Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of pH and dilution pretreatments and removal of water-soluble components on the functional… Tang, Karen Sze-Hang 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-0592.pdf [ 6.74MB ]
Metadata
JSON: 831-1.0089800.json
JSON-LD: 831-1.0089800-ld.json
RDF/XML (Pretty): 831-1.0089800-rdf.xml
RDF/JSON: 831-1.0089800-rdf.json
Turtle: 831-1.0089800-turtle.txt
N-Triples: 831-1.0089800-rdf-ntriples.txt
Original Record: 831-1.0089800-source.json
Full Text
831-1.0089800-fulltext.txt
Citation
831-1.0089800.ris

Full Text

THE EFFECTS OF pH AND DILUTION PRETREATMENTS AND REMOVAL OF WATER-SOLUBLE COMPONENTS ON THE FUNCTIONAL PROPERTIES OF SPRAY-DRIED EGG YOLK POWDER by KAREN SZE-HANG TANG B.Sc. (Hons.), The University of Toronto, 1996 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Food Science) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 2000 © Karen Sze-Hang Tang, 2000 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date QrJT { ?> , . DE-6 (2/88) ABSTRACT Spray-drying increases shelf-life and convenience of egg yolk, but may induce damage to its functional properties. The objectives of this study were to investigate the functional properties of spray-dried yolk powders as a function of different pH (3 to 9) and dilution (2- to 10-fold) pre treatments, and to study the impact of spray-drying on functional properties of yolk pellet, the fraction remaining after removal of water-soluble components. Protein solubility and emulsifying properties (Emulsifying Activity Index and Emulsion Stability Index) were considered. Pre-drying pH and dilution had significant effects on protein solubility of the spray-dried powders (p < 0.0001). Protein solubilities of liquid and spray-dried yolk were ~ 70 % and 50 %, respectively. Highest solubility (> 60 %) was obtained when pre-drying pH was between 5.5 to 9 and dilution between 4- to 9-fold, while emulsions of the yolk powders were most stable when pre-drying pH was between 8.0 to 10.0 and dilution between 2-to 4-fold. Based on the protein functionality test results, the best pretreatment condition would be pH 8.5 with 6 times dilution and the worst condition would be pH 3.0 with 6 times dilution. Pellet powder was relatively insoluble (25 % protein solubility) and its emulsion was less stable. Interestingly, liquid pellet gave better emulsion stabilization than commercial liquid yolk. Freeze-drying produced dried yolk with better emulsion stability than spray-drying. However, duration of storage and batch-to-batch variation of commercial yolk might have influenced the results. Differential Scanning Calorimetry showed similar thermograms for yolk powder with best pretreatment conditions, liquid yolk and spray-dried yolk control (Td ~ 80°C). Almost no denaturation peak was detected for pellet samples or yolk powder with the worst pretreatment ii conditions. Results showed that water-soluble fraction of yolk was the main contributor to the thermal behavior of yolk, and freeze-dried yolk samples were more labile to heat denaturation when reconstituted at pH 3 than at higher pHs. The Raman spectra of yolk samples were dominated by vibrational bands of the lipid components, and therefore possible differences in lipoprotein structure as a function of spray drying could not be detected. iii TABLE OF CONTENTS Page ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES viii LIST OF FIGURES x ACKNOWLEDGEMENTS xi Chapter 1. INTRODUCTION 1 Chapter 2. LITERATURE REVIEW 3 2.1 Physical Properties of Egg Yolk 3 2.1.1 Egg Yolk Composition 3 2.1.1.1 Lipids in Yolk 4 2.1.1.2 Proteins and Lipoproteins in Yolk 5 2.1.2 pH of Yolk 8 2.1.3 Color of Yolk 8 2.2 Functional Properties of Egg Yolk 9 2.2.1 Food Emulsions 9 2.2.2 Egg Yolk as an Emulsifier \ \ 2.2.3 Solubility of Yolk 16 2.2.4 Gelling Properties of Yolk 17 2.3 Processing of Egg Yolk 19 2.3.1 Spray-drying Process 19 2.3.2 Changes to the Functional Properties of Yolk after Dehydration 20 2.4 Immunoglobulin Y and Yolk Pellet 24 2.4.1 Immunoglobulin Y (IgY) 24 2.4.2 Yolk Pellet 25 2.5 Optimization Techniques 26 2.5.1 Response Surface Methodology (RSM) 26 2.5.2 Central Composite Rotatable Design (CCRD) 27 2.6 Structural Analysis Techniques 28 2.6.1 Differential Scanning Calorimetry (DSC) 28 2.6.2 Raman Spectroscopy 29 Chapter 3. THE STUDY 31 3.1 Background and Overall Objectives 31 3.2 Phase 1 - Optimization of Pretreatment Conditions 34 for Spray Drying of Yolk 3.2.1 Objectives in Phase 1 34 iv Page 3.2.2 Work Plan for Phase 1 35 3.2.3 Material and Methods for Phase 1 36 3.2.3.1 Sample Preparation 36 3.2.3.2 Pretreatments of Yolk Prior to Spray-drying 37 (1) Central Composite Rotatable Design 37 - Response Surface Methodology (CCRD-RSM) - Dilution and pH adjustment 37 (2) Preparation of Yolk Pellet 39 3.2.3.3 Spray-drying Conditions 39 3.2.3.4 Storage of Yolk Samples 40 3.2.3.5 Analyses of Yolk Samples 42 (1) Proximate Analyses 42 Moisture Content Determination 42 Protein Determination 43 Goldfish Crude Fat Determination 43 - "Total" Fat Determination 44 (2) Ffunterlab Color Analyses for Yolk Powders 46 (3) Functional Properties 47 - Protein Solubility 47 - Emulsifying Properties (EAI, ESI) 49 (4) Statistical Analyses 51 3.2.4 Results and Discussion for Phase 1 52 3.2.4.1 Pretreatment Conditions 52 3.2.4.2 Proximate Analyses 53 (1) Moisture (Solids) Content 53 (2) Protein Content (% N x 6.25) 56 (3) Fat Content ("Total" Fat and Crude Fat) 56 3.2.4.3 Hunterlab Color Analyses 59 3.2.4.4 Functional Properties of Yolk Samples 61 (1) Protein Solubility 61 (2) Emulsifying Properties of Yolk Samples 66 - Emulsifying Activity Index (EAI) 66 - Emulsion Stability Index (ESI) 69 3.2.4.5 Summary of Results in Phase I 73 3.3 Phase 2 - Further Investigation on Spray-dried Yolk Powders 75 3.3.1 Rationale 75 3.3.2 Objectives in Phase 2 76 3.3.3 Work Plan for Phase 2 77 3.3.4 Materials and Methods for Phase 2 79 3.3.4.1 Preparation of Hand-separated Egg Yolk 79 3.3.4.2 Freeze-drying of Yolk Samples 79 3.3.4.3 Analyses of Yolk Samples 80 (1) Proximate Analyses 80 Acid Hydrolysis Fat Determination 80 Water Activity of Yolk Powders 81 V Page (2) Conductivity of Yolk Preparations prior to Spray-drying 82 (3) Functional Properties 82 - EAI for yolk preparations with different 82 % solids with and without 0.5 M NaCI (4) Structural Analyses 83 - Differential Scanning Calorimetry (DSC) 83 - DSC on Freeze-dried Yolk Control and 84 Water-soluble Fraction at different pH's Raman Spectroscopy 84 (5) Statistical Analyses 85 3.3.5 Results and Discussion for Phase 2 86 3.3.5.1 Proximate Analyses 88 (1) Moisture (Solids) Content 88 (2) Water Activity of Yolk Powders in Phase 2 91 (3) Fat Content 94 Goldfish Crude Fat Determination 94 - Folch's Method of Fat Determination 95 Acid Hydrolysis Fat Determination 96 - Comparison of the 3 Fat Determination Methods 97 Ratios of Fat Content as obtained by different Fat 98 Determination Methods (4) Protein Content (% N x 6.25) 100 3.3.5.2 Hunterlab Color Analyses 102 (1) L- value 106 (2) a-value 106 (3) lvalue 107 3.3.5.3 Conductivity of Liquid Yolk Preparations 108 3.3.5.4 Functional Properties of Yolk Samples 109 (1) Protein Solubility 109 (2) Emulsifying Properties 112 - EAI and ESI of Yolk Emulsion at varying 112 % Solid Content with and without 0.5 M NaCI Emulsifying Tests on Selected Samples from 118 Phase 1 and 2 - Emulsifying Activity Index (EAI) 118 - Emulsion Stability Index (ESI) 121 Comparisons on the Emulsifying Properties of 124 Spray-dried and Freeze-dried Yolk and Pellet 3.3.5.5 Structural Analyses 126 (1) Differential Scanning Calorimetry (DSC) 126 Onset Temperature of Denaturation 129 Enthalpy of Endothermic Peak 129 - Denaturation Temperature 130 - Width at Half Peak Height 131 vi Page - DSC at different pH's for Freeze-dried Yolk 132 and Water-soluble Fraction - Summary of DSC results 134 (2) Raman Spectroscopy Analyses 136 Chapter 4. GENERAL DISCUSSION AND CONCLUSIONS 141 REFERENCES 145 APPENDIX A. Storage Information on Yolk Samples 154 APPENDIX B. List of Abbreviations 155 vii LIST OF TABLES Page Table 1 Conditions generated by CCRD-RSM for Pretreatment of Yolk prior 38 to Spray-drying with Liquid and Spray-dried Yolk and Yolk Pellet as Comparisons (Phase 1) Table 2 Proximate Analyses for Yolk Samples (Phase 1) 55 Table 3 Hunterlab Color Analyses for Yolk Samples (Phase 1) 60 Table 4 Protein Solubility of Yolk Samples in 0.1 M Sodium Phosphate Buffer 63 at pH 6.35 (Phase 1) Table 5 Emulsifying Activity Index (EAI) for Yolk Samples (Phase 1) 67 Table 6 Emulsion Stability Index (ESI) for Yolk Samples (Phase 1) 70 Table 7 Sample Designation for Phase 2 86 Table 8 Moisture Content of Yolk Samples (Phase 2) 90 Table 9a Water Activity of Selected Yolk Powders (Phase 1) 92 Table 9b Water Activity of Yolk Powders (Phase 2) 93 Table 10 Goldfish Fat Determination (Phase 2) 95 Table 11 Folch's Fat Determination (Phase 2) 95 Table 12 Acid Hydrolysis Fat Determination (Phase 2) 96 Table 13 Fat Determination by Different Methods 98 Table 14 Ratios of % Fat as determined by Different Fat Determination 99 Methods Table 15 Protein Content (% N x 6.25) (Phase 2) 101 Table 16 Hunterlab Color Analyses for Yolk Samples (Phase 2) 105 Table 17 Conductivity of Liquid Yolk Preparations (Phase 2) 108 Table 18 Protein Solubility of Spray-dried and Liquid Samples (Phase 2) 110 Table 19 Protein Solubility of Spray-dried and Freeze-dried Yolk and Pellet 111 Samples Table 20 EAI of Yolk Emulsion at Varying % Solid Content with and without 114 0.5 M NaCl viii Page Table 21 ESI of Yolk Emulsion at varying % Solid Content with and without 116 0.5MNaCl (Phase 2) Table 22 EAI of Selected Samples from Phase 1 and 2 120 Table 23 ESI of Selected Samples from Phase 1 and 2 123 Table 24 EAI and ESI of Spray-dried and Freeze-dried Yolk and Pellet Samples 125 Table 25 DSC Results for Selected Yolk Samples from Phase 1 and 2 127 Table 26 DSC Results for FYC and WSF solution (10 % w/v) at different pH's 133 ix LIST OF FIGURES Page Figure la Front View of Pilot Plant-scale Spray-dryer 41 Figure lb Side View of Pilot Plant-scale Spray-dryer 41 Figure 2 Protein & "Total" Fat Content of Yolk & Pellet Samples (Phase 1) 58 Figure 3 % Protein Solubility at pH 6.35 (Phase 1) 64 Figure 4 Contour Plot of Protein Solubility for CCRD-RSM Yolk Powders 65 Figure 5 EAI (m2/g protein) of Yolk Samples (Phase 1) 68 Figure 6 ESI (min) of Yolk Samples (Phase 1) 71 Figure 7 Contour Plot of ESI (min) for CCRD-RSM Yolk Powders 72 Figure 8 Liquid Yolk Control (Undiluted) and Liquid Pellet (Two-fold 103 dilution) Figure 9 Liquid Yolk Preparations (Preparation for "Worst" Powder, Yolk 103 Control and Preparation for "Best" Powder) Figure 10 Yolk Powder Samples (Spray-dried Yolk pellet, Spray-dried Yolk 104 Control and Commercial Spray-dried Yolk) Figure 11 Yolk Powder Samples ("Best" Powder, Yolk Control and "Worst" 104 Powder) Figure 12 EAI (m2/g solid) of Yolk Emulsion with 0.5 M NaCI 115 Figure 13 EAI (m2/g solid) of Yolk Emulsion 116 Figure 14 ESI (min) of Yolk Emulsion with 0.5 M NaCI 117 Figure 15 ESI (min) of Yolk Emulsion 117 Figure 16 Typical DSC Denaturation Peak of Yolk Samples 128 Figure 17 Raman Spectrum for YPC 138 Figure 18 Raman Spectrum for the oil extract of YPC 139 Figure 19 The difference in the Raman spectra for YPC and its corresponding 140 oil fraction X ACKNOWLEDGEMENTS My heartfelt thank you goes to Dr. E. Li-Chan. I really appreciate the dedication she put on her students, her advice and care are tremendous. Dr. Li-Chan, thank you for giving me many different opportunities to learn and to improve. It is my privilege to have you as my supervisor. I am thankful to Dr. T. Durance and Dr. E. Charter, members of my supervisory committee, for their valuable advice, and to Dr. J. Vanderstoep and Dr. K. Cheng for being in my examining committee. Thanks to everyone at Food Science for creating such a lively studying environment. Sherman and Val, thanks for attending to my many needs and questions in the lab. To everyone who helped with my project, I thank you all. Thank you to Professor Ching-Yung Ma for his guidance and hospitality during my stay in his lab at the University of Hong Kong. To Anna, Manoj, Sureka and Meng, at Professor Ma's lab, thank you for the friendship and assistance. I wish to show my appreciation to Vanderpols Eggs Ltd. and to Canadian Inovatech Inc. for making this project possible. Thanks to everyone who assisted me at Inovatech, especially to Mel and Cliff. Thanks also to Jack Losso for his advice. Last but not least, I would like to acknowledge my family, friends and relatives for their continuous support and love. You have all given me the courage to hold on. Thank you from the bottom of my heart. I promise I will not let you down. Mom and Dad, thanks for putting your trust in me, and Simon, thanks for all your encouragement. xi CHAPTER 1. INTRODUCTION Egg yolk is rich in both lipids and proteins and they are in close association with • each other. The yolk consists of lipoproteins, water-soluble proteins, phosphoproteins, as well as some minor components such as neutral lipids, carbohydrates, minerals and amino acids. The balance and interactions between these components give egg yolk a number of functional properties. Since egg yolk is well-known for its excellent emulsifying properties and for the fact that it is nutritious, it is a popular and useful ingredient in many daily dietary items, such as bakery products, cake mixes, salad dressings, noodles, baby foods, and military ration. Egg yolk can be made into a product of longer shelf-life and convenience by dehydration. The most common method used is spray-drying because of the economy and speed of the operation. Egg yolk is atomized and dried under a hot stream of air into a powder. Due to the heat involved in the drying process and the susceptibility of the components of yolk are susceptible to heat, there will be slight changes in the functional properties and also an increased chance of lipid oxidation in the yolk. Several studies in the past have shown the effects of pH on the functional properties of liquid yolk and the individual yolk components. It appears that the functionality of various components of yolk depends greatly on the pH environment they are under. However, little research has been reported on the effects of adjusting pH of liquid yolk prior to spray-drying on the properties of the subsequent yolk powder. Therefore, this study will look at the effects of different pH pretreatments prior to spray-1 drying on the functional properties of the resulting spray-dried yolk. In addition to adjusting the pH, it is also of interest to look at how yolk with different solid contents will behave upon spray-drying. As a result, the objective set forth here is to find an optimal pre-drying pH and solid content of the yolk, which will give the best yolk powder in terms of its functionality. Another aspect of yolk that will be investigated is the portion of yolk that remains after the extraction of the yolk antibodies, immunoglobulin Y (IgY). In recent years, there has been interest in using these antibodies for food and pharmaceutical applications. Research has been done to determine the best conditions for the extraction of IgY while best preserving the use of the remaining yolk fraction. Kwan et al. (1991) employed an IgY extraction method using minimal amount of organic solvents. This method extracts the IgY along with the water-soluble components from yolk, and the yolk fraction that is left is called the yolk pellet. The yolk pellet was found to have good emulsifying properties for mayonnaise production and is a potential emulsifying agent in food. Since commercial egg yolk is usually spray-dried to increase its storage life, it is interesting to find out whether this yolk pellet can undergo the same drying process while still retaining its functional properties. The results from this study should provide useful information on optimizing the conditions of the spray-drying process of yolk to retain or enhance their functional properties for applications in food systems. It would also verify the possible use of yolk pellet as an ingredient in food. 2 C H A P T E R 2. L I T E R A T U R E R E V I E W 2.1 Physical Properties of Egg Yolk 2.1.1 Egg Yolk Composition Egg yolk is a nutrient-dense food component. Although it is slightly over one-third of the edible portion of an egg, it yields about 78 % the total calories (Watkins, 1995). Yolk consists mainly of water, lipids and proteins. It is also a good source of phosphorus, iron, zinc, calcium, and several vitamins, such as vitamins B 6 , B12 and A, folic acid, pantothenic acid, thiamine and riboflavin (Watkins, 1995). It is also popular for its yellowish appearance. Therefore, egg yolk is a traditional food and is used as an ingredient in items such as bakery goods, cake mixes, salad dressings, noodles, baby foods, and military ration. In general, the total solid content of egg yolk is about 50 %. However, the value is affected by the age of the hens and the storage time of the shell eggs (Marion et al., 1965; Rose et al., 1966). Water migrates from the albumen to the yolk during storage of shell eggs, which lowers the solid content of yolk. Since albumen adheres to the vitelline membrane of the yolk, commercially prepared liquid yolk usually contains about 15-20 % albumen (Powrie and Nakai, 1985). On average, egg yolk contains about 31.8-35.5 % lipid, 15.7-16.6 % protein, 0.2-1.0 % carbohydrate and 1.1 % ash on a 50 % solid basis (Powrie and Nakai, 1985). 3 2.1.1.1 Lipids in Yolk Variability in total lipid content has been attributed primarily to strain rather than diet of the laying hens (Marion et al., 1965). Yolk lipid is composed of about 66 % triglyceride, 28 % phospholipid, and 5 % cholesterol (Powrie and Nakai, 1985) and a small amount of other lipids. Almost all of these lipids are bound to protein and are present as lipoproteins. Yolk phospholipid consists of 73.0 % phosphatidylcholine (lecithin), 15.0 % phosphatidylethanolamine (cephalin), 5.8 % lysophosphatidylcholine, 2.5 % sphingomyelin, 2.1 % lysophosphatidylethanolamine, 0.9 % plasmalogen, and 0.6 % inositol phospholipid (Rhodes and Lea, 1957). In terms of the fatty acids in the lipid fraction of the yolk, the saturated fatty acids, mainly palmitic and stearic acids, comprise between 30 to 38 % of the fatty acids in yolk lipid. These fatty acids represent about 30 % of the fatty acids in the triglyceride fraction, and about 49 % and 54 % of phosphatidycholine and phosphatidylethanolamine, respectively, in the phospholipid fraction. In contrasts, the percentages of oleic and linoleic acids are relatively higher in the triglyceride than in the phospholipid fraction. The fatty acid profile of egg yolk is influenced by a variety of biological and environmental factors, such as genetic strain of the birds, geographical influences, and composition of the feeding pellet (Nemecz and Mennear, 1995). The type of fat in the diets of the hens is the main determinant for the fatty acid composition (Cobos et al., 1995). Saturated fatty acids are less affected than the polyunsaturated fatty acids by the change in fatty acid composition in the diets of the laying hens. 4 Due to the high occurrence of multiple unsaturated carbon-carbon bonds, the yolk lipid is labile to oxidation because air is not excluded from the egg once it has been laid. However, the yolk itself does contain highly specialized chemical, structural, and physical properties to prevent oxidative damage and other changes. Egg yolk contains antioxidants, such as tocopherol and the yellow pigments, carotenoids. It has been suggested that the yolk protein, phosvitin, also acts as an antioxidant in egg (Lu and Baker, 1986). In addition, since all the lipids in yolk are bound, and are usually contained inside the lipoprotein particles, oxygen may have difficulty to access these sensitive lipids to cause damages (Burley, 1990). Martin et al. (1963) reported that more than 90 % of the total lipid present in egg yolk is associated with the low-density fraction lipoproteins. A water of hydration content between 10 and 30 % of the total weight of the lipoprotein micelle was suggested. 2.1.1.2 Proteins and Lipoproteins in Yolk Very little protein in yolk exists alone, as they are found mostly in close relations with lipids. Two-thirds of the yolk proteins are associated with lipids in the form of lipoproteins. Yolk may be described as a complex system containing a variety of particles suspended in a protein solution. The main types of particles are yolk spheres, free-floating drops of granules, low-density lipoprotein (LDL), and myelin figures (Li-Chan et al., 1995). Most of the yolk lipids are in the low-density lipoprotein fraction and present in the liquid yolk in a finely dispersed state. Schultz and Forsythe (1967) postulated that the triglycerides make up the inner core of the highly emulsified low-density lipoprotein, 5 which is surrounded by a phospholipid shell. The protein molecules are wrapped around the shell. In addition to yolk globules or spheres, whole yolk also contains large particles known as insoluble yolk globules, because they remain insoluble under concentrations of salt or urea that dissolve or disrupt other yolk structures (Vadehra et al., 1977). Upon high-speed centrifugation, yolk can be separated into sedimented granules and a clear fluid supernatant called the plasma (Schmidt et al., 1957). The major portion of yolk is the plasma which is about 78 % of the total liquid yolk (Li-Chan et al., 1995). The composition of plasma is near that of yolk but it contains significantly less protein and more lipids. Plasma is composed of livetins, which are lipid-free globular proteins, and of low-density lipoproteins (LDL). The livetins and low-density lipoproteins represent about 10.6 % and 66 %, respectively, of the total yolk solids (McCully et al., 1962). Minor components of the plasma includes riboflavin-binding protein and biotin-binding protein which have biologically active properties (Nakamura and Doi, 2000). The plasma lipoproteins fraction contains between 84 and 89 % lipids, which are made up of 75 % neutral lipids and 26 % phospholipids, the latter consisting of 71-76 % phosphatidylcholine, 16-20 % phosphatidylethanolamine, and 8-9 % sphingomyelin and lysophospholipids (Martin et al., 1963). LDL is considered to be a large globular molecule of 4-10 p,m diameter, consisting of a core of triglyceride and a surface layer of both phospholipids and protein with hydrophilic groups (Powrie and Nakai, 1986). As for the lipid-free livetins, there are different isomers of livetins (a-, P- and y-livetins). Livetins are presumably derived from the blood of the laying hens. Williams (1962) identified a-livetin as a serum albumin, (3-livetin as an o -^glycoprotein, and y-6 livetin as a y-globulin. The y-livetins, or IgY, have similar structure to mammalian IgE and IgM, and thus can be a potential source of antibodies. Chicken can be immunized to produce antigen-specific IgY for use as a therapeutic or prophylactic agent, as well as for fortifications of foods, especially of infant formulas. The granules, the sediment layer after centrifugation, make up about 19 to 23 % of the solids in yolk (Burley and Cook, 1961). Dried granules contain, by weight percentage, about half the lipids and cholesterol, and double the proteins of yolk itself (Anton and Gandemer, 1997). Divalent and trivalent cations, such as iron, are highly concentrated in the granules as shown by atomic absorption spectrophotometry (Causeret et al., 1991). According to Burley and Vadehra (1989), granules are composed of 70 % a- and |3-lipovitellins (collectively named as high-density lipoproteins, HDL), 16 % phosvitin, and 12 % low-density lipoprotein. Phosvitin is a non-lipid phosphoprotein, and the lipovitellins and low-density lipoproteins are lipid-protein complexes. Phosvitin and lipovitellins have an affinity for each other and phosvitin-lipovitellin complex is probably the basic unit of the granules (Burley and Cook, 1961). The low-density lipoprotein of granules is about 4 % of the total LDL of yolk and the properties of these LDL are almost of same as those of plasma LDL based on their similar molecular electrophoresis patterns and compositions (Nakamura and Doi, 2000). Phosvitin contains about 10 % phosphorus and represents about 80 % of the protein phosphorus in yolk. It has a capacity to bind iron strongly. Taborsky (1963) observed that phosvitin promotes a rapid oxidation of the ferrous ion and that it binds the ferric ion strongly and extensively. The antioxidant activity of phosvitin in an egg yolk phospholipid 7 emulsion has been investigated (Lu and Baker, 1986). Phosvitin was found to inhibit Fe z +-and Cu2+-catalysed oxidation (Lu and Baker, 1986). 2.1.2 pH of Yolk The pH of yolk in freshly laid eggs is generally about 6.0. During storage, the pH of yolk gradually increases to between 6.4 to 6.9 as a function of time and temperature (Hidalgo et al., 1996). Commercial separated liquid yolk usually has a higher pH of about 6.5 because of the presence of 10 to 15% of albumen (Powrie and Nakai, 1985), which is more alkaline than pure yolk. The isoelectric point of individual yolk components range from pi of 4 - 8.5, and the average pi of whole yolk is about 5.5 - 5.8 (Gallagher and Voss, 1970; Tenenhouse and Deutsch, 1966; Ternes, 1989). 2.1.3 Color of Yolk The naturally occurring pigments in chicken egg yolk are mainly alcohol-soluble xanthophylls, lutein and zeaxanthin (Yang and Baldwin, 1995). The yellowish color of the yolk is greatly influenced by the xanthophylls, also known as oxycarotenoids, in the hens diet (Karunajeewa et al., 1984). Xanophyll is the major component causing yolk to have a deep orange-yellow color (Watkins, 1995). Apparently, it has little or no nutritive value. However, yolk color does have influence on consumer preferences for the eggs. 8 2.2 Functional Properties of Egg Yolk As mentioned above, egg yolk is rich in both lipids and proteins. The balance and interactions between these components give egg yolk a number of functional properties and make it a useful ingredient in food systems. Functional properties such as emulsifying properties, solubility and gelling properties will be discussed. 2.2.1 Food Emulsions Food emulsions, are usually two-phase systems made up of immiscible liquids. One phase is in the form of finely divided droplets of diameters generally larger than 0.3 pm (Friberg, 1997). The other is a continuous or external phase in which the finely dispersed droplets, the internal or discontinuous phase, is suspended. The most common form of emulsion in food is the oil-in-water (O/W) emulsion, in which the dispersed phase is composed of oil droplets and water is the continuous phase. In the formation of an emulsion, the surface-active proteins form an interfacial film by orienting the polar portion of the molecule to the aqueous phase and the non-polar end to the oil phase, thus helping form a stabilized emulsion (Yang and Baldwin, 1995). An emulsion is intrinsically unstable and with time the droplets of the dispersed phase will come together and the emulsion will undergo processes which will result in the separation of oil and water phases. Oil-in-water emulsions may undergo flocculation, coalescence, creaming and phase inversion, which causes the emulsions to break. The stability of an emulsion depends on the balance of the attractive and repulsive forces as well as the steric forces that associate with the interfacial oil and water molecules. 