PHOTOOXIDATION AND PHOTOSENSITIZED OXIDATION OF LINOLEIC ACID, MILK, AND LARD

DISSERTATION

Presented in Partial Fulfillment of the Requirement for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By JaeHwan Lee, M.S.

$ * * * $

The Ohio State University 2002

Dissertation Committee: Approved by Dr. David B. Min, Adviser Dr. Howard Q. Zhang Dr. Polly D. Courtney Adviser Dr. Hua Wang Food Science and Nutrition UMI Number: 3081938

UMI UMI Microform 3081938 Copyright 2003 by ProQuest information and Learning Company. Aii rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest information and Learning Company 300 North Zeeb Road PO Box 1346 Ann Arbor, Mi 48106-1346 ABSTRACT

Photooxidation and photosensitized oxidation on the formation of volatile compounds in linoleic acid, milk, and lard were studied by a combination of solid-phase microextraction (SPME)-gas chromatography (GC)-mass spectrometry (MS) and headspace oxygen content.

Photooxidation is the oxidation under light in the absence of photosensitizers such as chlorophyll and riboflavin. Photosensitized oxidation is the oxidation under light in the presence of photosensitizers. Total volatile compounds in linoleic acid without added chlorophyll under light and in the dark did not increase for 48 hr at 4 °C. Total volatile compounds in linoleic acid with added 5 ppm chlorophyll under light at 4 °C for 0, 6, 12,

24, and 48 hr, increased from 8.9 to 11.6, 21.7, 26.1, 29.3 (IxlO"^) in electronic counts, respectively. 2-Pentylfuran, an undesirable reversion flavor compound in soybean oil, 2- oetene-l-ol, 2-heptenal, and l-oetene-3-ol were formed by photosensitized oxidation only. Light excited photosensitizers like chlorophyll can generate singlet oxygen from ordinary triplet oxygen. 2-Pentylfuran, 2-heptenal, and l-octene-3-ol can come from

CIO, C l2, and CIO hydroperoxide of linoleic acid, respectively, which can be formed by

singlet oxygen oxidation but not by triplet oxygen oxidation on linoleic acid. The singlet ii oxygen oxidation mechanisms for 2-pentylfuran, 2-heptenal, l-octene-3-ol, and 2-octene-

l-ol from linoleic acid were proposed.

Milk with or without added riboflavin, ascorbic acid, sodium azide, and hutylated hydroxyanisol (BHA) was stored at 4 °C under light and in the dark. Pentanal, dimethyl disulfides, hexanal, and heptanal were formed only in the light stored milk and increased

significantly as the added riboflavin concentration increased from 5 to 10, 50 ppm

(P<0.05). As fat content in milk increased from 0.5 to 1.0, 2.0, and 3.4%, pentanal, hexanal, and heptanal increased significantly (P<0.05) but dimethyl disulfide concentration did not change. BHA and ascorbic acid, hydrogen donating free radical

scavengers, reduced hexanal and heptanal formation. Sodium azide, a singlet oxygen quencher, prevented dimethyl disulfide formation. Formation of pentanal is different from that of hexanal and heptanal in milk. Singlet oxygen and free radicals play important roles in the formation of volatile compounds in photosensitized milk.

Volatile compounds and headspace oxygen content from lard containing 0 and 5 ppm chlorophyll in air-tight 10-mL bottles were studied at 55 °C under light and in the dark. Total volatile compounds in lard with 5 ppm chlorophyll under light were 19 times more than that in samples with 0 ppm chlorophyll for 48 hr. Headspace oxygen content in lard with 5 ppm chlorophyll under light changed from 21 to 15% for 48 hr but that in lard with 0 ppm chlorophyll did not change significantly (P>0.05). Photosensitizer-free lipid model system using lard was developed. Photooxidation mechanism seems to be a free radical chain reaction like autoxidation.

Photosensitizers like chlorophyll and riboflavin play important roles in the formation of volatile compounds. Singlet oxygen, which can he generated from triplet

iii oxygen by light excited photosensitizers, is involved in the formation of volatile compounds in linoleic acid, milk, and lard. Photosensitizer content in food should be minimized and food products should be packed with light impermeable package to prevent the photosensitized oxidation on lipids.

IV Dedicated to my mother, my wife, my daughter, and my late father.

V ACKNOWLEDGMENT

I wish to thank my adviser, Dr. David B. Min for his energetic guidance and enthusiasm which made this dissertation possible. He gave me a chance to study in this charming field and showed me a way to taste the world of knowledge on lipid and flavor.

It is my greatest pleasure to he his advisee in my academic study.

I also thank my committee members, Dr s. Ahmed E. Yousef, Howard Q. Zhang,

Polly D. Courtney, and Hua Wang for their valuable suggestions, consistent advisory and concerns throughout my graduate program.

I also thank my colleagues and graduate school friends, Jeff Boff, Raymond Diono, Hyun-Jung Chung, and Dr. TaeHyun Ji for their memorable friendship and assistance. Most importantly, I would like to express my greatest appreciation to my mother, my wife, my daughter, and my late father.

VI VITA

January 26, 1970 ------Bom in Seoul, Republic of Korea 1988-1994 ------B.S. Food Science and Technology Seoul National University, Seoul, Korea 1994-1996------M.S. Food Science and Technology Seoul National University, Seoul, Korea 1996-1999------CheilJedang Food Company, Korea 1999-2002------Graduate Research Associate The Ohio State University, Columbus, Ohio, U.S.A.

PUBLICATIONS

1. Steenson DF, Lee JH, and Min DB. Solid Phase Microextracion Analyses of Volatile Compounds of Soybean and Com Oils, J. Food Sci. 2002, 67(1), 71-76.

2. Chung MS, Lee JH and Min DB. Effects of Pseudomonas putrifaciens and Acinetobacter spp. on the Flavor Quality of Raw Ground Beef, J. Food Sci. 2002, 67(1), 77-83.

3. Yang WT, Lee JH and Min DB. Quenching Mechanisms and Kinetics of a-

Tocopherol and (3-Carotene on the Photosensitizing Effect of Synthetic Food Colorant FD & C Red No.3, J. Food Sci. 2002, 67 (2), 507-510.

V ll 4. Foley DM, Pickett K, Varon J, Lee JH, Min DB, Caporaso F and Prakash A. Pasteurization of Fresh Orange Juice using Gainnia Irradiation: Microbiological, Flavor and Sensory Analyses. J. Food Sci. 2002, 67(4),1495-1501.

5. Lee JH, Kang JH, and Min DB. Optimization of Solid Phase Microextraction on the Headspace Volatile Compounds in Kimchi, a Traditional Korean Fermented Vegetable Product. J. Food Sci.2002-0478 (In press).

6. Kang JH, Lee JH, Min S, and Min DB. Study of enzymatic and chemical changes of Kimchi, a Traditional Korean Fermented Vegetable Product. J. Food Sci. 2002- 0488 (In press).

7. Lee JY, Lee JH, Choi YO, and Min DB. Singlet oxygen oxidation and its effects on the soy flour off-flavor. J. AOCS (In press).

8. Lee JH, Ozcelik B, and DB Min. Electron donation mechanisms of (3-carotene as a free radical scavenger. J. Food Sci.2002-500 (In press).

9. Ozcelik B, Lee JH, and DB Min. Effects of Light, Oxygen and pH on the 2,2- Diphenyl-1 -Picrylhydrazyl (DPPH) method to evaluate antioxidants. J. Food Sci. (In press).

10. Lee JH, Diono R, Kim GY, and DB Min. Optimization of Solid Phase Microextraction on the Headspace Volatile Compounds in Parmesan Cheese. J. Agric. Food Chem. Jf025910+ (In press).

11. Kim GY, Lee JH, and DB Min. Study of Eight-Induced Volatile Compounds in Goat Cheese. J. Agric. Food Chem jf0259009a (In press).

vm PUBLISHED ABSTRACT

JH Lee and DB Min. Nuclear magnetic resonance (NMR) study on the electron donating ability of (3-carotene by using 2,2-diphenyl-1-picrylhydrazyl (DPPH). Institute of Food Technologist, Anaheim Ca. 2002

JH Kang, JH Lee, YT Ko and DB Min. Changes of headspace volatile compounds, oxygen, carbon dioxide and microorganisms in Kimchi stored at 5°c for 25 days. Institute of Food Technologist, Anaheim Ca. 2002

JH Lee, JH Kang, YT Ko and DB Min. Optimization of solid phase microextraction (SPME) on the headspace volatile compounds in Kimchi. Institute of Food Technologist, Anaheim Ca. 2002

JY Lee, JH Lee, EG Choe and DB Min. Analysis of volatile compounds and headspace oxygen and carbon dioxide in full fat soy flour. Institute of Food Technologist, Anaheim Ca. 2002

Ozcelik B, Lee JH and Min DB. The effects of chlorophyll, light and oxygen on the stability of 2,2-diphenyl-1-picrylhydrazyl radical in acetone and soybean oil. Institute of Food Technologist, New Orleans LA. 2001

Lee JH, Ozcelik B and Min DB. Evaluation of electron donating ability of (3-carotene by using soybean oil and 2,2-diphenyl-1-picrylhydrazyl (DPPH) in acetone system. Institute of Food Technologist, New Orleans LA. 2001

Diono R, Lee JH and Min DB. Solid phase microextraction-gas chromatographic analysis of cheese flavor compounds. Institute of Food Technologist, New Orleans LA. 2001

IX Chung MS, Lee JH and Min DB. Effects of Pseudomonas putrifaciens and Acinetobacter spp. on the Flavor Quality of Raw Ground Beef. Institute of Food Technologist, Dallas, TX. 2000

FIELD OF STUDY

Major Field : Food Science and Nutrition

X TABLE OF CONTENTS

Page A bstract ...... ii Dedication ...... v Acknowledgement ...... vi Vita ...... vii Eist of Tables ...... xiii Eist of Figures ...... xiv

Chapters: 1. Introduction ...... 1

2. Eiterature review ...... 4 2.1 Eipid Oxidation ...... 4 2.1.1 The chemistry of oxygen ...... 4 2.1.2 Formation of singlet oxygen ...... 8 2.1.3 Triplet oxygen oxidation with fatty acids ...... 11 2.1.4 Singlet oxygen oxidation with fatty acids ...... 14 2.1.5 Comparison of triplet and singlet oxygen oxidation ...... 16 2.1.6 Effects of singlet oxygen oxidation on food quality ...... 18 2.2 Photosensitizer ...... 21 2.2.1. Mechanisms of photosensitizer ...... 21 2.2.2. Riboflavin ...... 25 2.2.3 Chlorophyll ...... 29

3. Photooxidation and Photosensitized Oxidation of LinoleicAcid ...... 31 3.1 Abstract ...... 31 3.2 Introduction ...... 32

xi 3.3 Materials and Methods ...... 34 3.4 Results and Discussion ...... 37 3.5 Conclusion...... 48 3.6 Reference ...... 48

4. Photosensitized Oxidation of Milk ...... 52 4.1 Abstract ...... 4.2 Introduction ...... 4.3 Materials and Methods 4.4 Results and Discussion 4.5 Conclusion ...... 4.6 Reference ...... 73

Photooxidation and Photosensitized Oxidation of Lard ...... 76 5.1 Abstract ...... 5.2 Introduction ...... 77 5.3 Materials and Methods 5.4 Results and Discussion 5.5 Conclusion ...... 5.6 Reference ...... 94

6. Conclusion ...... 97

Bibliography ...... 99

APPENDIX A: Figures of sample preparation ...... 107

APPENDIX B: Figures of gas chromatograms ...... I l l

XII LIST OF TABLES

Table Page

1 Comparison of singlet and triplet oxygen ...... 8

2 Relative oxidation rates of triplet and singlet oxygen with oleate, linoleate and linolenate ...... 17

3 Hydroperoxides formed by singlet and triplet oxygen oxidation of oleic, linoleic and linolenic acid ...... 19

4 Effects of singlet oxygen on the food quality ...... 21

5 Function of various agents on the mechanism of photosensitization ...... 24

6 Volatile compounds identified from linoleic acid with or without added chlorophyll stored at 4°C for 48 hr in GC peak areas (1 xlO^) ...... 43

7 Effects of ascorbic acid, sodium azide, and BHA on the formation of dimethyl disulfide, hexanal, and heptanal in milk containing 50 ppm added riboflavin ...... 66

8 Effects of riboflavin, ascorbic acid, sodium azide, and BHA on the formation of pentanal in electronic counts (1x10'*) in milk stored at 4°C under fluorescent light for 4 hr ...... 71

9 Volatile compounds identified from lard with 0 and 5 ppm chlorophyll stored under light at 55 °C for 48 hr in GC peak areas (1 xlO'*) ...... 93 xiii LIST OF FIGURES

Figure Page

1 Molecular orbital of triplet oxygen ...... 5

2 Molecular orbital of singlet oxygen ...... 7

3 Singlet oxygen formation by chemical, photochemical and biological methods ...... 9

4 The chemical mechanism for the formation of singlet oxygen in the presence of sensitizer, light and triplet oxygen ...... 10

5 Energy required for hydrogen removal from linoleic acid ...... 11

6 General mechanisms of triplet oxygen oxidation with linoleic acid 13

7 Reactions of singlet oxygen with olefins by: 1,4-cycloaddition, the “ene” reaction, and 1,2-cycloaddtion ...... 15

8 Difference of triplet and singlet oxygen oxidation mechanisms ...... 16

9 Type I pathway of sensitizer ...... 23

10 Structure of riboflavin ...... 25

xiv 11 Effects of 0, 5 and 15 minutes illumination on electron spin resonance spectrum of 2,2,6,6-4-piperidone-N-oxyl in water solution of riboflavin and 2,2,6,6- tetramethyl-4- piperidone ...... 27

12 Structure of chlorophyll ...... 29

13 Gas chromatograms of headspace volatile compounds in linoleic acid containing chlorophyll stored under light for 48 hr ...... 38

14 The effects of chlorophyll on the volatile compounds in linoleic acid stored at 4 °C under light and dark for 48 hr ...... 40

15 Proposed formation mechanisms of 2-pentylfuran from linoleic acid by singlet oxygen oxidation ...... 45

16 Proposed formation mechanisms of 2-heptenal from linoleic acid by singlet oxygen oxidation ...... 46

17 Proposed formation mechanisms of 1 -octene-3-ol and 2-octene-1 -ol from linoleic acid by singlet oxygen oxidation ...... 47

18 Headspace volatile compounds in whole milk under the fluorescent light for 0 (a), 2 (b), 4 (c), and 8 hr (d) at 4 °C. 1 : pentanal, 2: dimethyl disulfide, 3: hexanal, 4: heptanal ...... 60

19 Headspace volatile compounds in whole milk stored under light or in the dark at 4 °C and control for 8 hr. 1: pentanal, 2: dimethyl disulfide, 3: hexanal, 4: heptanal ...... 61

XV 20 Effects of riboflavin on the formation of pentanal, dimethyl disulfide, hexanal and heptanal in photosensitized milk under light at 4 °C for 2hr. * indicates that the compound was not detected ...... 62

21 Effects of fat content on the formation of pentanal, heptanal, hexanal, and dimethyl disulfide in photosensitized milk under light for 8 hr ...... 64

22 The suggested formation mechanisms of hexanal in photosensitized milk containing riboflavin ...... 70

23 Gas chromatograms of headspace volatile compounds in lard with 0 and 5 ppm chlorophyll stored under light for 24 hr at 55 °C. Numbered volatiles are shown in Table 9 ...... 83

24 The effects of chlorophyll on the volatile compounds in lard stored

under light and dark for 48 hr at 55 ° C ...... 84

25 The effects of chlorophyll on the headspace oxygen content in lard stored

under light and dark for 48 hr at 55 °C ...... 86

26 The effects of chlorophyll on 2-hexenal, 2-heptenal, 2-pentylfuran, and 2-octenal in lard stored under light for 48 hr at 55 °C ...... 89

27 Proposed formation mechanisms of 2-hexenal, 2-heptenal, and l-octene-3-ol from photosensitized lard by singlet oxygen oxidation ...... 92

28 Solid phase microextraction ...... 107

XVI 29 Fluorescent light box used for the sample preparation of linoleic acid and m ilk ...... 108

30 Sample preparation of lard (Chapter 5 ) ...... 109

31 Light box used for lard samples (Chapter 5 ) ...... 110

32 Gas chromatogram of linoleic acid with added 5 ppm chlorophyll stored under light for 2 day (Chapter 3 ) ...... I l l

33 Gas chromatograms of milk with added 0 and 5 ppm riboflavin stored under light for 2 hr (Chapter 4 ) ...... 112

34 Gas chromatograms of milk with added 10 and 50 ppm riboflavin stored under ligbt for 2 hr (Chapter 4 ) ...... 113

35 Gas chromatograms of milk with 0.5 and 1.0% fat content stored under light for 8 hr (Chapter 4) ...... 114

36 Gas chromatograms of milk with 2.0 and 3.4% fat content stored under light for 8 hr (Chapter 4 ) ...... 115

37 Gas chromatograms of milk with added 50 ppm riboflavin with or without 100 ppm ascorbic acid stored under light for 4 hr (Chapter 4 ) ...... 116

38 Gas chromatograms of milk with added 50 ppm riboflavin with or without 100 ppm BHA stored under light for 4 hr (Chapter 4 ) ...... 117

xvii 39 Gas chromatograms of milk with added 50 ppm riboflavin with or without 1 GO ppm sodium azide stored under light for 4 hr (Chapter 4 ) ..... 118

40 Gas chromatograms of lard with 5 ppm chlorophyll stored under light for 0 and 6 hr (Chapter 5 ) ...... 119

41 Gas chromatograms of lard with 5 ppm chlorophyll stored under light for 12 and 24 hr (Chapter 5 ) ...... 120

42 Gas chromatograms of lard with 5 ppm chlorophyll stored under light for 36 and 48 hr (Chapter 5 ) ...... 121

43 Gas chromatograms of lard with 0 ppm chlorophyll stored under light for 0 and 6 hr (Chapter 5) ...... 122

44 Gas chromatograms of lard with 0 ppm chlorophyll stored under light for 12 and 24 hr (Chapter 5 ) ...... 123

45 Gas chromatograms of lard with 0 ppm chlorophyll stored under light for 36 and 48 hr (Chapter 5 ) ...... 124

X Vlll CHAPTER 1

INTRODUCTION

Lipid oxidation of unsaturated fatty acids can produce undesirable off-odor and toxic compounds, and cause loss of important nutrients in foods (Boff and Min 2002).

