. Sci. USA. 96: 10344-10348. Dodds, W.S., Stutzman L.F., Sollami B.J. 1956. Carbon dioxide solubility in water. Industrial and Engineering Chemistry. 1: 92-94 Dorgan, J.F., Sowell A., Swanson C.A., Potischman N., Miller R., Schussler N., Stephenson Jr. H.E. 1998. Relationship of serum carotenoids, retinal, -tocopherol and selenium with breast cancer risk: results from a prospective study in Columbia, Missouri. Cancer Causes Control. 9: 89-97. Doyle, M.P. 1983. Effect of carbon dioxide on toxin production by Clostridium botulium. Eur. J. Appl. Microbiol. Biotechnol. 17: 53-56. Elkashif, M.E., Huber D.J., Brecht J.K. 1989. Respiration and ethylene production in harvested watermelon fruit: Evidence for nonclimacteric respiration behaviour cultivars. Journal of the American Society for Horticultural Science. 144: 81-85. Elmstrom, G.W., Davis P.L. 1981. Sugars in developing and mature fruits of several watermelon cultivars. Journal of the American Society for Horticultural Science. 106: 330-333. Enfors, S.O., Molin G.1980. Effect of high concentrations of carbon dioxide on growth rate of Pseudomonas fragi, Bacillus cereus and Streptococcus cremoris. J. Appl. Bacteriol. 48: 409-416.

PAGE 104

91 Enomoto, A., Nakamura K., Hakoda M., Amaya N. 1997a. Lethal effect of high-pressure carbon dioxide on a bacterial spore. Journal of Fermentation and Bioengineering. 83: 305-307. Enomoto, A., Nakamura K., Nagai K., Hashimoto T., Hakoda M. 1997b. Inactivation of food microorganisms by high-pressure carbon dioxide treatment with or without explosive decompression. Biosci. Biotech. Biochem. 61: 1133-1137. Erkmen O. 2001. Effects of high pressure carbon dioxide on Escherichia coli in nutrient broth and milk. International Journal of Food Microbiology. 65: 131-135. Erkmen O. 2000a. Inactivation of Salmonella typhimurium by high pressure carbon dioxide. Food Microbiology. 17: 225-232. Erkmen O. 2000b. Effect of carbon dioxide pressure on Listeria monocytogenes in physiological saline and foods. Food Microbiology. 17: 589-596. Erkmen O. 2000c. Antimicrobial effect of pressurized carbon dioxide on Enterococcus faecalis in physiological saline and foods. Journal of the Science of Food and Agriculture. 80: 465-470. Erkmen O. 1997. Antimicrobial effect of pressurized carbon dioxide on Staphylococcus aureus in broth and milk. Lebensm.-Wiss. U.-Technol. 30: 826-829. FAOSTAT data, 2004. www.fao.org . Last viewed 3/7/05. Food and Drug Administration, FDA. 2001. Rockville MD. 21 CFR Part 120 Folkes, G. 2004. Pasteurization of beer by a continuous dense-phase CO2 system. Ph. D. Dissertation, University of Florida, Gainesville, FL. Foster, J.W., Cowan R.M., Maag T.A. 1962. Rupture of bacteria by explosive decompression. J. Bacteriol. 83: 330-334. Fraser, D. 1951. Bursting bacteria by release of gas pressure. Nature. 167: 33-34. Giovannucci, E., Ascherio A., Rimm E.B., Stampfer M.J., Colditz G.A., Willett W.C. 1995. Intake of carotenoids and retinal in relation to risk of prostate cancer. J. Natl. Cancer Inst. 97: 1767-1776. Gould, G.W. 1989. Mechanisms of Action of Food Preservation Procedures. Elsevier Science Publishers LTD., New York. Granado, R., Olmedilla B., Blanco I., Rojas-Hidalgo E. 1996. Major fruit and vegetable contributors to the main serum carotenoids in the Spanish diet. Eur. J. Clin. Nutr. 50: 246-250.