9 Several parameters are used to measure the emulsifying properties of an emulsion. Pearce and Kinsella (1978) stated that the ability of a protein to aid the formation of an emulsion is related to its ability to adsorb to and stabilize the oil-in-water interface. Therefore, the capacity of a protein to stabilize an emulsion may be related to the interfacial area that can be coated by the available protein. Thus, they define a property of emulsion, called the emulsifying activity, as the capacity of surface-active molecules (proteins or phospholipids) to cover the oil/water interface created by mechanical homogenization. Emulsifying Activity Index (EAI) is units of area of interface stabilized per unit weight of protein and is estimated from the turbidity of the emulsion at a wavelength of 500 nm. Emulsifying activity can also be measured by determining the particle size distribution of the dispersed phase. This can be done by microscopy Coulter counting or spectroturbidity estimations (Walstra, 1968). The higher the EAI, the larger the area of the lipid-protein interface is stabilized. Another important property of emulsions, which is the emulsion stability, is the capacity to avoid flocculation, creaming and/or coalescence (Anton and Gandemer, 1997). It is commonly measured in terms of the oil and/or cream separating from an emulsion during a certain period of time at a stated temperature and gravitational field (Hill, 1996). It can also be determined as the ratio of the volume of final to initial emulsion, after centrifugation at low speed or stood several hours with or without heating (Cheftel et al., 1985). The destructive technique of sampling layers, diluting them and measuring the turbidity is also carried out (Pearce and Kinsella, 1978). Pearce and Kinsella (1978) found that the Emulsion Stability Index (ESI) is dependent on the time interval chosen for the turbidity measurement. The decrease in absorbance with time is approximately first order 10 in absorbance. The emulsion breaks gradually and separates into two layers, the oil floats to the top and the bottom layer becomes a lot less turbid and therefore, turbidity decreases. Lastly, the emulsifying capacity, is define the capacity of the emulsion to hold oil and is expressed as the maximum amount of oil that is emulsified under specific conditions by a standard amount of protein before phase inversion (Cheftel et al., 1985). 2.2.2 Egg yolk as an emulsifier Egg yolk, by itself, is a natural oil-in-water emulsion and is also known to be an efficient emulsifying agent for other O/W emulsions. It consists of a dispersion of oil droplets in a continuous phase of aqueous components. Egg yolk contains various emulsifying agents such as hydrophobic and hydrophilic proteins, phospholipids, and cholesterol (Kiosseoglou and Sherman, 1983a; Carrillo and Kokini, 1988). The emulsification ability of yolk is achieved through a reduction in the surface tension at the interface of the emulsion. Due to its excellent emulsifying properties, egg yolk plays an important role in food emulsions such as mayonnaise, salad dressing, cake batter containing shortening, cream puffs, and hollandaise sauce (Yang and Baldwin, 1995). Egg yolk, whole egg and egg white are all good emulsifiers (surface-active agents) for food products. Egg yolk is rated as four times as effective as egg white as an emulsifier, and whole egg is intermediate. All constituents of yolk (HDL, LDL, phosvitin and livetin) can adsorb at the oil/water interface (Kiosseoglou and Sherman 1983a; Chung and Ferrier, 1991a; Davey et 11 al., 1969). The very excellent emulsifying properties of egg yolk were found to be due to the interactions between different yolk components. Very pure lecithin is ineffective as an emulsifier (Yeadon et al., 1958). Both the lipoprotein and livetin fractions contribute to the high surface activity of egg yolk by reducing the surface energy, which exerts a major influence on emulsion formation (Davey et al., 1969). All studies suggest that the emulsifying properties of egg yolk depend to a great extent on the lipoprotein fraction. Davey et al. (1969) suggested that LDL was the most important contributor to the emulsifying properties of egg yolk. Among yolk proteins, the low-density lipoproteins exhibit a higher adsorbing capacity than globular proteins, such as phosvitin and livetin, because they have a more flexible structure and a greater surface hydrophobicity. The excellent emulsifying ability of LDL was attributed to its high lipid binding ability (Mitzutani and Nakamura, 1985; Nakamura and Doi, 2000). It was reported that mean globule size in emulsions made with LDL was much smaller than that for emulsions made with bovine serum albumen (BSA ) (Mizutani and Nakamura, 1987). The smaller the globule size at the interface, the better the emulsifying properties. The proposed mechanism hypothesizes that lipoproteins structure become more open at the interface and then, apoproteins adsorb at the interface whereas neutral lipids coalesce with oil droplets (Kiosseoglou and Sherman, 1983b; Mine and Bergougnoux, 1998). Among those constituents of LDL, both protein and phospholipid are known to have excellent emulsifying and stabilizing properties (Mizutani and Nakamura, 1985; Yang and Baldwin, 1995). However, the contribution of proteins to emulsifying activity is higher than that of phospholipid (Kiosseoglou and Sherman, 1983b; Mizutani and Nakamura, 1984, 1985; 12 Bringe et al., 1996). Emulsions with LDL were more stable during storage than were those with BSA (Mitzutani and Nakamura, 1987). Both protein solubility and total emulsifying activity of the egg yolk protein concentrates decreased with repeated solvent (ethanol, hexane/isopropanol, chloroform/methanol) extractions for lipid extraction. This indicated that the soluble protein was probably a major contributor to emulsifying activity (Chung and Ferrier, 1991b). Insoluble proteins apparently played no important role in emulsification (Franzen and Kinsella, 1976; Holm and Eriksen, 1980). According to Powrie and Nakai (1985), commercially prepared liquid yolk usually contains 15-20 % of egg albumen, therefore it is interesting to know if albumen contamination will affect the yolk's emulsification ability. It was found that as the albumen content in yolk increases, the viscosity decreases by dilution and the emulsion stability decreases (Chang et al., 1970; Varadarajulu and Cunningham, 1972a). The reduced stability of the emulsion may be caused in part by the decrease in viscosity. However, the reduction in solids content and the interactions between albumen proteins and yolk fractions are probably the major factors. At 20 % albumen, emulsifying power of yolk was reduced by 50 % (Varadarajulu and Cunningham, 1972a). The ovalbumin present in albumen was found to be a poor emulsifying agent (Pearce and Kinsella, 1978; Cheftel et al., 1985) and its presence in commercial yolk can act as a destabilizer in emulsification. Thus, it was suggested that as much egg white should be removed from yolk as possible before drying if the egg yolk is intended to be an emulsifier (Chung and Ferrier, 1991a). 13 Emulsifying activity and emulsion stability increased with protein concentration, oil volume fraction and mixing speed. Mixing speed had the greatest influence and protein concentration had the least influence on emulsifying activity for phosvitin (Chung and Ferrier, 1991b). Varadarajulu and Cunningham (1972a) found that yolk emulsions were least stable at pH 5.6. The average particle (globule) size and concentration of lipoproteins at the interface were greater for emulsions made at pH 3.0 and 5.0 than at pH 7.0 and 9.0, resulting from the formation of lipoprotein dimers at acid pHs (Kisseoglou and Sherman, 1983b; Mine, 1998). The larger the particle size, the more unstable the resulting emulsion. Thus, the yolk emulsions are less stable at lower pH's. Modification of pH can induce a loss of the yolk granule structure by solubilization of proteins (Causeret et al., 1991). It was suggested that this was due to changes in the ionic bridges between bivalent cations and the phosphate groups of the phosphoseryl residues of phosphoproteins in the yolk. Similarly, increase in ionic strength has been shown to disrupt the granule structure (Causeret et al., 1991). However, there was a slight controversy about the effect of pH and salt on the emulsifying properties as it was reported by Mitzutani and Nakamura (1984, 1985) and Carrillo and Kokini (1988) that the emulsifying properties of egg yolk LDL were little affected by pH or salt concentration. Heat treatment also affects emulsifying properties. Varadarajulu and Cunningham (1972b) reported that pasteurization of liquid yolk at 61°C had no influence on emulsifying properties, but pasteurization at 63 or 65°C for four minutes significantly (p < 0.05) enhanced emulsifying capacity. The temperature at which a protein denatures, followed by aggregation, also depends on the pH and ionic 14 strength used. When the protein molecules are unfolded due to temperature or pH effect, more hydrophobic groups will be exposed and hydrophobic interactions can lead to aggregation followed by coagulation and precipitation (Englard and Seifter, 1990). Mayonnaise is an example of an oil-in-water (O/W) emulsion with egg yolk acting as the main surface-active ingredient. The viscoelasticity of this O/W emulsion stabilised with egg yolk is affected by pH, NaCI and temperature (Kiosseoglou and Sherman, 1983a). In terms of the possible difference in yolk fatty acid profiles in eggs, Jordan et al. (1962) found that the type of fat in the hens' diets produced no significant effect on emulsion separation. Pankey and Stadelman (1969) also found no significant differences in emulsification capacity of egg yolk from hens fed rations supplemented with corn, soybean, olive, safflower, or hydrogenated coconut oils. However, the age and strain of the laying hens may affect the emulsification capacity. For example, increase in the age of hens would cause an increase in solids and a decrease in emulsification ability of eggs. (Varadarajulu, 1971; Varadarajulu and Cunningham, 1972b). Eggs from Brown Leghorns had twice the emulsification capacity of eggs from White Leghorns. Low social dominant strains of both the Rhode Island Red and White Leghorn breeds produced eggs with greater emulsification ability than eggs from the high social dominant strains (Varadarajulu and Cunningham, 1972b). 15 2.2.3 Solubility of Yolk Among the functional properties of proteins, protein solubility appears to be the most important determinant because of its significance on the other functional properties of proteins (Hailing, 1981). In general, proteins need to have high solubility to provide good emulsion, foam, gelation and whipping properties (Vojdani, 1996). Protein solubility depends a lot on the surface-active properties of proteins. Solubility of yolk and yolk granules is related to ionic strength. An increase in ionic strength has been shown to disrupt the yolk granule structure making it more soluble (Causeret et al., 1991). Yolk and granules required an ionic strength > 0.3 M NaCI to become solubilized at pH 7.0, whereas plasma was solubilized at any ionic strength (Anton and Gandemer, 1997). The solubility of yolk and granules increased to 80 % when ionic strength increased to 0.3 M NaCI. At low ionic strength, granules were found to have low solubility due to the formation of non-soluble lipovitellin (HDL)-phosvitin complexes which can be precipitated by centrifugation (Causeret et al., 1991). When the solubility of the granules and plasma was related to the corresponding emulsifying properties, Dyer-Hurdon and Nnanna (1993) found that plasma had better emulsifying activity and emulsion stabilization property than granules at low ionic strength, < 0.3 M NaCI. As mentioned earlier, granules had low solubility below 0.3 M NaCI. At such ionic strength, the conformation of the lipoproteins in the granules is such that their hydrophobic groups were more exposed. However, it was found that granules exhibited a better emulsion stability • than plasma when both are at 80 % solubility in 0.5 M NaCI solution (Anton and 16 Gandemer, 1997). Therefore, it can be seen that the solubility of the components is intimately related to their emulsification properties. As mentioned above, protein solubility of the egg yolk protein concentrates can be decreased with repeated solvent extractions of lipids (Chung and Ferrier, 1991b). Moreover, it was found that modification of pH can induce the same loss of the yolk granule structure similar to the change in ionic strength by the solubilization of proteins (Causeret et al., 1991). Ohba et al. (1993) found that the isoelectric point of spray-dried egg yolk was in the vicinity of pH 4. Since water is intimately involved in the microstructure of the yolk, its removal could also alter the charge distribution of the yolk. Moreover, proteins are denatured by the effects of temperature on the non-covalent bonds involved in stabilization of secondary and tertiary structure. It was found previously by Le Demat et al. (1999), that the protein solubility of yolk and yolk plasma dropped sharply above 69°C because of aggregation of proteins. 2.2.4 Gelling Properties of Yolk Miller and Winter (1951) suggested that frozen yolk was a more efficient emulsifier in mayonnaise than fresh yolk. Yet frozen pure egg yolk is difficult to combine with other ingredients because of the irreversible gelation reaction, which occurs when yolk is frozen and thawed (Telis and Kieckbusch, 1997). Thus, freezing pure egg yolk has limited utility. This is apparently due to the aggregation of yolk lipoproteins because of 17 the imbalance and shift in water when yolk is frozen (Powrie et al., 1963). Gel formation can be defined as a protein aggregation phenomenon in which attractive and repulsive forces are so balanced that a well-ordered tertiary network or matrix, capable of hold much water, is formed (Matsumura and Mori , 1996). Low-density lipoproteins in yolk play a significant role in the gelation of egg yolk by freezing and thawing (Nakamura and Doi, 2000). Saari et al. (1964) reported that the LDLs of the yolk were specifically changed by freezing. The lipoprotein which was denatured by freezing was shown to contain a high percentage of free fatty acids. LDL gel is more stable than ovalbumin and serum albumin gels over a wide range of pH (pH 4-9) (Nakamura et al., 1982). The result is a product of greatly increased viscosity after freezing and thawing. Sodium chloride and sucrose, at a 10 % level are commonly added to egg yolk to inhibit gelation due to freezing. Since it was mentioned above that frozen yolk was found to be a more effective emulsifier than fresh yolk, this may imply that freeze-dried yolk may also have better emulsifying properties than fresh yolk. It was found that freeze-drying decreased the solubility of yolk as the freezing time increased (Sato et al., 1973). This could be due to the effects of dehydration as well as the influence of freezing. i 18 2.3 Processing of Egg Yolk Egg yolk can be made into a product of longer shelf-life and convenience by dehydration. The dehydrated yolk products are popular ingredients in many daily dietary items, such as bakery products, salad dressings, noodles, baby foods, cake mixes, military ration, etc. They can be stored and transported at low costs and they are convenient to use. Different processes can cause varying degrees of damage to the final products. Yang (1988) studied the microstructure of unprocessed and processed liquid egg yolk and found that the microstructure was affected by pH, freezing, manual homogenization, blending, dehydration and addition of egg white, salt or sugar. These microstructural changes included destruction of spheres or granules and alteration of the continuous background. 2.3.1 Spray-drying Process Spray-drying is the most common and economical method of producing dried egg products. The liquid sample is finely atomized into a chamber under a stream of hot air. Because of the enormous surface area created by atomization, evaporation of water is very rapid (Berquist, 1995). The dried particles, suspended in the air stream, flow into separation equipment where they are removed from the air and collected. During spray-drying, eggs are subjected to certain physical forces, for examples, pumping and atomization, that can affect their functional properties. These forces may result in surface denaturation and may cause some changes in the functional properties of egg (Berquist and 19 Stewart, 1952). Depending on the particle size of the dried powder, new and larger surfaces are formed, and the egg protein spreads as a monomolecular layer over these new surfaces and becomes irreversibly denatured. In addition, when water is removed, the structure of the other components which has close association with the water in the yolk will be disturbed. The heating of the liquid sample and the dehydrated material during the process will also have significant effects on the characteristics of the yolk. The outlet temperature, feeding rate of the sample into the drying chamber, nozzle size, current flow rate and compressor rates are the adjustable parameters in a spray-dryer. In general, the amount of heat involved in the process is minimized to prevent severe damage to the yolk properties. Most lab-scale research on spray-dried egg yolk uses inlet temperatures of about 200°C to give an outlet of around 100°C. 2.3.2 Changes to the Functional Properties of Yolk after Dehydration Whole egg and liquid yolk, because of the presence of lipids that are closely associated with the proteins, are much more resistant to damage by physical treatment than egg white, which consists mainly of proteins. The amount of heat the dried egg product absorbs during the final stages of drying is important and depends on the method of drying, the dryer design and conditions of its operation, and how rapidly the product is cooled after drying. Denaturation is a function of time and temperature. Yolk denatures in the temperature range from 63 to 70°C (145.4 tol58°F). Cunningham and Varadarajulu (1973) reported that spray-drying altered several yolk components. Yolk, which consists of mainly lipoproteins, cannot be returned to its original state after it has been dehydrated 20 (Cunningham and Varadarajulu, 1973). Franzen et al. (1970) studied the temperature stability of high-density lipoproteins in yolk and found that they were quite stable up to 60°C and experienced only reversible "structural loosing". But above 60°C, the lipoproteins denatured irreversibly. In another study, egg yolk low-density lipoprotein was heated at various temperatures from 55 to 100°C for 5 min. Above 70°C lipid extractability of the LDL gradually decreased with increase of treatment temperature. Thus, temperature affected the emulsifying properties and apparent viscosity of the LDL, which were significantly correlated with the extractability of lipid (Tsutsui, 1988). The results of the study by Dyer-Hurdon and Nnanna (1993) supported the hypothesis that protein-protein interaction, leading to protein aggregation, may occur during the spray-drying process, disrupting the natural conformation of LDL. During centrifugation, the aggregated LDL tends to deposit into the granule fraction. In addition, water is an essential part in the microstructure of the lipid and proteins in yolk. In plain-dried yolk, the removal of water irreversibly changes the structure of the low-density lipoproteins, when some of the bound lipids are released from these lipoproteins. Even mild conditions of drying, such as in freeze-drying, have been found to cause changes in the functional properties of whole egg and yolk (Rolfes et al., 1955). When yolk is co-dried with added carbohydrates, the carbohydrates partially protect the lipoproteins from this irreversible structural change, probably by replacing the water of hydration at its binding sites during the drying process (Schultz et al., 1968). Results of the spray-drying effects on emulsifying properties have been quite inconsistent. Schultz et al. (1966) reported that drying of yolk resulted in a rapid increase 21 in extractability of the "free lipids" which were extremely detrimental to the emulsifying capacity of yolk. Varadarajulu and Cunningham (1972b) found that emulsifying capacity of yolk was reduced on spray-drying. Lea and Hawke (1952) examined the effects of drying on the stability of lipovitellin preparation and found that the solubility was destroyed. On the other hand, Hassan et al. (1990) reported that dried eggs had greater emulsifying capacity than fresh eggs, but the emulsions obtained were less stable. Treatment of the yolk with different agents before or after drying may also affect emulsifying properties. Kim et al. (1982) looked at the emulsifying properties of yolk powder with pre-drying pH levels of 4.5, 7.0 and 10.0 and reported that the emulsion capacity and emulsion stability of dried yolk could be enhanced as the pre-drying pH increased. Carlin (1955) showed that rehydration of dried yolk with acetic acid instead of water apparently decreased its emulsification capacity. From these results, it can be suggested that a higher pH may have more favorable effect on emulsifying properties and greatest emulsifying capacity was observed at the normal pH of yolk (Zabik, 1969). The conditions under which the dried product is stored will also affect its functional properties. Berquist (1995) noted an increase in viscosity of egg yolk solids during storage. This was probably due to changes in, or denaturation of, lipoproteins. Long-term storage of dried egg solids at elevated temperature may result in rapid denaturation of lipoproteins and thus greatly increase the viscosity of solids (Lieu et al., 1978). During storage, spray-dried whole egg could undergo significant decrease in pH, solubility and heat-coagulated gel strength, but increases in browning value, emulsifying capacity and emulsion stability (Lee et al., 1991). 22 In terms of protein solubility, Hassan et al. (1990) reported that spray-drying resulted in a 27 % loss in protein solubility, for whole eggs. Wakamatu et al. (1982) found that the solubility was reduced when yolk and low-density lipoprotein were dehydrated to water contents of less than 0.19-0.22 % and total solids of 0.11-0.16 g/g of yolk. However, they found that the solubilities of the granule and water-soluble fractions were almost independent of water content. Generally, drying of eggs under normal conditions causes little, if any, loss of the nutritional properties of the eggs. Adverse drying conditions or poor storage conditions can however, damage nutritional properties. For example, Guardiola et al. (1995) found that higher spray-drying temperatures would led to a higher loss of many PUFA, especially of arachidonic acid (C20:4n6) and docosahexaenoic acid (C 22:6n3). 23 2.4 Immunoglobulin Y and Yolk Pellet 2.4.1 Immunoglobulin Y (IgY) Egg yolk contains y-livetins, or IgY, that have similar structure to mammalian IgM or IgE, which consist of four domains. These proteins are potential source of antibodies that can be used to fortify foods and applied in the pharmaceutical industry as a therapeutic agent. Therefore, there are methods set out to isolate this useful water-soluble protein. Many purification methods of IgY were used in the past to separate the water-soluble proteins from the lipids and lipoproteins in yolk. Methods such as ultracentrifugation to separate lipoproteins, delipidation by organic solvents, and precipitation of lipoproteins by polyethyleneglycol, sodium dextran sulfate, polyacryl acid resins had been used to isolate IgY from yolk (Hatta et al., 1997). However, these methods are costly and are not necessarily effective for isolation of IgY. In addition, methods based on organic solvent extraction to separate the lipid fraction are too harsh, such that the remaining yolk fractions cannot be used in the food system. Other less invasive and more economical methods were carried out for IgY purification. Methods that use food grade hydrocolloids, for example, sodium alginate, carrageenan and xanthan gums, were found to be highly effective as a precipitant of yolk "lipoprotein, possibly through ionic binding with the lipoproteins (Hatta et al., 1990). The purity of IgY thus obtained could be as high as 98 % and the recovery yield about 70 %. Another method by Kwan et al. (1991) employed a simple method using minimal amount of organic solvents and other additives to fractionate the water-soluble and water-insoluble fraction from egg yolk. Their method was based on centrifugation of yolk that was diluted 24 six- to tenfold with water and by adjusting the pH. This method was able to recover about 60-90 % of the immunological activity of yolk in the supernatant. 2.4.2 Yolk Pellet Although the purity of the IgY obtained by the gum precipitation methods was high, the usefulness of the fraction remaining after the extraction by these methods was questionable as there will be some gums left in the precipitate. On the other hand, after the removal of the water-soluble fraction containing IgY using the method by Kwan et al. (1991), the remaining water-insoluble fraction which is called the yolk pellet could still be used in food applications. It was found that this yolk pellet had excellent emulsifying properties for mayonnaise preparation. According to Kwan et al. (1991), the best condition for separation of lipids into the pellet and recovery of IgY in the supernatant was at pH 6.0 and a dilution ratio of 1:9, or ten times dilution. Under these conditions, and using eggs which had been stored at least 6 weeks after laying, over 90 % of the total lipids and phospholipids in the original egg yolk were recovered in the pellet. The recovery of lipids in the pellet declined to 70 % when eggs obtained shortly after laying were used in the process. Reasons for the more efficient separation of lipids after storage of the eggs are not known (Kwan et al., 1991). Possibly the loss of water through evaporation or loss of sulfhydryl (SH) groups which occur during cold storage (Burley and Vadehra, 1989) may affect separation of water-soluble and -insoluble fractions. Lipoproteins, which make up a large proportion of the egg yolk, are inherently unstable in the absence of water and contain a small proportion of sulfhydryl groups (Burley and Vadehra, 1989). 