The major mechanisms of lipid oxidation are autoxidation, lipoxygenase catalyzed oxidation, and light-induced oxidation (Chan 1977; Frankel 1985a, 1985b). A chemical reaction of atmospheric triplet oxygen and organic compounds, usually unsaturated compounds, is called as autoxidation (Chan 1987). Autoxidation of lipid is a free radical chain reaction and can be initiated by metal catalysis, high temperature, and UV irradiation (Frankel 1985b).

Light-induced oxidation causes the quality loss of foods and beverages, and makes the products less acceptable or unacceptable to consumers (Skibsted 2000). Many

studies have been reported on the formation of light-induced off-flavor in foods, including dairy products and oil products (Bradley 1980; Faria 1983; Kim and Morr

1996). Reversion flavor in soybean oil and sunlight flavor in milk are well-known examples of light-induced off-flavor (Min 1998; Jung et al. 1998). The mechanisms of light-induced lipid oxidation can be photosensitized oxidation and photooxidation depending on the presence or absence of photosensitizers, respectively. Photosensitizers in foods stored under light can accelerate lipid oxidation either by Type 1 or Type 11 mechanisms. Singlet state of photosensitizer can absorb light energy and become excited singlet state photosensitizer. The excited singlet state photosensitizer becomes to the excited triplet state photosensitizer by intersystem crossing mechanisms. The excited triplet state photosensitizer reacts with triplet oxygen to form singlet oxygen (Type II mechanism) or abstracts electron or hydrogen atom from a substrate to generate radicals (Type I mechanism) (Foote 1976; Bradely 1992; Lu et al.

1999).

Chlorophyll in plant materials, riboflavin in dairy foods, and myoglobin derivatives in meat products are well-known photosensitizers that can accelerate oxidation reactions (Lee and Min 1988; King 1996; Li and Min 1998). It was reported that photosensitizing function of riboflavin induced the decrease of the important nutrients including amino acids. Vitamin A, C and D (Jung et al. 1995; Li et al. 2000;

Whited et al. 2002).

The singlet oxygen oxidation (Type II mechanism of photosensitizers) is very important due to the greater reaction rate of singlet oxygen than that of triplet oxygen in foods even at very low temperatures (Raws and YanSanten 1970). Singlet oxygen oxidation can generate new compounds, which are not found in ordinary triplet oxygen oxidation in foods (Boff and Min 2002).

The objectives of this dissertation were (1) to study the formation of volatile compounds in photosensitized oxidation of linoleic acid model systems, (2) to study the

2 formation of volatile compounds in photosensitized oxidation of milk, (3) to develop a photosensitizer-free model system using lard, and (4) to propose the formation mechanisms of volatile compounds in linoleic acid, milk, and lard by photosensitized oxidation. CHAPTER 2

LITERATURE REVIEW

2.1. Lipid oxidation

2.1.1. The chemistry of oxygen

There are two types of oxygen in nature: one is triplet state and the other is singlet

state under magnetic field. Joseph Prestley discovered triplet oxygen, which is most abundant and in normal air we are breathing, in 1775. Avogadro reported in 1811 that oxygen is a diatomic molecule and Faraday reported paramagnetic characteristics of oxygen in 1848. The paramagnetic property of oxygen was due to the parallel spins of two outer electrons in the degenerate orbitals (Boff and Min 2002). The molecular orbital of triplet oxygen is shown in Figure 1. The spin multiplicity of molecule is defined as 2S+1, where S is the total spin quantum number. One spin is designated as

(+1/2). In case of ground state of oxygen, S is 1 (=l/2+l/2). The spin multiplicity of oxygen is 3 (=2*1+1), which can be shown under magnetic field as 3 distinctive energy

states. Therefore, ground state of oxygen is called triplet state and is a di-radical compound. Radical compounds can readily react with radical compounds and non-

4 radical compounds react with non-radical compounds. Therefore, triplet oxygen ean react with radical compounds.

Molecular

Atomic Atomic

2Px 2Py 2Pz 2Pz 2Py 2Px

Energy

Figure 1. Molecular orbital of triplet oxygen (Boff and Min 2002) Herzberg discovered singlet oxygen, which has a higher energy state using

spectroscopy in 1934. The importance of singlet oxygen was rediscovered by Foote and

Wexler (1964) and Corey and Taylor (1964). Singlet oxygen has been studied for the last

30 years in food chemistry, biochemistry, medicine, and organic chemistry.

The molecular orbital of singlet oxygen is shown in Figure 2. Molecular orbital of singlet oxygen is different from that of triplet oxygen, where the electrons are paired in the K antibonding orbital. Singlet oxygen is non-radical state. The total spin quantum number (S) of singlet oxygen is 0 (=+1/2-1/2, where each spin is in opposite position) and the multiplicity of state is 1 under magnetic field.

Singlet oxygen is ' A state and this type of singlet oxygen plays an important role in the oxidations in foods due to the relatively long lifetime. The energy state of

singlet oxygen is 22.5 kcal/mole and the lifetime is ranging 50 to 700 microseconds depending on the solvent types, which is long enough to react with other food components with electron rich moiety (Korycka-Dahl and Richradson 1978; Girotti

1998).

The activation energy of singlet oxygen is 0 to 6 kcal/mole, which is low enough for the oxidation reactions irrespective of temperature effects (Yang and Min 1994). The comparisons of singlet and triplet oxygen characteristics are shown in Tahle 1. Singlet oxygen has been studied in lipid and vitamin areas, which are most sensitive to the oxidative damage by singlet oxygen. Molecular

Atomic Atomic

2Px 2Py 2Pz 2Pz 2Py 2Px

Energy

Figure 2. Molecular orbital of singlet oxygen (Boff and Min 2002) 'O 2 'O 2

Energy Level 0 22.4 kcaFmole Diradical Nature Electron rich compounds Reaction Radical compound Non-radical, electrophilic

(Boff and Min 2002)

Table 1. Comparison of singlet and triplet oxygen

2.1.2 Formation of singlet oxygen

Singlet oxygen can be formed by the reactions of enzyme, chemical compounds, and sensitizers exposed under light (Krinsky 1977). The formation of singlet oxygen is

shown in Figure 3. Some formation mechanisms are not scientifically proven. ENZYMES RCOO* + RCOO* 'O2 + ^SENSITIZER

RC=0 ENDOPEROXIDES + SENSITIZER^ H2O2 + OCr RCOH PRODUCTS

PRODUCTS

OZONIDES ik

•OH 2H+.

O1 +Y+

Figure 3. Singlet oxygen formation by ehemical, photoehemical and biologieal methods

(Krinsky 1977)

The most well documented formation mechanism of singlet oxygen is the photochemistry using sensitizers. Chlorophyll, myoglobin, riboflavin, porphyrins, and

synthetic colorants are well-known photosensitizers, which can change triplet oxygen to

singlet oxygen (Foote and Denny 1968; Lledias and Hansberg 2000). The chemical mechanisms of formation of singlet oxygen by photosensitizers are shown in Figure 4.

9 Excited state- 'Sen* k = 1- 20 X 10 /s e c Intersyst& Crossing 'Sen* k = 2 X 10^ /sec hv + O2 Fluorescence k = 10 - 10 /sec. k = 1 - 3 X 10^ /sec Phosphor^cence Ground 'Sen 'O 9 state

Figure 4. The chemical mechanism for the formation of singlet oxygen in the presence of sensitizer, light and triplet oxygen.

Ground state of singlet sensitizer ('Sen) can become an excited singlet sensitizer ('Sen*).

'Sen* can return to 'Sen state by emitting fluorescence or change to excited triplet

sensitizer (^Sen*) by intersystem crossing mechanisms. ^Sen* loses the higher energy either by emitting phosphorescence or by reacting with triplet oxygen. Singlet oxygen can be formed through the energy transferring reaction from the ^Sen* to triplet oxygen by triplet-triplet annihilation mechanisms. The lifetime of^Sen* is longer than 'Sen*.

The returned 'Sen can begin another cycle of generation of singlet oxygen. Kochevar

10 and Redmond (2000) reported that a sensitizer molecule may generate 10’ to 10^ moleeules of singlet oxygen before becoming inactive.

2.1.3 Triplet oxygen oxidation with fatty acids

Triplet oxygen is a diradical compound, which can react with radical compounds.

Food components are not radicals. In order to react with triplet oxygen, one hydrogen atom should be removed from food components to become radicals. Energy required for hydrogen removal from linoleic acid is shown in Figure 5. The removal of hydrogen from saturated fatty acid requires approximately 100 kcal/mol but that of hydrogen at position 11 of linoleic acid is only about 50 kcal/mol (Boff and Min 2002). Therefore, hydrogen between double bonds is most easily removed due to low energy requirement.

100 kcal/mol 50 kcal/mol 75 kcal/mol I I I CH3— (C 1-12)2—CH2—CH2—CH=CH—CH2—CH—CH—CH2— (CH2)6—COOH

Figure 5. Energy required for hydrogen removal from linoleic acid.

(Boff and Min 2002)

11 General mechanism of triplet oxygen oxidation is shown in Figure 6. Triplet oxygen oxidation or autoxidation has 3 steps : Initiation, Propagation, and Termination (Min

1998). Initiation is a step for the formation of free alkyl radicals. The pentadienyl radical is formed from linoleic acid by hydrogen abstraction at Cl 1 position and rearranged to the C9 or 13 position of conjugated radicals at equal concentration. During rearrangement, cis double bond changes to trans double bond. Triplet oxygen can react with conjugated double bond radicals of linoleic acid and produce peroxyl radical at C9 and 13 positions. The peroxyl radicals abstract hydrogen from other fatty acids and become hydroperoxide and generate free alkyl radicals, which is called Propagation step.

The chain reactions of free alkyl radicals and peroxyl radicals accelerate the autoxidation.

Radicals react each other to form nonradical products, which is a Termination step.

Triplet oxygen oxidation produces only the conjugated diene hydroperoxides in linoleic and linolenic acids. The relative reaction ratio of triplet oxygen with oleie, linoleic, and linolenic acid are 1 : 12: 25 (Min et al. 1989). Linolenic acid reacts twice faster than linoleic acid due to the 2 pentadienyl groups. The number of double bonds does not change in triplet oxygen oxidation even though cis conformation of double bond changes to trans conformation. The changes during autoxidation are used for the measuring the degree of oxidation in oil and fat systems. Measuring ‘weight gain’,

‘Peroxide value’, and ‘conjugate diene’ by spectroscopy are common methods to determine the degree of oxidation before the decomposition of hydroperoxides steps, or primary changes.

12 14 13 12 II 10 9 CHj— (CH2)3- CH2~ CH=CH - CH j- CH=CH— CH2~R

INITIATION I - H • (METAL) *

CH3—(CH2)—CH-CH=CH-CH=CH— CH2-R

'O 2

CH3 — (CH2 )4 - ÇH - CH = CH - CH = CH — CH2 - R ? PROPAGATION 9 \ +H*

CH3—(CH2)r CH - CH=CH —CH=CH—CH2— R

HYDROPEROXIDE 9 DECOMPOSITION 0 i - »0H H

CH3—(CH2)—CH-CH=CH-CH=CH— CH2-R O

CH3-(CH2)3-CH2‘ + /C-CH=CH-CH=CH-CH2-R H TERMINATION I + H«

CH3 (CH2 )3 CH3 ^pg^YANE)

Figure 6. General mechanisms of triplet oxygen oxidation with linoleic acid

(Min 1998)

13 2-Thiobarbituric acid value, Oxirane value, p-Anisidien value, and measuring carbonyl groups, hydrocarbons, and fluorescent products are used for determining the degree of autoxidation after the decomposition of hydroperoxide, or secondary changes (Shahidi and Wanasundara 1998).

2.1.4 Singlet oxygen oxidation with fatty acids

Singlet oxygen is non-radical compounds and can directly react with non-radical compounds without radical intermediates like triplet oxygen oxidation.

One of the highest degenerate K antibonding molecular orbitals in singlet oxygen is vacant and singlet oxygen tries to fill the empty molecular orbital with electrons. Singlet oxygen can directly react with electron rich compounds such as unsaturated fatty acids with double bonds (Beutner et al. 2000). The reaction mechanisms of singlet oxygen with electron rich compounds are shown in Figure 7. Singlet oxygen can react with any compound with double bonds through the mechanisms of cycloaddition and “ene” reaction.

14 0 1,4- Cycloaddition o.1 Endoperoxide

“ene” Reaction o O' + I o o . Allyl Hydroperoxide

o CH2 1,2- Cycloaddition + O o CH2 & Dioxetane

Figure 7. Reactions of singlet oxygen with olefins by: 1,4-cycloaddition, the “ene” reaction, and 1,2-cycloaddtion (King 1996; Boff and Min 2002)

One thing should be noticed is singlet oxygen oxidation can break double bonds

and reduce the number of double bond in cycloaddition mechanisms while triplet oxygen

does not reduce the number of double bonds. The reaction rates of singlet oxygen with

oleic, linoleic, linolenic, and arachidonic acids are 0.7, 1.3, 1.9, and 2.4 xlO^ M ''Sec‘', respectively (Doleiden et al. 1974). Singlet oxygen oxidation depends on the number of

15 double bonds instead of types of double bonds, such as conjugated or nonconjugated double bonds.

2.1.5 Comparison of triplet and singlet oxygen oxidation

Triplet and singlet oxygen differ not only the reaction mechanisms but also reaction rates. Triplet oxygen can not directly react with double bonds but singlet oxygen can (Figure 8).

/ — \ / — \ ► / \// ^ Hydroperoxide R R' R ' q R H -O ' at 9, 10, 12, and 13

t /-=^ /=\ ► / \ // ^ ► Hydroperoxide R R. -H - R ^ R' 2 at 9 and 13

Figure 8. Differenee of triplet and singlet oxygen oxidation mechanisms

16 The reaction rates of triplet and singlet oxygen with linoleic acid are 8.9 x lO’ and 1.3 x

10^ M"'Sec'% respectively (Rawls and VanSanten 1970). Singlet oxygen can react 1450

times faster than triplet oxygen with linoleic acid. Relative reaction rates of singlet and

triplet oxygen are shown in Table 2.

C18:l C18:2 C18:3

Triplet Oxygen 1 27 77

Singlet Oxygen 3 x 10^ 4 X 10^ 7 x 10'^

(Min 1998)

Table 2. Relative oxidation rates of triplet and singlet oxygen with oleate, linoleate, and linolenate

Singlet oxygen can react readily with double bonds in unsaturated fatty acids due

to the low activation energy (0 to 6 kcal/mol). Temperature does not affect the reaction

rate of singlet oxygen but has a significant effect on the triple oxygen oxidation, which

requires high activation energy.

Hydroperoxide position formed from oleic,linoleic and linolenic acid by triplet

and singlet oxygen are shown in Table 3. (Frankel 1985). Singlet oxygen can form C9

17 and 10 trans hydroperoxide from oleic acid by “ene” reaction. Triplet oxygen can form

C8, 9, 10 and 11 position hydroperoxides from oleic acid. Frankel (1985) reported that

C8- and 11- hydroperoxides are formed in larger portion than inner hydroperoxide like

C9- and 10- hydroperoxides. In case of linoleic acid, singlet oxygen can form hydroperoxides at C9, 10, 12, and 13 positions while triplet oxygen can form C9 and 13 hydroperoxides. Hydroperoxides from triplet oxygen oxidations are all conjugated forms while singlet oxygen can produce both conjugated and nonconjugated hydroperoxides.

Singlet oxygen can form C9, 10, 12, 13, 15, and 16 positions of hydroperoxides from linolenic acid. Triplet oxygen can produce C9, 12, 13, and 16 conjugated hydroperoxides from linolenic acid. The concentration of hydroperoxides at C12 and 13 positions from linolenic acid is lower than C9 and 16 hydroperoxides due to the inner cyclization

(Frankel 1985). Hydrogen donor can prevent inner cyclization and produce equal amounts of hydroperoxides at C9, 12, 13, and 16 positions by triplet oxygen oxidation.

2.1.6 Effects of singlet oxygen oxidation on food quality

Singlet oxygen can readily react with electron rich moiety in food compounds including fatty acids, amino acids, and vitamins due to the low activation energy. The development of reversion flavor, known as beany or grassy, is a unique characteristic in

soybean oil, which prevents the wide consumption of soybean oil (Ho et al. 1978). 2-

Pentylfuran and 2-pentenyl furan, which are reported as reversion flavor compounds, can he formed in soybean oil containing chlorophyll stored under light only (Callison 2001).

18 Chlorophyll is a well-known photosensitizer in vegetable oils. Boff and Min (2002) reported the formation meehanisms of 2-pentylfuran and 2-pentenyl furan from linoleie and linolenic acids, respectively using singlet oxygen oxidation.

'O2 'O2

8-OOH Oleic acid 9-OOH 9-OOH lO-OOH lO-OOH 11-OOH

9-OOH Linoleic acid lO-OOH 9-OOH 12-OOH 13-OOH 13-OOH

9-OOH lO-OOH 9-OOH 12-OOH 12-OOH Linolenic acid 13-OOH 13-OOH 15-OOH 16-OOH 16-OOH

Table 3. Hydroperoxides formed by singlet and triplet oxygen oxidation of oleic,

linoleic and linolenic acid (Frankel 1985)

19 Singlet oxygen ean reaet with tryptophan, histidine, tyrosine, methionine, and eysteine (Foote 1976; Miehaeli and Feitelson 1997). The reaction of singlet oxygen and amino acids depends on pFI, solvent types, dielectric constants of medium, and the electron rich moiety in amino acids, including the presence of double bonds and sulfiir atom (Miskoski and Garcia 1993; Miehaeli and Feitelson 1994). Sunlight flavor in milk is one of important reactions between singlet oxygen and amino acid. Dimethyl disulfide, described as sunlight flavor in milk, can be formed in milk stored under light only (Jung et al. 1998). Jung et al. (1998) reported the formation mechanisms of dimethyl disulfide from singlet oxygen oxidation of methionine.