PAGE 105

92 Hartwig, P. and McDaniel M. 1995. Flavor characteristics of lactic, malic, citric, and acetic acids at various pH levels. Journal of Food Science. 60: 384-388. Haas, G.J., Prescott Jr. H.E., Dudley E., Dik R., Hintlian C., Keane L. 1989. Inactivation of microorganisms by carbon dioxide under pressure. Journal of Food Safety. 9: 253-265. Helzlsouer, K.J., Comstock G.W., Morris J.S. 1989. Selenium, lycopene, -tocopherol, -carotene, retinal, and subsequent bladder cancer. Cancer Res. 49: 6144-6148. Hochmuth, G.J., Stall W.M., Hewitt T.D., Ruppert K.C. 1997. Alternative opportunities for small farms: watermelon production review. Extension Administration Fact Sheet RF-AC029. Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. Hochmuth, G.J., Maynard D.N., Vavrina C.S., Stall W.M., Kucharek T.A., Stansly P.A., Taylor T.G., Smith S.A., Smajstrla A.G. 1999. Cucurbit production in Florida. Horticultural Sciences Department HS 725. Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. Holden, J.M., Eldridge A.L., Beecher G.R., Buzzard I.M., Bhagwat S., Davis C.S., Douglass L.W., Gebhardt S., Haytowitz D., Schakel S. 1999. Carotenoid content of U.S. foods: an update of the database. J. Food Compos. Anal. 12: 169-196. Hong S., Park W.S., Pyun Y.R. 1997. Inactivation of Lactobacillus sp. From kimchi by high-pressure carbon dioxide. Food Sci. Technol. 30: 681-685. Hong, S., Park W., Pyun Y. 1999. Non-thermal inactivation of Lactobacillus plantarum as influenced by pressure and temperature of pressurized carbon dioxide. International Journal of Food Science and Technology. 34: 125-130. Hong, S. and Pyun Y. 2001. Membrane damage and enzyme inactivation of Lactobacillus plantarum by high pressure CO2 treatment. International Journal of Food Microbiology. 63: 19-28. Hurst, A. 1977. Bacterial injury: a review. Can. J. Microbiol. 23: 936-944. Hutkins, R. W. and Nannen N. L. 1993. pH homeostasis in lactic acid bacteria. Journal of Dairy Science. 76: 2354-2365. Isenschmid, A., Marison I.W., von Stockar U. 1995. The influence of pressure and temperature of compressed CO2 on the survival of yeast cells. J. Biotechnol. 39: 229-237.

PAGE 106

93 Ishida, B.K., Ma J., and Chan B. 2001. A simple, rapid method for HPLC anaylsis of lycopene isomers. Phytochemical Analysis. 12: 194-198 Ishikawa , H., Shimoda M., Kawano T., Osajima Y. 1995a. Inactivation of Enzymes in an aqueous solution by micro-bubbles of supercritical carbon dioxide. Biosci. Biotech. Biochem. 59: 628-631. Ishikawa, H., Shimoda M., Shiratsuchi H., Osajima Y. 1995b. Sterilization of microorganisms by supercritical carbon dioxide micro-bubble method. Biosci. Biotech. Biochem. 59: 1949-1950. Ishikawa, H. Shimoda M., Tamaya K., Yonekura A., Kawano T., Osajima Y. 1997. Inactivation of Bacillus spores by supercritical carbon dioxide micro-bubble method. Biosci. Biotech. Biochem. 61: 1022-1023 Jones, R.P., Greenfield P.F. 1982. Effect of carbon dioxide on yeast growth and fermentation. Enzyme and Microbial Technology. 4: 210-223. Kamihira, M., Taniguchi M., Kobayashi T. 1987. Sterilization of microorganisms with supercritical carbon dioxide. Agric. Biol. Chem. 51: 407-412. Kashket, E.R. 1987. Bioenergetics of lactic acid bacteria: cytoplasmic pH and osmotolerance. FEMS Microbiol. 46: 233-244. Kincal, D. 2000. A continuous high pressure carbon dioxide system for cloud retention, microbial reduction and quality change in orange juice. Master’s Thesis. University of Florida, Gainesville. Kohlmeier, L., Kark J.D., Gomez-Garcia E., Martin B.C., Steck S.E., Kardinaal A.F.M., Rinstad J., Thamm M., Masaev V., Riemersma R., Martin-Moreno J.M., Huttunen J.K., Kok F.J. 1997. Lycopene and myocardial infarction risk in the EURAMIC study. Am. J. Epidemiol. 140: 618-626. Lin, H., Cao N., Chen L. 1994. Antimicrobial effect of pressurized carbon dioxide on Listeria monocytogenes. Journal of Food Science. 59: 657-659. Lin, H., Yang Z., Chen L. 1992a. Inactivation of Saccharomyces cerevisiae by supercritical and subcritical carbon dioxide. Biotechnol. Prog. 8: 458-461. Lin, H., Yang Z., Chen L. 1992b. An improved method for disruption of microbial cells with pressurized carbon dioxide. Biotechnol. Prog. 8: 165-166 Lin, H., Chan E., Chen C., Chen L. 1991. Disintegration of yeast cells by pressurized carbon dioxide. Biotechnol. Prog. 7: 201-204.