25 2.5 Optimization Techniques 2.5.1 Response Surface Methodology (RSM) Optimization is the goal to develop product of the highest quality. Classical approaches for optimization either includes modifying the variables one at a time or alternatively modifying the variables in the "back-and-forth" method. These methods are tedious and costly to perform and do not establish any equation which describes the relationship between the variables and responses to these variables. The Response Surface Methodology (RSM), in contrast, is an effective approach to product optimization. RSM can be defined as a statistical method that uses quantitative data from appropriate experimental designs to determine and simultaneously solve multivariate equations (Giovanni, 1983). The model is usually depicted graphically to allow visualization of the response surface. RSM is a four-step process. Firstly, two or three critical factors that are most important to the product or process under study are identified. Secondly, the range of factor levels which will determine the samples to be tested are defined. Thirdly, the conditions for the test samples are determined from the experimental design and then tested. These conditions are selected by using appropriate experimental design, such as Central Composite Rotatable Designs (CCRD) described below or Box-Behnken designs. These designs select a subset of samples to be tested from the set of all possible samples which could be tested. They emphasize those samples closest to the midpoints of the range of the factor levels specified, thereby decreasing the total number of samples which must be tested. Lastly, the data from these experiments are analyzed by RSM and interpreted. 26 RSM is dependent upon five assumptions: that the factors which are critical to the product are known; that the region of interest where the factor levels influence the product is known; that the factors vary continuously throughout the experimental range tested; that there exists a mathematical function which relates the factors to the measured response; and that the response which is defined by this function is a smooth surface. 2.5.2 Central Composite Rotatable Design (CCRD) The CCRD is a combination of the factorial and star designs, giving sufficient points to allow fitting of the data to a full, second-order polynomial model. The CCRD consists of "cube", "centre" and "axial" points. The "centre" point is usually replicated a number of times to allow for the evaluation of the "lack of fit" versus "pure error" components of the residual terms. The axial points make the design "rotatable". The designs are called rotatable because the standard error of the measured response is constant at equal distance from the centre of the experimental region and reasonably constant within a radius of 1 (in coded units). Rotatable response surfaces are both reliable and efficient, that is, they provide accurate results with the minimum number of experimental level combinations. For two variables, 9 level combinations are used. The centre point must be repeated five times. The variation found can be decomposed into three components, namely, the variation due to regression, the variation due to pure error (as evidenced by the variability of the five centre points), and the residual variation, which measures the inadequacy or lack of fit of the model (Mullen and Ennis, 1979). 27 2.6 Structural Analysis Techniques 2.6.1 Differential Scanning Calorimetry (DSC) When heat is applied to a sample, the sample may undergo a change in state or conformation, which results in an absorption or liberation of heat energy. Differential scanning calorimetry (DSC) is a technique in thermal analysis by which a sample and an inert reference are maintained at the same temperature and then both are gradually heated up at a programmed linear rate. DSC dynamically follows physiochemical changes that a substance undergoes during heating or cooling (Davis, 1994). It involves recording the difference in energy flux necessary to establish a zero temperature difference between a substance and reference material against either time or temperature when both are heated and cooled. Any thermally-induced changes to the sample are recorded as a differential heat flow in the form of a peak on a thermogram. The enthalpy change, denaturation or transition temperature, sample purity and reaction kinetics could be interpreted from the thermogram. For most proteins, the thermally-induced changes, such as structural melting or unfolding of the molecules, are usually detectable by DSC (Wright, 1986). DSC is useful in the study of the behavior of food proteins as a result of heat processing treatment. It can also be applied to food that has undergone other processing treatments (such as freezing, drying and mixing) indirectly by tracking the resultant changes in the thermal behaviour of the food proteins. The changes induced in food proteins by thermal treatments often lead to denaturation or unfolding of the native structure. Thermal denaturation is a highly cooperative phenomenon, which can be detected as an endothermic peak in the DSC thermograms, since the disruption of 28 intramolecular hydrogen bondsis an endothermic reaction (Ma and Harwalkar, 1990). The enthalpy (AH) is calculated from the area under the transition peak, and can provide an estimate of the thermal energy required to denature the protein. The temperature of denaturation (Td) can be estimated from the transition, generally as the peak temperature. It is important to analyze the samples at optimum concentration and heating rate. A change in the moisture content can drastically change the shape of the thermal curve and shift the transition temperatures. Very low heating rate results in "noisy" thermal scans, while high heating rates often result in a loss of resolution (Maurice and Page, 1983). 2.6.2 Raman Spectroscopy Raman spectroscopy is a light scattering technique which is applicable to the study of protein structure in solid and liquid food systems. It is a branch of vibrational spectroscopy which gives information on the vibrational motion of molecules. It is based on the shifts in wavelength or frequency of the exciting incident beam resulting from inelastic collisions with the sample molecules (Carey, 1982). Information obtained from Raman spectroscopy in the study of amino acid side chains of protein includes their exposure, state of ionization and interaction with other functional groups. The technique may also be used to estimate the secondary and tertiary structure of a protein molecule. An important advantage of this technique is its versatility in application to samples which may be in solution or solid, clear or turbid, in aqueous or organic solvent (Li-Chan, 1998). 29 The Raman effect is observed by the interaction of sample molecules with the incident radiation from a laser source, some energy was used from vibrational transitions and the remaining is detected as scattered radiation. The frequency of the scattered radiation is shifted i from the incident radiation, resulting in Stokes (v0-Vj) lines and anti-Stokes (v 0+Vi) lines, where i corresponds to a vibrational transition. In Raman spectroscopy, vibrational transitions are measured in terms of the shift in frequency or wavenumber (cm"1), V j , from the incident wavenumber, v0. Raman scattering depends on changes in the polarizability of the molecule (Li-Chan et al., 1994). One disadvantage of Raman spectroscopy is the problem of fluorescence from the chromophores that absorb in the visible region. The fluorescence depends on the wavelength of the exciting beam. Resonance Raman spectroscopy may be able to correct this problem since it usually uses UV excitation and the problem of fluorescence by chromophores absorbing in the visible region would be eliminated. With this kind of spectroscopy, only aromatic residues are observed because they can absorb energy in the UV region. Alternatively, FT-Raman, which utilizes near-infrared excitation at 1064 nm, can also eliminate the fluorescent problems in a similar way. However, when the wavelength of the excitation beam increases, the corresponding energy of the beam is reduced, which causes the lower sensitivity of FT-Raman with near-infrared excitation. As a result, this kind of Raman spectroscopy is better applied to samples of higher protein concentrations. 30 CHAPTER 3. THE STUDY 3.1 Background and Overall Objectives Spray-drying involves the atomizing and heating of the sample. Changes in the components of egg yolk can occur upon spray-drying, which may subsequently affect the functional properties of the resulting yolk powder. In the past, salts and sugars have been added to help maintain the functionality of yolk after spray-drying. However, the flavor of the yolk powder could be altered. Alternate ways to preserve the functionality of yolk after spray-drying are in demand. In this study, it is hypothesized that by adjusting the pH and dilution of the liquid yolk to the optimal levels prior to spray-drying, the solubility and emulsifying properties of the resulting yolk powder will be better than the spray-dried powder from the untreated yolk. There may exist an optimal combination of pH and dilution of yolk such that the resulting spray-dried yolk powder will have better solubility and emulsifying properties than if untreated. pH and dilution were the factors to be optimized because, firstly, egg yolk microstructure is altered by pH, which in turn, affects the functional properties of the yolk. Yolk adjusted to lower than its original pH tends to have poor emulsifying properties and vice versa. Secondly, while spray-drying involves the removal of the water in the sample, the increase in water content of the yolk preparation prior to spray-drying may affect the rate of drying and the degree of damage from the heat in the spray-dryer on yolk components. Thus, optimizing the pH and water content of yolk prior to spray-drying may enhance or, at least, preserve the functional properties, such as protein solubility and emulsifying properties, of the reconstituted yolk. The liquid egg yolk is subjected to 31 different combinations of pH and dilution pretreatment conditions, spray-dried and the resulting powders are studied for their solubility and emulsifying properties. In conjunction with the pH and dilution pretreatments, the effect of removal of the water-soluble fraction from the commercial liquid yolk prior to spray-drying was also investigated. Kwan et al. (1991) used hand-separated yolk for the preparation of the water-soluble-fraction containing the useful IgY. Interestingly, they found that the remaining yolk pellet had good emulsifying properties. In order to increase the shelf-life of the yolk pellet, spray-drying could be a choice of preservation. Therefore, the process was carried out in this study. Overall, the objectives of the study were to find the optimum level of those pretreatment factors for yolk which would give the best spray-dried yolk powder in terms of functionality and to study the difference in the properties of spray-dried yolk powder after different pretreatments. The study was carried out in two phases. In Phase 1, the Response Surface Methodology was used to optimize the pH and dilution of yolk prior to spray-drying with respect to the resulting functional properties the yolk powders. The effects of spray-drying on yolk pellet after the removal of the water-soluble fraction was also investigated. To validate the results obtained in Phase 1, the "best" and "worst" sets of pretreatment conditions as well as the removal of the water-soluble fraction were repeated on the yolk before spray-drying in Phase 2 and the resulting yolk powders were analyzed. Further investigation on the structural difference between the yolk and yolk pellet samples was also carried out. 32 Most studies in the past have used either hand-separated yolk or yolk components for studying the effect of spray-drying on yolk, however, little has been done on the actual commercially-separated egg yolk, which is expected to have albumen contamination. Thus, for practical reasons, this study focused on the actual commercial liquid yolk and to look at the changes to its functionality after different pretreatments followed by spray-drying. 33 3.2 Phase 1 - Optimization of pretreatment conditions for spray-drying of yolk To test the effects of pretreatment conditions (pH and dilution) prior to spray-drying on the resulting yolk powder, the pH ranging from 3.0 to 9.0 and dilution from 2 to 10 times were chosen as the limits for the CCRD-RSM. pH within the range of 3.0 to 9.0 were set, as proteins outside this range are normally irreversibly damaged. Chung and Ferrier (1992) noticed a marked decrease in the emulsion stability for yolk phosvitin at pH 10 which was attributed to the denaturation of the yolk protein. In terms of the dilution of yolk, a lower limit of 2 times dilution was set because it is a usual practice at Canadian Inovatech Inc. to dilute commercial yolk with a little volume of water to give a consistency that can be pumped effectively into a pilot-scale spray-dryer. However, due to the size of operation, there is generally no dilution needed for the yolk to be pumped efficiently into a commercial-scale spray dryer. 3.2.1 Objectives in Phase 1 (1) to determine the optimal pretreatment conditions (pH and dilution) for yolk which will give the best emulsifying properties and protein solubility after spray-drying with the use of a Central Composite Rotatable Design (CCRD) and the Response Surface Methodology (RSM) (2) to investigate the emulsifying properties and protein solubility of yolk spray-dried after the removal of the water-soluble fraction 34 (3) to compare the functional properties (solubility and emulsifying properties) of spray-dried yolk and yolk pellet with their liquid counterparts 3.2.2 Work Plan for Phase 1 Commercial Liquid Yolk Yolk Water-soluble fraction Yolk Pellet No pretreatments Liquid Pretreatments CCRD-RSM (13 sets: D H , dilution) No pretreatments Liquid Spray-dried (constant conditions) Spray-dried (constant conditions) Both liquid and dried products were analyzed for: (1) Proximate Composition • Solid Content <Vacuum Oven Moisture Determination> • Protein Content <Leco Nitrogen Determination> 35 • Fat Content <Goldfish & Folch's Methods of Fat Determination> (2) Color of the Spray-dried Yolk Powders <Hunterlat» (3) Functional Properties • Protein Solubility • Emulsifying Activity Index (EAI) and Emulsion Stability Index (ESI) (4) Statistical Differences • Response Surface Regression (RSR) • One-way ANOVA • Tukey's or Fisher's Pairwise Comparison (p = 0.05) 3.2.3 Materials and Methods for Phase 1 3.2.3.1 Sample Preparation Commercial liquid egg yolk and commercial spray-dried yolk powder were donated by Vanderpols Eggs Ltd. (Abbotsford, B.C.). Spray-drying was performed on the commercial egg yolk and the detailed method is provided below. pH measurements of the liquid yolk samples were taken as soon as they were collected. 0.05 % sodium azide was added to all liquid samples to prevent possible microbial deterioration during storage. 36 3.2.3.2 Pretreatments of Yolk Prior to Spray-drying (1) Central Composite Rotatable Design and Response Surface Methodology (CCRD-RSM) The CCRD-RSM (Minitab Statistical Software, Release 12.1) was used to generate experimental conditions to optimize the 2 factors - pH and dilution - for the commercial egg yolk prior to spray-drying. The range of pH was from 3.0 to 9.0 and the dilution ranged from 2 times to 10 times. The CCRD-RSM generated 13 sets of pH and dilution for yolk pretreatment - 9 unique sets (A to I) with 5 replicates (1-1 to 1-5) of the center conditions (Table 1). Liquid commercial yolk, liquid yolk pellet and their spray-dried forms were included for comparison with the 13 CCRD samples. Each batch of yolk was about 2 L and was obtained during the summer of 1999. Dilution and pH adjustments About 200 mL of liquid yolk was used for each pretreatment. With respect to the conditions generated from the CCRD-RSM, a volume of deionized distilled water was added to the yolk and subsequently, NaOH or HC1 of concentrations from 0.1 N to 1.0 N was added gradually with stirring to the diluted yolk to obtain the required pH. The total volume of the base or acid used was recorded. Finally, the yolk solution was brought to the desired dilution with deionized distilled water. 37 Table 1. Conditions generated by CCRD-RSM for pretreatment of yolk prior to spray-drying, with liquid and spray-dried yolk and yolk pellet as comparisons (Phase 1) Sample Designator Sample #** Pretreatment ToutletCQ*** Batch**** Dilution pH Yolk powder 1 2.00 6.52 80 a Pellet powder 2 2.00 6.58 85-95 a Liquid yolk 16 (ND) 6.52 a Liquid pellet 17 (ND) 2.00 6.58 a A 7 10.00 6.00 80 b B 8 2.00 6.03 95 b C 9 8.83 3.88 96 c D 10 3.17 3.88 90 c E 11 8.83 8.11 80 d F 12 3.17 8.11 95 d G 13 6.00 3.01 75 e H 14 6.00 9.01 80 e 1-1* 3 6.00 6.00 95 f 1-2* 4 6.00 6.00 86 f 1-3* 5 6.00 6.00 80-85 g 1-4* 6 6.00 6.00 85 g 1-5* 15 6.00 6.00 80 h * Replicates of the centre conditions. ** Sample # denotes the order in which the yolk samples with different pretreatment were spray-dried; ND = not dried *** T o u t i e t (°C) is the outlet temperature or outlet temperature range of the spray-dryer during the dehydration of the sample **** Samples with different lowercase letter designators were from different batches of commercial liquid yolk. 38 (2) Preparation of Yolk Pellet The objective of this procedure was to separate the water-soluble proteins, including the IgY, from the liquid yolk and to recover the insoluble fraction termed yolk pellet. According to the method of Akita and Nakai (1992), 400 mL of commercial liquid yolk was diluted with about 3.3 L of deionized distilled water. The diluted yolk was acidified to pH 5.2 using 0.1 N HC1. The total volume of HC1 added was recorded and the yolk solution was brought to a final volume of 4 L, which was a ten times dilution of the original yolk. After being held overnight at 4°C, the mixture separated into two layers, a translucent top layer and an opaque, light yellow bottom layer. The top layer was drawn off carefully with a pipette and the bottom layer was centrifuged at 10,000 x g at 4°C for 25 minutes. The supernatant, which contained the water-soluble fraction, was removed and the bottom pellet was pooled. To obtain a more liquid consistency for pH measurements, about 100 mL of deionized distilled water was added to the pellet and stirred. The slurry was then adjusted using 2.0 M NaOH to the pH of the original liquid yolk, which was about pH 6.5. Lastly, the pellet was reconstituted to a total volume of 400 mL with deionized distilled water. The reconstituted pellet was diluted two times to a volume of 800 mL for spray-drying. 3.2.3.3 Spray-drying Conditions All spray-drying was carried out at Canadian Inovatech Inc. (Abbotsford, B.C.) using a pilot plant-scale Niro Mobile Minor spray-dryer (Type 53, Copenhagen, Denmark) with Econ II vari-flow pump. The pump was set to deliver about 30-50 mL of sample per 39 2 minute into the drying chamber. The pressure of the atomizer was at about 4 kg/cm and the thermostat was set at 4.5 of a full-scale of 10. Inlet temperature was estimated to be in the range of 180-220°C while outlet temperature was maintained at between 75-95°C by adjusting the pumping rate and the speed of the atomizer. The sample preparation was stirred occasionally during drying to allow for consistent sample delivery into the drying chamber. For the pellet preparation, since the pellet was quite insoluble in water and clumps were developed, the mixture needed to be strained with a No. 16 strainer before spray-drying. The strainer was thoroughly scraped to minimize the loss of sample after straining. Figures la and lb shows the front and side views of the lab-scale spray-dryer. 3.2.3.4 Storage of Yolk Samples 0.05 % sodium azide was added to all liquid yolk samples. The liquid samples were kept in tightly sealed high-density polyethylene (FLDPE) bottles, whereas the dried yolk powders were sealed in bags made with polypropylene (PP) film and stored in a plastic storage container. All samples were kept refrigerated at 4 ° C prior to analyses. All the analyses in Phase 1 were carried out within 2 months after sample collection. For details on the storage of samples prior to spray-drying, please refer to Appendix A. 40 Figure la. Front view of the pilot plant-scale spray-dryer 41 3.2.3.5 Analyses of the Yolk Samples Since there was an estimated 5 % of yolk powder adhering to the drying chamber during the drying of each sample which needed to be manually air blown into the collecting container at the end of each run, the last portion collected during each run was not used in the analyses. Instead, the middle portion of spray-dried yolk, which was about two-thirds of the total yield, during each run was used for all the analyses, with the assumption that this portion would best represent each of the spray-dried samples. (1) Proximate Analyses Moisture Content Determination The moisture content of the yolk samples was determined according to the vacuum oven procedure, AOAC 925.30 (Helrich, 1990). Liquid samples of about 5 g each and powder samples of about 1.5 g each were weighed onto pre-weighed aluminum dishes. The samples in the dishes were dried overnight to constant weight in a vacuum oven at 80°C, with an internal pressure of 26 in. Hg. The aluminum dishes with the samples were transferred to a dessicator for the samples to cool to room temperature. The differences in the weight before and after oven drying were then used to calculate the % moisture or % solids of the samples. Aluminum dishes were handled with tongs at all times. The analyses were done in triplicate. % solids = (final weight of the sample/initial weight of the sample) x 100 % % moisture = 100 % - % solids 42 Protein Determination (Leco Nitrogen Determination) The total nitrogen content of the samples was measured using the Dumas combustion method with the Leco Instrument (Model LECO FP-428, manufactured by LECO Corporation, Joseph, MI, USA). The combustion temperature was 950°C. Oxygen gas was used in the combustion process while Helium was the carrier gas. Both gases were Ultra High Pure (UHP) grade. The weight of the sample used for combustion was about 200 to 300 mg. EDTA (9.59 % Nitrogen) was used to standardize the instrument. The conversion factor used for the egg yolk samples was 6.25, that is, % protein = 6.25 x % nitrogen. Due to the high precision of the equipment (SD = 0.5 %) and the large sample size facilitating representative sampling for a single measurement, duplicates were only done on several yolk samples and the others were conducted with a single measurement. Goldfish Crude Fat Determination The Labconco-Goldfish method, AO AC 920.39 (Helrich, 1990), was used for the determination of crude fat in the samples. Since this method uses petroleum ether, which is a relatively non-polar solvent, tightly bound fat or highly polar fat would not be released from the sample or be extracted. Therefore, this analysis was considered to give the amount of crude fat in the samples. All liquid samples were freeze-dried prior to this analysis. About 2 g each of the dried sample was weighed into extraction thimbles and then, about 60 mL of petroleum ether was added to each of the pre-weighed beakers. The condenser of the Labconco Goldfish apparatus (Serial No. 17751, Labconco Co., Kansas City, Missouri, USA) was 43 kept cool with cold running water. After assembling the unit, the petroleum ether was heated to just boiling at the "High" heat setting and then the heat was brought to the "Low" setting. The samples were kept under reflux in the apparatus overnight for about 16 hours, allowing the fat from the samples to be extracted and dissolved into the petroleum ether. After overnight extraction, the apparatus was dismantled and the petroleum ether was evaporated. The beakers were allowed to cool in a dessicator before weighing. The difference between the weight of the beaker before and after the extraction gave the weight of crude fat, which was reported as % crude fat in the sample. The analyses were done in triplicate. "Total" Fat Determination (Folch et al., 1957) The method by Folch et al. (1957) was designed for determination of animal lipoproteins and was used for the determination of "total" fat of the yolk samples in Phase 1. This method uses chloroform as the solvent to extract the fat. Due to the polarity of chloroform, C H C I 3 , it was expected that the bound and polar fat of the samples could be released and extracted. About 2 g of yolk powder or 5 g of liquid yolk was weighed into a dried 125 mL Erlenmeyer flask. After the addition of 50 mL of Folch's solution #1 (CHC13:CH30H, 2:1) to the flask, the mixture was stirred with a glass rod and the flask was covered with aluminum foil and left overnight to extract the fat. Then, the mixture was filtered through a fluted filter paper into a glass stoppered graduated cylinder. The Erlenmeyer flask was 44 rinsed with 10-20 mL of Folch's #1 solution and the washing was added to the filter. A volume of 0.88 % (w/v) sodium chloride solution, which was about 0.2 times the final volume of the solution in the graduated cylinder, was added and the cylinder was stoppered and tilted two times for mixing. The cylinder was left overnight until the mixture separated into two layers, the bottom chloroform layer, which was yellowish, and a clear methanol layer on the top. Upon separation, the methanol layer was drawn off carefully with a pipette. As a washing step, 10 mL of Folch's #2 solution (CHC13:CH30H:H20, 3:47:48) was added to the chloroform layer and the cylinder was stoppered and tilted for two times. The mixture was again allowed to separate into two layers overnight. The final volume (Vf) of bottom chloroform-lipid layer was recorded and the top methanol layer was drawn off carefully. Exactly 5 mL of the bottom chloroform layer was pipetted onto a pre-weighed aluminum dish. The chloroform was evaporated, by warming the aluminum dish on a hot plate under very low heat and under nitrogen. The aluminum dish was cooled in a dessicator before weighing. The difference between the final and initial weight of the aluminum dish gave the weight of fat in 5 mL of the chloroform layer. The analyses were repeated in duplicate. % "total" fat = V f x (weight of fat/5 mL) x 100% weight of yolk sample where Vf = final volume of the bottom chloroform-lipid layer 45 (2) Hunterlab Color Analyses for Yolk Powders The Labscan Hunterlab device (ISSN-13685, Hunter Associates Laboratory, Inc., Reston, Virginia, USA) was used to determine the L, a and b values of the yolk powder samples. After calibration of the device using white and black tiles, yolk samples were spread uniformly (approximately 0.3 cm in depth) onto a petridish of 8.65 cm diameter. The petridish was then placed on top of the detector with an aperture of 0.63 cm diameter for reading. Each petridish was read four times by rotating the dish by 90° each time and triplicate petridishes were read for each sample. 