Vitamin D, which is an important nutrient for normal mineralization and growth of bones, can be oxidized by singlet oxygen oxidation at the rate of 2.2 xlO^ M'^ s’^ (King and Min 1998; Li et al. 2000). Oxidation of vitamin D in model system did not occur in the absence of either photosensitizer or light. Ascorbic acid, which is an important nutrient in foods, can be destroyed easily by oxidation. Jung et al. (1995) reported ascorbic acid was oxidized by singlet oxygen at the rate of 5.77 x 10* M'^ s'^ in pH 6 potassium buffer at 20 °C. Yang (1994) reported the reaction rate for ascorbic acid with

singlet oxygen as 3.08 xlO* M '‘ s'' in an aqueous solution of pH 7 at 25 °C. The

summary of effects of singlet oxygen on food quality is shown in Table 4.

20 Food types Quality changes Involved chemical compounds

Soybean oil Reversion flavor 2-pentylfuran 2-(2-pentenly) furan

Milk Sunlight flavor Methionine Dimethyl disulfie

Proteins Oxidation of nutrients Tryptophan, Histidine, Tyrosine, Methionine, Cysteins

Vitamins Oxidation of nutrients Vitamin C Vitamin D

Meat Flavor and color change Myoglobin Hemoproteins

(Boff and Min 2002)

Table 4. Effects of singlet oxygen on food quality

2.2 Photosensitizer

2.2.1 Mechanisms of photosensitizer

The reaction mechanisms of photosensitizer with triplet oxygen or other

substrates are shown in Figure 9. The excited triplet sensitizer can react directly with compounds hy donating and abstracting hydrogen or electrons to produce free radicals or free radical ion, which is known as Type I pathway. The hydrogen abstraction of

sensitizer produces free radicals and transferring electron from sensitizer to triplet oxygen

21 produces radical anion like superoxide anion. The rate of the Type I pathway depends on the type and concentration of sensitizers and substrates. Phenols and reducible compounds favor Type I pathway (Korycka-Dahl and Richardson 1978).

Type I mechanism can be explain using one-electron reduction potentials. In case of riboflavin, low (-0.3 V) potential of ground riboflavin is excited to high (1.7 V) triplet flavin, which is higher than that of target compounds like Try (1.5 V), Tyr (0.93 V) or polyunsaturated fatty acids (0.6V) (Buettner 1993; Decker 1998; Lu et al. 1999).

Excited triplet riboflavin can take electron, or hydrogen atom from substrates easily making cation radical, or neutral radical, respectively, due to its relatively high one- electron reduction potential. These radical compounds can initiate radical reactions with each other or triplet oxygen, which is di-radical.

Type 11 mechanism happens when triplet oxygen exist abundantly. Energy transfer from excited triplet sensitizer to triplet oxygen makes singlet oxygen (Foote

1976). Triplet photosensitizer transfer energy to dissolved oxygen using a dye-oxygen complex or energized singlet oxygen. More than 99% of the reaction between triplet oxygen and excited triplet sensitizer produces singlet oxygen.

The rate of Type II pathway depends on the solubility and concentration of oxygen in systems. Oil system with chlorophyll, where triplet oxygen is more soluble, favors Type 11 pathway while aqueous system like milk favors Type 1 pathway due to the limitation of triplet oxygen. The result compounds from Type I and Type 11 pathway accelerate the oxidation reactions by providing reactive free radical species and singlet oxygen.

22 H transfer (Hydrogen atom) , back reaction Sen* + RH ------> «Sens H + R« ^ Sens + RH

electron transfer

’Sen* + R ------> Sens '. + R* . back reaction^ ^

Sens '# + R^ • > O2 * or HO 2 # or «ROi ------> H2 O2

+ O2

Sen H • or R^ • or Sen • ^ O 2 * or HO 2 * ------> H2 O2

Figure 9. Type I pathway of sensitizer

Addition of different scavengers and singlet oxygen quenchers (Table 5) can help to understand the mechanisms of kinetics and types of photosensitizers (Edwards and

Silva 2001(43)). Nitrogen or oxygen gas bubbling can be another technique to understand the radical process in photosensitized system.

23 Agents Chemical compounds Effects

Singlet oxygen quencher NaN], sodium azide, (0.1-5 Type II confirm. If mM), Histidine (55.0 mM) effected, 'O 2 is involved.

Singlet oxygen enhancer D2O If effected, 'O 2 is involved.

Electron acceptor K3Fe(CN)6 (0.1 mM) If effected, electron is involved. Type 1

H2 O2 trap Catalase (10.0 U Sigam/ml) Type I

Superoxide trap •O 2 ’ Superoxide dismutase (SOD) 2 .O 2 + 2H + ^H 2Û 2 + 3.5 X 10"^ M O2 , Type I

H.O 2 ' + # 0 2 ' + H2 O2 + UV ^ 2 .0 H HO2 ^ O2 ' + H+ H2 O2 + O2 (pH dependent)

•OFI trap Sodium benzoate (1.0 mM) Type I Mannitol, Salicylic acid

Cu(II) chelator Neocuproine (lOOpM) Copper chelating and can be used for antioxidant.

(Edwards and Silva 2001)

Table 5. Function of various agents on the mechanism of photosensitization

24 2.2.2 Riboflavin

Riboflavin is a 7,8-dimethyl-lO-(r-D-ribityl) isoalloxazine (Figure 10).

Riboflavin is present as free form or flavin mononucleotide (FMN) or flavin adenine dinucleotide (FAD) in all aerobic cells. FAD and FMN can act as oxidases or dehydrogenases to transfer electrons in oxidation-reduction reactions in biological

systems. A deficiency in riboflavin may affect the metabolisms of glucose, fatty acids, and amino acids. The excess of riboflavin did not produce toxic effects (King 1996).

CH 2 - (CH0H)3 - CH 2 O H N, ,N.r ^ ,NH

o

Figure 10. Structure of riboflavin

The recommended daily allowance (RDA) for riboflavin is 0.6 mg (Cooperman and

Lopez 1984). The concentration of riboflavin is reported 0.17 mg/100 g in milk and 3.5 mg/ lOOg in liver. The stability of riboflavin is influenced by pH, heat, and light.

Riboflavin is stable at acidic and neutral conditions but less stable in basic environment.

25 Riboflavin can undergo redox cycling easily and one hydrogen reduced form, flavosemiquinone radical and two hydrogen reduced compound, flavohydoquinone can be formed. Riboflavin is very sensitive to light exposure, especially 450 nm wavelength.

The loss of riboflavin was as high as 30% during a 30 min fluorescence light exposure at room temperature, while only a 12% loss occurred in boiling processing. The reduction of riboflavin during light exposure is due to the abstraction of hydrogen atom from coplane of rings.

Riboflavin has been known as a photosensitizer which can be excited by light and react with triplet oxygen (Type II) or by substrates (Type I). Type I reaction of riboflavin is favored due to the low concentration of oxygen in water matrix and easy oxidation- reduction property of riboflavin (McGinnis et al. 1999; Gutierrez et al 2001). However, the generation of singlet oxygen from riboflavin has been confirmed using Electron Spin

Resonance (ESR) and singlet oxygen trapping compound like 2,2,6,6-tetramethyl-4- piperidone (Bradley 1991) (Figure 11). The photosensitizing effect of riboflavin on the oxidation of many compounds, including vitamin D 2 , hydroxypyridine, has been reported

(Li et al. 2000; Pajares 2001). The destruction of riboflavin under light in nitrogen gas is less than in oxygen gas, which suggested that not only light energy but also singlet oxygen produced by riboflavin itself can be related with the loss of riboflavin.

26 0 MINUTES

5 MINUTES

15 MINUTES

f

3 3 7 0 G 3390 G 3 4 1 0 G

Figure 11. Effects of 0, 5 and 15 min illumination on electron spin resonance spectrum of 2,2,6,6-4-piperidone-N-oxyl in water solution of riboflavin and 2,2,6,6- tetramethyl-4- piperidone (Bradley 1991)

27 Riboflavin and FMN are photosensitizers and induee the modification of other eompounds by transferring energy of visible light. Inactivation effects of riboflavin photosensitization (446-450- 452 nm) were studied on many enzymes including catalase, horse-radish peroxidase, lysozme, and aromatic amino acids like Trp and Tyr (Edward and Silva 2001).

Ascorbic acid protects riboflavin and other target amino acids by being oxidized first in the presence of oxygen (Lee at al. 1998). The oxidation of ascorbic acid was accelerated by higher light intensity and more riboflavin content. Cysteine showed

strong antioxidant activity on the riboflavin-sensitized oxidation on ascorbic acid (Jung et al. 1995).

Riboflavin can be found in dairy products without exceptions and induce the changes of food quality. Riboflavin oxidized various substrates such as malondialdehyde, tryptophan, ethylene glycol, and ascorbic acid (King 1996; Jung et al.

1995; Edwards and Silva 2001). Riboflavin ean induce sunlight flavor from light exposed milk by singlet oxygen oxidation (Jung et al. 1998; Skibsted 2000). Riboflavin can photooxidize ascorbic acid 1.4 times and 21 times faster than methylene blue and protoporphyrin, respectively (Jung et al. 1995).

28 2.2.3 Chlorophyll

Chlorophylls are magnesium complexes derived from porphins with tetrapyrrole rings (Figure 12). Chlorophyll a and b, which have different substituents, are found in green plants in an approximate ratio of 3:1 (Elbe and Schwartz 1996). Many chlorophyll derivatives have been identified from plants, marine algae, and photosynthetic bacteria.

H

chi a R : CHg

Mg. chib R:CHO

CH

Figure 12. Structure of chlorophyll

29 One of the most important characteristics of chlorophyll and its derivatives is the absorption of visible light between 400-500 nm and 600-700nm. Chlorophyll a and b in ethyl ether showed the maximum absorption at 660.5 and 642 nm, respectively, in the red region and 428.5 and 452.5 nm in the blue region, respectively.

Chlorophyll is light sensitive and easily degraded into colorless photodegradation compounds including methyl ethyl maleimide and glycerol (Jen and Mackinney 1970;

Schwartz and Lorenzo 1990). Photodegradation of chlorophyll results in the opening of tetrapyrrole ring. Singlet oxygen and hydroxyl radicals, which can be formed from chlorophyll exposed to light in the presence of oxygen, can react with tetrapyrroles of chlorophyll and destroy the porphyrins of chlorophyll and become colorless compounds

(Elbe and Schwartz 1996).

Chlorophyll has been used to accelerate the lipid oxidation in olive oil (Kiritsakis and Dugan 1985), soybean oil (Jung and Min 1991; Lee et al.l997), and fatty acid model

systems (Rawls and VanSanten 1970; Jung et al. 1999). Rawls and VanSanten (1970) reported that chlorophyll works as a photosensitizer through mostly Type II mechanism

(singlet oxygen generation) while Type I mechanism of chlorophyll is involved in some degree.

30 CHAPTER 3

Photooxidation and Photosensitized Oxidation of Linoleic Acid

3.1 Abstract

Volatile compounds of linoleic acid with or without added 5 ppm chlorophyll under light and dark were studied by solid phase microextraction (SPME)-gas chromatography (GC)-mass spectrometry (MS). Total volatile compounds in linoleic acid without added chlorophyll under light and in the dark did not increase and the same for 48 hr at 4 °C. Total volatile compounds in linoleic acid with added 5 ppm chlorophyll under light at 4 °C for 0, 6, 12, 24, and 48 hr, increased from 8.9 to 11.6,

21.7, 26.1, 29.3 (1x10'*) in electronic counts, respectively. 2-Pentylfuran, an undesirable reversion flavor compound in soybean oil, 2-octene-l-ol, 2-heptenal and l-octene-3-ol were formed by photosensitized oxidation only. Light excited photosensitizers like chlorophyll can generate singlet oxygen from ordinary triplet oxygen. 2-Pentylfuran, 2- heptenal, and l-octene-3-ol can come from CIO, C l2, and CIO hydroperoxide of linoleic acid, respectively, which can be formed by singlet oxygen oxidation but not by triplet 31 oxygen oxidation on linoleic acid. The singlet oxygen oxidation mechanisms for 2- pentylfuran 2-heptenal, l-octene-3-ol, and 2-octene-l-ol from linoleic acid were proposed.

3.2 Introduction

The oxidation of lipid containing foods causes off-flavor, loss of nutrients and production of undesirable compounds during manufacturing and storage, which decreases food quality and makes the foods less acceptable or unacceptable to consumers. Lipid can be oxidized by free radicals, lipoxygenase, and singlet oxygen (Chan 1977; Frankel

1985; Zhuang et al. 1998).

Autoxidation or triplet oxygen oxidation of lipids is a free radical chain reaction on unsaturated fatty acids. Free radical chain reaction can be initiated by metal catalysis heat, or ultraviolet irradiation, and produce hydroperoxides of unsaturated fatty acids

(Frankel 1985). The decomposition of lipid hydroperoxides produces volatile compounds, including aldehydes, alcohols, ketones, and hydrocarbons, which are responsible for the increase of off-odor and the decrease of food qualities (Min 1998).

Light-induced lipid oxidation has been studied on unsaturated fatty acids and vegetable oils with or without addition of photosensitizers (Frankel et al. 1976; Frankel et al. 1981; Faria 1983). Even though the energy of visible light is not absorbed by unsaturated fatty acids, photooxidation can induce oxidation in linoleic acid and

safdower oil exposed to ultraviolet (UV), incandescent, and fluorescent light, which may

32 be due to the presence of impurities ineluding photosensitizers (Faria 1983). UV light photooxidation may aceelerate lipid oxidation by generating the alkoxyl radicals and hydroxyl radicals from hydroperoxides (Faria 1983).

Photosensitizers, including chlorophyll, riboflavin, and methylene blue, can accelerate lipid oxidation (Yang and Min 1994). Singlet state of photosensitizer can absorb light energy and become excited singlet state photosensitizer. The excited singlet

state photosensitizer becomes to the excited triplet state photosensitizer by intersystem crossing mechanisms. The excited triplet state photosensitizer reacts with triplet oxygen to form singlet oxygen (Type II mechanism) or abstracts electron or hydrogen atom from a substrate to generate radicals (Type I mechanism) (Foote 1976; Bradely 1992).

Singlet oxygen can play important roles in the photosensitized oxidations in foods due to the relatively long lifetime. The energy state of singlet oxygen is 22.4 kcal/mole and the lifetime is ranging 50 to 700 microseconds depending on the solvent types, which is long enough to react with other food components with electron rich moiety (Korycka-

Dahl and Richradson 1978; Girotti 1998). The activation energy of singlet oxygen is 0 to

6 kcal/mole, which is low enough for the oxidation reactions (Yang and Min 1994). The reaction rates of triplet and singlet oxygen with linoleic acid are 8.9 x lO' and 1.3 x 10^

M ''s’', respectively (Rawls and VanSanten 1970).

Linoleic acid is one of the major unsaturated fatty acids in vegetable oils including soybean oil (51%), com oil (61%), and sunflower oil (67%) (Chu and Kung

1998). The isomeric stmctures of hydroperoxides from fatty acids by autoxidation and photosensitized oxidation have been studied extensively (Frankel et al. 1981; Frankel

1985). However, the effects of photooxidation and chlorophyll photosensitized oxidation

33 on the formation of volatile compounds from linoleic acid and the formation mechanisms of light-induced volatiles were not fully understood.

The objectives of this work were (1) to study the effects of light and chlorophyll on the formation of volatile compounds from linoleic acid model system and (2) to propose the formation of volatile compounds from linoleic acid by photosensitized oxidation.

3.3 Materials and Method

Materials

Chlorophyll b, linoleic acid (99%), 1-pentanol, hexanal, 2-hexenal, 2-heptenal, 1- octene-3-ol, 2,4 heptadienal, 2,4 octadienal, 2-nonenal, octanoic acid, and 2,4 decadienal were purchased from Sigma Chemical Co. (St. Louis, MO., U.S.A.). Teflon-coated rubber septa, aluminum caps, serum bottles, glass liners, the fiber assembly holder, and

65 pm Polydimethylsiloxane/Divinylbenzene (PDMS/DVB) were purchased from

Supelco, Inc. (Bellefonte, PA., U.S.A.). 2-Pentylfuran was purchased from Karl

Industries Inc. (Aurora, OH., U.S.A.).

Sample preparation for the light and dark storage o f linoleic acid model system

Chlorophyll was dissolved in linoleic acid to obtain 5 ppm (weight/volume) while magnetic stirring. Samples of 0.10 g linoleic acid with or without added 5 ppm chlorophyll were put in 10-mL serum vials (25 x 40 mm, 20 mm diameter from Supelco,

34 Inc.) and sealed air-tight with Teflon-eoated rubber septa and aluminum caps. Sample bottles of lionleie acid with or without added 5 ppm ebloropbyll were kept at 4 °C in a fluorescent light box in a cold room of the Department of Food Science and Technology

of the Ohio State University (Lee and Min 1988). One set of sample bottles was wrapped with aluminum foil to make a dark eondition. Fluorescent light souree was four Sylvania

fluoreseent lamps (GE, Cleveland, OH., U.S.A.) with 1,500 lux light intensity. Volatile

compounds of sample bottles were analyzed at 0, 6, 12, 24, and 48 hr by SPME. Sample bottles were prepared in duplicate at each sample analysis.

Analysis of volatile compounds by SPME

The headspace volatile eompounds in air-tight sealed sample bottles were isolated by 65 jam PDMS/DVB of SPME solid phase. Sample bottles were put in a 30 °C water bath for 30 min while 65 |am PDMS/DVB of SPME solid phase was exposed to the headspace. The isolated volatile compounds by SPME solid phase were separated in a

gas chromatograph (GC) equipped with a flame ionization detector.