PAGE 107

94 Lucier, G., Lin B-H. 2001. Factors affecting watermelon consumption in the United States. Vegetables and Specialties Situation and Outlook, VGS-287, Economic Research Service, USDA. Meyssami, B., Balaban M., Teixeira A. 1992. Prediction of pH in Model Systems Pressurized with Carbon Dioxide. Biotechnology. 8: 149-154 Nakamura, K., Enomoto A., Fukushima H., Nagai K., Hakoda M. 1994. Disruption of microbial cells by the flash discharge of high-pressure carbon dioxide. Biosci, Biotech, Biochem. 58: 1297-1301. O’Mahony, M. 1995. Sensory measurements in food science: fitting methods to goals. Food Technology. Apr: 72-82. Pratt, H.K. 1971. Melons. In Biochemistry of Fruits and their Products, ed A.C. Hulme, vol 2, Academic Press, London. Rice-Evans, C.A., Sampson J., Bramley P.M., Holloway D.E. 1997. Why do we expect carotenoids to be antioxidants in vivo? Free Rad. Res. 26: 381-398. Sargent, S. 2000. Handling Florida vegetables: watermelon. SS-VEC-934, Department of Horticultural Sciences, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. Schamp, N. and Dirinck P. 1982. The use of headspace concentration on Tenax for objective evaluation of fresh fruits. In Chemistry of Foods and Beverages, Academic Press, NY. Song, J., Gardner B.D., Holland J.F., Beaudry R.M. 1997. Rapid analysis of volatile flavor compounds in apple fruit using SPME and GC/time of flight spectrometry. Journal of Agric. Food Chemistry. 45: 1801-1807 Spilimbergo, S., Elvassore N., Bertucco A. 2003. Inactivation of microorganisms by supercritical CO2 in a semi-continuous process. Italian Journal of Food Science. 15:115-124. Spilimbergo, S. and Bertucco A. 2003. Non-thermal bacteria inactivation with dense CO2. Biotechnology and Bioengineering. Vol 84, no 6: 627-638. Stahl, W., Sies H. 1996. Lycopene: a biological important carotenoid for humans? Arch Biochem Biophys. 336: 1-9. Teranishi, R. and Kint S. 1993. Sample preparation. In Flavor Science: Sensible Principles and Techniques, ACS Books, Washington DC.

PAGE 108

95 Toews, K.L., Shroll R.M., Wai C.M., Smart N.G. 1995. pH-defining equilibrium between water and supercritical CO2. Influence on SFE of organics and metal chelates. Anal. Chem. 67: 4040-4043. United States Department of Agriculture, USDA. 1994. Watermelons: An Economic Assessment of the Feasibility of Providing Multiple-Peril Crop Insurance, Economic Research Service, USDA in cooperation with the University of California for the Federal Crop Insurance Corporation November 22, 1994 United States Department of Agriculture, USDA. 2002. Vegetables 2002 Summary, National Agricultural Statistics Service. Washington, D.C. Vg 1-2 (03) United States Department of Agriculture, USDA. 2004. ARS Agricultural Research Service. Nutrient Data Laboratory. http://www.nal.usda.gov/fnic/foodcomp/search/ . Last viewed 3/7/05. Vasquez, B.C., O.N. Nesheim. 2000. Florida crop/pest management profiles: watermelon., CIR 1236, University of Florida, IFAS extension. Young, A., Britton G. 1993. Carotenoids in Photosynthesis. Chapman & Hall, New York. Watanabe, K., Saito T., Hirota S., Takahashi B. 1987. Carotenoid pigments in red, orange and yellow-fleshed fruits of watermelon (Citrullus vulgaris). Journal of Japanese Society of Horticultural Sciences. 56: 45-50. Wei, C.I., Balaban M.O., Fernando S.Y., Peplow A.J. 1991. Bacterial effect of high pressure CO2 treatment on foods spiked with Listeria or Salmonella. Journal of Food Protection. 54: 189-193. Zeuthen, P. 1984. Thermal Processing and Quality of Foods. Elsevier Applied Science, London.

PAGE 109

BIOGRAPHICAL SKETCH Matthew Lecky was born in Summerside, Prince Edward Island, Canada, on June 10 th , 1979. His family moved to Florida when he was 4 years old. He received his B.S. degree in food science and human nutrition and a minor in business administration in May 2002, at the University of Florida, Gainesville, Florida. In August 2002, he started the graduate program in the Food Science and Human Nutrition Department at the University of Florida. He is expected to receive his M.S. degree and a minor in food and resource economics in August 2005. 96