46 (3) Functional Properties Protein Solubility Usually for protein solubility testing, the concentration of protein needs to be in the range of 1 - 25 mg protein/mL or else if the solubility product constant is exceeded, some of the protein will be insoluble and will sediment by centrifugation (Vojdani, 1996). In this study, the protein concentration was 2.5 mg protein/mL. In general, the lowest possible concentration of buffer required for maintaining the pH is used to minimize the non-specific ionic strength effect. A phosphate buffer of pH 6.35, the measured pH of commercial egg yolk, was chosen as the standard solution to test for the protein solubility and the emulsifying properties in the next section. The solubility of the samples was measured by a method modified from Anton and Gandemer (1997). Yolk powder of about 375 mg was dissolved in 50 mL of 0.1 M sodium phosphate buffer at pH 6.35 with 0.05 % sodium azide to give a protein concentration of 2.5 mg/mL. The preparations were shaken by hand and then equilibrated overnight at 4°C. Before centrifuging, 1.0 mL of the sample mixture was taken out. The mixture was then centrifuged for 30 minutes at 20,000 x g. The supernatant was filtered through Whatman #1 filter paper. The protein concentrations of the sample before centrifugation and of the supernatant were determined with the Lowry Protein Assay Kit (Sigma Diagnostics, Procedure No. P5656, St. Louis, MO, USA). Bovine serum albumin, fraction V (BSA) having a concentration of 400 ug/mL, was used to make the dilutions for the protein standard curve. The range of the concentration of protein used for the standard 47 curve was from 50 to 400 pg/mL. 1.0 mL of deionized distilled water was added to the tube labeled for the blank. The Lowry Reagent solution (1.0 mL) was added to the standard, blank and sample tubes and mixed well. After standing at room temperature for exactly 20 minutes, 0.5 mL of Folin and Ciocalteu's Phenol Reagent working solution was added to the test tubes with rapid and immediate mixing. A blue color developed after 30 minutes. The solutions were transferred to cuvettes and the absorbances of the standard and sample tubes vs. blank were measured at 750 nm with the ATI UNICAM UV/Vis Spectrometer (Division of Analytical Technology, Inc., Cambridge, UK). The absorbance readings of the BSA standard vs. their corresponding protein concentrations were used to prepare a calibration curve. The protein concentrations of the samples were found from interpolation of the calibration curve. After multiplying the results by the appropriate dilution factor, the protein concentrations of the original supernatant of the samples were determined. Since the measured protein concentration of the yolk preparations before centrifugation was not consistent within duplicates, possibly due to sedimentation of the yolk components even after stirring, the theoretical protein concentration of the yolk preparations was used in the analysis instead. The theoretical protein concentration was calculated based on the Leco Nitrogen (Protein) results and weight of samples in the preparation. By comparing the theoretical protein concentration of the sample preparation and that of the corresponding supernatant, the % protein solubility of the yolk sample was determined. 48 Emulsifying Properties (EAI, ESI) In this study, a buffer having the same pH as the commercial yolk was used for the preparation of a 6 % yolk solid solution. Canola oil, which is a common oil ingredient in mayonnaise was used for preparation of emulsions. In the literature, corn oil is also commonly used for the emulsifying tests (Hill, 1996). However, it should be noted that emulsions made from different oil type with different blenders may differ in turbidity (Hill, 1996). Emulsifying activity index (EAI) and emulsion stability index (ESI) were determined by a method modified from Pearce and Kinsella (1978). Egg yolk powder or liquid was diluted with 0.1 M sodium phosphate buffer at pH 6.35 with 0.05 % sodium azide to yield a 6 %-solids mixture. Bovine serum albumin, BSA (Fraction V, Sigma Chemical Co., St. Louis, MO, USA), was used as a control for the emulsifying tests. It was diluted with the 0.1 M sodium phosphate buffer to make a 0.5 %-protein mixture. All mixtures were left overnight at 4°C to allow for dissolution. ,3.0 mL of the egg yolk mixture (or BSA standard) and 1.0 mL of Crisco canola oil (purchased at local supermarket, Safeway, Vancouver, B.C.) were pipetted into small cups designed for the Sorvall Omnimixer microattachment assembly (Serial No. 217, OMNI International, Waterbury, CT, USA and Ivan Sorvall, Inc., Norwalk, Conn, USA). The cup was attached to the mixer and the mixture homogenized for exactly 1 minute at 13500 rpm. To avoid heating of the sample due to high-speed mixing, the microattachment cup was immersed in an ice bath during the homogenization. At 30 seconds after the homogenizer was turned off, a 0.05 mL-sample was taken from the bottom of the emulsion into a test tube with 9.95 49 mL of 0.3 % sodium dodecyl sulfate (SDS) solution. The solution was mixed well by inversion. The absorbance (turbidity) of the emulsion-SDS mixture was measured at 500 nm with the ATI UNICAM UV/Vis Spectrometer (Division of Analytical Technology, Inc., Cambridge, UK) and the measurement was used to calculate the EAI, using the formula given below. EAI = (2T)/[(0)(C)] = 2(2.303)(A5oo x D)/[(0)([protein] g/mL)(10"2m)(l mL/10"6m3)] where A5 0o = absorbance at 500 nm; D = dilution = 200; (|> = oil volume fraction = 1 mL oil/(l mL oil + 3 mL protein concentrate) = 0.25; [protein] = concentration of protein in yolk-buffer preparation; For ESI determination, 0.05 mL-aliquots were taken out from the bottom of the emulsion at specific time intervals after homogenization. Again, the sample was mixed well with 9.95 mL of 0.3 % SDS solution in a test tube. The absorbance of the mixture was measured at 500 nm with the ATI UNICAM UV/Vis spectrometer (Division of Analytical Technology, Inc., Cambridge, UK). ESI was defined as the time when the absorbance of the emulsion-SDS mixture decreased to half the original value and was found from the plot of absorbance against time. 0.3 % SDS solution was used as the blank solution for all absorbance reading. All analyses were done at least in duplicate. 50 (4) Statistical Analyses Minitab Statistical Software (Release 12.1) was used for the statistical analyses. The CCRD-RSM was used to generate the experimental conditions as mentioned in the objectives and the Response Surface Regression was used to determine the relationship between the pretreatment conditions and the different parameters of the yolk powders. The results of replicate data are reported as mean ± standard deviation values for samples with the number of replicates n > 3, and as the means with the individual values shown in parentheses for duplicate data (n = 2). The differences between the samples were compared using the one-way analysis of variance (ANOVA) followed by Tukey's pairwise comparison with the family error rate set at 5 (i.e. p = 0.05). Fisher's test, which is based on the individual rather than family error rate, was conducted for pairwise comparison of samples with unequal number of replicates (especially Leco protein determination values). 51 3.2.4 Results and Discussion for Phase 1 3.2.4.1 Pretreatment Conditions To test the effects of pretreatment conditions (pH and dilution) prior to spray-drying on the resulting yolk powder, the pH ranging from 3.0 to 9.0 and dilution from 2 to 10 times were chosen as the limits for the CCRD-RSM. The CCRD-RSM generated 13 sets of pretreatment conditions, 9 unique sets with 5 replicates of the centre conditions. Besides having the pH and dilution pretreatments, the effect of removal of the water-soluble fraction from the commercial liquid yolk prior to spray-drying was also investigated. The yolk pellet was collected, pH-adjusted to the original yolk pH and reconstituted. It was diluted two times before spray-drying. However, it was noticed that the margarine-like pellet was not able to dissolve well in water, mostly because of the removal of the water-soluble components, but also possibly caused by the changes to the yolk components during pH adjustments. The pretreatment conditions assigned to the samples are given in Table 1 (Section 3.2.3.2). 52 3.2.4.2 Proximate Analyses on Yolk Samples (1) Moisture (Solids) Content As shown in Table 2, spray-dried yolk and yolk pellet powder had very similar % solids (about 99 % solids). Reconstituted liquid yolk pellet with a dilution of 2 times had about 18 % solids, which means that the reconstituted but undiluted yolk pellet would have roughly 36 % solids, which was lower than that of liquid yolk (45 % solids). The difference was due to the removal of the water-soluble fraction, which, according to Osuga and Feeny (1977), is about 42.4 % of the total protein in egg yolk. Due to the presence of albumen in commercial liquid yolk, its % solids was found to be lower than that of pure yolk (52-53%) (Powrie and Nakai, 1985). The Response Surface Regression (RSR) analysis using a full quadratic model did not yield a significant model for describing % solids as a function of pH or dilution. This means that the extent of removal of water from yolk by the spray-drying process could not be predicted from the pH and dilution of the yolk prior to drying. As expected, Tukey's pairwise comparison test showed that the liquid samples were significantly different from all dried samples (p < 0.05). The slight but significant differences between the dried yolk samples were probably due to the sampling of the yolk powder. As mentioned in the methodology section, the yolk sample from the middle portion of each'spray-drying run was used for analyses. The slight differences in the feeding rate and the temperature of the drying chamber as well as the atmospheric humidity during the summer when the spray-drying was done may also contribute to the difference in moisture of the dried samples. Sample G was significantly lower in % solids than all other dried samples. This could be due to the precipitation seen as the pre-drying pH was very low (pH 3.01). Some of the 53 water bound within the microstructure could have been released and thus, the % solids was lower for this sample. However, when this result was combined with the results of all the samples, the regression analysis did not yield significant effects of pH on the % solids. Moisture content of the yolk powders could be affected by: the feeding rate of the sample into the drying chamber; the inlet and outlet temperatures; the pressure of the atomizer; the current flow in the drying chamber; and the amount of bound water present in the yolk particles, which could be measured as the water activity of the sample (refer to the Results and Discussion section for Phase 2). 54 < u oo m O H 'Si ii cd 00 E oo <D OO c < X s CL, CN 2 03 H ' 1 ,1 in in cn oo in NT p m 3 g ' 5 3 o l-l OH E 00 Si cu |3 c o l o ^ o o +1 +! 0 0 ly-, °°. ON O N VO VO OO • w X> ' VO VO O N in vo in o' in vo H 0 0 0 0 vo m *o 0 0 « ? ON o « cn<N in in t-~. Tt O c n t n 00 m VO cn CN CN 0 0 CN CO OV v 7 \ © +1 o +1 o r-; O N O N 0 0 m Tt- r-ON ON T j - — 1 CN 0 0 CN 0 0 in in in in VO VO VO VO O O O O CN CN O o CN s i O Cu $ E CM T3 3 '3 a l ~ 3 O N VO O cn O © +1 +1 o 0 0 0 6 O N TT VO VO ~ CN 2? cn 0 © © o ° ' +1 +1 +1 +1 +1 CN cn ON ^ r-CN rr. ON ^ ON t~T" 0 0 ON 0 0 O ON T l " 1 0 +1 +1 0 0 O N in ON I— J ON r~ cn cn 0 6 v d vo in in in in in in ON cn ©" cn 06 r-^  O in in 0 0 vo in vo in in in in 0 0 -rj- in CN O N 0 0 O N TJ- 0 6 VO in in m in m in O N a ^ V O ^ in in cn O N 0 0 VO T t + | in m T l w w 0 0 °* « in' m m in in m D X> X £> X" X X> o vo o TJ- 0 0 vo O N cn cn CN r-; CN r-^  p vq in ^'cNinin'cNcncNCN cncncncncncncncn +1 cn cn ho x) in 0 0 O © © © +1 +1 cn vo 0 0 " " O TT CN O CN O N O N O N O N O © +1 CN ON <N © +1 TT 0 0 © © +1 © +1 vo r-m vo o +1 CN ON ON 0 0 ON ON ON ON VO ON 0 0 ON ON ON ON O cn 0 0 0 0 O © 0 0 0 0 © O O vovocncnoooocnONVO cn 0 0 r~- cn r- © © © <-H 0 0 O O © cNoocnoocnvovovo o cn -a o 4—» o 3 TJ c o o (U o c -a T3 C ID 03 -*—> O +-» 03 O e o s 2 o-, cn II c «T "o 03 «3 03 T3 <U 1/1 a CU X ID <U o3 00 —^> a. o o X m •c T3 oi <D 1-«s 03 § D *-* 03 3 cn a, 11 2H C O 03 C CO S 3 5 0< CN H (3 in U-i a T3 <D 00 "c3 c 03 00 1-< u o3 -4—» o VO II C G ' 5 3 o I H in E, II 03 o3 & 5 3 p 03 ca II 00 F—H O 00 00 <U -4—> cd o "E. H e -*-» c < u o m < u o 00 •*-» 00 00 C o o E in © o A a. -a 00 m © g ' 5 3 4-> o i-P H 4—» o o o -a < u 00 3 00 03 -4—» C/) (/) 1=1 s o o 00 00 •g - q c3 03 cu a, S I o o o o (U u 00 00 03 03 CU CU CJ (D .2 3 00 o c I c 6 3 E o a I 00 (U -4—> g 00 +-> & 4 •q o 00 (U 3 00 CD 03 00 U 4—> 60 3 00 55 (2) Protein Content (% Nitrogen x 6.25) The nitrogen content of the yolk samples was determined by the Leco combustion method. Due to the precision of the instrument, duplicates were only done for sample 1-5 and the two liquid samples, e.g. for liquid yolk, the protein content (db) was 33.28 % (33.37 %, 33.18 %). After multiplying the nitrogen content by a factor of 6.25, the protein content of all dried yolk powders was found to be about 33 % which was almost the same as that for liquid yolk on a dry weight basis (Table 2). This also corresponds well to the literature value of yolk protein of about 32 % (Powrie and Nakai, 1985). The protein (dry basis) contents of the dried and liquid yolk pellet were lower (28.05 % and 26.71 % respectively), which confirmed the extraction of yolk water-soluble proteins during the pellet preparation. As expected, the RSR analysis using a full quadratic model did not yield a significant model for describing % protein as a function of pH or dilution. The amount of protein in the sample was not affected by either pretreatments or spray-drying. (2) Fat Content ("Total" Fat and Crude Fat) The "total" fat content for all samples as obtained by the method by Folch et al. (1957) was significantly higher than the crude fat content as obtained by the Goldfish extraction method (p < 0.05). As described in the materials and methods section, the Folch's method uses chloroform, which is a more polar solvent than the petroleum ether used in the Goldfish method. Thus, the Folch's method can not only extract the surface lipid, it can also access the core lipids by causing the lipoprotein to unfold. On a dry basis, the % "total" fat for dried yolk pellet (67.6 %) and liquid yolk pellet (62.5 %) was 56 significantly higher (p < 0.05) compared to those of dried yolk and liquid yolk, having 57.0 and 55.6 % "total" fat respectively (Table 2). Kwan et al. (1991) also reported about 58 % fat in their yolk control. Yolk pellet contains more fat than yolk itself because the pellet consists of relatively less protein (Figure 2). This could also be visually observed from the margarine appearance of the liquid pellet and the greasiness of the pellet powder. The % total fat was not significantly affected by pH and dilution, whereas the % crude fat was slightly affected by pH 2 (p = 0.054), which was a quadratic relationship. It was found that the sample with the lowest pH pretreatment (sample G, pH 3.01) had the highest % crude fat while sample C with pretreatment of pH 3.88 had the lowest % crude fat. Samples with higher pH pretreatment had higher % crude fat than sample C. Adjusting the pH to either the lower end (pH 3) or higher end (pH 9) could have altered the structure of the lipoprotein, exposing more of the buried lipids in the lipoproteins, thus rendering them more extractable even with a less polar solvent. This could also be due to the more severe damage to the sample during spray-drying. Lipoproteins, which make up a large proportion of the egg yolk, are inherently unstable in the absence of water and contain a small proportion of sulfhydryl groups (Burley and Vadehra, 1989). Therefore, the changes in water content and/or sulfhydryl groups that occur during spray-drying could affect the separation of yolk lipids (Burley and Vahedra, 1989). On the other hand, since the pis of the components in yolk ranges from pH 4-8.5 (Gallagher and Voss, 1970;Tenehouse and Deutsch, 1966; Ternes, 1989), it could be speculated that the pH pretreatments would affect the conformation of the protein components and the ionic interactions within the yolk's microstructure, thus exposing the lipids for extraction to a different degree. 57 Figure 2. Protein and "Total" Fat Content of Yolk and Pellet Samples (Phase 1) | 0% Protein H%Fat Liquid Pellet Liquid Yolk Pellet Powder Yolk Powder 0 20 40 60 80 100 percent dry basis (% db) Figure 2 depicts the difference between the protein and "total" fat content of the yolk and yolk pellet samples. The fat contents of the samples were those determined by the Folch's method. Liquid and powder form of yolk pellet had relatively lower % protein and higher % fat than the yolk samples. Spray-drying did not affect the composition of yolk samples, as there were no significant differences between the protein and fat content of the spray-dried and liquid form of either the yolk control or the yolk pellet. On a dry basis, the sum of % protein and % "total" fat of the pellet samples appeared to be higher than that of the yolk sample, both in the liquid and powder form, because some of the water-soluble components in yolk, for example, carbohydrates may be removed along with the water-soluble proteins. This was also indicated by the lower solid content of the pellet 58 samples. Consequently, the sum of protein and fat would be proportionately higher for the pellet samples than for the yolk samples. 3.2.4.3 Hunterlab Color Analyses for Yolk Powders Since notable differences were visually observed in the color of the samples, Hunterlab measurements were conducted to quantitate those differences, though yolk color can also be qualitatively scored with the Roche Color Fan (Marusich and Bauernfeind, 1970). The three parameters of Hunterlab, L, a, b were measured. L is the degree of lightness and darkness, a is the degree of redness and greenness and b is the degree of yellowness and blueness of the sample. The RSR analysis using a full quadratic model did not yield a significant model for describing L and b as a function of pH or dilution. However, there was a slight pH effect on a (p = 0.058). Tukey's pairwise comparison of variance showed differences in L, a, b of the samples. L was higher for samples C and D, both having a lower pH pretreatment. These samples also had lower a values. Sample I showed a lower b value. On the other hand, samples B and G and pellet powder had relatively higher b values. The differences between the color of the samples could be due to the changes to the color pigments in the yolk. Oxycarotenoids are the major class of pigment in the yolk that gives the yolk its yellow and red appearance. Although these pigments are quite heat-stable, they are sensitive to oxidation. Spray-drying increases the susceptibility of these components to oxidative damage. Once exposed to atmospheric oxygen, the carotenoids are oxidized very quickly, the rate depending on light, heat, and the presence of pro- and 59 antioxidants (Francis, 1985). In addition, the change in pH and the removal of the water-soluble fraction may affect the structure and concentration of carotenoids. Carotenoids are a group of mainly lipid-soluble cbmpounds, when the concentration of lipid increases, the concentration of carotenoids will also increase. This may explain why the yolk pellet powder appeared to be more orange than the yolk powder control. There may also have been browning reactions involving the sugars present in yolk. Carbohydrates are present in the yolk at 0.2 - 1.0 % and most of these carbohydrates are bound to protein in the form of mannose-glucosamine polysaccharides (Powrie and Nakai, 1985). Moreover, the reducing sugars may occur in combination with carotenoids in the yolk via a glycosidic bond (Francis, 1985). Table 3. Hunterlab color analyses for yolk samples (L, a, b) (Phase 1) Sample Pretreatment Hunterlab Dilution pH L a b Yolk powder 2.00 6.52 84.55 ±0.87 a 2.31 ± 0 . 2 3 a b c 22.09 ± 0.63a Pellet powder 2.00 6.58 84.58 ± 0.43a 1.97 ± 0.15ab 27.93 ± 0.96c A 10.00 6.00 85.47 ±0 .71 a 3 .12±0 .37 b c d 23.82 ± 0.97ab B 2.00 6.03 84.87 ± 0.96a 3 .65±0 .61 c d 26.78 ± 0.73bc C 8.83 3.88 87.87 ± 1.29bc 1.19± 0.30a 22.13 ±0 .94 a D 3.17 3.88 89.08 ±0.31° 1.05±0.05 a 23.86 ± 0.30ab E 8.83 8.11 84.91 ±0.67 a 3.41 ± 0.29cd 24.11 ±0 .35 a b F 3.17 8.11 83.57 ± 1.22a 3.42 ± 0.45cd 23.30 ± 0.43ab G 6.00 3.01 84.17 ±0.39 a 2.91 ± 0 . 3 3 b c d 28.13 ± 1.30c H 6.00 9.01 83.77 ± 0.40a 3 .74±0 .21 d 23.82 ± 0 . 5 3 a b 1 6.00 6.00 86.21 ± 0.93ab 2.63 ± 0.66bc 21.91 ± 1.83a L = + lightness / - darkness; a = + redness / - greenness; b = + yellowness / - blueness. Sample I consists of the 5 centre replicates, n = 15; all other samples, n = 3 Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 60 3.2.4.4 Functional Properties of Yolk Samples (1) Protein Solubility The protein molecules would need to have a certain degree of solubility in an environment in order to be able to interact with other components in their surrounding. The proportion and distribution of surface hydrophilic and hydrophobic patches are the main factors in determining the degree of solubility of protein. Solubility of protein can be improved by increasing the number of hydrophilic groups present. pH, ionic strength, temperature, solvent components and other food components affect solubility. At the pi, a protein usually has the least solubility. Protein-protein interactions increase because electrostatic forces of molecules are at minimum and protein aggregation occurs. At pH values above and below the pi, where a protein has a net negative or positive charge, more water interacts with the protein charges. Net charges and charge repulsion contribute to greater protein solubility and the protein may stay in solution. Furthermore, at extreme acidic or basic pH values, the protein may unfold, exposing more hydrophobic groups. However, at too high or too low pH values, due to a decrease in electrostatic bonds, the protein undergoes denaturation followed by aggregation and precipitation. The pi values of several yolk proteins fall in the range of 5.5 - 5.8 (Gallagher and Voss, 1970; Tenenhouse and Deutsch, 1966; Ternes, 1989). As shown in Table 4 and Figure 3, the protein solubility in 0.1 M phosphate buffer at pH 6.35 of liquid yolk was about 70 %, while liquid pellet was only 39 % soluble. The loss of about two-fifth of the solubility in the yolk pellet was contributed by the removal of the water-soluble fraction. Except for samples H and I, spray-drying reduced the protein 61 solubility of all yolk samples compared to commercial liquid yolk. Dried yolk with different pretreatments had protein solubility ranging between 43 - 62 % in 0.1 M sodium phosphate buffer of pH 6.35. Yolk powder control had a protein solubility of 51.77 %. The standard deviation (13.02 %) was quite large for this sample, which was likely due to < possible contamination of the supernatant fraction by the bottom precipitate for one of the replicates during its preparation, thus overestimating the protein solubility of that replicate. Another one of the replicates however had low protein solubility, possibly because the sample drawn from the original yolk preparation into the centrifuge tube was not homogenous, thus gave an underestimated protein solubility for that sample. Dried yolk pellet had the lowest solubility of 24.16 %. Response surface regression analysis showed both pH (p = 0.000) and dilution2 (D2) (p = 0.001) effects on the % protein solubility of the yolk powder as described by the following equation: % protein solubility = 69.12 +1.79D + 6.833pH - 5.47 D 2 + 6.16pH2+ 1.142pH*D (Ra d j 2= 89.2 %, p < 0.001, n = 26). The result of the regression analysis is expressed in form of a contour plot (Figure 4). Estimating from Figure 4, the highest solubility (> 60 %) was seen between pH 5.5 - 9 and dilution of 4 - 9 times. Tukey's pairwise comparison showed that dried pellet had the lowest solubility while liquid pellet was the second lowest in solubility. Samples C, D and G, which had the pre-drying pH lower than pH 4.0, also showed lower solubilities than other yolk samples. 62 Table 4. Protein solubility of yolk samples in 0.1 M sodium phosphate buffer at pH 6.35 (Phase 1) Sample Pretreatment % Protein Solubility Dilution PH Yolk powder 2.00 6.52 51.77 ± 13.02cde Pellet powder 2.00 6.58 24.16 ± 3.17a Liquid yolk 6.52 69.78 ± 1.29g Liquid pellet 2.00 6.58 38.68 ±2 .35 b A 10.00 6.00 56.52 ± 4.38et B 2.00 6.03 50.97 ± 3.76cde C 8.83 3.88 43.51 ± 1.66bcd D 3.17 3.88 43.22 ± 1.59bc E 8.83 8.11 58.69 ± 3.28ef F 3.17 8.11 54.31 ± 2 . 3 0 d e f G 6.00 3.01 43.50 ± 2.67bcd H 6.00 9.01 60.94 ± 3.59efg • I 6.00 6.00 62.44 ± 4.24fg Sample I consists of the 5 center replicates, n = 20; all other samples n = 4 Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 63 Figure 3. % Protein Solubility at pH 6.35 (Phase 1) 100 • CCRD-RSM samples 64 Figure 4. Contour Plot of % Protein Solubility for CCRD-RSM Yolk Powders 2 3 4 5 6 7 8 9 10 Dilution 65 (2) Emulsifying Properties of Yolk Samples Egg yolk is notable for its emulsifying properties because of the presence of lipoproteins and other components. It is a popular emulsifying agent in food products. Several parameters are used to define emulsifying properties and Emulsifying activity Index and Emulsion Stability Index (ESI) were measured in this study. Emulsifying Activity Index (EAI) Table 5 shows the EAI values for the yolk samples measured in duplicate. The EAI values obtained for the replicates for dried yolk, liquid and dried yolk pellet showed poor precision. It was noticed that the spinning rate of the blade of the homogenizer was not very constant, judging from the noise generated when homogenizer was running, even with the equipment strobed before and after the experiment. EAI of the yolk samples ranged from 12-25 m2/g protein for the yolk powder samples, when the yolk preparation was at 6 % solids. On the other hand, the EAI of BSA, at 0.5 % protein concentration, was found to be between 51-54 m2/g protein. Kwan et al. (1991) reported that the EAI (m2/g solids) for liquid yolk and yolk pellet was significantly different (5.0 m /g solids and 5.8 m2/g solids respectively). These EAI values were slightly lower than those found in this study. In addition, the EAI (m2/g solids) of liquid yolk and yolk pellet were not significantly (p < 0.05) different from each other in this study. Albumen contamination, yolk composition and the precision of the equipment might all have contributed to the differences found between studies. 