Gas chromatography condition

A Hewlett-Packard 5890 gas chromatograph was equipped with a 0.75 mm ID

glass injection liner, a flame ionization detector, and a 30 m x 0.32 mm ID, 0.25 |im film,

HP-5, from Agilent Technologies (Palo Alto, CA., U.S.A.). The oven temperature was held at 40 °C for 2 min and increased from 40 to 160 °C at 6 °C/min and from 160 to 220

°C at the rate of 10 °C/min. The temperatures of injector and detector were 250 and 300

°C, respectively. The flow rafe of nifrogen carrier gas was 1.0 mL/min. The isolated 35 volatile compounds in solid phase of SPME were desorbed at 250 °C of GC injector for 2 min.

Identification o f volatile compounds

A Hewlett-Packard 5971A mass selective detector (MS) equipped with a Hewlett-

Packard 59822 B ionization gauge controller was used. All mass spectra were obtained at 70 eV and 220 °C ion source temperature. The identification of compounds was made by the combination of NIST Mass Spectra and gas chromatographic retention times of

standard compounds. Helium carrier gas at 0.9 ml/min and a 30m x 0.25 mm i.d., 0.25

|im film thickness, HP-5 from Agilent Technologies (Palo Alto, CA., U.S.A.) were used.

The GC conditions for GC-MS were the same as the gas chromatographic analysis conditions described previously.

Statistical analysis

One-way analysis of variance, Tukey’s multiple comparisons method, and general linear model were used to analyze the data. A P value of 0.05 or less was considered

significant. All statistical analyses were conducted with Minitab 12.1 (Minitab Inc., State

College, PA, U.S.A.)

36 3.4 Results and Discussion

SPME analysis o f volatile compounds from linoleic acid

Gas chromatograms of headspace volatile compounds from linoleic acid with added 5 ppm chlorophyll stored at 4 °C for 48 hr by SPME are shown in Figure 13. The coefficient of variation in total GC peak areas of gas chromatograms was 1.8% as shown in Figure 13, which indicates that SPME method on the headspace volatile compounds in linoleic acid model systems is reproducible. SPME has been used to analyze the headspace volatile eompounds in linseed and sunflower oil (Jelen et al. 2000), and

soybean and eanola oil (Steenson et al. 2002).

SPME solid phase of PDMS/DVB was seleeted due to the high sensitivity on the headspace volatile compounds. Preliminary study showed that PDMS/DVB and

CAR/PDMS had better sensitivity than polydimethylsiloxane (PDMS) solid phase on headspace volatiles from soybean oils. Jelen et al. (2000) reported that SPME solid phases containing Divinylbenzene (DVB), carboxen™,and PDMS showed the best

sensitivity on the volatile compounds from oxidized linseed oil and sunflower oil.

Divinylbenzene (DVB) and carhoxen™ are porous materials that can retain chemical compounds inside the pore (Shirey 1999), which can increase the sensitivity compared to the non-porous materials including PDMS and polyacrylate.

37 Total peak area : 2.93 X 10^ in electronic counts

Total peak area : 2.87 X 10^ in electronic counts

Total peak area : 2.98 X 10^ in electronic counts

0 10 155 20 Time (min)

Figure 13. Gas chromatograms of headspace volatile compounds in linoleic acid containing chlorophyll stored under light for 48 hr. 38 Volatiles from linoleic acid model system

Changes of headspace volatile compounds in total GC peak areas from linoleic acid with or without added chlorophyll stored at 4 °C under light and dark for 48 hr are

shown in Figure 14. Total volatile compounds in linoleic acid without added chlorophyll under light and in the dark did not increase and the same for 48 hr at 4 °C. Total volatile compounds from linoleic acid with added chlorophyll under light were significantly higher than linoleic acid without added chlorophyll stored under light for 48 hr (P<0.05).

Total volatile compounds in linoleic acid with added 5 ppm chlorophyll under light at 4

°C for 0, 6, 12, 24, and 48 hr, increased from 8.9 to 11.6, 21.7, 26.1, 29.3 (IxlO"^) in GC electronic counts, respectively. Lee and Min (1988) reported that headspace volatile compounds in soybean oil increased only in the presence of chlorophyll and light.

The photosensitizing effects of chlorophyll can explain the rapid increase of volatile compounds from linoleic acid with added chlorophyll under light for 48 hr at 4

°C. Photosensitizers exposed to light can oxidize substrates through Type 1 or Type II mechanisms depending on the solubility and concentration of oxygen. The solubility of oxygen is higher in lipid and nonpolar solvents than in water (Korycka-Dahl and

Richardson 1978). Singlet oxygen generation (Type II mechanism) is preferred in linoleic acid by the light-excited chlorophyll due to the high solubility of triplet oxygen in lipid phase (Bradely and Min 1992).

39 3.5

3.0

a 3 uo 2.5

3 . With chlorophyll I 2.0 under light O o .2 5 Without chlorophyll 1.5 under light

s a 1.0 o H Linoleic acid 0.5 in the dark

0.0 0 12 24 36 48 Time (hr)

Figure 14. The effects of chlorophyll on the volatile compounds in linoleic acid stored at 4 °C under light and dark for 48 hr

40 Identified volatile compounds in linoleic acid model systems

Volatile compounds identified from linoleic acid with or without added chlorophyll stored at 4 °C under light for 0 and 48 hr are shown in Table 6. Some degree of oxidation already occurred in the linoleic acid samples used in this study. Pentane, chloroform, pentanal, 1-pentanol, hexanal, 2-octene, 2-butyl furan, limonene, 2-nonenal, and 2,4 decadienal were identified from 0 hr sample of linoleic acid.

Linoleic acid with added chlorophyll stored at 4 °C under light for 48 hr produced

2-hexenal, 2-heptenal, l-octene-3-ol, 2-pentylfuran, 2,4 heptadienal, 3-octene-2-one, 2- octenal, 2-octene-l-ol, 2,4 octadienal, methyl octanoic acid, and octanoic acid, which were not identified in linoleic acid without added chlorophyll and the linoleic acid stored under dark at 4 °C. 2-Pentylfuran is an undesirable reversion flavor in soybean oil.

Callison (2001) identified 2-pentylfuran in soybean oil containing 5 ppm chlorophyll

stored under light. However, 2-pentylfuran was not identified in soybean oil without added chlorophyll or without light conditions.

l-Octene-3-ol, 2-octenal, and 2-heptenal were detected in the highest abundance from linoleic acid with added chlorophyll. Frankel (1985) reported that 2-heptenal and 1- octene-3-ol were formed in trace concentration from autoxidation of methyl linoleate but photooxidation of methyl linoleate produced 9.9 and 1.9 relative percentage of 2-heptenal and l-octene-3-ol, respectively. Callison (2001) reported 2-heptenal was formed in

soybean oil only in the presence of chlorophyll and light.

The formation of light-induced volatile compound in chlorophyll photosensitized oxidation of linoleic acid can be explained by singlet oxygen oxidation. The reaction mechanisms of singlet and triplet oxygen on unsaturated fatty acids are different (Frankel 41 1985). Triplet oxygen, which is a di-radical, needs radical forms of unsaturated fatty acids to form hydroperoxides. The radicals of linoleic acid can be formed by removing hydrogen atom at Cl 1 position in pentadiene structure of linoleic acid with high temperature, metal catalysis or UV irradiation. Singlet oxygen does not need radical forms of unsaturated fatty acids. Singlet oxygen can directly react with electron rich double bonds of unsaturated fatty acids through 1, 4 cycloaddition, “ene” reaction, and 1,

2 cycloaddition and generate hydroperoxides (Boff and Min 2002).

Singlet oxygen can form hydroperoxide at C9, 10, 12, or 13 positions in linoleic acid while triplet oxygen can form hydroperoxide at C9 or 13 positions in linoleic acid

(Frankel 1985). To produce 2-pentylfuran, 2-heptenal, and l-octene-3-ol, hydroperoxide of linoleic acid should be formed at CIO, Cl 2, and CIO position, respectively, which can he formed hy singlet oxygen oxidation but not by triplet oxygen oxidation on linoleic acid. 2-Octene-l-ol can come from the cleavage of Cl 1 hydroperoxide of linoleic acid, which can he formed hy singlet oxygen oxidation on the double bond of either C9 or C13 hydroperoxide.

42 Volatile 0 hr sample Light sample Light sample Compounds without added with added chlorophyll chlorophyll Pentane" 1.2 5.2 7.0 C h lo r o f o rm 0.3 0.3 0.3 Pentanal 6.3 144 14.6 1-Pentanol 0.1 2.9 2.0 Hexanal 19.8 2L8 15.3 2-Octene 1.4 0.7 0.6 2-Hexenal n.d. n.d. 3.5 2-Butyl furan 0.5 1.1 0.9 2-Heptenal n.d. n.d. 184 l-Octene-3-ol n.d. n.d. 1394 2-Pentylfuran n.d. * n.d. 5.5 2,4 Heptadienal n.d. n.d. 1.0 Limonene 0.4 0.4 0.4 3-Octene-2-one n.d. n.d. 2.1 2-Octenal n.d. n.d. 12.1 2-Octene-l-ol n.d. n.d. 0.7 2,4 Octadienal n.d. n.d. 0.6 2-Nonenal 0.5 0.5 0.5 Octanoic acid n.d. n.d. 2.1 2,4 Decadienal 0.1 0.2 1.4

“ : Bold character represents a compound, which was not identified in photooxidation but identified in photosensitized oxidation of linoleic acid, n.d. : Not detected ; a compound identified by GC-MS only. The other compounds were identified by both GC retention time of standard compound and GC-MS. Data did not vary more than 5% of coefficient of variation.

Table 6. Volatile compounds identified from linoleie acid with or without added

chlorophyll stored under light at 4°C for 0 and 48 hr in GC peak areas (1 xlO ) 43 Proposed formation mechanisms of 2-pentylfuran from linoleic acid by singlet oxygen oxidation are shown in Figure 15. Singlet oxygen oxidation on linoleic acid produces CIO hydroperoxide with nonconjugated diene. The cleavage of C9-C10 bond of CIO linoleic acid hydroperoxide produces 3-nonenal, which can react with singlet oxygen and form 2-pentylfuran (Figure 15).

Proposed formation mechanisms of 2-heptenal from linoleic acid by singlet oxygen oxidation are shown in Figure 16. 2-FIeptenal has been identified as a unique product of photosensitized oxidation in methyl linoleate (Frankel et al. 1979; Franekel

1985). The presence of 2-heptenal suggested that singlet oxygen oxidation was taken place in the sample. Singlet oxygen can form nonconjugated hydroperoxide at C12 position in linoleic acid. 2-Heptanal is produced from the cleavage of C11-C12 bond of linoleic acid hydroperoxide with C12 (Figure 16).

Proposed formation mechanisms of l-octene-3-ol and 2-octene-l-ol from linoleic acid by singlet oxygen oxidation are shown in Figure 17. Singlet oxygen produces C9 hydroperoxide with conjugated diene or CIO hydroperoxide with nonconjugated diene from linoleic acid. The Cl 1 hydroperoxide of linoleic acid is formed by 1,2 cycloaddition of singlet oxygen oxidation on the C9 hydroperoxide of conjugated diene.

2-Octene-l-ol, which was identified only in chlorophyll sensitized linoleic acid, can be formed from the cleavage of ClO-Cl 1 bond of Cl 1 linoleic acid hydroperoxides. The cleavage of ClO-Cl 1 bond of CIO linoleic acid hydroperoxide with nonconjugated diene produces 2-octene, which can react with singlet oxygen through “ene” reaction and produce l-octene-3-ol (Figure 17).

44 CH]— (CH2)4 - c h = c h - CH 2 —CH= CH - CH2 — (CH2)6—COOH

i 'O 2

CH]— (CH2)4 - c h = c h - CH 2 - CH - CH= CH - (CHi)^—COOH b

CH]— (CH2)4 - CH= c h - CH 2 - CH

i 'O 2 "

CH]— (CH2)4 - CH - CH - CH2 - CH O — O O

CH]— (CH2)4 - CH - ÇH - CH2 - CH O— o* o

l +2RH

CH]— (CH2)4 - CH - CH] - CH2 - CH + 2 R. 0 O 1 I OH [

CH]— (CH2)4 - CH - CH] - CH] - CH O O

CH]— (CH])4 - C - CH] - CH] - CH O I O

CH]—(CH])4-C=CH-CH=CH OH OH I -H ]0

CH]—(CH])4— / I O

Figure 15. Proposed formation mechanisms of 2-pentylfuran from linoleic acid by singlet oxygen oxidation 45 CH3— (CH2)3— CHj —C H = C H -C H 2 — C H = C H -C H 2 — (CH2)6— COOH

+ ^Ü2

CH3 — (CH2 ) 3 — CH= CH - CH - CH2 — CH = CH - CH2 — (CH2 )6 —COOH 0 1 OH OH

CH3 — (CH2 ) 3 — C H = C H - CH f CH2 — CH = CH - CH2 — (CH2 )g— COOH I o

CH3 —(CH2 ) 3 —CH=CH-CH + .CH 2 —CH = CH-CH 2 — (CH2 ) 6 —COOH O

CH] — (CH2 ) 3 — C H = CH - CH o

2-Heptenal

Figure 16. Proposed formation mechanisms of 2-heptenal from linoleic acid by singlet oxygen oxidation

46 R-CH2-CH=CH-CH2-CH=CH-CH2-R'

' 0 2

R-CH2-CH=CH-CH=CH-CH-CH2-R R-CH2-CH=CH-CH2-CH-CH=CH-R OOH OOH

I + 'O 2 , 1,2 c y c lo a d d itio n I -.O H

R-CH 2 —CH-CH-ÇH—ÇH-ÇH-CH 2 —R' R -C H 2 —C H = C H -C H 2 -|-CH-CH=CH—R' I I I I I O O OOH O

+ 2H. + O 2 , “ e n e ” re a c tio n

R-CH2-CH=CH-CH-CH2-CH-CH2-R + 2H. OOH OOH R-CH2-CH-CH=CH2 I -.O H OOH

-.OH R-CH2-CH=CH-CH-I-CH2-CH-CH2-R i ■* O ' OOH R-CH2-CH-CH=CH2 2H. o

j + H . R-CH2-CH=CH-CH2 OH R-CH2-CH-CH=CH2 OH 2-Octene -l-ol l-Octene-3-ol R C H 3 — ( C H 2 ) 3 R: (CH2)6-C00H

Figure 17. Proposed formation mechanisms of l-octene-3-ol and 2-octene-l-ol from linoleic acid by singlet oxygen oxidation

47 3.5 Conclusion

Formation of volatile compounds from linoleic acid by photosensitized oxidation and photooxidation at 4 °C were studied. Photosensitizers like chlorophyll play important roles in the formation of volatile compounds. Singlet oxygen oxidation is mainly responsible for the formation of 2-pentylfuran, 2-heptenal, l-octene-3-ol, and 2- octene-l-ol in photosensitized oxidation of linoleic acid. Photosensitizer content in food

should he minimized and food products should be packed with light impermeable package to prevent the photosensitized oxidation on lipids.

3.6 References

Boff JM, Min DB. 2002. Chemistry and reaction of singlet oxygen in foods.

Comprehensive review in Food Science and Food Safety l(l):58-72.

Bradley DG. 1991. Riboflavin photosensitized singlet oxygen oxidation in milk products.

Dissertation in the Ohio State University. 109 pp.

Bradley DG, Min DB. 1992. Singlet oxygen oxidation of foods. Grit Rev Food Sci Nutr

31:211-236.

48 Callison AL. 2001. Singlet oxygen lipid oxidation mechanisms for reversion flavor formation in soybean oil [MS thesis]. Columbus, Ohio: The Ohio State University,

Columbus, p 50-86.

Chan HWS. 1977. Photo-sensitized oxidation of unsaturated fatty acid methyl esters. The identification of different pathways. JAOCS. 54(3): 100-104.

Chu YH, Kung YL. 1998. A study on vegetable oil blends. Food Chem 62(2): 191-195.

Faria JAF, Mukai MK. 1983. Use of gas chromatographic reactor to study lipid photooxidation. JAOCS 60(1): 77-81.

Foote CS. 1976. Photosensitized oxidation and singlet oxygen: consequences in biological systems. In: Pryor, WA, editor. Free Radicals in Biology. Vol. 2. New York:

Academic Press, p 85

Frankel EN. 1985. Chemistry of autoxidation: Mechanism, Products and Flavor

Significance. In: Min DB, Smouse TH, editors. Flavor chemistry of fats and oils.

Champaign, IL: AOCS. p. 1-37.

Frankel EN, Neff WE, Bessler TR. 1979. Analysis of autoxidized fats by gas chromatography-mass spectrometry: V. Photosensitized oxidation. Lipids 14(12):961-

967.

49 Frankel EN, Neff W, S elk E. 1981. Analysis of autoxidized fats by gas chromatography- mass spectrometry: VIL Volatile thermal decomposition products of pure hydroperoxides from autoxidzed and photosensitized oxidized methyl oleate, linoleate and linolenate.

Lipid 16(5):279-285.

Girotti AW. 1998. Lipid hydroperoxide generation, turnover, and effector action in biological systems. J Lipid Res 39:1529-1542.

Jelen HH, Obuchowska M, Zawirska-Wojtasiak R, Wasowicz E. 2000. Headspace Solid-

Phase Microextraction Use for the Characterization of Volatile Compounds in Vegetable

Oils of Different Sensory Quality. J Agrie Food Chem 48(6): 2360-2367.

Koryeka-Dahl MB, Riehardson T. 1978. Activated oxygen species and oxidation of food constituents. Crit Rev Food Sci Nutr 10:209-240.

Lee EC, Min DB. 1988. Quenching mechanism of ^-carotene on the chlorophyll

sensitized photooxidation of soybean oil. J Food Sci 53(6): 1894-1895.

Min DB. 1998. Lipid oxidation of edible oil. In: Akoh CC, Min DB, editors. Food Lipids: chemistry, nutrition, biotechnology. New York : Marcel Dekker. p. 283-296.