66 The response surface regression analysis showed a slight but not significant pH effect (p-value of about 0.100). With higher pH pretreatment, EAI tends to be higher. Tukey's pairwise comparison indicated that sample G, having the lowest pH pretreatment, had the lowest EAI. No trends could be determined for the dilution effect. EAI of liquid pellet and pellet powder appears to be similar to the other yolk samples with no significant difference. Table 5. Emulsifying Activity Index (EAI) for Yolk Samples (Phase 1) Samples Pretreatment EAI (m /g protein) EAI (m /g solid) Dilution pH BSA 50.96 (46.50,55.42) Yolk powder 2.00 6.52 21.85bc (24.36,19.33) 7.24b (8.07,6.41) Pellet powder 2.00 6.58 24.89c (22.92,26.86) 7.00b (6.44,7.55) Liquid yolk 6.52 19.89bc (19.55,20.22) 6.62b (6.51,6.73) Liquid pellet 2.00 6.58 23.17bc (21.25,25.09) 6.19ab (5.68,6.70) A 10.00 6.00 16.52ab (17.61,15.42) 5.64ab (6.01,5.26) B 2.00 6.03 17.30abc (17.93,16.67) 5.67ab (5.87,5.46) C 8.83 3.88 19.45bc (19.19,19.70) 6.89b (6.80,6.97) D 3.17 3.88 19.46bc (20.36,18.56) 6.86b (7.17,6.54) E 8.83 8.11 22.49bc (21.76,23.22) 7.37b (7.13,7.61) F 3.17 8.11 21.39be (21.26,21.52) 7.07b (7.03,7.11) G 6.00 3.01 12.21a (11.20,13.22) 3.99a (3.66,4.32) H 6.00 9.01 20.89bc (20.66,21.11) 6.80b (6.72,6.87) I 6.00 6.00 18. 37 ± 2.49b 6.34±0.68 b BSA at 0.5 % protein was used as a control of the method Sample I consists of the 5 centre replicates, n = 10 Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 67 Figure 5. EAI (m2/g protein) of Yolk Samples (Phase 1) 68 Emulsion Stability Index (ESI) The ESI of all the yolk powders were similar, except for the slightly lower ESI for low pH pretreated sample and the pellet powder (Table 6). Sample G (pH 3.01, 6 x dilution) had the lowest ESI (2 min) compared to other samples. It was interesting that liquid yolk pellet had the best ESI (> 20 min) and it gave a more stable emulsion than commercial liquid yolk and the difference was significant (p < 0.05). However, upon spray-drying, the emulsion for yolk pellet was not as stable as that of spray-dried yolk control (Figure 6). Spray-drying negatively affected the structure of yolk pellet more drastically than yolk control. In addition, since yolk pellet possesses proportionately more insoluble proteins, the emulsion stability could apparently be reduced (Franzen and Kinsella, 1976; Holm andEriksen, 1980). Response surface regression analysis shows significant pH effects (p = 0.013) on the resulting spray-dried yolk, but the effects of dilution were not significant. The interaction between pH and dilution (pH*D) on the spray-dried yolk was also significant (p = 0.028). Figure 7 illustrates the contour plot of the result of the response surface regression analysis. The highest ESI (> 6 minutes) was found for samples with pretreatment pH of 8 - 9 at dilution between two to four times. 69 Table 6. Emulsion Stability Index (ESI) for Yolk Samples (Phase 1) Samples Pretreatment ESI (min) Dilution pH BSA 23.5 Yolk powder 2.00 6.52 5.5,g Pellet powder 2.00 6.58 3.2b Liquid yolk 6.52 5.7fg Liquid pellet 2.00 6.58 > 20.0h A 10.00 6.00 5.9g B 2.00 6.03 4 2 c d C 8.83 3.88 5.4fg D 3.17 3.88 2.5a E 8.83 8.11 4.0C F 3.17 8.11 5.2ef G 6.00 3.01 2.0a H 6.00 9.01 6.0s I 6.00 6.00 4.7 ± 0.6de Sample I consists of the 5 centre replicates N.B. ESI = Time when absorbance reaches Vi of the original absorbance at time = 0 min Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 70 Figure 6. ESI (min) of Yolk Samples (Phase 1) Liquid 71 Figure 7. Contour Plot of ESI (min) for CCRD-RSM Yolk Powders 3.2.4.4 Summary of Results in Phase 1 The % solids, % protein and % total fat remained relatively constant after spray-drying of liquid yolk subjected to different pretreatment conditions. However, the extractability of the crude fat from the yolk powder was affected by both spray-drying and pH treatment. Aggregation of the lipoproteins could have occurred when the yolk was pH adjusted and during atomization and heating upon spray-drying. Solubility of the yolk proteins in phosphate buffer at pH 6.35 was reduced about one-fifth after spray-drying based on the comparison of the liquid yolk and spray-dried yolk control. The % protein solubility was even lower for both liquid yolk pellet and dried pellet powder because of the initial removal of soluble protein from the yolk. Liquid yolk pellet had about two-fifth less protein solubility than liquid yolk. Indeed, this could be deduced from the lower protein content and higher total fat content of the pellet samples. Protein solubility of the yolk powder tended to be higher when the pre-drying pH was at yolk's natural pH or at slightly alkaline pH. When the yolk was diluted to about 4-9 times before spray-drying, the solubility of the resulting powder was higher. Diluting the yolk may ease in the feeding of the sample into the atomizer and may reduce the heat-induced damage to the proteins as the protein molecules are more "protected" by an increased amount of surrounding water molecules. No particularly significant trends in the effects of pretreatments on emulsifying properties were observed, except that at a pretreatment of about pH 3.0, the emulsion activity was lowest and the emulsion stability was worst. Yolk pellet, on the other hand, was found to have similar emulsifying activity to the yolk control, both in the liquid and 73 powder form. It had better emulsion stability than liquid yolk control, but the stability was lost dramatically after spray-drying. Overall, the pre-drying pH of yolk seemed to have greater effects on the spray-dried yolk samples than the pre-drying dilution. Further studies on the structural changes in the yolk, subjected to different pretreatments, upon spray-drying were needed to help explain all of the above observations during analyses on the yolk samples. 74 3.3 Phase 2 - Further Investigation on the Spray-dried Yolk Powders 3.3.1 Rationale In Phase ,1, the ranges of pH and dilution of yolk which give the "best" and "worst" spray-dried yolk powder in terms of functionality were generated from the regression analysis of the results from the CCRD-RSM samples. In terms of optimizing both the protein solubility and emulsifying properties of yolk powder, the pretreatment conditions were chosen based on the central point of the optimal region found in Phase 1 and the pretreatments of pH between 8.0 - 8.5 and dilution of about 6 times was found to give the "best" powder. On the contrary, low pH of about pH 3.0 and dilution of about 6 times gave the "worst" powder. As for the yolk pellet, it was found to have better emulsifying properties than yolk itself when in the liquid form, but not in the spray-dried form. For confirmation of the results in Phase 1 and a continuation of the investigation on spray-dried yolk with different pretreatments, spray-drying was done on yolk using the best and worst sets of pH and dilution pretreatments determined from Phase 1 and on the yolk pellet. For comparisons, liquid yolk, yolk pellet, as well as hand-separated liquid yolk and commercial yolk powder were also included in the analyses. Freeze-dried yolk, yolk pellet and water-soluble fraction were also included in this Phase. Freeze-drying which is an expensive method of dehydration has been reported to cause minimal heat damage to the samples, therefore, by comparison of the spray-dried samples with those from freeze-drying the effect of spray-drying could be better interpreted. All proximate analyses were carried out on the new samples. In addition, acid hydrolysis was used for fat determination, water activity of the yolk powders was recorded 75 and the conductivity of the liquid yolk preparations before spray-drying was determined. Solubility and emulsifying tests were also done on the samples. Furthermore, to investigate the structural changes in the yolk samples, Differential Scanning Calorimetry (DSC) and Raman spectroscopy were carried out on the samples. 3.3.2 Objectives in Phase 2 (1) To confirm the results of the proximate analyses, solubility and emulsifying tests on the yolk powders as determined in Phase 1, by repeating spray-drying under conditions which gave the best and worst protein functionality in Phase 1 and on yolk pellet. (2) To compare the results for the spray-dried yolk samples with commercial liquid yolk, liquid pellet, hand-separated yolk, as well as with commercial yolk powder. (3) To investigate the structural changes to the yolk powders after different pretreatments conditions and spray-drying Differential Scanning Calorimetry and Raman Spectroscopy. 76 3.3.3 Work Plan for Phase 2 Commercial Liquid Yolk Water-soluble fraction Yolk Liquid Freeze-dried Yolk Pellet y Y No pretreatments No pretreatments Pretreatments ("best"&"worst") Liquid Liquid 1 J Freeze-dried Spray-dried (constant conditions) Freeze-dried Spray-dried (constant conditions) Commercial Spray-dried Yolk Powder Hand-separated Yolk r Liquid 1 Freeze-dried Both liquid and dried products were analyzed for: (1) Proximate Composition • Solid Content <Vacuum Oven Moisture Determination> • Protein Content <Leco Nitrogen Determination> 77 • Fat Content <Goldfish, Folch's and Acid Hydrolysis Methods for Fat Determination (2) Water Activity of Yolk Powders (3) Conductivity of Yolk Preparations Prior to Spray-drying (4) Color of the Spray-dried Yolk Powders <Hunterlat» (5) Functional Properties • Protein Solubility • Emulsifying Activity Index (EAI) and Emulsion Stability Index (ESI) (6) Structural Changes • Differential Scanning Calorimetry (DSC) • Raman Spectroscopy <FT-Raman> (7) Statistical Differences • One-way ANOVA • Tukey's or Fisher's Pairwise Comparison (p = 0.05) 78 3.3.4 Materials and Methods for Phase 2 In addition to the materials and methods used in Phase 1, the following were added to the sample preparation and analyses in Phase 2. Commercial yolk powder was provided from Vanderpol Eggs Ltd. (Abbotsford, B.C.) which was transported from their spray-drying plant in Winnipeg, Manitoba. 3.3.4.1 Preparation of hand-separated egg yolk Hand-separated egg yolk was obtained from Lucerne large-size eggs purchased from a local supermarket (Safeway, Vancouver, B.C.). Twelve yolks were separated manually from the albumen and rolled on a paper towel to remove as much of the albumen as possible. The yolks were punctured and pooled together. 0.05 % sodium azide was added to preserve the yolk from microbial deterioration and the mixture was refrigerated at 4°C prior to analyses. 3.3.4.2 Freeze-drying of Yolk Samples Freeze-dried commercial yolk, yolk pellet and water-soluble fraction and hand-separated yolk samples were prepared as comparisons to the spray-dried samples. A volume (about 200 mL) of the commercial liquid yolk sample was frozen at -35°C overnight in a flat bottom container before being put into the freeze dryer (Model No. 75018, Labconco Corp., Kansas City, MO) for drying. Depending on the moisture content and the volume of the starting liquid, the dehydration process took between 4-6 days. The 79 dried samples were ground into powder form using a mortar and pestle, transferred into bags made with polypropylene film and stored in the same container with the spray-dried samples. The duration of storage prior to analyses could be found in Appendix A. 3.3.4.3 Analyses of the Yolk Samples (1) Proximate Analyses Acid Hydrolysis for Total Fat Determination In addition to the Goldfish and Folch et al. (1957) fat extraction methods, the official AOAC method for determining the total fat in yolk was also employed in Phase 2. According to the AOAC 925.32 (Helrich, 1990) method and with modifications, a liquid yolk sample of about 2 g was weighed into a 20 mL micro-Kjeldahl flask. With vigorous shaking, 10 mL of concentrated HC1 was added slowly into the flask. For the yolk powder samples, a sample of about 1 g was weighed into the micro-Kjeldahl flask. Again with vigorous shaking, 10 mL of HCkPLO mixture (4:1) was added slowly, washing any egg particles adhering to the sides of the flask. A blank with only 10 mL HC1 in the flask was also prepared. The flasks were placed in a 70°C water bath and heated to boiling. The samples were left to boil for 30 minutes, with shaking of the flasks every 5 minutes. The flasks were removed and cooled to room temperature. The acid-treated samples were transferred to Mojonnier fat-extraction tubes and the micro-Kjeldahl flasks were washed several times with a small amount of deionized distilled water and the wash added to the 80 Mojonnier tubes accordingly. The bulbs of the Mojonnier tubes were then nearly filled up with deionized distilled water. Exactly 25 mL each of the diethyl ether and petroleum ether was added to the treated samples and mixed. The tubes were allowed to stand until the top ether-fat layer was clear. The ether-fat solution was drawn off as much as possible into weighed 125 mL Erlenmeyer flasks containing boiling chips. The liquid remaining in the tubes was re-extracted twice using 15 mL each of the ethers at a time. The samples were well shaken upon addition of each of the ethers. The ether-fat solution was drawn off into the same Erlenmeyer flask as before and the ethers were slowly evaporated in a 70°C water bath. The fat remaining in the flask was then dried in an oven at 100°C for about 90 minutes. After removal from the oven, the flasks were allowed to cool at room temperature for about 30 minutes before weighing. A blank with no sample in the flask was prepared in the same way. It was found to have no fat. Thus there was no need to correct the weight with the blank. The results were reported as % fat by acid hydrolysis. The analyses were done in duplicate. Water Activity of Yolk Powders The Rotronic Hygroskop DT water activity meter (Serial No. 299101/1; Rotronic Instrument Corp., Switzerland) was calibrated using a 20 % relative humidity standard. Small samples of yolk powder (0.3 cm in depth) were placed and spread out evenly onto the miniature plastic dishes (approximately 4 cm in diameter) designed for the water activity meter. Samples were placed in turn into the chamber of the water activity meter 81 and the water activity and temperature were recorded once equilibrium was reached (about 20-30 minutes). Each sample was analyzed in duplicate. (2) Conductivity of Yolk Preparations prior to Spray-drying The conductivity of pH-and dilution-adjusted egg yolk solutions prior to spray-drying was measured using a Model 31 conductivity bridge with a #3403 conductivity cell (Yellow Springs Instrument Co. Ltd., Yellow Springs, OH, USA). The solutions were allowed to equilibrate to room temperature before measurement. The conductivity measurements were corrected for a temperature of 18°C. Duplicate measurements were done for each sample. (3) Functional Properties Protein solubility and Emulsifying tests were done as described in Phase 1. Emulsifying Activity Index for yolk preparations with different % solids with and without 0.5 M NaCI In order to investigate the effects of solids content of yolk and the addition of salt on the emulsifying properties (EAI and ESI) of the emulsion, yolk preparations ranging from 0.5 to 6.0 % solids were prepared in 0.1 M sodium phosphate buffers at pH 6.35, with or without 0.5 M NaCI. Only the spray-dried yolk powder control was used as the sample in this experiment. The yolk powder prepared in the phosphate buffers was allowed to 82 dissolve overnight at 4°C. EAI and ESI were determined according to the methods described in Phase 1. The samples were analyzed in duplicate. (4) Structural Analyses Differential Scanning Calorimetry (DSC) DSC was carried out at the Department of Botany, University of Hong Kong. All samples were packed in a styrofoam box with ice-packs and were transported by air to Hong Kong. The samples were stored at 4°C upon arrival at the destination. A TA Instruments Thermal Analysis - DSC standard cell (Instrument 2920 MDSC V2.2A) was used. About 1 mg of the dried sample was weighed onto an aluminum hermetic pan. Distilled water was added to the sample and stirred with a pin to make a 10 % (w/w) slurry. For each liquid sample, 10 pi of the sample was loaded onto the pan. Each pan was tightly sealed and loaded onto the DSC cell with an empty sealed aluminum pan as reference. The pans were heated from 10°C to 130°C in the DSC at a programmed rate of 10°C/min. The thermograms were analysed by the Universal Analysis program (Universal VI.9D). The initial temperature of the appearance of the endothermic peak, the enthalpy of the endothermic peak, the denaturation temperature and the temperature range at half of the peak area were identified for each sample. Triplicate analyses were done for each sample. 83 DSC on freeze-dried yolk control and freeze-dried water-soluble fraction at different pH's About 1 mg of the freeze-dried yolk control, or freeze-dried water-soluble fraction was weighed onto an aluminum hermetic pan. To test for the effects of pH on the samples, distilled water, 0.1 M citric acid buffer at pH 3.0 or 0.1 M sodium phosphate buffer at pH 8.0 was added accordingly to the sample to make a 10 % (w/w) slurry. Each pan was tightly sealed and loaded onto the DSC cell with an empty sealed aluminum pan as reference. The analyses of the samples were carried out as described before. Triplicate analyses were done for each sample. Raman Spectroscopy Raman spectroscopy was carried out at the Department of Chemistry, University of Hong Kong. All samples were packed in a styrofoam box with ice-packs and were transported by air to Hong Kong. The samples were stored at 4°C upon arrival at the destination. Raman spectra were collected on a Bio-Rad FT Raman spectrometer equipped with Nd:YAG laser at 1064 nm (Bio-Rad Lab., Cambridge, MA, USA). Liquid or powdered egg samples, egg oils extracted using the Folch et al. (1957) method and a blank consisting of 0.05 % sodium azide were packed into glass capillary tubes. Raman spectra were recorded at room temperature under the following conditions: laser power, 500 mW; spectral resolution, 4 cm"1; number of scans, 260 for solid samples and 1000 for liquid samples. The spectral data were baseline-corrected and normalized to the intensity of the phenylalanine band at 1004 ± 1 cm"1. To normalize for egg oils, the spectra of the oil extracted from the yolk samples were recorded with 260 scans. The oil was extracted 84 using the method by Folch et al. (1957). The chloroform was evaporated and the extracted oils from the yolk samples were sealed in screw-capped test tubes under nitrogen and stored at 4°C before analysis. The data was analyzed using GRAMS/386 Level I v3.02 (Galactic Industries Corporation). Duplicate were done for both liquid and powder yolk and sodium azide solution. Yolk lipid was analyzed only once. (5) Statistical Analyses Minitab Statistical Software (Release 12.1) was used for the statistical analyses. Details of the methods used were described in Section 3.2.3.5 (Phase 1). 85 3.3.5 Results and Discussion for Phase 2 Based on the results from Phase 1, a set of conditions was chosen as pretreatments for yolk to further investigate the effects of spray-drying in Phase 2. Several samples were also added as comparisons to the spray-dried samples to help better understand the possible causes of changes to the egg yolk during spray-drying. The samples in Phase 2 are listed in Table 7. Table 7. Sample Designation for Phase 2 Sample Code Description Pretreatment Toutlet Batch** (°C)* pH Dilution 1 YPC Spray-dried yolk powder (control) 6.56 2.0 76-80 X 2 PPC Spray-dried pellet powder 6.56 2.0 82-90 y 3 YPB Spray-dried yolk; the "best" powder 8.51 6.0 70 X 4 YPW Spray-dried yolk; the "worst powder 3.00 6.0 80 X 5 COM Commercial spray-dried yolk — — — z 6 LYC Liquid yolk 6.56 — — X 7 LPC Liquid pellet 6.56 2.0 — y 8 HSY Liquid hand-separated yolk — — — — 9 FYC Freeze-dried yolk 6.56 — — X 10 FPC Freeze-dried pellet 6.56 . . . — y 11 FHY Freeze-dried hand-separated yolk — — — 12 WSF Freeze-dried water-soluble fraction 5.20 — ~ y * Toutiet (°C) is the outlet temperature or outlet temperature range of the spray-dryer during the dehydration of the sample ** Samples with different lowercase letter designators were from different batches of commercial liquid yolk 86 Specifically, commercial spray-dried yolk was included to compare the characteristics of yolk obtained from pilot plant-scale and commercial-scale spray-drying. Liquid hand-separated yolk was used to verify the effects of albumen on yolk samples. Lastly, freeze-dried samples were included to distinguish the effects of drying with spray-dryer and drying per se on yolk. Freeze-drying is a more expensive operation than spray-drying. It is a dehydration method which is achieved by reduction in water activity without heating the food, and nutritional and sensory qualities are consequently better retained. However, the effects of freeze-drying may be complicated by the gelation phenomenon induced in yolk by freezing. 87 3.3.5.1 Proximate Analyses on Yolk Samples (1) Moisture (Solids) Content Again, spray-drying was carried out on a pilot plant-scale spray dryer and about 2 L of liquid yolk preparation was dried each time. Table 8 illustrates the results for the solid content of the yolk samples. Among the dried samples, freeze-dried yolk samples had the highest % solids which could be due to more complete drying as freeze-drying involves the sublimation of water from ice in the already frozen sample rather than the evaporation of water from the surface of food in spray-drying. The yolk powder control (YPC) was the second highest in % solids, followed by pellet powder (PPC), commercial yolk powder (COM), the "worst" powder (YPW) and the "best" powder (YPB). The variability in the % solids of the yolk samples may be the effects of the pretreatment. The pH pretreatment could have affected the removal of water from the sample. Both pH-treated samples had more moisture when spray-dried than that of the yolk control (p < 0.05). Higher pH pretreatment might have led to a more enclosed structure thus more water is retained after spray-drying of the sample. However, the slight differences in the drying parameters, for example, inlet or outlet temperature and the sample delivery rate and the atmospheric humidity on the day of the drying might also have contributed to the variations in % solids of the samples. When the solid contents of the yolk samples were compared to those obtained in Phase 1 (Table 2), it could be seen that the variations in % solids were similar for the liquid and powdered yolk control and yolk pellet. However, when the % solids of YPB was compared to those samples having similar pH pretreatment in Phase 1 (Samples D and F), it was noticed that YPB had a significantly lower % solids (p < 0.05). This was not likely 88 due to the variation between batches of yolk as YPB was from the same batch as the yolk control, which had similar % solids to the yolk control in Phase 1. Instead, the differences might depend more on how representative the sample was, the settings of the spray-dryer, as the outlet temperature for drying YPB was about 70°C which was lower than the outlet temperature for the other samples which was above 76°C. Since the drying temperature was lower, YPB was not as dry as the other samples. This higher moisture content of YPB could in part affect the resulting functionality discussed in this part of the study. As for the liquid samples, commercial liquid yolk had about 42 % solids and was comparable to that found in Phase 1. Hand-separated yolk, which should contain a minimal amount of albumen, had significantly higher % solids (49.60 %) than commercial liquid yolk (42 %) as expected (p < 0.05). Reconstituted liquid yolk pellet in a two-fold dilution had about 16.74 % solids. This means that the undiluted and reconstituted yolk pellet would have about 33 % solids which was similar to that in Phase 1 which was about 36 %. 89 Table 8. Moisture Content of Yolk Samples (Phase 2) Sample Code Preatment % moisture (wb) % solid (wb) pH Dilution 1 YPC 6.56 2.0 0.61+0.1 l b 99.39±0.11 b 2 PPC 6.56 2.0 0.65±0.06b 99.35±0.06 b 3 YPB 8.51 6.0 4.32±0.22 e 95.68±0.22 e 4 YPW, 3.00 6.0 1.22±0.04c 98.78±0.04 c 5 COM — — 2.24±0.09d 97.76±0.09 d 6 LYC 6.56 — 57.7110.128 42.29±0.12 g 7 LPC 6.56 2.0 83.26±0.02h 16.74±0.02h 8 HSY — — 50.40±0.06 f 49.60±0.06 f 9 FYC 6.56 — 0.31±0.17 a b 99.69±0.17 a b 10 FPC 6.56 — 0.18±0.17a 99.82±0.17 a 11 FHY — -- 0.46±0.15 a b 99.54±0.15 a b Results bearing the same superscripts within the same column are not significantly different (p > 0.05) For all samples, n = 3 90 (2) Water Activity for Yolk Powders in Phase 1 and 2 Water activity of the selected spray-dried samples from Phase 1 and the five spray-dried yolk powder samples from Phase 2 was recorded. The results are given in Table 9a and 9b respectively. In Phase 1, the samples were stored for 6 months before the water activity was measured. During the storage period, the samples could have undergone chemical changes or taken in moisture, which could affect the water activity measurement. When the results in Table 9a were compared to the corresponding moisture content of the samples in Table 2, it was noticed that sample G, which had similar pretreatment conditions to the "worst" powder in Phase 2, had the highest % moisture had also had highest water activity. In Phase 2, the water activity of the five samples were all significantly different from each other (p < 0.05), with the "best" powder having the highest water activity (aw= 0.52), followed by commercial yolk powder (aw = 0.39), the "worst" powder (aw = 0.26), pellet powder (aw = 0.16) and yolk powder control (aw = 0.13). These results showed similar trend to that obtained from the moisture determination of the powders. Those samples with higher moisture content were found to have higher water activity. The water activity of sample G in Phase 1 (similar to the "worst' powder in Phase 2) was higher than the rest of the selected samples and yet in Phase 2, the water activity of the "best" powder was higher than that of the "worst" powder. These results suggest that within the range of water activity of the powders obtained, there was no correlation with the water activity and the functional properties. 