50 Rawls HR, VanSanten PJ. 1970. A possible role for singlet oxidation in the initiation of fatty aeid autoxidation. J Am Oil Chem Soc 47:121-125.

Shirey RE. 1999. SPME fibers and selection for specific applications. In: Wercinski SAS, editor. Solid Phase Microextraction. A practical guide. New York: Marcel Dekker. p 59-

110.

Steenson DF, Fee JH, Min DB. 2002. Solid phase mieroextraetion of volatile soybean oil and com oil compounds. J Food Sci 67(l):71-76.

Yang WTS, Min DB. 1994. Chemistry of singlet oxygen oxidation of foods. In: Ho CT and Hartmand TG, editors. Fipids in Food Flavors. Washington DC: American Chemical

Society, p 15-29.

Zhuang H, Barth MM, Hildebrand DF. 1998. Fatty acid oxidation in plant tissues. In:

Akoh CC, Min DB, editors. Food Fipids: chemistry, nutrition, biotechnology. New York

: Marcel Dekker. p. 333-376.

51 CHAPTER 4

Photosensitized Oxidation of Milk

4.1 Abstract

Milk with or without added riboflavin, ascorbic acid, sodium azide, and butylated hydroxyanisol (BHA) was stored at 4 °C under light and in the dark. Pentanal, dimethyl disulfides, hexanal, and heptanal were formed only in milk under light and increased

significantly as the added riboflavin concentration increased from 5 to 10, 50 ppm

(P<0.05). As fat content in milk increased from 0.5 to 1.0, 2.0, and 3.4%, pentanal, hexanal, and heptanal increased significantly (P<0.05) while dimethyl disulfide concentration did not change. BHA and ascorbic acid, hydrogen donating free radical

scavengers, reduced the hexanal and heptanal formation. Sodium azide, a singlet oxygen quencher, prevented dimethyl disulfide formation. Formation of pentanal is different from that of hexanal and heptanal in milk. Singlet oxygen and free radicals play important roles in the formation of the light-induced volatile compounds in milk.

52 4.2 Introduction

Off-flavors in dairy products can be formed from microbial spoilage, packaging materials, and chemical reactions including lipid or amino acid oxidation caused by light exposure or pro-oxidant metals (Marsili 1999a). Light exposure can decrease the nutritional value and the flavor quality of dairy products (Whited et al. 2002), and this light-induced off-flavor makes the dairy products less acceptable to consumers.

Light exposure in milk can induce two distinctive off-flavors. One is “sunlight” flavor, which gives a burnt and oxidized odor predominant for two or three days in milk under light. The other is “cardboard-like or metallic” flavor, which develops in milk with prolonged light exposure and does not dissipate (Skibsted 2000). Dimethyl disulfide and methional, which can be formed from the oxidation of sulfur containing amino acids like methionine, are responsible for the “sunlight” flavor (Gaafar and Gaber 1992; Jung et al.

1998). The “cardboard-like or metallic” flavor comes from the secondary lipid oxidation products including hexanal, pentanal, ketones, alcohols and hydrocarbons (Skibsted

2000).

Riboflavin is a well-known photosensitizer, which can absorb light energy and react with non-photo reactive compounds. Riboflavin plays important roles in the formation of light-induced off-flavor in dairy products (Edwards and Silva 2001; Boff and Min 2002). Ground state of singlet riboflavin becomes an excited single state by light exposure and this excited singlet riboflavin changes to the excited triplet riboflavin by the intersystem crossing mechanisms. The excited triplet riboflavin has lifetime long enough to react with triplet oxygen (Type II mechanism) or to abstract electron or

53 hydrogen from a substrate to generate radicals (Type I mechanisms) (Foote 1976). The reduction potential of triplet riboflavin is about 1.7 V at pH 7 (Lu et al. 1999), which is high enough to abstract electron or hydrogen atom from food compounds including polyunsaturated fatty acids, ascorbic acid, and tocopherol (Decker 1998). The presence of singlet oxygen was reported in the sunlight or fluorescent light exposed milk by electron spin resonance spectroscopy (Bradley 1991). Singlet oxygen in milk, which is generated from triplet oxygen by the excited triplet riboflavin (Type II mechanisms), can initiate and accelerate the lipid oxidation due to the low activation energy of singlet oxygen (Min 1998).

Dimethyl disulfides, pentanal, hexanal and heptanal were detected from the light exposed milk by dynamic headspace method (Kim and Morr 1996) or by SPME method in parts per billion levels (Marsili 1999b). However, the formation mechanisms of light- induced volatile compounds in milk are not fully understood. The effects of fat content, riboflavin content, singlet oxygen quencher or free radical scavengers on the light- induced volatiles in milk are not reported in the literature.

The objectives of this work were (1) to study the effects of riboflavin and fat content on the fluorescent light-induced volatile compounds in milk and ( 2 ) to determine the formation of volatile compounds using singlet oxygen quencher and free radical

scavengers in the photosensitized milk.

54 4.3 Materials and Method

Materials

Commercial whole milk, 1% fat milk, 2% fat milk, and skim milk were purehased from local grocery store (Meijer, Columbus, OH., U.S.A.). Teflon-eoated rubber septa, aluminum caps, serum bottles, glass liners, the fiber assembly holder, and 75 |im

Carbowen/polydimethylsiloxane (CAR/PDMS) were purchased from Supeleo, Ine.

(Bellefonte, PA., U.S.A.). Riboflavin, ascorbic acid, sodium azide, BHA, dimethyl disulfide, heptanal, and hexanal were purchased from Sigma Chemical Co.(St. Louis,

MO., U.S.A.).

Riboflavin effects on the light-induced volatile compounds in whole milk

Riboflavin was added to the 500 mL whole milk in an aluminum foil wrapped 1-L bottle with a magnetic stirring bar to obatin 5, 10, or 50 ppm, respectively. Five mL of whole milk with added 5, 10 or 50 ppm riboflavin was put in a 20-mL semm bottle (22.6

X 75 mm from Supelco Inc.) and sealed air-tight with a Teflon-coated rubber septum and an aluminum cap. Whole milk without added riboflavin was used as eontrol sample.

Sample bottles were kept at 4 °C in the dark and under light in a fluorescentlight box for

8 hr in a cold room of the Department of Food Science Technology in the Ohio State

University. Light souree was four Sylvania fluorescent lamps (General Electrics,

Cleveland, OH., U.S.A.) with 1,500 lux light intensity. One set of sample bottles was wrapped with aluminum foil to make a dark eondition. Volatile compounds of samples

55 were measured at 0, 1, 2, 4, and 8 hr by CAR/PDMS. Samples were prepared in triplicate at each sample analysis.

Effects of fat content in milk on the light-induced volatile compounds

Five mL of skim milk, 1% fat milk, 2% fat milk and whole milk was put in a 20- mL serum vial, respectively, and air-tight sealed with a Teflon-coated rubber septum and an aluminum cap. Samples were prepared in triplicate. Sample bottles were kept in a

light box for 8 hr at 4 °C in the dark and under light. Volatile compounds of samples were measured at 0, 1, 2, 4, and 8 hr by CAR/PDMS.

Formation mechanisms study o f light-induced volatile compounds in milk using ascorbic acid, sodium azide, and BHA

Ascorbic acid, sodium azide, or BHA was added to whole milk with added 50 ppm riboflavin to obtain 100 ppm, respectively. Five mL of each sample was put in a 20- mL serum bottle sealed air-tight with a Teflon-coated rubber septum and an aluminum cap. Whole milk with added 50 ppm riboflavin without ascorbic acid, sodium azide, and

BHA was used as eontrol. Samples were prepared in triplicate. Sample bottles were kept at 4 °C in a light box in the dark and under light. Volatile eompounds of samples were measured at 0, 1, 2, and 4 hr by CAR/PDMS.

Analysis of headspace volatile compounds in milk by SPME

Sample bottles were kept in a 30 °C water bath for 10 min to aehieve the equilibrium of volatile compounds between headspace and milk liquid in the sample 56 bottles. The headspace volatile compounds in air-tight sealed sample bottles were isolated by 75 |im CAR/PDMS of SPME solid phase. Sample bottles were put in a 30 °C water bath for 30 min while CAR/PDMS of SPME solid phase was exposed to the headspace. The isolated volatile compounds by SPME solid phase were separated in a gas chromatograph (GC) equipped with a flame ionization detector.

Conditions o f gas chromatography

A Hewlett-Packard 5890 gas chromatograph was equipped with a 0.75 mm ID glass injection liner, a flame ionization detector, and a 30 m x 0.32 mm ID, 0.25 qm film,

HP-5, from Agilent Technologies (Palo Alto, CA., U.S.A.). The oven temperature was held at 40 °C for 2 min and increased from 40 to 160 °C at 6 °C/min and from 160 to 220

°C at the rate of 10 °C/min. The temperatures of injector and detector were 250 and 300

°C, respectively. The flow rate of nitrogen carrier gas was 1.0 mL/min. The isolated volatile compounds in solid phase of SPME were desorbed at 250 °C of GC injector for 2 min.

Identification o f volatile compounds in milk

A Hewlett-Packard 5971A mass selective detector (MS) equipped with a Hewlett-

Packard 59822 B ionization gauge controller was used. All mass spectra were obtained at 70eV and 220 °C ion source temperature. The identification of compounds was made by the combination of NIST Mass Spectra and gas chromatographic retention times of

standard compounds. Helium carrier gas at 0.9 ml/min and a 30m x 0.25 mm i.d., 0.25

57 |im film thickness, HP-5 from Agilent Technologies (Palo Alto, CA., U.S.A.) were used.

The GC conditions for GC-MS were the same as the gas chromatographic analysis conditions described previously.

Statistical analysis

One-way analysis of variance and Tukey's multiple comparison method were used to analyze the data. A P value of 0.05 or less was considered significant. All statistical analyses were conducted with Minitab 12.1 (Minitab Inc., State College, Pa., U.S.A.).

4.3 Results and Discussion

Formation o f light-induced volatile compounds in milk

Gas chromatograms of headspace volatile compounds in whole milk without

added riboflavin stored at 4 °C under fluorescent light for 0 (a), 2 (b), 4 (c), and 8 hr (d) are shown in Figure 18. As storage time increased from 0 to 2, 4, and 8 hr, pentanal, hexanal, heptanal, and dimethyl disulfide were formed and increased. From these gas chromatograms, pentanal and hexanal were formed before 2 hr, heptanal was formed

before 4 hr, and dimethyl disulfide was formed before 8 hr storage at 4 °C under fluorescent light. Marsili (1999b) reported that pentanal and hexanal were formed and increased before dimethyl disulfide was detected in the light exposed milk. Preliminary

study showed that hexanal was the first light-induced volatile compound formed in milk at 4 °C under fluorescent light.

58 Gas chromatograms of headspace volatile compounds in whole milk without added rihoflavin stored at 4 °C for 0 hr, for 8 hr under light or in the dark are shown in

Figure 19. Fleadspaee volatile eompounds in milk samples stored at 4 °C in the dark for

8 hr were not different from those in 0 hr samples in quantitatively and qualitatively

(Figure 19). Pentanal, dimethyl disulfide, hexanal, and heptanal, which were formed in

whole milk stored at 4 °C under light, were not formed in whole milk in the dark for 8 hr.

This result shows that light exposure plays important roles in the formation of pentanal, dimethyl disulfide, hexanal, and heptanal in milk.

Effects of riboflavin content on the light-induced volatile compounds in milk

Effects of riboflavin content on the formation of pentanal, dimethyl disulfide, hexanal, and heptanal in milk stored at 4 °C under light for 2 hr are shown in Figure 20.

As the concentration of added rihoflavin increased from 0 to 5, 10, and 50 ppm, the peak area of pentanal and hexanal increased hy 35 and 150% for 2 hr, respectively. Dimethyl disulfide and heptanal, which were not detected in milk with added 0 and 5 ppm riboflavin stored at 4 °C for 2 hr, were formed in 10 or 50 ppm riboflavin added milk.

Riboflavin is an important factor for the formation of light-induced off-flavor in milk

(Skibsted 2000; Boff and Min 2002). The concentration of riboflavin in whole milk used in this study was 1.5 |lg/mL by the analysis of high performance liquid chromatography

(data not shown). Riboflavin in milk accelerates the formation of light-induced volatile compounds in milk.

59 IL

m

Figure 18. Headspace volatile compounds in whole milk under the fluorescent light for 0 (a), 2 (b), 4 (c), and 8 hr (d) at 4°C. 1 : pentanal, 2: dimethyl disulfide, 3: hexanal, 4: heptanal

60 Control

m

4 °C in the dark

4 °C under light

Figure 19. Headspace volatile compounds in whole milk stored under light or in the dark at 4 °C and control for 8 hr. 1 : pentanal, 2: dimethyl disulfide, 3 : hexanal, 4: heptanal

61 2.5

S Pentanal H Dimentyl disulfide 2.0 □ Hexanal * uni Heptanal £3 O u 1.5 £3

H "3 X .S ^ 1.0 * CQ k 1:3 :3 &» Qh 0.5

0.0 0 10 50

Added riboflavin in milk (ppm)

Figure 20. Effects of riboflavin on the formation of pentanal, dimethyl disulfide, hexanal and heptanal in photosensitized milk under light at 4 °C for 2 hr. * indicates that the compound was not detected.

62 Effects of fat content on the light-induced volatile compounds in milk

Effects of fat content on the formation of pentanal, hexanal, heptanal, and dimethyl disulfide in milk stored at 4 °C under light for 8 hr are shown in Figure 21. The fat content of skim milk, 1% fat milk, 2% fat milk, and whole milk was 0.5, 1.0, 2.0, and

3.4%, respectively byusing Bobcoek method (Case et al. 1985). Major fatty acid compositions of milk fat are myristic (12.8%), palmitic (43.7%), stearic (11.3%), oleic

(11.3%), and linoleic acids (1.5%) (Swaisgood 1996). As fat content in milk increased from 0.5 to 1.0, 2.0, and 3.4%, pentanal, heptanal, and hexanal increased by 560, 130, and 120%, respectively for 8 hr. The concentration of dimethyl disulfide did not change

significantly as the fat content increased (P>0.05). Formation of aldehydes including pentanal, hexanal, and heptanal depends on the fat content in milk while dimethyl disulfide does not depend on the fat content in milk. Jung et al. (1998) reported that dimethyl disulfide can be formed from sulfur containing amino acid like methionine by

singlet oxygen oxidation. Cadwallader and Howard (1998) reported that as the fat level of the milk increased from skim to 2 % fat and whole milk, light-induced flavor increased.

Marsili (1999b) reported that as light exposure time increased from 0 to 48 hr, hexanal increased linearly while pentanal showed second-order polynomial increases in skim milk and 2 .0 % fat milk, and the author suggested that this difference might be due to the availability of substrates for pentanal and hexanal.

63 30

g 0.5% 25 TT □ 1.0% O X in 20% □ 20 24% a a o u u a 15 0 u "03 .S 10 a r f k a 1 ^ f

iAÈ r I r

Pentanal Heptanal Hexanal Dimethyl disulfide

Figure 21. Effects of fat contents on the formation of pentanal, heptanal, hexanal,

and dimethyl disulfide in photosensitized milk under light for 8 hr.

64 Formation mechanisms of light-induced volatile compounds in milk using ascorbic acid, sodium azide, and BHA

Effects of 100 ppm ascorbic acid, sodium azide, and BHA on the formation of dimethyl disulfide, hexanal, and heptanal in milk added 50 ppm riboflavin stored at 4 °C under fluorescent light for 4 hr are shown in Table 7.

Sodium azide, a singlet oxygen quencher, prevented the formation of dimethyl disulfide in milk completely. Jung et al. (1998) proposed the formation mechanism of dimethyl disulfide in milk from methionine by singlet oxygen oxidation. The complete prevention of dimethyl disulfide formation in milk by sodium azide clearly indicates that dimethyl disulfide is formed by singlet oxygen oxidation.

BHA and ascorbic acid donate hydrogen atom to free radicals and work as antioxidant. BHA and ascorbic acid reduced the dimethyl disulfide formation in milk

significantly (P<0.05). Jung et al. (1998) reported that the addition of ascorbic acid lowered the formation of dimethyl disulfide and reduced off-flavor in skim milk. This

study shows that even though singlet oxygen and free radicals are involved in the formation of dimethyl disulfide, dimethyl disulfide does not form in milk stored at 4°C under light without singlet oxygen.

Ascorbic acid, BHA, and sodium azide reduced the formations of hexanal and heptanal in milk stored at 4°C under light for 4 hr significantly, compared to those in control samples (P<0.05) (Table 7).

65 Volatile Time Ascorbic Sodium Compounds (hr) Control acid azide BHA

0 n.d. n.d. n.d. n.d.

Dimethyl 1 236 n.d. n.d. n.d.

disulfide 2 4.74 335 n.d. 295

4 1330 736 n.d. 170

0 n.d. n.d. n.d. n.d.

Hexanal 1 21.18 10.50 10.92 331

(lig/1) 2 4198 12.94 2637 198

4 5736 12.90 4138 11.49

0 n.d. n.d. n.d. n.d.

Heptanal 1 187 n.d. n.d. n.d.

(|ig/ml) 2 1L80 n.d. n.d. n.d.

4 19.61 n.d. 639 n.d.

n.d. : not detected Values are mean of triplicate Data did not vary more than 5% of coefficient of variation.

Table 7. Effects of ascorbic acid, sodium azide, and BHA on the formation of dimethyl disulfide, hexanal, and heptanal in milk containing 50 ppm added rihoflavin stored at 4 °C under fluorescent light for 4 hr.

66 The formation of hexanal and heptanal in milk containing riboflavin stored under light can be explained using combination of Type 1 and Type II photosensitization mechanisms of riboflavin. Riboflavin has both Type I and Type II mechanisms, which are competitive depending on the solubility and concentration of triplet oxygen (Edward and Silva 2001).