91 Water activity is an indicator of the amount of accessible water that can allow for degradative activities, such as growth of microorganisms and hydrolytic chemical reactions. It takes into account the water that associates strongly with non-aqueous constituents (Fennema, 1985). Water activity of a sample may be quite different from its water content depending on the amount of bound water in the sample. It is defined as the ratio of partial pressure of water in the food to the vapour pressure of pure water at a given temperature (Bradley Jr., 1994). Table 9a. Water Activity for Selected Yolk Powders (Phase 1) Sample Code Pretreatment Temp(°C) Water activity (aw) pH Dilution a-1 a-YPC 6.52 2.00 23.0 (22.4,23.6) 0.4276 (0.431,0.423) a-2 a-PPC 6.58 2.00 23.2 (22.7,23.7) 0.663b (0.660,0.665) a-9 C 3.88 8.83 23.3 (22.9,23.7) 0.453d (0.454,0.451) a-11 E 8.11 8.83 23.5 (23.2,23.8) 0.468c (0.465,0.471) a-13 G 3.01 6.00 23.5 (23.2,23.8) 0.686a (0.686,0.686) a-14 H 9.01 6.00 23.6 (23.3,23.8) 0.476c (0.477,0.475) Water activity was measured for selected Phase 1 samples after 6 months of storage at 4°C "a-" corresponds to samples from Phase 1; the number following the letter is the designated # of that sample in Phase 1 (refer to Table 1) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 92 Table 9b. Water Activity for Yolk Powders (Phase 2) Sample Code Pretreatment Temp (°C) Water activity (aw) pH Dilution b-1 b-YPC 6.56 2.0 22.0 (22.1,21.9) 0.134a (0.132,0.136) b-2 b-PPC 6.56 2.0 22.1 (22.2,21.9) 0.160b (0.162,0.158) b-3 YPB 8.51 6.0 22.0 (22.2,21.7) 0.5206 (0.521,0.518) b-4 YPW 3.00 6.0 22.1 (22.3,21.9) 0.256c (0.256,0.255) b-5 COM — — 22.0 (22.0,21.9) 0.385d (0.382,0.388) "b-" corresponds to samples from Phase 2; the number following the letter is the designated # of that sample in Phase 2 (refer to Table 7) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 93 (3) Fat Content Goldfish Crude Fat Determination The Labconco-Goldfish method estimates the amount of crude ("free") lipid in the yolk samples. Similar to the results in Phase 1, liquid and powdered pellet had significantly higher % crude fat (db) (p < 0.05) than all other samples because of the relatively lower protein in these samples after removal of the water-soluble fraction (Table 10). Hand-separated yolk also had higher % crude fat because of the lower albumen content than commercial liquid yolk. Since liquid yolk had significantly higher % crude fat than its spray-dried counterparts (p < 0.05), spray-drying could have been the cause of a decrease in extractability of fat from the yolk's surface. This effect was not seen when the water-soluble fraction is removed as the liquid and powdered pellet had quite similar crude fat. YPB had a significantly higher % crude fat than YPC, which could mean that adjusting the pre-drying pH to 8.5 may cause more destruction to the surface structure and lipoproteins of the yolk after spray-drying that made the fat more easily extractable. Pre-drying pH at 3.0 was not as "destructive" to the surface structure and lipoproteins as the % crude fat of YPW was not significantly different from that of YPC. However, these observations were opposite to those found for similar samples in Phase 1. These contradictions could be explained by the minor variations in the composition of the components between different batches of yolk or the effect of moisture in the powders or temperature during spray-drying (as indicated by outlet temperature) on the results of crude fat extractibility. 94 Table 10. Goldfish Fat Determination (Phase 2) Sample Code Pretreatment Goldfish pH Dilution % Fat (wb) % Fat (db) 1 YPC 6.56 2.0 47.30±0.73 47.5910.743 2 PPC 6.56 2.0 55.08±0.25 55.4410.25cd 3 YPB 8.51 6.0 47.40±0.58 49.5410.60" 4 YPW 3.00 6.0 47.33±0.90 47.91+0.91ab 5 COM — — 47.41±0.98 48.49il.00b 6 LYC 6.56 — 48.05±1.18 48.2011.19b 7 LPC 6.56 2.0 56.5710.10 56.6710.10d 8 HSY — — 53.3510.04 53.6010.04° Results bearing the same superscripts within the same column are not significantly different (p > 0.05) Folch's Method of Fat Determination As in Phase 1, liquid and powder pellet have the highest % fat (db) of about 63 % using the Folch's fat extraction method. All other samples have lower % fat (db) which were not significantly different from each other. Table 11 Folch's Fat Determination (Phase 2) Sample Code Pretreatment Folch 's pH Dilution % Fat (wb) % Fat (db) 1 YPC 6.56 2.0 56.9 (56.9,57.0) 57.3ab (57.2,57.3) 2 PPC 6.56 2.0 62.0 (62.1,62.0) 62.4bc (62.5,62.4) 3 YPB 8.51 6.0 52.0 (52.7,51.2) 54.3a (55.1,53.5) 4 YPW 3.00 6.0 53.6 (53.4,53.8) 54.3a (54.0,54.5) 5 COM — — 56.4 (57.7,55.1) 57.7ab (59.0,56.3) 6 LYC 6.56 — 22.2 (21.6,22.8) 52.5a (51.0,54.0) 7 LPC 6.56 2.0 11.1 (11.5,10.8) 66.5C (68.7,64.3) 8 HSY — 28.9 (29.7,28.0) 58.2ab (60.0,56.4) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 95 Acid Hydrolysis Fat Determination Acid hydrolysis is the official AOAC fat determination method for eggs. The sample was subjected to acid digestion to release all the bound fat. Similar to the results from the Folch's method for fat determination, liquid and powder pellet have the highest % fat (db) which was about 64 %. Yolk powder control and hand-separated yolk also have high % fat (59 % and 61 %, respectively), but were only significantly higher (p < 0.05) than commercial yolk powder (55 %). Both liquid and powder form of either the yolk control or yolk pellet had similar % fat. Table 12. Acid Hydrolysis Fat Determination (Phase 2) Sample Code Pretreatment Acid Hydrolysis pH Dilution % Fat (wb) % Fat (db) 1 YPC 6.56 2.0 57.95,58.34 58.50bc (58.31,58.70) 2 PPC 6.56 2.0 65.38,65.02 65.63d (65.81,65.44) 3 YPB 8.51 6.0 53.34,53.42 55.84ab (55.75,55.94) 4 YPW 3.00 6.0 56.45,55.84 56.84abc (57.15,56.53) 5 COM — — 55.17,53.14 55.40a (56.44,54.36) 6 LYC 6.56 — 24.41,24.32 57.61abc (57.71,57.50) 7 LPC 6.56 2.0 10.58,10.80 63.85d (63.20,64.50) 8 HSY ~ — 29.85,30.53 60.87c (60.18,61.55) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 96 Comparison of the 3 Fat Determination Methods Similar results for fat content for individual yolk samples were obtained by the AOAC acid hydrolysis method and the Folch's fat extraction method. Moreover, the fat content of both liquid and powder forms of the yolk control or yolk pellet are comparable. Thus, either method would be a good choice for determining the total fat in egg yolk. However, acid hydrolysis seems to be superior to Folch's method in determining the amount of total fat as the % fat for PPC, YPW and LYC were significantly higher as determined by the acid hydrolysis method. Greater precision of the duplicate analyses was also obtained using the acid hydrolysis method than Folch's method. In terms of the ease of extraction, the AOAC method seemed to be less time consuming, due to the use of strong acid. On the other hand, the Goldfish fat extraction method gave significantly lower % fat (db) for all samples. The petroleum ether, which is relatively non-polar, might not be able to release most of the bound fat from the lipoproteins in the yolk. Therefore, it is considered as a method for determining the crude (or free) fat in egg yolk. It is important to have this method of extraction in order to monitor whether the pretreatment and spray-drying would have effects on the degree of extractable crude fat in yolk. Table 13 summarizes the amount of fat in yolk samples as analyzed by the three different methods. 97 Table 13. Fat Determination by different methods (Phase 2) Sample Code %fat (db) - Goldfish %fat (db)-Folch's % fat (db)- Acid Hydrolysis 1 YPC 47.59±0.74 a x 57.3aby (57.2,57.3) 58.50bcy (58.31,58.70) 2 PPC 55.44±0.25 c d x 62.4bcy (62.5,62.4) 65.63dz (65.81,65.44) 3 YPB 49.54±0.60 b x 54.3ay (55.1,53.5) 55.84aby (55.75,55.94) 4 YPW 47.91±0.91 a b x 54.3ay (54.0,54.5) 56.84abcz (57.15,56.53) 5 COM 48.49±1.00 b x 57.7aby (59.0,56.3) 55.40ay (56.44,54.36) 6 LYC 48.20±1.19 b x 52.5ax (51.0,54.0) • 57.61abcy (57.71,57.50) 7 LPC 56.6710.10dx 66.5cy (68.7,64.3) 63.85dy (63.20,64.50) 8 HSY 53.60±0.04 c x 58.2aby (60.0,56.4) 60.87cy (60.18,61.55) Results bearing the same superscripts (a c) within the same column are not significantly different (p > 0.05) Results bearing the same superscripts (xyz) within the same row are not significantly different (p > 0.05) Ratios of Fat Content as obtained by the Different Fat Determination Methods When the % fat determined from the three different methods were expressed as % fat ratios (Table 14), the difference in the proportion of fat extracted from each method could be seen. The ratios of the fat determined by the Folch's method and by the acid hydrolysis were close to 1, which means that the two methods are very comparable. Acid hydrolysis gave higher % fat for most of the samples. When the crude fat was compared to the total fat in terms of the G/A ratio, it could be seen that the crude fat was relatively higher for YPB than the spray-dried yolk powder control. However, with reference to Table 13, the relationship between the pre-drying pH and the extractability of fat from the yolk powders could not be drawn here as these ratios did not seem to be significantly different. 98 Table 14. Ratios of % Fat as obtained by Different Fat Determination Methods Sample Code Pretreatment Ratio pH Dilution av. G/F av. G/A av.F/A 1 YPC 6.56 2.0 0.83 0.81 0.98 2 PPC 6.56 2.0 0.89 0.84 0.95 3 YPB 8.51 6.0 0.91 0.89 0.97 4 YPW 3.00 6.0 0.88 0.84 0.95 5 COM — — 0.84 0.88 1.04 6 LYC 6.56 — 0.92 0.84 0.91 7 LPC 6.56 2.0 0.85 0.89 1.04 8 HSY — — 0.92 0.88 0.96 G = Goldfish Fat Extraction Method F = Folch's Fat Extraction Method A = AOAC Acid Hydrolysis Method 99 (4) Protein Content (% Nitrogen x 6.25) Table 15 shows the protein content of the samples in Phase 2. Excluding the water-soluble fraction and pellet samples, all yolk samples had protein content of about 33 -35 % (dry basis). Liquid yolk and freeze-dried yolk control had significantly higher % protein (db) than the spray-dried yolk control. Liquid yolk contains about the same % protein as freeze-dried yolk (about 35 %). This could mean that there was a loss of protein after the spray-drying process. However, this is unlikely because the liquid, freeze-dried and spray-dried pellet was found to have similar protein content. On the other hand, the spray-dried yolk powder used for the analysis might not be totally representative, as only the middle portion of the spray-drying of each sample was used for all the analyses. Due to the removal of the water-soluble protein, liquid and powder pellet have lower protein content than all other samples, as in Phase 1. On the other hand, freeze-dried water-soluble fraction of the yolk has the highest protein content of about 74 %. Hand-separated yolk has slightly less protein than commercial liquid yolk because of lower albumen contamination in the hand-separated yolk sample. Overall, dehydration, specifically spray-drying, did not seem to change and should not have affected the protein content of the dried products. too Table 15. Protein Content (% N x 6.25) (Phase 2) Sample Code %N(wb) % Protein (wb) % Protein (db) 1 YPC 5.167 32.29 • 32.49cd 2 PPC 4.557 28.48 28.67a 3 YPB 5.198 32.49 33.95de 4 YPW 5.402 33.76 34.18de 5 COM 5.490 (5.435,5.544) 34.17 (33.97,34.65) 35.096ef (34.75,35.44) 6 LYC 2.428±0.038 15.17±0.24 35.88±0.57 f 7 LPC 0.755 (0.760,0.749) 4.72 (4.75,4.68) 28.173 (28.38,27.96) 8 HSY 2.543±0.015 15.90±0.09 32.05±0.19 c 9 FYC 5.644 35.28 35.39ef 10 FPC 4.592 28.70 28.75a 11 FHY 5.008 31.30 31.45bc 12 WSF 11.758 73.42 73.57g Samples # 5 & 7 were analyzed in duplicate; samples # 6 & 8 were analyzed in triplicate Fisher's pairwise comparison test was used (p = 0.05) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 101 3.3.5.2 Hunterlab Color Analyses for Yolk Powders (Phase 2) Based on visual observations, yolk powder control, pellet powder and the "best" powder had quite similar colors. Commercial yolk powder had a more orange appearance. The "worst" powder was the palest sample. The difference in color between the "worst" sample and other yolk powders was detected even in the liquid preparation prior to drying. When the yolk was adjusted to below pH 4, the yolk solution became more opaque and lighter in color. Spray-drying per se might may contribute to the variation in the color of the powders with different pretreatments as discussed in the Phase 1, however, the effects of the pH on the yolk preparation itself should not be overlooked. Figures 8-11 illustrate the difference in the color between the liquid yolk samples and between the spray-dried samples. The results of Hunterlab color measurements are shown in Table 16. 102 Figure 8. Liquid Yolk Control (undiluted) and Liquid Pellet (two-fold dilution) 103 Figure 10. Yolk Powder Samples (Spray-dried Yolk Pellet, Spray-dried Yolk Control and Commercial Spray-dried Yolk) 104 Table 16. Hunterlab Color Analyses for Yolk Samples (Phase 2) Sample Code Pretreatment Hunterlab PH Dilution L a b 1 YPC 6.56 2.0 86.46±0.13' 2.36±0.05 e 22.70±0.18d 2 PPC 6.56 2.0 85.46±0.22h 3.34±0.18 f 24.57±0.37 e 3 YPB 8.51 6.0 82.78±0.30 g 3.88±0.31 g 25.43±0.v78 e f 4 YPW 3.00 6.0 90.03+0.71j 0.68±0.06 c 18.29±0.53c 5 COM — ~ 81.57±0.18 f 5.02±0.03h 30.42±0,06 g 6 LYC 6.56 ~ 55.45±0.18 c 6.39±0.04 j 29.90±0.04 g 7 LPC 6.56 2.0 80.52±0.07 e 1.64±0.12d 26.06±0.25 f 8 HSY ~ — 52.08±0.24b 4.03±0.03 g 29.55±0.07 g L3 L-YPB 8.51 6.0 56.75±0.02d -1.99±0.03 b 15.37±0.02b L4 L-YPW 3.00 6.0 43.38±0.17 a -5.51±0.03 a 5.86±0.27a L3 and L4 were made from LYC adjusted to pH 8.51 & 6x dilution for L3 and pH 3.00 & 6x dilution for L4 (i.e. the conditions of YPB and YPW before drying) L = + lightness /- darkness; a = + redness /- greenness; b = + yellow /-blueness Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 105 (1) L-value L is a measurement of the lightness (+) and darkness (-) in appearance of the samples, the lighter the sample, the higher the L. The samples all had L values significantly different from each other (p < 0.05). YPW had the highest L value indicating that it is the lightest in color among all the dried samples. LPC was the lightest among all the liquid samples, possibly due to the fact that it was diluted while LYC and HSY were not. Dilution of the samples could also have made the samples appear lighter. (2) a-value a is a measurement of the redness (+) and greenness (-) of the sample. Among the dried samples, COM had the highest <a-value which confirm the visual observation that it was the most orange sample. YPB and PPC had higher a-values than YPC and YPW had the least fl-value. For the liquid samples, liquid yolk and hand-separated yolk were more "red" than liquid pellet, again, this might have been the combination effect of dilution and the absence water-soluble components for the liquid pellet. The a-values for the liquid preparations of YPB and YPW prior to spray-drying (L-YPB and L-YPW) were below zero. These samples might be too dilute (6 times dilution) for the accurate color determination by the Hunterlab instrument which is based on the reflectance of the samples. 106 (3) 6-value b is the measurement of the yellowness (+) and blueness (-) of the samples. COM had the highest 6-value, followed by YPB, PPC and YPC. YPW had the lowest 6-value among the dried samples and its liquid counterpart also had lower 6-value than the yolk control and hand-separated yolk. Comparing to the results from Phase 1, the results were comparable for samples with similar pretreatment conditions. Based on the Hunterlab analysis, the amount of water present in the sample and the removal of the water-soluble fraction appeared to affect the a- and 6-values of the yolk. When these results were compared with the results obtained in Phase 1, it was seen that when the pH was adjusted to below 4.0, the dried samples would always appear to be lighter, less reddish and yellowish and its liquid counterpart would always be more opaque than other samples. In contrast, increasing the pH up to pH 8.5 did not seem to affect color. It is important to note that color of liquids measured by Hunterlab L, a, b may not be reliable, unless the liquid is totally opaque, since the method is based on reflectance. The amount of water present affects the measurement drastically. 107 3.3.5.3 Conductivity of Liquid Yolk Preparations Prior to Spray-drying To compare the relative ionic strength of the liquid yolk preparations prior to spray-drying, the conductivity of the liquid preparations was measured. The more charges present in a sample, the higher its conductivity. In this study, all liquid yolk preparations had significantly different conductivity (Table 17). The "worst" yolk preparation (pH 3.0, 6x dilution) had the highest conductivity (2.99 ds/m), followed by the yolk control (2x dilution), and the "best" yolk preparation (pH 8.5, 6x dilution) which had a conductivity of 1.14 ds/m. Liquid pellet had the lowest conductivity (0.51 ds/m). When considering the protein solubility and in the next sections, the conductivity of the yolk preparations did not seem to contribute to the degree of protein solubility of their corresponding yolk powders. On the other hand, the emulsifying activity index (EAI) of the powder samples seemed to be inversely related to the conductivity of the corresponding yolk preparations. Table 17. Conductivity of Liquid Yolk Preparations (Phase 2) Sample Code Pretreatment Conductivity (ds/m) PH DUution 1 YPC 6.56 2.0 2.54c (2.59,2.49) 2 PPC 6.56 2.0 0.5 la (0.50,0.52) 3 YPB 8.51 6.0 1.14b (1.08,1.20) 4 YPW 3.00 6.0 2.99d (2.93,3.05) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 108 3.3.5.4 Functional Properties of Yolk Samples (1) Protein Solubility Table 18 illustrates the protein solubility of the spray-dried and liquid samples in Phase 2. Similar to the results found in Phase 1; the pellet powder had the lowest protein solubility (21 %) in phosphate buffer of pH 6.35. The yolk powder control and the "best" yolk powder had the highest protein solubilities among the powder samples and were about 67 %. The "worst" yolk powder had lower solubility of about 55 % and the commercial yolk powder had 48 %. Since the storage condition for the commercial yolk powder was unknown prior to delivery for its inclusion in this study, the poor solubility of the commercial yolk powder (COM) could have been temperature-abuse during its transportation from Winnipeg to Abbotsford. In addition, COM was produced about 2 months before the pilot plant-scale spray-drying was done in this Phase of the study. The poorer solubility of COM is unlikely to be due to the longer storage time because Parkinson (1977) found that the changes to spray-dried yolk during storage involved a transformation of some of the high lipid lipoprotein in the yolk LDL from an insoluble to a soluble form. In addition, commercial spray-drying differs from pilot plant-scale drying in that the inlet temperature of the dryer may be slightly lower (about 150°C) than the inlet temperature (about 200°C) for the pilot plant-scale dryer (Masters, 1991) and that the particle size of the powder from commercial spray-drying may be finer than that obtained from pilot plant-scale spray-dryer. Although the difference between the protein solubility of YPC and LYC was not significant, all liquid samples appeared to have higher solubility than their spray-dried 109 counterparts, which indicates that spray-drying affected the protein solubility of yolk. Spray-drying contributed to the changes in the surface activity of yolk molecules, making them less soluble at pH 6.35. The presence of albumen and the water-soluble fraction in the yolk contributes to greater protein solubility at pH 6.35 as hand-separated yolk and liquid pellet had lower solubility than the commercial yolk, which had 75.20 % protein solubility. Spray-drying appeared to have a more severe effect on yolk pellet than on the liquid whole yolk, as the solubility of liquid and powder yolk pellet were significantly different while no significant difference was noted between liquid and powdered yolk control. Table 18. Protein Solubility of Spray-dried and Liquid Samples (Phase 2) Sample Code Protein Solubility (%) 1 YPC 69.01de (71.06, 66.96) 2 PPC 21.44a (21.28,21.60) 3 YPB 67.65d (71.04, 64.27) 4 YPW 55.31c (56.29, 54.33) 5 COM 47.53b (49.41,45.66) 6 LYC 75.20e (77.12, 73.28) 7 LPC 48.22bc (50.72, 45.72) 8 HSY 66.91d (69.50, 64.31) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) When the results were compared to those obtained from Phase 1 (Table 4), it was found that all the protein solubilities obtained in Phase 2 were higher than in Phase 1. The variation in the composition of the samples between batches could be one possible reason. Lower temperatures during spray-drying of Phase 2 than Phase 1 samples, as indicated by 110 the outlet temperatures (T o u t i e t , Tables 1 and 7), could also have been an important factor. The average of the results from the 2 phases showed the following trend of protein solubility: Liquid yolk > hand-separated yolk (in Phase 2 only) > "best" powder > yolk control powder > "worst" powder > liquid pellet > pellet powder. In addition to the protein solubility test on spray-dried yolk powder, the test was also carried out for the freeze-dried samples to study the effect of the form of dehydration towards protein solubility. The results are depicted in Table 19. Table 19. Protein Solubility of Spray-dried and Freeze-dried Yolk and Pellet Samples* Code Description Protein Solubility (%) YPC FYC PPC FPC Spray-dried yolk control Freeze-dried yolk control Spray-dried pellet Freeze-dried pellet 75.28d (75.24, 75.33) 64.63c (65.28, 63.98) 13.86a (14.59, 13.14) 20.97b (22.63, 19.31) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) *Solubility tests were conducted on dried samples after 6 months of storage at 4°C Freeze-drying of the liquid samples were carried out about 2 weeks after the collection of the spray-dried samples (Appendix A). Prior to freeze-drying, the liquid yolk was kept at 4°C with the addition of 0.05 % sodium azide. The freeze-dried samples were also stored under the same condition as the spray-dried samples at 4°C. Since the protein solubility test in this part of the study was done about 6 months later than the previous analysis, the duration of storage at 4°C on the samples could have affected the solubility of i l l the samples. It was found that the solubility of the spray-dried yolk control was about 15 % higher than before (from 69.01 to 75.28 %), which was in agreement with the results reported by Parkinson (1977), while that for spray-dried pellet decreased from 21.44 % to 13.86 %. Thus, the effect of storage could be different for yolk samples with different composition. As pellet contains more lipids, the effect of possible lipid oxidation would be more prominent to the pellet samples. Freeze-drying and spray-drying seemed to have different trends on the protein solubility of yolk and pellet samples. The composition of the sample might have been the determinants of how the different dehydration might affect its solubility. Freeze-drying, although carried out at a lower temperature, may not necessarily result in better solubility of the dried sample in the case of yolk due to the gelation phenomenon that occurs during the freezing step. (2) Emulsifying Properties EAI and ESI of Yolk Emulsion at Varying % Solid Content with and without 0.5 M NaCI In order to establish the effects of concentration of yolk solids in the phosphate buffer preparation on the emulsifying test, yolk preparations with different % solids were prepared and the effects on emulsifying properties were noted. Moreover, as an aside, the effect of ionic strength was tested by addition of 0.5 M NaCI into the phosphate buffer preparations for one set of samples. The results are shown in Tables 20-21 and Figures 12-15. 112 An exponential relationship (with R 2 = 0.97) for each of the emulsions prepared in the presence or absence of 0.5 M NaCl was found between the solid content of the preparation and the resulting EAI of the emulsion (Figures 12 and 13). As the concentration of yolk solids in the emulsion increases, the EAI decreases. This kind of relationship was also observed for emulsifying properties of food muscle samples (Cofrades et al., 1996). This kind of inverse relationship, however, contradicts the findings of Chung and Ferrier (1991b) who suggested that the emulsifying activity of phosvitin increased with protein concentration. However, Chung and Ferrier (1991b) defined emulsifying activity simply as the absorbance at 500 nm, whereas the emulsifying activity expressed here was on a basis of the surface stabilized per gram protein or per gram solid of the samples. Even though absorbance increases as concentration increases, in fact, the EAI (m2/g) would be decreased as concentration increases. The presence of 0.5 M NaCl was observed to give emulsions with higher EAI. Kiosseoglou and Sherman (1983a) found that adding salt to egg yolk solution when preparing the emulsions disrupts the granules and provides additional surface-active material for adsorption. In terms of the stability of the emulsion, an increase in % solids seems to have a positive relationship with the ESI of the corresponding emulsion. A linear relationship (R2 = 0.99) was seen for ESI as a function of % solids for the emulsion with NaCl (Figure 14) and an exponential relationship (R2 = 0.99) was seen for the ESI as a function of % solids for the emulsion in the absence of NaCl (Figure 15). These observations suggest that moderate ionic strength may enhance the stability of emulsions made from yolk powder. With a purpose to best distinguish between the stability of the 113 emulsions made from the actual samples, a yolk concentration of 6 %-solids was chosen for the preparation of samples in phosphate buffer of pH 6.35, without NaCI, for making the emulsions. Table 20. EAI of Yolk Emulsion at Varying % Solid Content with and without 0.5 M NaCI With 0.5 M NaCI % Solids in Sample EAI (m /g protein) EAI (m /g solid) 0.5 % S 108.23 35.17 1.0 % S 87.11 28.30 3.0 % S 42.02 13.65 6.0 % S 22.37 7.27 DUt NaCI % Solids in Sample EAI (m /g protein) EAI(m2/g solid) 0.5 % 101.61 33.01 1.0% 66.97 21.76 3.0% 39.20 12.74 6.0% 18.80 6.11 Yolk preparations were prepared from yolk powder control (YPC) 0.5 % S corresponds to a yolk preparation containing 0.5 % yolk solids with 0.5 M NaCI 0.5 % corresponds to a yolk preparation containing 0.5 % yolk solids 114 Figure 12. EAI (m/g solid) of Yolk Emulsion with 0.5M NaCI 40.00 35.00 30.00 1 25.00 w ^ 20.00 E ^ 15.00 LU 10.00 5.00 0.00 • "V. y = 37.44e 2 8 - 4 8 x ^ ^ ^ R 2 = 0.9773 • ^ ^ ^ ^ ^ 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% % solids in yolk preparation 6.0% 7.0% 115 Figure 13. EAI (m 2/g solid) of Yolk Emulsion 35.00 - i 0.00 "I 1 r 1 1 1 1 1 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0% % solids in yolk preparation <4 Table 21. ESI of Yolk Emulsion at Varying % Solid Content with and without 0.5 M NaCl % solids in Sample ESI (min) 0.5 % S 2.4 1.0 % S 3.3 3.0 % S 27.6 6.0 % S 55.9 0.5 % 0.5 1.0% 0.6 3.0% 1.3 6.0% 7.1 Yolk preparations were prepared from yolk powder control (YPC) 0.5 % S corresponds to a yolk preparation containing 0.5 % yolk solids with 0.5 M NaCl 0.5 % corresponds to a yolk preparation containing 0.5 % yolk solids 116 Figure 14. ESI (min) of Yolk Emulsion with 0.5M NaCI 60.0 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0% % solids in yolk preparation Figure 15. ESI (min) of Yo lk E m u l s i o n % sol ids in yolk preparation 117 Emulsifying Tests on Selected Samples from Phase 1 and Phase 2 To confirm the results found in Phase 1, selected samples from Phase 1 were analysed again after 6 months storage with the samples from Phase 2 in terms of EAI (Table 22) and ESI (Table 23). The liquid samples from Phase 1 (a-LYC and a-LPC) had probably gone through some chemical changes as sodium azide added was only able to prevent bacterial deterioration. Emulsifying Activity Index (EAI) The EAI of BSA obtained in Phase 1 and Phase 2 were comparable (50.96 and 47.68 m2/g protein, respectively), indicating that the results obtained in the two Phases could be compared together. .