Singlet oxygen quencher and free radical scavengers work differently in milk containing riboflavin. Singlet oxygen, which is formed by Type II mechanisms of riboflavin, can directly react with double bonds of unsaturated fatty acids and form hydroperoxide. Therefore, sodium azide inhibits the formation of hydroperoxides of unsaturated fatty acids by quenching singlet oxygen.

Type I mechanism of riboflavin can generate free radicals of unsaturated fatty acid by abstracting hydrogen atom. The excited triplet riboflavin can abstract hydrogen atom directly from Cl 1 position in linoleic acid due to the lower activation energy of hydrogen abstraction (50 kcakmol) than that of C 8 or C14 (75kcal/mol) or CIS (100 kcal/mol) (Boff and Min 2002). Radical at Cl 1 can be moved at C9 or 13 position after rearrangement. Triplet oxygen can react with free radicals of unsaturated fatty acids and form peroxyl radicals, which become hydroperoxides by abstracting hydrogen atom from unsaturated fatty acids. Hydroperoxides formation is a slow step in lipid oxidation and the concentration of peroxyl radicals is found the greatest of all fatty acid radicals

(Decker 1998). Standard one-electron reduction potentials of alkyl, peroxyl and alkoxyl radicals are 600, 1000, and 1600 mV, respectively (Buettner 1993). The reduction potential of ascorbic acid is 282 mV (Buettner 1993) and ascorbic acid can prevent hydroperoxide formation by donating a hydrogen atom to the peroxyl radicals faster than

67 polyunsaturated fatty acids donate hydrogen atom to the peroxyl radicals. Also, BHA and aseorhie aeid can donate hydrogen atom to the excited triplet riboflavin and reduce the excited triplet riboflavin.

Ascorbic acid and BHA prevented the hexanal and heptanal formation better than

sodium azide in milk stored at 4 °C under light for 4 hr, which indicates that Type I mechanisms of riboflavin plays more important roles in the formation of hexanal and heptanal in milk than Type II mechanisms.

Type II mechanism Type I mechanism

R-CH2—CH = CH-CH 2 —CH = CH-CH 2 -R'

______I______- H- BHA, Reduced riboflavin Excited Riboflavin ascorbic acid

R-CH2—CH=CH-CH—CH=CH-CH 2-R

+ '02 Sodiumazide

R-CH2—CH-CH=CH—CH = CH-CH 2 -R ’

+ ^02

R-CH2—CH-CH = CH—CH = CH-CH 2-R ' I 00 - + H- BHA, ascorbic acid

R-CH2—CH-CH = CH—CH = CH-CH 2 - R OOH

68 The suggested formation mechanisms of hexanal from linoleic acid by a combination of Type I and Type II mechanisms of riboflavin are shown in Figure 22. To form hexanal from linoleic acid, hydroperoxide should be formed at C13 position, which can be done by singlet oxygen and triplet oxygen oxidation (Frankel 1985). Singlet oxygen can directly react with the double bond of C12 and C13 in linoleic acid with

“ene” reaction to form linoleic acid hydroperoxide at C13 position. Light excited riboflavin can abstract hydrogen atom at Cl 1 position and the radical of Cl 1 position can move to C9 or C13 position. Triplet oxygen can react with radical of C13 position and form hydroperoxide of linoleic acid at C13 position with conjugated diene (Min 1998).

The cleavage between C12 and C13 of linoleic acid hydroperoxides at C13 generates hexanal (Figure 22).

Effects of riboflavin, ascorbic acid, sodium azide, and BHA on the formation of pentanal are shown in Table 8. As the added riboflavin content increased from 0 to 5, 10, and 50 ppm, formation of pentanal increased with second-order polynomial correlation, which agrees with the previous report of Marsili (1999b). Riboflavin plays an important role in the formation of pentanal.

Ascorbic acid and BHA decreased the formation of pentanal but sodium azide increased the formation of pentanal in milk stored at 4 °C under light for 4 hr, compared to the control samples of whole milk with added 50 ppm riboflavin. The formation mechanism of pentanal in milk is not the same as those of hexanal and heptanal, which agrees with previous reports of Bassette (1976) and Marsili (1999b). Bassette (1976) reported that the changes of pentanal concentration were not the same as those of the other carbonyl compounds in the light exposed milk.

69 14 13 12 11 10 9 R-CH2—CH=CH-CH2-CH=CH-R'

- .H + O2 Excited triplet Reduced triplet riboflavin riboflavin

R-CH2-CH-CH=CH-CH=CH-R R-CH2—CH=CH-CH-CH=CH-R 0 1 OH 'O 2 .H - .OH

R-CH2-CH-CH=CH-CH=CH-R 0 1 R-CH2-CH^CH=CH-CH=CH-R OH I I - .OH O ■

R-CH2-CHyCH=CH-CH=CH-R I I o ■

CH3— (CH2)4-CH R:CH3-(CH2)3 O R: (CH2)7-C00H

Figure 22. Suggested formation mechanisms of hexanal in photosensitized milk

containing riboflavin

70 Type I mechanisms of riboflavin may play more important role in the formation pentanal in milk than singlet oxygen oxidation. Ascorbic acid and BHA can donate hydrogen atom to the excited triplet riboflavin and peroxyl radicals of unsaturated fatty acids and reduce the pentanal formation in milk. The difference of the pentanal and hexanal concentration in light exposed milk from Marsili (1999b) may be due to the difference in the formation mechanisms.

Whole milk with the addition Sodium Time of riboflavin in ppm Ascorbic BHA acid azide (hr) 0 5 10 50

0 n.d. n.d. n.d. n.d. n.d. n.d. n.d.

1 0.67 0.72 0.86 0.90 0.50 1.71 0.55

2 1.36 1.46 1.76 1.81 1.15 1.99 1.41

4 2.65 2.06 2.13 1.68 0.70 2.29 1.67

n.d. : not detected Values are mean of triplicate Ascorbic acid, sodium azide, and BHA in 50 ppm riboflavin added milk Data did not vary more than 5% of coefficient of variation.

Table 8. Effects of riboflavin, ascorbic acid, sodium azide, and BHA on the formation of pentanal in electronic counts (1x1in milk stored at 4°C under fluorescent light for 4 hr. 71 4.5 Conclusion

Formation of light-induced volatile compounds in milk stored at 4 °C is influenced by the riboflavin content and fat content in milk. Singlet oxygen and free radicals formed hy light excited riboflavin are involved in the formation of light-induced volatile compounds in milk. To prevent the formation of light-induced volatile compounds, milk should be packed in light and oxygen impermeable packages to minimize the exposure of light.

72 4.6 References

Bassette R. 1976. Effects of light on concentrations of some volatile materials in milk. J

Milk Food Technol 39(1): 10-12.

Boff JM, Min DB. 2002. Chemistry and reaction of singlet oxygen in foods.

Comprehensive review in Food Science and Food Safety l(l):58-72.

Bradley DG. 1991. Riboflavin photosensitized singlet oxygen oxidation in milk products.

Dissertation in the Ohio State University. 109 pp.

Buettner OR. 1993. The pecking order of free radicals and antioxidants: lipid peroxidation, a-tocopherol and ascorbate. Archives of Biochem Biophy 300(2):535-543.

Cadwallader KR, Floward CL. 1998. Analysis of aroma-active components of light- activated milk. ACS Symposium Series 705(Flavor Analysis):343-358.

Case RA, Bradley RL, Williams RR. 1985. Chemical and physical methods. In :

Richardson GH, editor. Standard methods for the examination of dairy products.

Washington, D.C.: American Public Health Association, p 347-350.

Decker EA. 1998. Antioxidant mechanisms. In: Akoh CC, Min DB, editors. Food lipids- chemistry, nutrition, and biotechnology. New York: Marcel Dekker. p 397-401.

73 Edwards AM, Silva E. 2001. Effect of visible light on selected enzymes, vitamins and amino acids. J Photochem Photobiol B 63(1-3):126-131.

Foote CS. 1976. Photosensitized oxidation and singlet oxygen: consequences in biological systems. In: Pryor, WA, editor. Free Radicals in Biology. Vol. 2. New York:

Academic Press, p 85

Frankel EN. 1985. Chemistry of autoxidation: Mechanism, Products and Flavor

Significance. In: Min DB, Smouse TH, editors. Flavor chemistry of fats and oils.

Champaign, IE: AOCS. p. 1-37.

Gaafar AM, Gaber FI. 1992. Volatile flavor compounds of sunlight-exposed milk. J

Dairy Sci 20(I):III-II5.

Jung MY, Yoon SH, Lee HO, Min DB. 1998. Singlet oxygen and ascorbic acid effects on dimethyl disulfide and off-flavor in skim milk exposed to light. J Food Sci 63(3):408-

412.

Kim YD, Morr CV. 1996. Dynamic headspace analysis of light-activated flavor in milk.

Int Dairy J6(2):185-93.

74 Lu CY, Wang WF, Lin WZ, Fian ZH, Yao SD, Lin NY. 1999. Generation and photosensitization properties of oxidized radical of riboflavin: a laser flash photolysis

study. J Photochem Photobiol BiBiol 52(1-3): 111-116.

Marsili RT. 1999a. SPME-MS-MYA as an electronic nose for the study of off-flavors in milk. J Agric Food Chem 47(2):648-654.

Marsili RT. 1999h. Comparison of solid-phase microextraction and dynamic headspace methods for the GC-MS analysis of light-induced lipid oxidation products in milk. J

Chromatogr Sci 37(1): 17-23.

Min DB. 1998. Lipid oxidation of edible oil. In: Akoh CC, Min DB, editors. Food lipids- chemistry, nutrition, and biotechnology. New York: Marcel Dekker. p 283-296.

Skibsted LH. 2000. Light-induced changes in dairy products. Bulletin of the International

Dairy Federation 346:4-9.

Swaisgood HE. 1996. Characteristics of milk. In: Fennema OR, editor. Food chemistry

3"^. New York: Marcel Dekker. p 850-851.

Whited LJ, Hammond BH, Chapman KW, Boor KJ. 2002. Vitamin A degradation and light-oxidized flavor defects in milk. J Dairy Sci 85(2):351-354.

75 CHAPTER 5

Photooxidation and Photosensitized Oxidation of Lard

5.1 Abstract

Volatile compounds and headspace oxygen content from lard containing 0 and 5 ppm chlorophyll in air-tight 10-mL bottles under light and dark at 55 °C were studied.

Total volatile compounds in lard with 5 ppm chlorophyll under light were 19 times more than that in samples with 0 ppm chlorophyll for 48 hr. Headspace oxygen content in lard with 5 ppm chlorophyll under light changed from 21 to 15% for 48 hr but that in lard with 0 ppm chlorophyll did not change significantly (P>0.05). Photosensitizer-free lipid model system using lard was developed. Photooxidation mechanism seems to be a free radical chain reaction like autoxidation. Singlet oxygen oxidation mechanisms for the volatile formations in lard were proposed.

76 5.2 Introduction

The lipid oxidation in foods causes nutritional losses and decreases the food qualities during manufacturing and storage. Autoxidation, lipoxygenase catalyzed oxidation, and light-induced oxidation are major lipid oxidation mechanisms in foods

(Chan 1977; Frankel 1985; Zhuang et al. 1998).

Autoxidation or triplet oxygen oxidation of lipids is a free radical chain reaction of unsaturated fatty acids. Autoxidation of lipids can be initiated by metal catalysis or heat, and produce hydroperoxides of unsaturated fatty acids (Frankel 1985). The decomposition of lipid hydroperoxides produces volatile compounds, including aldehydes, alcohols, ketones, and hydrocarbons, which are responsible for the increase of off-odor and the decrease of food qualities (Min 1998).

Light-induced lipid oxidation can be either photooxidation or photosensitized oxidation depending on the absence or presence of photosensitizers in samples, respectively (Chan 1977). Unsaturated fatty acids do not absorb the energy of visible light (Faria 1983) but ultraviolet irradiation can produce free radicals in oils (Faria 1983;

Frankel 1985; Yang and Min 1994). Several authors reported that light exposure accelerates lipid oxidation in foods and model systems, which may be due to the presence of coloring materials such as chlorophyll and riboflavin (Edwards and Silva 2001; Thron et al. 2001).

Photosensitizers can accelerate lipid oxidation either by Type I or Type II mechanisms. Singlet state of photosensitizer can absorb light energy and become excited

singlet state photosensitizer. The excited singlet state photosensitizer becomes to the

77 excited triplet state photosensitizer by intersystem crossing mechanisms. The excited triplet state photosensitizer reacts with triplet oxygen to form singlet oxygen (Type II mechanism) or abstracts electron or hydrogen atom from a substrate to generate radicals

(Type I mechanism) depending on the solubility and availability of triplet oxygen and

substrates (Foote 1976; Bradely 1992).

Study of photooxidation on lipid is not easy due to the difficulty of removing photosensitizers completely in foods and model systems. A combination of adsorbents including activated silica gel, Celite, powder sugar, and activated charcoalhas been used to remove the impurities txom oils using solvent under nitrogen gas (Lee and Min 1988;

Kulas and Ackman 2001). However, the complete removal of photosensitizer in

adsorbent treated oils can not be guaranteed and the residual solvent in samples may

interfere the analysis of headspacevolatile eompounds hy SPME method.

Lard is animal lipid and has been used in deep fat frying and emulsified cake

shortening (Kines 1985). Even though myoglobin in musele-based foods can induce lipid

oxidation as a photosensitizer, myoglobin could not accelerate the oxidation in lard due to the low solubility of myoglobin in lard (Pokorny et al. 1966; Baron and Andersen 2002).

Therefore, lard can be a suitable substrate for the study of photooxidation and photosensitized oxidation.

The objectives of this work were (1) to develop a photosensitizer-free lipid model

system using lard and (2) to study the volatile compounds in photosensitized oxidation of

lard.

78 5.3 Materials and Method

Materials

Pork meat containing lard was purehased from a local groeery store (Meijer,

Columbus, OH., U.S.A.). Chlorophyll b, 1-pentanol, hexanal, 2-hexenal, heptanal, 2- heptenal, l-oetene-3-ol, 2,4 heptadienal, 2-nonenal, octanoic aeid, and 2,4 decadienal were purchased from Sigma Chemical Co. (St. Louis, MO., U.S.A.). Teflon-coated rubber septa, aluminum caps, serum bottles, glass liners, the fiber assembly holder, and

65 |xm Polydimefhylsiloxane/Divinylbenzene (PDMS/DVB) were purehased from

Supelco, Inc. (Bellefonfe, PA., U.S.A.). 2-Penfylfuran was purchased from Karl

Industries Inc. (Aurora, OH., U.S.A.).

Sample preparation for the light and dark storage o f lard

White lard was carefully separafed from pork meat to avoid containing muscle tissue with red pigments. Thirty gram of lard was eut into small pieces approximately 0.5 cm diameter and put into a 100-mL serum bottle sealed air-tight with Teflon-eoated rubber septa and aluminum caps. Headspace air in the bottles of lard was replaced with nitrogen gas for 30 sec. Bottles of lard was put in a 70 °C oven to melt lard, and the oils from lard was collected in a 10-mL serum bottle (25 x 40 mm, 20 mm diameter from

Supelco, Inc.) wrapped with aluminum foil under nitrogen gas flow.

Chlorophyll was dissolved in the oils from lard in a 10-mL bottle (25 x 40 mm,

20 mm diameter) with a magnetic stirring bar (10 X 3mm) to obtain 0 and 5 ppm

(weight/volume) in a 55 °C water bath while magnetic stirring. The temperature of 55 °C

79 , which is the melting point of stearic acid, was chosen due to the highest melting point of fatty acids in lard. Samples of 0.1 g lard with 0 or 5 ppm chlorophyll were put in 10-mL

serum vials and sealed air-tight with Teflon-coated rubber septa and aluminum caps.

Sample bottles of the lard were kept at 55 °C in a light box. Light source was 2 tungsten light bulbs of 100 watts from Topco Association. (Smokie, IL., U.S.A.) with 4,000 lux light intensity (Lee and Min 1988). One set of sample bottles was wrapped with aluminum foil to produce a dark condition. Volatile compounds and headspace oxygen content in sample bottles were analyzed at 0, 6, 12, 24, 36, and 48 hr. Sample bottles were prepared in duplicate at each sample analysis.

Analysis of volatile compounds by SPME

The headspace volatile compounds in air-tight sealed sample bottles were isolated by 65 pm PDMS/DVB of SPME solid phase. Sample bottles of lard were put in a 55 °C water bath for 30 min while 65 pm PDMS/DVB of SPME solid phase was exposed to the headspace. The isolated volatile compounds by SPME solid phase were separated in a gas chromatograph (GC) equipped with a flame ionization detector.

The content of headspace oxygen was determined by injecting 100 pE headspace gas of sample bottles into a gas chromatograph (GC) equipped with thermal conductivity detector. A stainless steel column (1.8 m x 0.32 cm) packed with 60/80 Molecular Sieve

13 X (Alltech Assoc., Inc. Deerfield, 11., U.S.A.) was used. The flow rate of helium gas

80 was 20 mL/min. Temperatures of oven, injector, and thermal conductivity detector were

40, 120 and 150 °C, respectively.

Gas chromatography condition

A Hewlett-Packard 5890 gas chromatograph was equipped with a 0.75 mm ID glass injection liner, a flame ionization detector, and a 30 m x 0.32 mm ID, 0.25 qm film,

HP-5, from Agilent Technologies (Palo Alto, CA., U.S.A.). The oven temperature was held at 40 °C for 2 min and increased from 40 to 160 °C at 6°C/min and from 160 to 220

°C at the rate of 10 °C/min. The temperatures of injector and detector were 250 and 300

°C, respectively. The flow rate of nifrogen carrier gas was 1.0 mL/min and GC was operated in splitless mode. The isolated volatile compounds in solid phase of SPME were desorbed at 250 °C of GC injector for 2 min.

Identification o f volatile compounds

A Hewlett-Packard 5971A mass selective detector (MS) equipped with a Hewlett-

Packard 59822 B ionization gauge controller was used. All mass spectra were obtained at 70eV and 220 °C ion source temperature. The identification of compounds was made by the combination of NIST Mass Spectra and gas chromatographic retention times of

standard compounds. Helium carrier gas at 0.9 ml/min and a 30m x 0.25 mm i.d., 0.25

|im film thickness, HP-5 from Agilent Technologies (Palo Alto, CA., U.S.A.) were used.