-For Phase 1 samples: On a per gram solid basis, sample G (pH 3.01, 6x dilution) had significantly lower EAI (p < 0.05), while the other powder samples had similar EAI to the liquid yolk control. Liquid pellet had a lower EAI than liquid yolk on a per gram solid basis, but the difference was not significant. . For Phase 2 samples: On a per gram solid basis, the "worst" powder had a significantly lower EAI than all the other yolk powder samples (p < 0.05). The "best" powder had a significantly higher EAI than the worst powder. Among the liquid samples, hand-separated yolk had a significantly lower (p < 0.05) EAI than the liquid pellet and its EAI was comparable to that of the "worst" powder. 118 The samples with similar pretreatments from Phase 1 and 2 had very similar EAI. It could be seen clearly that at low pH pretreatments, EAI of the resulting emulsion seemed to be lower, indicating that with low pH pretreatments and with combination of spray-drying, the surface activity of yolk was damaged, as a result, the amount of emulsion stabilized was smaller. With reference to the yolk control and yolk pellet samples, the EAI values on a per gram protein basis and on a per gram solid basis appeared to show slightly different trends. Although not significant at p < 0.05, the EAI of the powder samples, when expressed as unit per gram protein, was higher than that of their liquid counterparts, whereas when the EAI was expressed as unit per gram solid, the reverse trend was noticed. The lipid that is also associated with the protein may contribute to the EAI as well as the protein, therefore, it might be more appropriate to use the EAI expressed as m per gram solids. When the trend of the EAI for the yolk samples was compared to that obtained for the protein solubility of the samples, a relationship could not be drawn between the two functional parameters of yolk samples. 119 Table 22. EAI of Selected Samples from Phase 1 and 2 Sample Code pH Dilution EAI (m /g protein) EAI (ml/g solid) BSA BSA — 47.68 a-1 a-YPC 6.52 2.00 22.39±2.69 b c d e (21.85) 7.42±0.89 c d(7.24) a-2 a-PPC 6.58 2.00 25.78±2.49 d e (24.89) 7.24±0.70 c d(7.00) a-11 E 8.11 8.83 22.58±0.75 b c d e (22.49) 7.40±0.25 c d(7.37) a-13 G 3.01 6.00 11.04±2.26a (12.21) 3.61±0.74 a(3.99) a-14 H 9.01 6.00 23.84±5.13 c d e (20.89) 7.76±1.67 c d(6.80) a-16 a-LYC 6.52 — 19.90±0.34 b c d e (19.89) 6.62±0.11 c d(6.62) a-17 a-LPC 6.58 2.00 21.80±3.05 b c d e (23.17) 5.83±0.81 a b c d(6.19) b-1 b-YPC 6.56 2.00 24.26±2.97 c d e 7.91±0.96 c d b-2 b-PPC 6.56 2.00 27.74±0.91 e 7.95±0.26 c d b-3 YPB 8.51 6.00 22.99±0.99 c d e 7.80±0.34 c d b-4 YPW 3.00 6.00 14.76±2.55 a b 5.04±0.87 a b b-5 COM — — 23.88±1.74 c d e 8.38±0.61d b-6 b-LYC 6.56 — 17.54±1.23 a b c d 6.30±0.44 b c d b-7 b-LPC 6.56 2.00 26.68±5.22 e 7.51±1.47 c d b-8 HSY — — 16.87±3.45 a b c 5 .41± l . l l a b c "a-" corresponds to samples from Phase 1 analyzed after 6 months of storage (numbers in parentheses were analyzed soon after spray-drying, as shown in Table 5); "b-" corresponds to samples from Phase 2; the number following the letter is the designated # of that sample in that Phase (refer to Table 1 and Table 7) YPC, PPC, LYC, LPC stand for the yolk powder control, pellet powder, liquid yolk, liquid pellet and the letter preceding the codes indicate which Phase the samples were from All other samples had the same codes as before All analyses were done in triplicate Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 120 Emulsion Stability Index (ESI) As shown in Table 23, for Phase 1 samples: liquid pellet had the highest ESI of more than 35 minutes, followed by that of liquid yolk, yolk powder control and sample E (pH 8.8, 6x dilution). Pellet powder and sample G (pH 3.01, 6x dilution) had the least ESI, which was about 2 minutes. For Phase 2 samples: the "best" powder (pH 8.5, 6x dilution) had the highest ESI of about 30 minutes. Yolk powder control had an ESI of about 9.5 minutes. Commercial yolk powder and pellet powder had ESI below 4 minutes and the "worst" powder (pH 3.0, 6 x dilution) had the least ESI of about 2 minutes. Liquid pellet also had a high ESI of about 15 minutes, followed by that of liquid yolk, which was about 14 minutes. Based on the literature, it could be expected that ESI of commercial liquid yolk would be lower than that of hand-separated yolk because it was reported that albumen, specifically ovalbumin, contamination would reduce the stability of the emulsion (Pearce and Kinsella, 1978; Cheftel et al., 1985; Chung and Ferrier, 1991). Yet in this study, liquid yolk control was found to have an ESI of 14 minutes which was comparatively higher than that of hand-separated yolk (8 minutes). The ESI of liquid yolk control analyzed in Phase 1 (Table 6) was found to be 5.7 minutes which was lower than the ESI of 14 minutes found here, indicating that there could be variations between different batches of samples or inconsistencies in the operation of the homogenizer. It should be noted also that differences between commercial liquid yolk and hand-separated yolk in the present study may not be solely caused by the presence of albumen, since the yolk preparations were from different batches and sources. 121 Although the samples from Phase 1 and Phase 2 had quite different ESI, which might be due to the different duration of the sample in storage, similar trends were seen for samples with similar pretreatment conditions. In summary, the samples with low pH pretreatments (pH 3.0) had very poor emulsion stability while those samples with higher pH than original yolk (pH 8.0) had equivalent or better stability than the control. Liquid yolkpellet had superior stability to liquid commercial yolk, however, the stability of the pellet was lost to a greater extent after spray-drying, while that of the yolk control was not much affected after spray-drying. As described before, protein solubility of the yolk samples did not directly correlate with the EAI of the corresponding emulsion, nor did it correlate with the stability of the emulsion. In addition, the sample's EAI did not reflect on the degree of stability of the corresponding emulsion. 122 Table 23. ESI of Selected Samples from Phase 1 and 2 Sample Code Pretreatment ESI (min) P H Dilution BSA BSA — 38.8 (23.5) a-1 a-YPC 6.52 2.00 12(5.5) a-2 a-PPC 6.58 2.00 2.1 (3.2) a-11 E 8.11 8.83 8.4 (4.0) a-13 G 3.01 6.00 2.1 (2.0) a-14 H 9.01 6.00 4.5 (6.0) a-16 a-LYC 6.52 — 14 (5.7) a-17 a-LPC 6.58 2.00 > 35 (> 20) b-1 b-YPC 6.56 2.00 9.5 b-2 b-PPC 6.56 2.00 3.3 b-3 YPB 8.51 6.00 28 b-4 YPW 3.00 6.00 2 b-5 COM — — 4.2 b-6 b-LYC 6.56 — 13.6 b-7 b-LPC 6.56 2.00 15 b-8 HSY — — 8 "a-" corresponds to samples from Phase 1 analyzed after 6 months of storage (numbers in parentheses were analyzed soon after spray-drying, as shown in Table 6); "b-" corresponds to samples from Phase 2; the number following the letter is the designated # of that sample in that Phase (refer to Table 1 and Table 7) YPC, PPC, LYC, LPC stand for the yolk powder control, pellet powder, liquid yolk, liquid pellet and the letter preceding the codes indicate which Phase the samples were from All other samples had the same codes as before 123 Comparisons on the Emulsifying Properties of Spray-dried and Freeze-dried Yolk and Pellet To investigate whether the difference in the EAI was mainly due to the effect of dehydration or was specifically due to spray-drying, the freeze-dried and spray-dried forms of the yolk control and yolk pellet from Phase 2 were compared. According to the principles of freeze-drying and to previous literature, the yolk powder produced from this method is not exposed to high heat and thus, the protein structure could be better maintained. Since both spray-dried and freeze-dried samples were all stored for about 6 months at 4°C prior to this analysis, it was assumed that the possible effect of storage on these samples would be similar. As depicted in Table 24, the EAI of all the samples were not significantly different from each other. However, the stability of the yolk emulsions made from the spray-dried and freeze-dried yolk control were different. Freeze-dried yolk control gave a more stable emulsion with ESI of 40.0 minutes, while the ESI for the emulsion made from the spray-dried yolk control was significantly lower (p < 0.05). The pellet samples had similarly poor ESI of about 3 minutes, which means no matter what the method of dehydration was, the removal of water from the pellet samples had detrimental effects on its emulsion stability. As seen in Phase 1 and Phase 2, yolk pellet had lower ESI than yolk control after dehydration. Freeze-drying appeared to be a better method of choice if the stability of the emulsion made from the yolk control powder is of concern. The more stable emulsion may be due to the presence of more free fatty acids in the yolk sample after freezing (Nakamura et al., 1982). 124 Although the EAI values for BSA were different in Table 22 and 24, the EAI obtained for the spray-dried and freeze-dried yolk and pellet samples did not differ too much with those of the corresponding samples in Table 22. When the ESI values in Table 24 were compared to those in Table 23, it was noticed that the ESI for spray-dried yolk control was a lot higher than that in Table 23. This increase in ESI could be due to the impact of storage of the sample at 4°C for 6 months. On the other hand, storage did not have much impact on the ESI of the spray-dried pellet sample. Table 24. EAI and ESI of Spray-dried and Freeze-dried Yolk and Pellet samples (Phase 2) Code Description EAI (m /g protein) EAI (m /g solid) ESI (min) BSA Bovine serum albumin 56.23 (54.07,58.40) 18.2 (16.4, 20.0) YPC FYC PPC FPC Spray-dried yolk control Freeze-dried yolk control Spray-dried pellet Freeze-dried pellet 21.68a(22.65,20.70) 7.09a(7.41,6.77) 20.59a(20.92,20.26) 7.31a(7.42,7.19) 26.53a(26.62,26.43) 7.65a(7.68,7.63) 27.38a(29.86,24.90) 7.89a(8.60,7.17) 25.9b (25.0, 26.7) 40.0a (37.3, 42.7) 2.2C (2.1, 2.3) 3.9C (3.8, 4.0) Emulsifying tests were conducted on dried samples after 6 months of storage at 4°C Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 125 3.3.5.5 Structural Analyses (1) Differential Scanning Calorimetry (DSC) Table 25 shows the results of the Differential Scanning Calorimetry (DSC). DSC is a particularly useful method for gaining information on protein structure and the effects of processing on the protein in the samples. However, the reproducibility of the DSC experiments on yolk sample was relatively poor. The relatively high fat content and the close association of the fat with the protein in the yolk samples may affect the dissolution of the dried powders in water and subsequently the process of protein denaturation and thus the result of DSC. Sample G (pH 3.01, 6x dilution) and pellet powder from Phase 1, as well as the "worst" sample (YPW) from Phase 2 did not show any denaturation peak within the temperature range that the other samples showed denaturation. The specific proteins that gave the denaturation peak in the other yolk samples, could have already been denatured in these three samples and thus did not show any changes at the denaturation range. Therefore, these three samples were excluded from further discussion on the DSC results of the other yolk samples below. The pellet powder in Phase 2 showed a denaturation peak. The difference seen between the thermal profiles of pellet powders from the 2 phases could be attributed to the effect of storage. The different parameters measure for the samples are illustrated in Table 25. Figure 16 shows a typical denaturation peak of yolk samples in the DSC profile. 126 C N X J CO CO CO t CS E >< X i CD 4-» co 13 C/3 3 CO u Q in CN co 1 ft, C N 5 I: I ^3 to r-C N x cn X V O r-d +i o V O d vo oo C N S-i V O > 7 in ^ oo V O 0 0 in OO in cn I H O i2 2 V O C N V O cn 0 0 oq cn CN C N in o *^  -a oo vo CM o U 03 ^ C N O N 1^ <—< r^, C N •a in 0 0 o C N cn ^ oo C N O O —I - H C C N - 0 1J m TT -Cod o\ c O N o X XI r- oo oo r~-cn • . ON av O N ^ - N C N oo _ r-°°. cn ^ oo H ""1 cn O N vo" C N P r- T - H ^ s i c ^ •§ S x oo 0 0 Tt* ON O N C N oo CN On p oo cn C N O ON r- o u cn CN oo oo ^ cn" cn" O C N cN i ! vd £ SSoo oo cn ^ V O o <^  )S oo oo in cn in oo oo^  oo^  oo" O N O N oo in oq cn T f oo oo oo X 0 0 O C N +1 C N in cn x x x oo vo in O N p vo O N in cn oo oo oo £ CN ^-t oo "1 oo" CN CN oo ^ o ^ vd oo oo^  22 CN cn vd ^ oo oo . "1 ON CN vo cn oo oo oo oo r-r - H T j - V O ° ^ C N ^ +\ ^ ^ C N w w ^ x x m O N in r~ vo cn oo d +1 r-cn oo in in cn r~ cn" C N TT S - ^ ft C N O N +1 in _ _ oo r - H r~- C N 1 > ^ , ON" cn cn ° . s 6 2 a x x " C N O N f > O N 0 0 S^vd vd oo o o in vo '. m C N V O cn p C N cn „ m r- "S-H H (S 5 CN V O d +1 O N Tt* -a a> CN t— t— cn -^5 CN C N o ^ in ^ d 'fr o" E: o m ^ O N O O B „ o —I V O Tt in ^ o r-i O ^ 1 o vo s ^ ^ H 0 0 C N ^ as " ^ S *^ O C N co +1 r-M I T , O J vo O N C N ^ ^ O N CN O C N O N T 3 cn p +1 in in r-o" •<* T t r-. oo p in r--O N o ON r- « ^ S 0 0 P-c ^ o oo oo oo vo 0 0 3 cn = r-in ^ • vo +1 oq in vd vd oo" vo" K vd c^  ^. ° ^ s, cn cn ^ ' O N 5 r~ r~ vo C N cn ^ C N cn r-cn vq r-^  vd r- x k. cn cn u X ^ C N O N <^  ^ 2 ^ CS CS CN cn T t i i i -P X> £> m vo i i I .o 0 0 O N I I XJ -P C N -P -P X) c '33 2 X3 s 3 * J 03 53 P H a c o ts l-l 3 ts n <D T3 ( U -3 •a H T3 C *4—» CS u 3 CS 3 <U -a »M .2 2 3 ^ CS 1-<U I t o C o bO 3 CS <U 3 cs c s: •4—1 1-o T 3 D CO "cs 3 CS <U •~-D ^ O Cl, 3 C c S CO ( U o — ^ U o O cn o cn c CO CO > 4—> CO X) co 00 3 c S u co 3 4-» CS i-CO P H 3 3 CO 4 - J <u x; 4—» u O co CO nal U c S o 2 O cn CO 1 o C N 1 X) en CO •4-* | <D X) CO 4-> 03 o 3 T3 co 3 O -a 2 CO _C0 " P 4 I CO l - i CO X ! co ts o 'E, co 3 O CO O N I X) X J CO CO ts o E 4 2 co > 3 O CO 3 O X J CO c S •ti in in o XJ co c o 3 c o c S X ) c o 3 O CO a, 6 o co CO c o c o "SH CO X ! CO in o d A Cu 3 2 .co x) x> 4—> CO 1 o 3 I _3 E CO co I co co X ! 4—> 3 2 P H •c o CO 1-CO P -3 c/3 co £ c S CO CO X ! •4-1 ' c S ^ P - CO X) c o CO Pi 127 128 Onset Temperature of Denaturation (°C) Freeze-dried hand-separated yolk, the "best" powder and yolk powder control had a significantly higher onset temperature Tj (about 79°C, p < 0.05) than freeze-dried pellet (73.8°C), liquid yolk (74.3°C) and freeze-dried water-soluble fraction (74.3°C). It was reported by Ozawa (1986) that raw egg yolk had an endotherm between 69-90°C. However, the specific components contributing to the endotherm were unknown. The samples in this study seem to have a higher onset temperature than that reported by Ozawa (1986), which is likely due to the different instrument or conditions, e.g. heating rate, used for the analysis. Enthalpy of Endothermic Peak (J/g sample) Liquid yolk pellet had the lowest enthalpy of 0.27 J/ g sample. Sample E (pH 8.8, 8x dilution) from Phase 1 had significantly higher enthalpy (4.12 J/g sample) than hand-separated yolk and liquid and powdered yolk pellet. Commercial yolk powder had a significantly higher enthalpy (3.68 J/g sample) than liquid yolk pellet. Freeze-dried water-soluble fraction had significantly higher enthalpy (9.24 J/g sample) than all other samples. When calculated on a per gram protein basis, there was no significant difference found among the spray-dried samples because of the large standard deviation between replicate analyses. However these samples all had significantly lower enthalpies than the liquid yolk control and freeze-dried WSF. Based on the averages of the results, sample E 129 from Phase 1 and the freeze-dried water-soluble fraction (WSF) had significantly higher enthalpies than other samples. Pellet powder had the lowest enthalpy. Donovan et al. (1975) found that the enthalpy of denaturation of egg white equals the sum of the enthalpies of denaturation of its component proteins. Thus, it could be speculated that the same relationship would apply for the yolk samples in this study, that is, the enthalpy of individual samples corresponds to the their composition. Due to the removal of water-soluble proteins by preparation of the yolk pellet, some of the protein that contributes to the overall enthalpies of the yolk samples may be lost in the yolk pellet sample. In addition, the more stabilized the native protein structure, the higher the apparent enthalpy on the DSC curve (Tanford, 1970). This indicates that the freeze-dried water-soluble fraction has the most stabilized protein upon heating while yolk pellet is more prone to heat damage among the samples. Denaturation Temperature (°C) Ozawa (1986) reported that the denaturation temperature for raw egg yolk was 81°C. In the present study, the denaturation temperature of the samples was found to be between 82°C to 87°C, which was slightly higher that that reported for raw yolk with minimal amount of albumen. There was no significant difference in the denaturation temperature of the samples except freeze-dried yolk pellet showed significantly lower denaturation temperature (82.21°C) than the yolk powder spray-dried under the best conditions (86.04°C) and yolk powder control (85.55°C) (p < 0.05). The components in yolk contributing to the denaturation peak are unknown as the results found here seems to 130 be different from the denaturation temperature of yolk components in model solubiotns reported in the literature. Chung and Ferrier (1995) found that the denaturation of phosvitin in water at pH 7.0 had a denaturation temperature of about 78 - 80°C and the denaturation temperature of native LDL was found to be about 77°C (Mine, 1997). IgY that is present in the water-soluble fraction of yolk had a denaturation temperature of about 74°C (Hatta et al., 1993). The denaturation temperature observed here could be the additive results of the denaturation of individual components in yolk. Width at Half Peak Height (°C) The width at half of the peak height for all the yolk samples, whether liquid or powder, were not significantly different. This means that the denaturation of liquid samples was similar to that of the yolk powder obtained from different dehydration methods. 131 DSC at different pH's for Freeze-dried Yolk and Water-soluble Fraction It was suspected that the denaturation peak noted above was mainly due to the behavior of the water-soluble proteins in the yolk sample because WSF had the highest enthalpy among the yolk samples and the yolk pellet samples had relatively lower enthalpy at T d. These water-soluble proteins that contributed much to the peak were extracted during preparation of the yolk pellet. Moreover, it was suspected that there were changes to the yolk samples while the pH of the yolk was adjusted as the pretreatment, which contribute more than spray-drying per se on the structural properties of the resulting powder. Thus, to test these hypotheses, freeze-dried samples, which were thought to be less damaged than spray-dried yolk, were reconstituted in buffers of different pH's. Specifically, freeze-dried yolk control and freeze-dried water-soluble fraction were studied. The samples were made into a 10 % slurry with buffers of pH 3 or pH 8 to simulate the pre-drying condition prior to spray-drying and analyzed by DSC and the results were given in Table 26. The denaturation peak was again detected at about 70-85°C depending on the pH of the samples. The values of Tj and Td were at the lower end of the temperature range for both FYC and WSF at pH 3, compared to the samples prepared with distilled water and with pH 8 buffer. This could mean that there was structural alteration at pH 3 which causes the proteins in yolk to denature at lower temperature. On the other hand, the proteins were quite similar at yolk's original pH and at pH 8. As the water-soluble fraction was probably about 42.4 % of the total proteins in FYC (Osuga and Feeney, 1977), and the enthalpies as well as width at half peak height were lower for FYC than for WSF at all pH 132 tested, this indicates that the water-soluble fraction could have been the main component contributing to the denaturation peak in the yolk samples and the thermal stability of yolk depends very much on the presence of the water-soluble components. Table 26. DSC Results for FYC and WSF solutions (10 % w/v) at different pH's Sample Buffer Ti(°C) Enthalpy Enthalpy Td(°C) Width at 1/2 (J/g sample) (J/g protein) PeakHt(°C) FYC pH3 64.81a 0.57a 1.63a 68.42a 6.55a (67.22,62.39) (0.49,0.66) (1.39,1.87) (69.03,67.80) (4.26,8.83) FYC pH8 76.92c 2.34a 6.64b 84.80c 10.81bc (76.30,77.53) (2.80,1.87) (7.95, 5.30) (84.30,85.29) (11.47,10.15) FYC dH 20 77.6011.85c 2.17±0.68a 6.15±1.92b 84.76±2.51 c 9.10±0.80 a b WSF pH3 69.26ab 2.14a 2.92ab 74.34b 7.45ab (67:13,71.38) (1.34,2.95) (1.82,4.01) (74.30,74.37) (7.25,7.65) WSF pH8 72.14bc 2.30a 4.08ab 77.71b 14.20c (70.93,73.35) (4.63,1.36) (6.31,1.85) (77.86,77.55) (16.41,11.98) WSF dH 20 74.27c 9.24b 12.59c 83.93c 11.48bc (74.34,74.20) (9.02,9.46) (12.29,12.88) (82.38,85.48) (11.23,11.72) Except for FYC with dFEO having n=3, all other samples had n=2 Fisher's pairwise comparison test was used (Individual error rate = 0.05) Results bearing the same superscripts within the same column are not significantly different (p > 0.05) 133 Summary of DSC Results The yolk powder spray-dried under the "worst" pretreatment condition, that is, pH 3.0 and 6x dilution, gave no denaturation peak on the DSC thermogram. This means that the proteins in the powder obtained under these conditions were completely denatured prior to DSC analysis. Freeze-dried hand-separated yolk, the "best" powder (pH 8.5, 6x dilution) and the spray-dried yolk control had a higher denaturation temperature (Td) than freeze-dried pellet, and these 3 samples also had higher onset temperature (Ti) of denaturation than liquid yolk, freeze-dried pellet and freeze-dried water-soluble fraction. Ffigher T; and T d indicate that the proteins in the samples were relatively more resistant to heat denaturation, also, there could be interactions between the lipid and proteins in the samples affecting the process of denaturation, therefore, a higher temperature was needed before the denaturation peak was reached. Spray-dried and liquid pellet generally had lower enthalpies on a per gram sample basis as well as on a per gram protein basis. Hand-separated yolk had lower enthalpy than liquid yolk control and freeze-dried water-soluble fraction. The highest enthalpy on a per gram protein basis was observed for the freeze-dried water-soluble fraction and the liquid yolk control. Since the liquid yolk control contains albumen, these observations suggest that the proteins in the albumen and water-soluble fraction played a dominant role in the DSC profile. When FYC and WSF were reconstituted in different pH buffers, it was noticed that at pH 3, FYC had lower values for all the measured parameters than all the other samples. The proteins in FYC could be more denatured or more labile to 134 denaturation at pH 3. At this pH, the proteins in WSF were more stable than FYC based on the higher Td of the former than the latter. In addition, it was observed that the enthalpies were higher for WSF than FYC. In general, WSF was more stable, with higher Tj, T d and enthalpies in water than in 0.1 M buffers at pH 3 and 8. This could be due to the different pH effect on the sample or to the effect of ionic strength on stability of the proteins. At pH 3 or in water, FYC had lower Td than WSF. This could mean that the proteins in yolk were less stable than in the WSF at this particular pH. However, at pH 8, the reversed was observed. The T d of FYC was greater than that of WSF, meaning that the proteins in FYC might be more stable than in WSF at pH 8. This might be related to the observed results of RSM-CCRD yolk powder of having better retention of solubility or emulsifying properties with pretreatments of pH 8-10. -135 (2) Raman Spectroscopy Analyses The FT-Raman Spectrometer was chosen to study the protein structure of the yolk samples due to its advantages of not having problems with sample fluorescence and being applicable to both liquid and solid samples. Any changes in the yolk samples due to pretreatments and spray-drying was expected to be found by comparing the profile of the yolk powder with the control. Indeed, the instrument was able to give clear spectra for the yolk samples with low number of scans. The Raman profiles were obtained for all the spray-dried samples in Phase 2, the liquid yolk control, liquid pellet and hand-separated yolk as well as the freeze-dried yolk, yolk pellet and water-soluble fraction. Since the yolk contains relatively high proportion of fat, the corresponding oil for the yolk samples extracted with the Folch's method were also analysed. As a blank, the profile of 0.05 % sodium azide was also obtained. The Raman spectra for the yolk powder control, the oil extract of the yolk powder control were illustrated in Figures 17 and 18, respectively. Since the changes to the protein was expected after the pretreatments and spray-drying, it was necessary to eliminate the peaks due to the lipid. Therefore, the spectra for the oil was subtracted from the yolk sample spectra and the difference spectra between samples were compared. Figure 19 shows the difference between the two spectra. The spectra for the yolk powder control and for its oil extract showed several prominent bands were detected at the regions: 1100, 1200, 1700 and 3000 cm"1 for all the yolk samples, which were possibly the bands for oil and carotenoids (Li-Chan, 1996). Unfortunately, when the difference of the spectra were taken, it was found that the major peaks disappeared and the remaining difference 136 spectrum had too small variation that could not be interpreted. Similar profiles were obtained from the other yolk samples. The difficulties encountered in interpreting the data was due to the much stronger Raman scattering of the lipid components than the protein components. The resulting difference spectra, intended to investigate the protein components, showed poor signal-to-noise ratio, making it difficult to compare the different spectra between samples. The FT-Raman analysis for the yolk samples was found to be inconclusive at this point which in part was due to the overwhelming effects of the lipid from lipoproteins on the spectra. 