The GC conditions for GC-MS were the same as the gas chromatographic analysis conditions described previously.

81 Statistical analysis

One-way analysis of variance, Tukey’s multiple eomparisons method, and general linear model were used to analyze the data. A P value of 0.05 or less was considered

significant. All statistical analyses were conducted with Minitab 12.1 (Minitab Inc., State

College, Pa., U.S.A.)

5.4 Results and Discussion

Effect o f photosensitized oxidation on the headspace volatile compounds

Gas chromatograms of headspace volatile compounds in lard containing 0 and 5 ppm chlorophyll stored under light at 55 °C for 24 hr are shown in Figure 23. Gas chromatogram of lard with 5 ppm chlorophyll is different quantitatively and qualitatively from that of lard with 0 ppm chlorophyll under light for 24 hr.

Changes of headspace volatile compounds in total GC peak areas from lard with 0 and 5 ppm chlorophyll stored under light and dark for 48 hr are shown in Figure 24.

Total volatile compounds in lard with 0 ppm chlorophyll stored under light are not different significantly from those in lard stored in the dark for 48 hr (P>0.05), which indicates that lard with 0 ppm chlorophyll under light was oxidized to the similar degree to the lard in the dark. As storage time increased from 0 to 6, 12, 24, 36, and 48 hr, total volatile compounds in lard with 5 ppm chlorophyll stored under light were more than those in lard with 0 ppm chlorophyll under light by 460, 820, 1470, 2150, and 1930%, respectively, which shows that chlorophyll accelerated lipid oxidation in lard.

82 0 hr sample

Total peak areas : 0.3 X 10

Sample with 0 ppm chlorophyll 1 3 15 Total peak areas : 0.5 X 10^

Sample with 5 ppm chlorophyll

Total peak areas : 8.2 x 10

1 8 1 9

Time (min)

Figure 23. Gas chromatograms of headspace volatile compounds in lard with 0 and 5 ppm chlorophyll stored under light for 24 hr at 55 °C. Numbered volatiles are shown in Table 9. 83 20

With 5 ppm chlorophyll under 16 light

12

8 With 0 ppm chlorophyll under light 4 Samples in the dark

0 10 20 30 40 50 Time (hr)

Figure 24. The effects of chlorophyll on the volatile compounds in lard stored under light and dark for 48 hr at 55 °C

84 oxWa/zo» OM ^Ae Aeaùkpace o;g;geM coM^e»^

Changes of headspace oxygen content in lard sample bottles with 0 and 5 ppm chlorophyll stored under light and dark for 48 hr are shown in Figure 25. Headspace oxygen analysis has been used to determine the degree of oxidation in oils (Lee and Min

1988). Headspace oxygen content in lard sample bottles with 0 ppm chlorophyll stored under light was not significantly different from that in lard sample bottles stored in the dark for 48 hr (P>0.05), which indicates that lard samples used in this study did not contain any photosensitizers, and photooxidation decreased the headspace oxygen content as the same rate as autoxidation did.

Headspace oxygen content in lard sample bottles with 5 ppm chlorophyll stored under light was significantly different from that in lard sample bottles stored in the dark and lard sample bottles with 0 ppm chlorophyll for 48 hr (P<0.05). As storage time increased from 0 to 6, 12, 24, 36, and 48 hr, headspace oxygen content in lard sample bottles with 5 ppm chlorophyll stored under light was lower than that in lard sample bottles with 0 ppm chlorophyll by 0.5, 1.0, 2.5, 4.4, and 6.2%, respectively, which shows that chlorophyll accelerated the headspace oxygen depletion and the lipid oxidation of lard.

85 Samples in the dark

With 0 ppm chlorophyll under light

With 5 ppm chlorophyll under light

0 10 20 30 40 50 Time (hr)

Figure 25. The effects of chlorophyll on the headspace oxygen content in lard stored under light and dark for 48 hr at 55 °C

86 Identified volatile compounds in lard

Volatile compounds identified from lard containing 0 and 5 ppm chlorophyll

stored under light at 55 °C for 0 and 48 hr are shown in Table 9. Pentane, pentanal, hexanal, heptanal, and nonanal were identified in lard both with 0 and 5 ppm chlorophyll

stored under light, which indicates that pentane, pentanal, hexanal, heptanal, and nonanal can be formed without photosensitizers. Frankel (1985) reported that pentane, pentanal, and hexanal can be formed from methyl linoleate, and heptanal and nonanal from methyl oleate by autoxidation. Lipid oxidation mechanism of photooxidation seems to be a free radical chain reaction like autoxidation. Volatile compounds, which were identified in photooxidation of lard, were found in lard stored at 55 °C in the dark (data not shown).

GC peak area of pentane, pentanal, hexanal, heptanal, and, nonanal in lard with 5 ppm chlorophyll stored under light for 48 hr was higher than that in lard with 0 ppm chlorophyll by 1100, 2200, 300, 480, and 230%, respectively (Table 9), which indicates that chlorophyll photosensitized oxidation accelerated the formation of volatile compounds.

l-Penten-3-ol, 1-pentanol, 2-hexenal, 2-heptenal, l-octene-3-ol, 2-pentylfuran, octanal, 2,4 heptadienal, 3-octene-2-one, 2-octenal, octanoic acid, 2,4 decadienal, and 2- undecenal, which were not identified in lard with 0 ppm chlorophyll under light, were identified in lard with 5 ppm chlorophyll at 55 °C under light for 48 hr (Table 9). The changes of 2-hexenal, 2-heptenal, 2-pentylfuran, and 2-octenal in lard with 5 ppm chlorophyll stored under light are shown in Figure 26. As storage time increased from 0 to 48 hr, 2-hexenal, 2- heptenal, 2-pentylfuran, and 2-octenal in lard with 5 ppm chlorophyll stored under light were formed and increased for 48 hr. Callison (2001) 87 identified 2-pentylfuran, a reversion flavor eompound, in soybean oil containing 5 ppm ehlorophyll during storage under light. However, 2-pentylfuran was not identified in

soybean oil without chlorophyll nor without light. Frankel (1985) reported that 2- heptenal and l-octene-3-ol were formed in trace concentration from autoxidation of methyl linoleate but photooxidation of methyl linoleate produced 9.9 and 1.9 relative percentage of 2-heptenal and l-octene-3-ol, respectively. Callison (2001) reported 2- heptenal was formed in soybean oil only in the presence of chlorophyll and light.

The photosensitizing effects of chlorophyll can explain the significant increase of volatile compounds and the significant decrease in headspace oxygen content in lard with

5 ppm chlorophyll under light. The major fatty acids of lard are myristic (1-4%), palmitic (20-28%), stearic (5-14%), oleic (41-51%), linoleic (2-15%), linoleic (trace-

0.1%), and arachidonic acids (0.3-1.0%) (Kincs 1985). Photosensitizers exposed to light can oxidize unsaturated fatty acids in lard through Type 1 and Type II mechanisms depending on the solubility and concentration of oxygen and substrates. The solubility of oxygen is higher in lipid and nonpolar solvents than in water. Singlet oxygen generation

(Type II mechanism) is preferred in lipid phase due to the high solubility of triplet oxygen (Bradely and Min 1992). Rawls and VanSanten (1970) reported chlorophyll accelerated the oxidation of methyl linoleate approximately 80% through Type II mechanism and 20% through Type I mechanism.

88 300 —*— 2-Hexenal

a 250 2-Heptenal 3 u0 u 2-Pentylfuran '3 200 1 2-Octenal 4> rO - s 150 .5 X

u a 100 a 0» a U 50 O

0 10 20 30 40 50 Time (hr)

Figure 26. The effects of chlorophyll on 2-hexenal, 2-heptenal, 2-pentylfiiran, and 2- octenal in lard stored under light for 48 hr at 55 °C

89 The reaction mechanisms of singlet and triplet oxygen on unsaturated fatty acids are different (Frankel 1985). Triplet oxygen, which is a di-radical, needs radical forms of unsaturated fatty acids to form hydroperoxides. Singlet oxygen does not need radical forms of unsaturated fatty acids. Singlet oxygen can directly react with electron rich double bonds of unsaturated fatty acids through 1, 4 cycloaddition, “ene” reaction, and 1,

2 cycloaddition and generate hydroperoxides (Foote 1976; Boff and Min 2002).

0 O' 1,4- Cycloaddition 4 I 0 O.

0 “ene” Reaction 4 - I 0 J)

0 CH2 4 ?■ 1,2 - Cycloaddition O- 0 CHi

Proposed formation mechanisms of 2-hexenal, 2-heptenal, and l-octen-3-ol from lard by singlet oxygen oxidation are shown in Figure 27. Singlet oxygen can form nonconjugated hydroperoxide at C l2 position in linoleic acid, which can not be formed from triplet oxygen oxidation in linoleic acid (Frankel 1985). 2-FIeptenal is produced from the cleavage of Cl 1-C12 bond of C12 hydroperoxide of linoleic acid. Singlet oxygen directly reacts with double bond of C13-C14 by “ene” reaction and hydroperoxide at C13 position with double bond at C14-C15 can be formed. 2-lTexenal is formed from cleavage of C12-C13 bond of linoleic acid hydroperoxide with C13 position. The CIO hydroperoxide of linoleic acid with nonconjugated diene is formed by

singlet oxygen oxidation. The cleavage of ClO-Cl 1 bond of CIO linoleic acid

90 hydroperoxide with noneonjugated diene produees 2-oetene, whieh can reaet with singlet oxygen through “ene” reaction and produee l-octene-3-ol (Figure 27).

5.5 Conclusion

Photosensitizer-free lipid model system was developed using lard, which can he used for the study of photosensitization and photooxidation on lipids. Volatile eompounds and headspace oxygen eontent from lard by photosensitized oxidation and photooxidation were determined. Oxidation mechanisms of photooxidation and photosensitized oxidation are different. The oxidation mechanism of photooxidation

seems to be a free radical chain reaction like autoxidation. Singlet oxygen can explain the formation of volatile compounds and headspace oxygen depletion in photosensitized oxidation of lard.

91 R-CH2-CH2-CH=CH-CH2-CH=CH-R'

I + ‘02

2-Heptenal

R-CH2—CH=CH-CHtCH 2 -CH=CH-R' R-CH2 —CH2 —CH=CH-CH 2 - C H - C H = R ? ' ? OH OH

-.OH + '02

R-CH=CH-CH-CH-CH 2 - C H = C H - R ' R-CH2—CH2— CH=CH-CH2-|-CH-CH = R' 0I o I ' o. 1 I OH OH + O2, “ene” reaction

.OH + 2 H.

R-CH2—CH2—CH-CH=CH2 R-CH = CH-CH-+CH-CH 2 -C H = CH-R' III 0 o o 1 I OH OH ..OH

R-CH2 —CH2 — CH -CH = CH 2

r - c h = c h - c h O oII 2-Hexenal + H .

R-CH 2— CH2 —CH -CH = CH2

R:CH3-(CH2h OH R : (CH2)7 -C 0 0 H 1- Octene-3-ol

Figure 27. Proposed formation mechanisms of 2-hexenal, 2-heptenal, and 1- octene-3-ol from photosensitized lard hy singlet oxygen oxidation

92 Volatile 0 hr sample Light sample Light sample No. Compounds with 0 ppm with 5 ppm chlorophyll chlorophyll 1“ Pentane n.d. ^ 0.1 4.9 2 l-Penten-3-ol n.d. n.d. 1.0 3 Pentanal n.d. 0.5 10.4 4 1-Pentanol n.d. n.d. 0.8 5 Hexanal 1.4 6.7 26.8 6 2-Hexenal n.d. n.d. 0.9 7 Heptanal n.d. 0.7 4.0 8 2-Heptenal n.d. n.d. 25.6 9 l-Octene-3-ol n.d. n.d. 2.0 10 2-Pentylfuran n.d. n.d. 7.4 11 Octanal n.d. n.d. 4.6 12 2,4 Heptadienal n.d. n.d. 5.4 13 3-Octene-2-one n.d. n.d. 1.3 14 2-Oetenal n.d. n.d. 16.7 15 Nonanal n.d. 0.8 2.6 16 2-Nonenal n.d. n.d. 6.6 17 Octanoic acid n.d. n.d. 0.7 18 2,4 Decadienal n.d. n.d. 1.0 19 2-Undecenal n.d. n.d. 1.9

“ : No. indicates the numbered volatiles in Figure 22. * : n.d. : Not deteeted ‘ : Bold charaeter represents a compound, which is identified in lard with 5 ppm chlorophyll but not identified in lard with 0 ppm chlorophyll stored under light. .' a compound identified by GC-MS only. The other compounds were identified by both GC retention time of standard eompound and GC-MS.

Table 9. Volatile compounds identified from lard with 0 and 5 ppm chlorophyll stored under light at 55 °C for 0 and 48 hr in GC peak areas (1 xlO^)

93 5.6 References

Baron CP, Andersen H. 2002. Myoglobin-induced lipid oxidation. A review. J Agric

Food Chem 50(14):3887-3897.

Boff JM, Min DB. 2002. Chemistry and reaction of singlet oxygen in foods.

Comprehensive review in Food Science and Food Safety l(l):58-72.

Bradley DG. 1991. Riboflavin photosensitized singlet oxygen oxidation in milk products.

Dissertation in the Ohio State University. 109 pp.

Bradley DG, Min DB. 1992. Singlet oxygen oxidation of foods. Critical Reviews in Food

Science and Nutrition. 31(3):211-236.

Callison AL. 2001. Singlet oxygen lipid oxidation mechanisms for reversion flavor foramtion in soybean oil [MS thesis]. Columbus, Ohio: The Ohio State University,

Columbus, p 50-86.

Chan HWS. 1977. Photo-sensitized oxidation of unsaturated fatty aeid methyl esters. The identification of different pathways. JAOCS. 54(3): 100-104.

Edwards AM, Silva E. 2001. Effect of visible light on selected enzymes, vitamins and amino acids. J Photochem Photobiol B 63(1-3): 126-131.

94 Faria JAF, Mukai MK. 1983. Use of gas chromatographic reactor to study lipid photooxidation. JAOCS 60(1): 77-81.

Foote CS. 1976. Photosensitized oxidation and singlet oxygen: consequences in biological systems. In: Pryor, WA, editor. Free Radicals in Biology. Vol. 2. New York:

Academic Press, p 85-133.

Frankel EN. 1985. Chemistry of autoxidation: Mechanism, Products and Flavor

Significance. In: Min DB, Smouse TH, editors. Flavor chemistry of fats and oils.

Champaign, IL: AOCS. p. 1-37.

Kincs FR. 1985. Meat fat formulation. JAOCS 62(4):815-818.

King JM. 1996. Chlorophyll photosensitized singlet oxygen oxidation of vitamin D (Ph.

D dissertation}. Columbus, Ohio: The Ohio State University, Columbus. P 65-105.

Kulas E, Ackman RG. 2001. Properties of a-, y-, and a- tocopherol in purified fish oil triacylglycerols. JAOCS 78(4):361-367.

Lee EC, Min DB. 1988. Quenching mechanism of p-carotene on the chlorophyll

sensitized photooxidation of soybean oil. J Food Sci 53(6): 1894-1895.

95 Min DB. 1998. Lipid oxidation of edible oil. In: Akoh CC, Min DB, editors. Food Lipids: chemistry, nutrition, biotechnology. New York : Marcel Dekker. p. 283-296.

Pokomy J, Zwain H, Pazlar M. 1966. Effect of myoglobin on the autoxidation of lard in nonaqueous medium. Nahrung 10(7):730-731.

Rawls FIR, Van Santen PJ. 1970. A possible role for singlet oxidation in the initiation of fatty acid autoxidation. J Amer Oil Chem Soc 47:121-125.

Thron M, Eichner K, Ziegleder G. 2001. The influence of light of different wavelengths on chlorophyll-containing foods. Eebensm-Wiss u Techno1 34(8):542-548.

Yang WTS, Min DB. 1994. Chemistry of singlet oxygen oxidation of foods. In: Ho CT and Hartmand TO, editors. Lipids in Food Flavors. Washington DC: American Chemical

Society, p 15-29.

Zhuang H, Barth MM, Hildebrand DF. 1998. Fatty acid oxidation in plant tissues. In:

Akoh CC, Min DB, editors. Food Lipids: chemistry, nutrition, biotechnology. New York

: Marcel Dekker. p. 333-376.

96 CONCLUSION

1) The photosensitized oxidation of linoleic acid, which is one of the major fatty

acids in vegetable oils, produced significantly more volatile compounds than the

photooxidation of linoleic acid at 4 °C for 48 hr (P<0.05).

2) Singlet oxygen oxidation can explain the formation of volatiles, including 2-

pentylfuran, 2-heptenal l-octene-3-ol, and 2-octene-l-ol from the photosensitized

oxidation of linoleic acid.

3) Pentanal, dimethyl disulfides, hexanal, and heptanal were formed only in the

photosensitized milk at 4 °C for 8 hr and increased significantly as the riboflavin

concentration increased from 5 to 10, 50 ppm (P<0.05).

4) As fat content in milk increased, pentanal, hexanal, and heptanal formation

increased but dimethyl disulfide formation was not changed significantly in the

photosensitized milk at 4 °C for 8 hr (P>0.05).

5) BHA and ascorbic acid, which are free radical scavengers, reduced hexanal and

heptanal formation in the photosensitized milk. Sodium azide, a singlet oxygen

quencher, prevented dimethyl disulfide formation in the photosensitized milk.

97 6) Pentanal formation was dependent on the riboflavin content. The formation

mechanism of pentanal seems to be different from that of hexanal and heptanal in

the photosensitized milk.

7) Singlet oxygen and free radicals play important roles in the formation of volatile

compounds in the photosensitized milk.

8) Lard was a photosensitizer-free lipid model system, which can he used for the

photooxidation and photosensitized oxidation study.