137 Figure 17. Raman Spectrum for YPC 15 10 OH 3000 2500 2000 1500 1000 Arbitrary Y / Raman Shift (cm-1) Paged Y-Zoom C U R S O R 138 Figure 18. Raman Spectrum for the Oil Extract of YPC 3500 3000 2500 2000 1500 1000 500 Arbitrary Y / Raman Shift (cm-1) Paged Y-Zoom C U R S O R 139 Figure 19. The difference in the Raman Spectra for YPC and its corresponding Oil Fraction ' CHAPTER 4. GENERAL DISCUSSION AND CONCLUSIONS In Phase 1 of the study, the effects of pH, dilution and removal of the water-soluble fraction of the yolk prior to spray-drying were studied. The pH and dilution of the yolk were optimized using the CCRD-RSM in terms of the protein solubility and emulsifying properties of the resulting powders. Results from the response surface regression shows that the % solids, % protein and % total fat remained relatively constant for the yolk samples after spray-drying. However, the extractability of crude fat from the yolk powder was affected by both spray-drying and pH treatment. Aggregation of the lipoproteins could have occurred when the yolk was pH adjusted and during atomization and heating upon spray-drying. Solubility of the yolk proteins in phosphate buffer at pH 6.35 was reduced from about 70 to 52 % after spray-drying. The % protein solubility was even lower for liquid yolk pellet and dried pellet powder which is likely caused by the removal of the water-soluble proteins from the yolk during preparation of the pellet. Protein solubility of the samples appears to be more affected by the pre-drying pH than the dilution of the yolk. Yolk powders had higher protein solubility when the pre-drying pH was at yolk's natural pH and at slightly alkaline pH. Due to the inconsistency of the homogenizer, no particular trends of the effect of pretreatments on emulsifying properties were observed, though a slight pH effect was seen on Emulsion Stability Index (ESI) of the yolk powders. As a result of the optimization study in Phase 1, it was concluded that the "best" powder and the "worst" powder could be obtained by spray-drying with pretreatments of pH 8.5 with 6 times dilution and pH 3.0 with 6 times dilution, respectively. Yolk pellet was found to 141 have poor solubility in both liquid and solid form, yet, it was found to have better emulsifying properties, particularly emulsion stabilization, in the liquid form than yolk. In Phase 2 of the study, it was confirmed that the "best" powder with a pretreatment of pH 8.5 and 6 times dilution, was at least as good as the yolk powder control in terms of the protein functionality. Indeed, it was found to have better emulsion stabilization ability than the yolk powder control, as shown in its higher ESI of 28 minutes compared to an ESI of 9.5 minutes for the yolk powder control. The protein solubility of the "best" powder and yolk powder control was similar in Phase 2 (67.65 % and 69.01 %, respectively) in contrast to the higher solubility of the "best" powder in Phase 1. The "worst" powder had similarly poor protein functionality as pellet powder. It should be noted that solubilities of all the powders were relatively higher in Phase 2 than in Phase 1, possibly a reflection of the lower temperatures encountered during spray-drying as evidenced by the outlet temperatures. The overall trend for protein solubility from both phases of this study at pH 6.35 from the highest to the lowest were: liquid yolk control > hand-separated yolk > "best" powder > yolk powder control > "worst" powder > liquid pellet > pellet powder control. In terms of the emulsifying properties, the EAI (m2/g protein) of the samples from the highest to the lowest were: pellet powder > yolk powder control = "best" powder = liquid pellet > liquid yolk > "worst" powder. As for the ESI of the samples, the samples from highest to lowest were: liquid pellet > liquid yolk = yolk powder control = "best" powder > pellet powder = "worst" powder. 142 From the thermograms obtained by Differential Scanning Calorimetry, the samples with the "worst" pretreatment conditions and the pellet samples showed very small or no denaturation peak as the proteins were probably denatured during the process of spray-drying. Water-soluble fraction of yolk was found to be the major component in yolk contributing to the DSC profile and its structural changes due to different pH could be clearly detected. The thermal profile of the freeze-dried water-soluble fraction, which includes the albumen from commercial yolk, was generally more stable than that of freeze-dried yolk, as indicated by the higher T;, Td and enthalpies in different pH environment. The proteins in the water-soluble fraction, which were in fact derived from the same commercial yolk as the freeze-dried yolk, may govern how commercial yolk behaves at different pH environment. The proteins in the water-soluble fraction were more stable than those in freeze-dried yolk in pH 3 buffer and in water, judging from the higher T d in the thermal profile. However, the reverse was observed when these yolk powders were reconstituted at pH 8. Raman Spectroscopy did not yield conclusive results on protein structure due to the dominance of the lipid components in the spectra. The difference spectra obtained for the yolk samples and their corresponding yolk oil had high signal-to-noise ratios. The effects of freeze-drying and spray-drying on the functional properties of the resulting yolk powders'were also compared. Freeze-dried yolk had lower protein solubility than spray-dried yolk, yet the emulsion stability was higher for the freeze-dried sample. Freeze-dried yolk has been reported to have more free fatty acids which may contribute to the lower protein solubility and better hydrophobic and lipophilic balance of the emulsion 143 observed. On the other hand, the protein solubility of spray-dried pellet was lower than that of freeze-dried pellet while the ESI was not significantly different. The protein and lipid interactions in the yolk pellet could have already been affected during the extraction procedure, and spray-drying may have had greater detrimental effect on the solubility of the dried pellet then freeze-drying. In addition, the storage of these samples for 6 months at 4°C could also contribute to the difference in functionality of the samples, perhaps as a result of lipid oxidation. The presence of albumen in the yolk samples may contribute to the better protein solubility and higher ESI of commercial liquid yolk when compared to those parameters of hand-separated yolk. However, other factors contributing to these differences should not be overlooked as there might have been batch differences between the commercial liquid yolk and the hand-separated yolk. It was interesting to note in this study that the emulsion stabilization property of liquid yolk pellet was superior to that of yolk. The removal of the water-soluble fraction could have removed the proteins which negatively affect the stability of yolk emulsion. In addition, the process of the pellet preparation, including acidification and dilution, centrifugation and subsequent neutralization, could have caused some conformational changes in the proteins or lipoproteins in the pellet. These changes might have affected the hydrophobic and lipophilic balance giving yolk pellet higher ESI properties. The yolk pellet is worth further pursuing in the future and perhaps, a preservation method other than spray-drying could be applied to the product. 144 REFERENCES Akita EM and Nakai S (1992) Immunoglobulins from egg yolk: Isolation and purification. J. Food Sci. (57) 629-634 Anton M and Gandemer G (1997) Composition, solubility and emulsifying properties of granules and plasma of egg yolk. J. Food Sci. (62) 484-487 Berquist DH and Stewart GF (1952) Atomization as a factor affecting the quality in spray dried albumen. Food Technol. (6) 201-203 Berquist DH (1995) Chp 14. Egg Dehydration. In: Stadelman WJ and Cotterill OJ (eds.) Egg Science and Technology, 4th ed. Haworth Press, Inc., New York. pp. 335-376 Bradley RL Jr. (1994) Chp 7. Moisture and Total Solids. In: Neilsen SS (ed.) Introduction to the Chemical Analysis of Foods. Jones and Bartlett Publishers, Inc. Boston, MA. pp. 108-109 Bringe NA, Howard DB and Clark DR (1996) Emulsifying properties of low-fat, low-cholesterol egg yolk prepared by supercritical CO2 extraction. J. Food Sci. (61) 19-23,43 Burley RW (1990) The Hen's Egg as a model for food technology. Food Research Quarterly. (50) 42-27 Burley RW and Cook WH (1961) Isolation and composition of avian egg yolk granules and their constituent and lipovitellins. Can. J. Biochem. Physiol. (39) 1295-1307 Burley RW and Vadehra D.V. (1989) Chp 7. Egg Yolk: Structure and Properties. In: The Avian Egg - Chemistry and Biology. John Wiley & Sons, Inc. USA. pp. 171-233 Carey PR (1982) Chp 2. Principles of Raman Spectroscopy. In: Carey PR (ed.) Biochemical applications of Raman and Resonance Raman spectroscopies. Academic Press, New York. pp. 11-47 Carrillo AR and Kokini JL (1988) Effect of egg yolk and egg yolk + salt on rheological properties and particles size distribution of model oil-in-water salad dressing emulsions. J. Food Sci. (53) 1352-1355, 1366 Carlin AF (1955) Relation of acid to the emulsifying action of dried yolk in the preparation of mayonnaise. Poultry Sci. (34) 1186-1187 (abs.) Causeret D, Matringe E and Lorient D (1991) Ionic strength and pH effects on composition and microstructure of yolk granules. J. Food Sci. (56) 1532-1536 145 Chang PK, Powrie WD and Fennema O (1970) Effect of heat treatment on viscosity of yolk. J. Food Sci. (35) 864-867 Cheftel JC, Cuq J and Lorient D (1985) Chp 5. Amino acids, peptides, and proteins. In: Fennama O (ed.) Food Chemistry, 2nded. Marcel Dekker, Inc. New York. pp. 245-370 Chung SL and Ferrier LK (1991a) Conditions affecting emulsifying properties of egg yolk phosvitin. J. Food Sci. (56) 1259-1262 Chung SL and Ferrier LK (1991b) Partial lipid extraction of egg yolk powder: effects on emulsifying properties and soluble protein fraction. J. Food Sci. (56) 1255-1258 Chung SL and Ferrier LK (1992) pH and sodium chloride effects on emulsifying properties of egg yolk phosvitin. J. Food Sci. (57) 40-42 Chung SL and Ferrier LK (1995) Heat denaturation and emulsifying properties of egg yolk phosvitin. J. Food Sci. (60) 906-908, 916 Cobos A, de la Hoz L, Cambero MI and Ordonez JA (1995) Dietary modification and hen strain dependence of egg yolk lipids. Food Research International (28) 71-76 Cofrades S, Carballo J and Careche M (1996) Emulsifying properties of actomyosin from several species. Lebensm.- Wiss. u.- Technol. (29) 379-383 Cunningham FE and Varadarajulu P (1973) Electrophoretic study of fresh, heated, spray dried, and sonicated egg yolk. Poultry Sci. (52) 365-369 Davey EM, Aabbik ME, and Dawson LE (1969) Fresh and frozen egg yolk protein fractions: Emulsion stabilizing power, viscosity and electrophoretic patterns. Poultry Sci. (48)251-160 Davis EA (1994) Chp 34 Thermal Analysis. In: Introduction to the Chemical Analysis of Foods. Nielsen SS (ed.) Jones and Bartlett Publishers, Inc. Boston, USA. pp.505-515 Donovan JW, Mapes CJ, Davis JG and Garibaldi JA (1975) A differential scanning calorimetric study of the stability of egg white to heat denaturation. J. Sci. Food Agric. (26) 73-83 Dyer-Hurdon JN and Nnanna IA (1993) Cholesterol content and functionality of plasma and granules fractionated from egg yolk. J. Food Sci. (58) 1277-1281 Englard S and Seifter S (1990) Precipitation techniques. In: Deutscher MP (ed.) Methods in Enzymology, vol. 182. Academic Press Inc., New York. pp. 285-300 146 Fennama O (1985) Chp 2. Water and Ice. In: Fennama O (ed.) Food Chemistry, 2n a ed. Marcel Dekker, Inc., New York. pp. 23-68 Folch J, Lees M and Sloane-Stanley GH (1957) A simple method for the isolation and purification of total lipids from animal tissue. J. Biol. Chem. (226) 497-509 Francis FJ (1985) Chp 8. Pigments and other colorants. In: Fennama O (ed.) Food Chemistry, 2nd (ed.) Marcel Dekker, Inc. New York. pp. 545-584 Franzen JS, Kuo I and Bobik CM (1970) The temperature stability of egg yolk high density lipoprotein. Biochim. Biophys. Acta (200) 566-569 Franzen KL and Kinsella JE (1976) Functional properties of succinylated and acetylated leaf protein. J. Agric. Food Chem. (24) 914-919 Friberg SE (1997) Chp 1. Emulsion Stability. In: Friberg SE and Larsson K (eds.) Food Emulsions, 3r ed. Marcel Dekker, Inc. New York. pp. 1-56 Gallagher JS and Voss EW Jr. (1970) Immune precipitation of purified chicken antibody at low pH. Immunochem. (7) 771-785 Giovanni M (1983) Response Surface Methodology and Product Optimization. Food Technol. (11)41-45,83 Guardiola F, Codony R, Manich A, Rafecas M and Boatella J (1995) Stability of polyunsaturated fatty acids in egg powder processed and stored under various conditions. J. Agric. Food.Chem. (43) 2254-2259 Hailing PJ (1981) Protein-stabilized foams and emulsions. CRC Crit. Rev. Food Sci. Nutr. (15)155-203 Hassan BVI, Heikal YA, El Shahrigi AF, Mohamed AA (1990) Properties and utilization of spray dried liquid whole egg. Annals of Agric. Sci. (35) 345-358 (abs.) Hatta H, Kim M and Yamamoto T (1990) A novel isolation method for hen egg yolk antibody, "IgY". Agric. Biol. Chem. (54) 2531-2535 Hatta H, Ozeki M and Tsuda K (1997) Chp 11. Egg yolk antibody IgY and its application. In: Yamamoto T, Juneja LR, Hatta H and Kim M (eds.) Hen Eggs - Their basic and applied science. CRC Press, Inc., Florida, pp. 151-178 Hatta H, Tsuda K, Akachi S, Kim M and Yamamoto T (1993) Productivity and some properties of egg yolk antibody (IgY) against human Rotavirus compared with rabbit IgG. Biosci. Biotech. Biochem. (57) 450-454 147 Helrich K (ed.) (1990) Official Methods of Analysis of the Association of Official Analytical Chemists, 15th ed., AOAC Inc., Arlington, Virginia. Methods 920.39, 925.30, 925.32 Hidalgo A, Lucisano M, Comelli EM and Pompei C (1996) Evolution of chemical and physical yolk characteristics during storage of shell eggs. J. Agric. Food Chem. (44) 1447-1452 Hill SE (1996) Chp 6. Emulsions. In: Hall GM (ed.) Methods of testing protein functionality. Chapman and Hall, Great Britain, pp. 153-185 Holm R and Eriksen S (1980) Emulsifying properties of undenatured potato protein concentrates. J. Food Technol. (15) 71-83 Jordan R, Vail GE, Rogler JC and Stadelman WJ (1962) Further studies on eggs from hens on diets differing in fat content. Food Technol. (6) 118-120 Karunajeewa H, Jughes RJ, McDonald MW and Shenstone FS (1984) A review of factors influencing pigmentation of egg yolks. World's Poultry Sci. (40) 52-65 Kim KS, Song IS, Kang TS, Min BY, Suh KB and Song KW (1982) Studies on egg powder (denaturation of egg protein during dehydration). Korean J. Animal Sci. (24)445-451 (abs.) Kiosseoglou VD and Sherman P (1983a) Influence of egg yolk lipoproteins on the rheology and stability of O/W emulsions and mayonnaise. 1. Viscoelasticity of groundnut oil-in-water emulsions and mayonnaise. J. Texture Stud. (14) 397-417 Kiosseoglou VD and Sherman P (1983b) The influence of egg yolk lipoproteins on the rheology and stability of O/W emulsions and mayonnaise. Colloid Polymer Sci. (261) 502-507 Kwan L, Li-Chan E, Helbig N and Nakai S (1991) Fractionation of water-soluble and -insoluble components from egg yolk with minimum use of organic solvents. J. Food Sci. (56) 1537-1541 Lea CH and Hawke JC (1952) Lipovitellin. 2. The influence of water on the stability of lipovitellin and the effects of freezing and drying. Biochem. J. (52) 105-114 Lee CJ, Chang HS, Sheen HS and Hung LT (1991) Studies on storage of dried whole egg. J. Chinese Soc. Animal Sci. (20) 521-530 (abs.) Le Denmat M, Anton M and Gandemer G (1999) Protein denaturation and emulsifying properties of plasma and granules of egg yolk as related to heat treatment. J. Food Sci. (64) 194-197 148 Li-Chan ECY (1998) Methods to monitor process-induced changes in food proteins. In: Shahidi et al. (ed.) Process-Induced Chemical Changes in Foods. Plenum Press, New York. pp. 5-23 Li-Chan E, Nakai S, Hirotsuka M (1994) Chp 8. Raman Spectroscopy as a probe of protein structure in food systems. In: Yada RY, Jackman RL and Smith JL (eds.) Protein Structure - Function Relationships in Foods. Blackie Academic & Professional. New York. pp. 164-169 Li-Chan ECY, Powrie WD, Nakai S (1995) Chp 6. The Chemistry of Eggs and Egg Products. In: Stadelman WJ and Cotterill OJ (eds.) Egg Science and Technology, 4th ed. Haworth Press, Inc., New York. pp. 105-175 Lieu EH, Froning GW and Dam R (1978) Effect of storage on lipid composition and functional properties of dried egg products. Poultry Sci. (57) 912-923 Lu CK and Baker RC (1986) Characteristics of egg yolk phosvitin as an antioxidant for inhibiting metal-catalyzed phospholipid oxidation. Poultry Sci. (65) 2065-2070 Ma CY and Harwalkar VR (1990) Chp 1. Instrumentation and techniques of thermal analysis in food research. In: Harwalkar VR and Ma CY (eds.) Thermal Analysis of Foods. Elsevier Science Publisher Ltd., England, pp.1-15 Marusich WL and Bauernfeind JC (1970) Oxycarotenoids in poultry pigmentation. I. Yolk studies. Poultry Sci. (49) 1555-1566 Marion JE, Woodroof JG and Cook RE (1965) Some physical and chemical properties of eggs from hens of five different stocks. Poultry Sci. (44) 529-534 Martin WG, Tattrie NH and Cook WH (1963) Lipid extraction and distribution studies of egg yolk lipoproteins. Can. J. Biochem. Physiol. (41) 657-666 Masters K (1991) Spray-drying Handbook, 5th ed. Longman Scientific & Technical, England, pp. 159-161, 617-620 Matsumura Y and Mori T (1996) Chp 4. Gelation. In: Hall GM (ed.) Methods of testing protein functionality. Cahapman and Hall. Great Britain, pp. 76-109 Maurice TM and Page CM (1983) Characterize foods with thermal analysis. Canadian Research (12) 19-21 McCully KA, Mok CC and Common RH (1962) Paper electrophoretic characterization of proteins and lipoproteins of hen's egg yolk. Can. J. Biochem. Physiol. (40) 937-952 149 Miller C and Winter AR (1951) Pasteurized frozen whole egg and whole egg and yolk for mayonnaise production. Food Res. (16) 43-49 Mine Y and Bergougnoux M (1998) Adsorption properties of cholesterol-reduced egg yolk low-density lipoprotein at oil-in-water interfaces. J. Agric. Food Chem. (46) 2153-2158 Mine Y (1997) Structural and functional changes of hen's egg yolk low-density lipoproteins with phospholipase A2. J. Agric. Food Chem. (45) 4558-4563 Mine Y (1998) Adsorption behavior of egg yolk low-density lipoproteins in oil-in-water emulsions. J. Agric. Food Chem. (46) 36-41 Mizutani R and Nakamura R (1984) Emulsifying properties of egg yolk low density lipoprotein (LDL): comparison with bovine serum albumin and egg lecithin. Lebensm.- Wiss. u.- Technol. (17) 213-216 Mitzutani R and Nakamura R (1985) The contribution of polypeptide moiety on the emulsifying properties of egg yolk low density lipoprotein (LDL). Lebensm.- Wiss. u.-Technol. (18) 60-63 Mitzutani R and Nakamura R (1987) Emulsifying properties of a complex between apoprotein from hen's egg yolk low density lipoprotein and egg yolk lecithin. Agric. Biol. Chem. (51) 1115-1119 Mullen K and Ennis DM (1979) Rotatable Designs in Product Development. Food Technol. (7) 74-80 Nakamura R and Doi E (2000) Chp 4. Egg processing. In: Nakai S and Modler HW (eds.) Food Proteins - processing applications. Wiley-VCH, Inc. New York. pp. 171-207 Nakamura R, Fukano T and Taniguchi M (1982) Heat-induced gelation of hen's egg yolk low density lipoprotein (LDL) dispersion. J. Food Sci. (47) 1449-1453 Nemecz G and Mennear JH (1995) Phospholipid degradation is induced by heat in cc-tocopherol-enriched eggs. Poultry Sci. (74) 1520-1526 Ohba R, Teramoto Y and Ueda S (1993) Clarification of spray-dried egg yolk suspensions and solubilization of proteins from lipoproteins. J. Food Sci. (58) 307-309 Osuga DT and Teeney RE (1977) Chp 8. Egg Proteins. In: Whitaker JR and Tannenbaum SR (eds.) Food Proteins, AVI Publishing Co., Westport, CT. pp. 209-266 Ozawa Y (1986) Thermal analysis of raw egg. Nippon Shokuhin Kogyo Gakkaishi. (33) 813-820 150 Pankey RD and Stadelman WJ (1969) Effect of dietary fats on some chemical and functional properties of eggs. J. Food Sci. (34) 312-317 Parkinson TL (1977) Effect of storage on the proteins of frozen lipid egg and spray dried-eggs. J. Sci. Food Agric. (28) 811-821 Pearce KN and Kinsella JE (1978) Emulsifying properties of proteins: Evaluation of a turbirrietric technique. J. Agric. Food Chem. (26) 716-723 Powrie WD, Little H and Lopez A (1963) Gelation of egg yolk. J. Food Sci. (28) 38-46 Powrie WD and Nakai S (1985) Chp 14. Characteristics of edible fluids of animal origin: Eggs. In: Fennema OR (ed.) Food Chemistry, 2nd ed. Marcel Dekker, New York, pp. 829-856 Powrie WD and Nakai S (1986) Chemistry of eggs and egg products. In: Stadelman WJ and Cotterill (eds.) Egg Science and Technology, AVI Publishing, Westport, CT, pp. 97-139 Rhodes MB and Lea CH (1957) Phospholipides.4. On the composition of hen's egg phospholipids. Biochem. J. (54) 526-533 Rolfes T, Clements P and Winter AR (1955) The physical and functional properties of lyophilized whole egg, yolk, and white. Food Technol. (18) 1485-1488 Rose D, Gridgeman NT and Fletcher DA (1966) Solids content of eggs. Poultry Sci. (45) 221-226 Saari A, Powrie WD and Fennema O (1964) Isolation and characterization of low-density lipoproteins in native egg yolk plasma. J. Food Sci. (29) 307-315 Sato Y, Nakamura R and Haruta T (1973) Lowering the solubility of freeze-dried egg yolk. Nippon Shokuhin Kogyo Gakkaishi (20) 369-373 (abs.) Schmidt G, Bessman MJ, Flickey MD and Thannhause SJ (1957) The concentrations of some constituents of egg yolk in its soluble phase. J. Biochem. Physiol. (223) 1027-1031 Schultz JR and Forsythe RH (1967) The influence of egg yolk lipoprotein-carbohydrate interaction on baking performance. Baker's Dig. (41) 56-57, 60-62 (abs.) Schultz JR, Snyder HE and Forsythe RH (1966) Interactions of carbohydrates with egg lipoproteins. Poultry Sci. (45) 1122 (abs.) 151 Schultz JR, Synder HE and Forsythe RH (1968) Co-dried carbohydrates effect on the performance of egg yolk solids. J. Food Sci. (33) 507-513 Taborsky G (1963) Interaction between phosvitin and iron and its effect on a rearrangement of phosvitin structure. Biochemistry (2) 266-271 Tanford C (1970) Protein denaturation. Part C. Theoretical models for the mechanism of denaturation. Adv. Protein Chem. (24) 1-93 (abs.) Telis VRN and Kieckbusch TG (1997) Viscoelasticity of frozen/thawed egg yolk. J. Food Sci. (62) 548-550 Tenenhouse HS and Deutsch HE (1966) Some physical-chemical properties of chicken gamma-globulins and their pepsin and papain digestion products. Immunochem. (3) 11-20 Ternes W (1989) A research note: Characterization of water-soluble egg yolk proteins with isoelectric focussing. J. Food Sci. (54) 764-765 Tsuitsui T (1988) Functional properties of heat-treated egg yolk low density lipoprotein. J. Food Sci. (53) 1103-1106 Vadehra DV, Bain JM and Burley RW (1977) Lipid-protein globules of avian egg yolk. Biochem. J. (166)619-624 Varadarajulu P and Cunningham RE (1972a) A study of selected characteristics of hen's egg yolk. 1. Influence of albumen and selected additives. Poultry Sci. (51) 542-546 Varadarajulu P and Cunningham FE (1972b) A study of selected characteristics of hens' egg yolk. 3. Influence of breed, strain, and age of bird. Poultry Sci. (51) 1327-1331 Vojdani F (1996) Chp 2. Solubility. In: Hall GM (ed.) Methods of testing protein functionality. Chapman and Hall, Great Britain, pp. 11-60 Wakamatu T, Sato Y and Saito Y (1982) Relationship between solubility change during dehydration and unfreezable water in egg yolk and low density lipoprotein. J. Agric. Chem. Soc. Japan. (56) 117-122 (abs.) Walstra P (1968) Estimating globule-size distribution of oil-in-water emulsions by spectroturbidity. J. Colloid Interface. (27) 493-398 Watkins BA (1995) Chp 7. Nutritive Value of the Egg. In: Stadelman WJ and Cotterill OJ (eds.) Egg Science and Technology, 4th ed. The Haworth Press, Inc., Binghamton, New York. pp. 177-194 152 Williams J (1962) Serum proteins and the livetins of hen's egg yolk. Biochem. J. (83) 346-355 Wright DJ (1986) Application of Differential Scanning Calorimetry to the study of food behaviour. Analytical Proceedings (23) 389-390 Yang SC (1988) Physical, chemical, functional and microstructural characteristics of egg yolk containing salt or sugar. Dissertation-Abstract-International,-B. (48) 2835 (abs.) Yang SC and Baldwin RE (1995) Chp 16. Functional Properties of Eggs in Foods. In: Stadelman WJ and Cotterill OJ (eds.) Egg Science and Technology, 4th ed. The Haworth Press, Inc., Binghamton, New York. pp. 405-463 Yeadon AD, Goldblatt LA and Altschul HM (1958) Lecithin in oil-in-water emulsions. J. Am. Oil Chemists' Soc. (35) 435-438 Zabik ME (1969) Comparison of frozen, foam-spray-dried, freeze-dried, and spray-dried eggs. 6. Emulsifying properties at three pH levels. Food Tech. (23) 130-132 153 APPENDIX A. Storage Information on Yolk Samples All dried samples were stored in bags made from polypropylene films at 4°C until analyzed. The liquid samples were stored in HDPE bottles with 0.05 % (w/v) sodium azide added to prevent bacterial deterioration. PHASE 1 Phase 1 samples were collected and spray-dried between May and June, 99 Analyses in Phase 1 were carried out between June and September, 99 Phase 1 samples were re-analyzed along with Phase 2 samples for the emulsifying tests in January, 00 PHASE 2 Phase 2 samples were collected and spray-dried in late October, 99 Freeze-drying of the samples were carried out in early November, 99 - DSC and Raman spectroscopy were done in February, 00 in Hong Kong - The solubility and emulsifying tests comparing the spray-dried and freeze-dried yolk samples were done in May, 00 (results in Table 19 and 24) All other analyses in Phase 2 were carried out between November, 99 and January, 00 154 APPENDIX B. List of Abbreviations ANOVA Analysis of variance LDL a w Water activity LPC BSA Bovine serum albumin L Y C CCRD Central composite rotatable ND design O/W COM Commercial spray-dried PP yolk powder PPC DSC Differential Scanning RSM Calorimetry EAI Emulsifying activity index RSR EDTA Ethylene diamine tetraacetic acid SD ESI Emulsion stability index SDS FHY Freeze-dried hand- SH separated yolk powder Td FPC Freeze-dried pellet powder Temp. FYC Freeze-dried yolk control Tj FT Fourier Transform T o u ti et HDL High-density lipoprotein HDPE High-density polyethylene UHP HSY Hand-separated yolk WSF Ht Height AH Enthalpy YPB IgE Immunoglobulin E YPC IgG Immunoglobulin G YPW IgM Immunoglobulin M IgY Immunoglobulin Y Low-density lipoprotein Liquid pellet Liquid yolk control Not dried Oil-in-water Polypropylene Pellet powder Response surface methodology Response surface regression Standard deviation Sodium dodecyl sulfate Sulfhydryl Denaturation temperature Temperature Onset of denaturation Outlet temperature of spray-dryer Ultra high pure Freeze-dried water-soluble fraction "best" powder (pH 8.5, 6x) Yolk powder control "worst" powder (pH 3.0, 6x) 155 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0089800/manifest

Comment

Related Items