9) Total peak areas in the photosensitized lard were 19 times more than that in the

photooxidized lard for 48 hr. Headspace oxygen content in the photosensitized

lard changed from 21 to 15% for 48 hr but that in the photooxidized lard did not

change significantly (P>0.05).

10) Photooxidation mechanism seems to be a free radical chain reaction like

autoxidation.

98 BIBLIOGRAPHY

Bassette R. 1976. Effects of light on concentrations of some volatile materials in milk. J Milk Food Technol 39(1): 10-12.

Baron CP, Andersen H. 2002. Myoglobin-induced lipid oxidation. A review. J Agric Food Chem 50(14):3887-3897.

Beutner S, Bloedom B, Hoffman T, Martin HD. 2000. Synthetic singlet oxygen quenchers. In: Packer L and Sies H, editors: Methods in Enzymology Volume 319. New York: Academic Press. 226-241.

Buettner GR. 1993. The pecking order of free radicals and antioxidants: lipid peroxidation, a-tocopherol and ascorbate. Archives of Biochem Biophy 300(2):535-543.

Boff JM, Min DB. 2002. Chemistry and reaction of singlet oxygen in foods. Comprehensive review in Food Science and Food Safety I(I):58-72.

Bradley DG. 1991. Riboflavin photosensitized singlet oxygen oxidation in milk products. Dissertation in the Ohio State University. 109 pp.

Bradley DG, Min DB. 1992. Singlet oxygen oxidation of foods. Grit Rev Food Sci Nutr 31:211-236.

Cadwallader KR, Howard CL. 1998. Analysis of aroma-active components of light- activated milk. ACS Symposium Series 705 (Flavor Analysis):343-358.

Callison AL. 2001. Singlet oxygen lipid oxidation mechanisms for reversion flavor formation in soybean oil. M.S. Thesis. The Ohio State University, Columbus.

99 Case RA, Bradley RL, Williams RR. 1985. Chemical and physical methods. In : Richardson GH, editor. Standard methods for the examination of dairy products. Washington, D.C.: American Public Health Assoeiation. p 347-350.

Chan HWS. 1977. Photo-sensitized oxidation of unsaturated fatty aeid methyl esters. The identification of different pathways. JAOCS. 54(3): 100-104.

Chan HWS. 1987. The meehanism of autoxidation. In: Chan HWS, editor. Autoxidation of unsaturated lipids. London: Academic Press, p 1-16.

Chu YH, Rung YL. 1998. A study on vegetable oil blends. Food Chem 62(2): 191-195.

Cooperman JM, Lopez R. 1984. Riboflavin In: Machlin LJ editor. Handbook of vitamins, nutritional biochemical and clinical aspects. New York: Mareel Dekker. p 299-327.

Corey EJ, Taylor WC. 1964. A study of the peroxidation of organic compounds by externally generated singlet oxygen molecules. J Am Chem Soc, 86:3881.

Decker EA. 1998. Antioxidant mechanisms. In: Akoh CC, Min DB, editors. Food lipids- chemistry, nutrition, and biotechnology. New York: Marcel Dekker. p 397-401.

Doleiden FH, Farenholtz SR, Lamola AA, Trozzolo AM. 1974. Reactivity of cholesterol and some fatty acids toward singlet oxygen. Photochem Photobiol 20:519-521.

Edwards AM, Silva E. 2001. Effect of visible light on selected enzymes, vitamins and amino acids. J Photochem Photobiol B 63(1-3):126-131.

Elbe JH, Schwartz SJ. 1996. Colorants. In: Fennema OR, editor. Food chemistry 3'^'^. New York: Marcel Dekker. p 659-673.

Faria JAF, Mukai MK. 1983. Use of gas ehromatographie reactor to study lipid photooxidation. JAOCS 60(1): 77-81.

100 Foote CS. 1976. Photosensitized oxidation and singlet oxygen: consequences in biological systems. In: Pryor, WA, editor. Free Radicals in Biology. Vol. 2. New York: Academic Press, p 85.

Foote CS, Wexler S. 1964. Olefin oxidations with excited singlet molecular oxygen. J Am Chem Soc, 86:3879.

Foote CS, Denny RW. 1968. Chemistry of singlet oxygen quenching by 6-carotene. J Am Chem Soc 90:6232-6238.

Frankel EN, Neff WE, Bessler TR. 1979. Analysis of autoxidized fats by gas chromatography-mass spectrometry: V. Photosensitized oxidation. Lipids 14(12):961- 967.

Frankel EN, Neff W, Selk E. 1981. Analysis of autoxidized fats by gas chromatography- mass spectrometry: VII. Volatile thermal decomposition products of pure hydroperoxides from autoxidzed and photosensitized oxidized methyl oleate, linoleate and linolenate. Lipid 16(5):279-285.

Frankel EN. 1985. Chemistry of tree radical and singlet oxidation of lipids. Prog Lipid Res 23:197-221.

Frankel EN. 1985. Chemistry of autoxidation: Mechanism, Products and Flavor Significance. In: Min DB, Smouse TH, editors. Flavor chemistry of fats and oils. Champaign, IL: AOCS. p. 1-37.

Gaafar AM, Gaber FI. 1992. Volatile flavor compounds of sunlight-exposed milk. J Dairy Sci 20(1): 111-115.

Girotti AW. 1998. Lipid hydroperoxide generation, turnover, and effector action in biological systems. J Lipid Res 39:1529-1542.

Gomes AM, Malcata FX. 1998. Development of probiotic cheese manufactured from goat milk: response surface analysis via technological manipulation. J Dairy Sci 81(6): 1492-1507.

101 Gutierrez I, Criado S, Bertolotti S, Gareia N. 2001. Dark and photoindueed interaction between Trolox, aploar-solvent-soluble model for vitamin E, and riboflavin. J Photocehm Photobiol 62(1):133-139.

Ho CT, Smagula MS, and Chang SS. 1978. The synthesis of 2-(l-pentenyl) furan and its relationship to the reversion flavor of soybean oil. J Am Oil Chem Soc 55:233-237.

Ho CT, Chen Q. 1994. Lipids in food flavors. An overview. In: Ho CT, Hartman TG, editors. Lipids in food flavors-ACS symposium series 558. Washington DC: American Chemical Society, p. 2-14.

Jelen HH, Obuchowska M, Zawirska-Wojtasiak R, Wasowicz L. 2000. Headspace Solid- Phase Microextraction Use for the Charaeterization of Volatile Compounds in Vegetable Oils of Different Sensory Quality. J Agric Food Chem 48(6): 2360-2367.

Jen JJS, Mackinney G. 1970. Photodecomposition of chlorophyll in vitro. II. Intermediates and breakdown products. Photochem Photobiol 11(5): 303-308.

Jung MY, Kim SK, and Kim SY. 1995. Riboflavin-sensitized photoxidation of ascorbic acid: kinetics and amino acid effects. Food Chem 53:397-403.

Jung MY, Yoon SH, Lee HO, Min DB. 1998. Singlet oxygen and ascorbic acid effects on dimethyl disulfide and o ff flavor in skim milk exposed to light. J Food Sci 63(3):408- 412.

Kim YD, Morr CV. 1996. Dynamic headspace analysis of light-activated flavor in milk. Inf Dairy J 6(2):185-93.

Kincs FR. 1985. Meat fat formulation. JAOCS 62(4):815-818.

King JM. 1996. Riboflavin photosensitized singlet oxygen oxidation of vitamin D (Ph. D dissertation}. Columbus, Ohio: The Ohio State University, Columbus. P 65-105.

King JM and Min DB. 1998. Riboflavin photosensitized singlet oxygen oxidation of vitamin D. J Food Sci 63:31-34.

102 Kulas E, Ackman RG. 2001. Properties of a-, y-, and a- tocopherol in purified fish oil triacylglycerols. JAOCS 78(4):361-367.

Kochevar IE, Redmond RW. 2000. Photosensitized production of singlet oxygen. In: Packer L and Sies H, editors: Methods in Enzymology Volume 319. New York: Academic Press. 20-28.

Korycka-Dahl MB, Richardson T. 1978. Activated oxygen species and oxidation of food constituents. Grit Rev Food Sci Nutr 10:209-240.

Krinsky NI. 1977. Singlet oxygen in biological systems. Trends Biochem Sci 2:35-38.

Lee KH, Jung MY, Kim SY. 1998. Effects of ascorbic acid on the light-induced riboflavin degradation and color changes in milk. J Agric Food Chem 46(2):407-410.

Lee EC, Min DB. 1988. Quenching mechanism of P-carotene on the chlorophyll sensitized photooxidation of soybean oil. J Food Sci 53(6): 1894-1895.

Li TL, King JM, Min DB. 2000. Quenching mechanisms and kinetics of carotenoids in riboflavin photosensitized singlet oxygen oxidation of vitamin D 2 . J Food Biochem 24 (6):477-492.

Liu TT, Yang TS. 2002. Optimization of solid-phase microextraction analysis for studying changes of headspace flavor compounds of banana during ripening. J Agric Food Chem 50(4):653-657.

Lledias F, Hansberg W. 2000. Catalase modification as a marker for singlet oxygen. In: Packer L and Sies H, editors: Methods in Enzymology Volume 319. New York: Academic Press. 110-119.

Lu CY, Wang WF, Lin WZ, Han ZH, Yao SD, Lin NY. 1999. Generation and photosensitization properties of oxidized radical of riboflavin: a laser flash photolysis study. J Photochem Photobiol B:Biol 52(1-3): 111-116.

103 Marsili RT. 1999a. SPME-MS-MVA as an electronic nose for the study of off-ftavors in milk. J Agric Food Chem 47(2):648-654.

Marsili RT. 1999b. Comparison of solid-phase microextraction and dynamic headspace methods for the GC-MS analysis of light-induced lipid oxidation products in milk. J Chromatogr Sci 37(1): 17-23.

McGinnis BD, Adams VD, Middlebrooks EJ. 1999. Evaluation of methylene blue and riboflavin for the photosensitized degradation of ethylene glycol. Environmental International 25 (8):953-959.

Michaeli A., Feitelson J. 1994. Reactivity of singlet oxygen toward amino acids and peptides. Photochem Photobiol 59:284-289.

Michaeli A and Feitelson J. 1997. Reactivity of singlet oxygen toward proteins: the effect of structure in basic pancreatic trypsin inhibitor and in ribonuclease A. Photochem Photobiol 65:309-315.

Min DB. 1998. Lipid oxidation of edible oil. In: Akoh CC, Min DB, editors. Food lipids- chemistry, nutrition, and biotechnology. New York: Marcel Dekker. p 283-296.

Min DB. 2000. Recent developments in lipid oxidation and antioxidants. Canadian section of the American Oil Chemists’ Society Meeting. November, 2000. Quebec, Ontario.

Min DB, Lee HO. 1999. Chemistry of lipid oxidation. In: Teranishi R, Wick EL and Homstein 1, editors. Flavor Chemistry: Thirty Years of Progress. New York: Kluwer Academic/Plenum Publishers, p 175-187.

Min DB, Lee HO. 1996. Chemistry of lipid oxidation. In: McDonald RE and Min DB, editors. Food Lipids and Health. New York: Marcel Dekker, Inc. p 241-268.

Min DB, Lee SH, Lee EC. 1989. Singlet oxygen oxidation of vegetable oils. In: Min DB and Smouse TH, editors: Flavor Chemistry of Lipid Foods. American Oil Chemists’ Society, Champaign, IL. p 57-97.

104 Miskoski S and Garcia NA. 1993. Influence of the peptide bond on the singlet molecular oxygen-initiated (02[^Ag]) photooxidation of histidine and methionine dipeptides. A kinetic study. Photochem Photobiol 57:447-452.

Nawar WW. 1998. Lipids. In: Fennema OR, editor. Food Chemistry. 3"^. New York: Marcel Dekker. p 225-320.

Olsen JR. Ashoor SH. 1987. An assessment of light-induced off-flavors in retail milk. J Dairy Sci 70(7): 1362-1370.

Pajares A, Gianotti J, Stettler G, Bertolotti S, Criado S, Posadaz A, Amat-Guerri F, Garcia NA. 2001. Modelling the natural photodegradation of water contaminants a kinetic study on the light-induced aerobic interactions between riboflavin and 4- hydroxypyridine. J Photochem Photobiol 139(2-3): 199-204.

Parkes GD. 1967. Oxygen. In: Parker GD, editor. Mellor’s modem inorganic chemistry. New York: John Wiley and Sons, Inc. P 331-355.

Penton A. 1999. Method development with solid phase microextraction. In: Wercinski S AS, editor. Solid Phase Microextraction. A practical guide. New York: Marcel Dekker. p 27-58.

Pokomy J, Zwain H, Pazlar M. 1966. Effect of myoglobin on the autoxidation of lard in nonaqueous medium. Nahrang 10(7):730-731.

Rawls HR, VanSanten PJ. 1970. A possible role for singlet oxidation in the initiation of fatty acid autoxidation. J Amer Oil Chem Soc 47:121-125.

Sawaya WN, Khalil JK, Al-Shalhat AF. 1984. Mineral and vitamin content of goafs milk. J Am Diet Assoc 84(4): 43 3-43 5.

Schwartz SJ, Lorenzo TV. 1990. Chlorophylls in foods. Crit Rev Food Sci Nutr 29(1): 1- 17.

105 Shahidi F, Wanasundara UN. 1998. Methods of measuring oxidative raneidity in fats ad oils. In: Akoh CC, Min DB, editors. Food lipids-chemistry, nutrition, and biotechnology. New York: Mareel Dekker. p 377-396.

Shirey RE. 1999. SPME fibers and selection for specific applications. In: Wercinski SAS, editor. Solid Phase Microextraction. A practical guide. New York: Marcel Dekker. p 59- 110.

Skibsted LH. 2000. Eight-induced changes in dairy products. Bulletin of the International Dairy Federation 346:4-9.

Steenson DF, Eee JH, Min DB. 2002. Solid phase mieroextraetion of volatile soybean oil and com oil compounds. J Food Sci 67(l):71-76.

Swaisgood HE. 1996. Characteristics of milk. In: Eennema OR, editor. Food chemistry 3"^^. New York: Marcel Dekker. p 850-851.

Thron M, Eichner K, Ziegleder G. 2001. The influenee of light of different wavelengths on chlorophyll-containing foods. Lebensm-Wiss u Technol 34(8):542-548.

Whited EJ, Hammond BH, Chapman KW, Boor KJ. 2002. Vitamin A degradation and light-oxidized flavor defects in milk. J Dairy Sci 85(2):351-354.

Yang WTS. 1994. Effects of synthetic food colorants on singlet oxygen oxidation of foods. Ph.D. Dissertation. The Ohio State University, Columbus.

Yang WTS, Min DB. 1994. Chemistry of singlet oxygen oxidation of foods. In: Ho CT and Hartmand TG, editors. Eipids in Food Flavors. Washington DC: American Chemical Society, p 15-29.

Zhuang H, Barth MM, Hildebrand DF. 1998. Fatty acid oxidation in plant tissues. In: Akoh CC, Min DB, editors. Food Lipids: chemistry, nutrition, biotechnology. New York : Marcel Dekker. p. 333-376.

106 APPENDIX A

Figures of sample preparation

Plunger

Barrel

Stationary phase (fiber)

Figure 28. Solid phase microextraction

107 Wire net shelf Sample bottles

n n

4 fluorescent light lamp

Figure 29. Fluorescent light box used for the sample preparation of linoleic acid and milk

108 Nitrogen gas flushing

Cut the fat Keep at 70 °C part only to melt lard and drain oils Pork meat

Figure 30. Sample preparation of lard (Chapter 5)

109 Figure 31. Light box used for lard samples (Chapter 5)

110 APPENDIX B

Figures of gas chromatograms

Figure 32. Gas chromatogram of linoleic acid with added 5 ppm chlorophyll stored under light for 2 day (Chapter 3)

111 0 ppm

5 ppm

JL

Figure 33. Gas chromatograms of milk with added 0 and 5 ppm riboflavin stored under light for 2 hr (Chapter 4)

112 10 ppm

50 ppm

Figure 34. Gas chromatograms of milk with added 10 and 50 ppm riboflavin stored under light for 2 hr (Chapter 4)

113 0.5%

1.0%

Figure 35. Gas chromatograms of milk with 0.5 and 1.0% fat content stored under light for 8 hr (Chapter 4)

114 2.0%

3.4%

______L

Figure 36. Gas chromatograms of milk with 2.0 and 3.4% fat content stored under light for 8 hr (Chapter 4)

115 Without ascorbic acid

With ascorbic acid

Figure 37. Gas chromatograms of milk with added 50 ppm riboflavin with or without 100 ppm ascorbic acid stored under light for 4 hr (Chapter 4)

116 Without BHA

With BHA

JL

Figure 38. Gas chromatograms of milk with added 50 ppm riboflavin with or without 100 ppm BHA stored under light for 4 hr (Chapter 4)

117 Without Sodium azide

With Sodium azide

1

Figure 39. Gas chromatograms of milk with added 50 ppm riboflavin with or without 100 ppm sodium azide stored under light for 4 hr (Chapter 4)

118 0 hr

6 hr

> Lfy-

Figure 40. Gas chromatograms of lard with 5 ppm chlorophyll stored under light for 0 and 6 hr (Chapter 5)

119 12 hr

jLj J

24 hr

Figure 41. Gas chromatograms of lard with 5 ppm chlorophyll stored under light for 12 and 24 hr (Chapter 5)

120 36 hr

48 hr

Figure 42. Gas chromatograms of lard with 5 ppm chlorophyll stored under light for 36 and 48 hr (Chapter 5) 121 Ohr

6 hr

itP -J:

_ruL_ JL

Figure 43. Gas chromatograms of lard with 0 ppm chlorophyll stored under light for 0 and 6 hr (Chapter 5)

122 12 hr

24 hr

- I N _

Figure 44. Gas chromatograms of lard with 0 ppm chlorophyll stored under light for 12 and 24 hr (Chapter 5)

123 36 hr

■t. L

48 hr

Figure 45. Gas chromatograms of lard with 0 ppm chlorophyll stored under light for 36 and 48 hr (Chapter 5)

124